diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9408 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1338,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002242152466367713,
+      "grad_norm": 5.668336391448975,
+      "learning_rate": 7.462686567164179e-08,
+      "loss": 0.8487,
+      "step": 1
+    },
+    {
+      "epoch": 0.004484304932735426,
+      "grad_norm": 5.69871187210083,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 0.8597,
+      "step": 2
+    },
+    {
+      "epoch": 0.006726457399103139,
+      "grad_norm": 5.451218605041504,
+      "learning_rate": 2.2388059701492537e-07,
+      "loss": 0.8949,
+      "step": 3
+    },
+    {
+      "epoch": 0.008968609865470852,
+      "grad_norm": 5.7345781326293945,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 0.869,
+      "step": 4
+    },
+    {
+      "epoch": 0.011210762331838564,
+      "grad_norm": 5.690176963806152,
+      "learning_rate": 3.7313432835820895e-07,
+      "loss": 0.886,
+      "step": 5
+    },
+    {
+      "epoch": 0.013452914798206279,
+      "grad_norm": 5.48387336730957,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 0.8755,
+      "step": 6
+    },
+    {
+      "epoch": 0.01569506726457399,
+      "grad_norm": 4.854045391082764,
+      "learning_rate": 5.223880597014925e-07,
+      "loss": 0.8782,
+      "step": 7
+    },
+    {
+      "epoch": 0.017937219730941704,
+      "grad_norm": 5.143599510192871,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 0.8564,
+      "step": 8
+    },
+    {
+      "epoch": 0.020179372197309416,
+      "grad_norm": 3.4202659130096436,
+      "learning_rate": 6.716417910447762e-07,
+      "loss": 0.8308,
+      "step": 9
+    },
+    {
+      "epoch": 0.02242152466367713,
+      "grad_norm": 3.3882153034210205,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 0.8554,
+      "step": 10
+    },
+    {
+      "epoch": 0.02466367713004484,
+      "grad_norm": 3.180309295654297,
+      "learning_rate": 8.208955223880598e-07,
+      "loss": 0.8213,
+      "step": 11
+    },
+    {
+      "epoch": 0.026905829596412557,
+      "grad_norm": 2.136509895324707,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 0.7869,
+      "step": 12
+    },
+    {
+      "epoch": 0.02914798206278027,
+      "grad_norm": 2.1281492710113525,
+      "learning_rate": 9.701492537313434e-07,
+      "loss": 0.7711,
+      "step": 13
+    },
+    {
+      "epoch": 0.03139013452914798,
+      "grad_norm": 1.8773627281188965,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 0.7711,
+      "step": 14
+    },
+    {
+      "epoch": 0.033632286995515695,
+      "grad_norm": 1.7584943771362305,
+      "learning_rate": 1.119402985074627e-06,
+      "loss": 0.7641,
+      "step": 15
+    },
+    {
+      "epoch": 0.03587443946188341,
+      "grad_norm": 1.4110311269760132,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 0.7291,
+      "step": 16
+    },
+    {
+      "epoch": 0.03811659192825112,
+      "grad_norm": 1.217228651046753,
+      "learning_rate": 1.2686567164179105e-06,
+      "loss": 0.7166,
+      "step": 17
+    },
+    {
+      "epoch": 0.04035874439461883,
+      "grad_norm": 1.1579238176345825,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 0.714,
+      "step": 18
+    },
+    {
+      "epoch": 0.042600896860986545,
+      "grad_norm": 1.0090633630752563,
+      "learning_rate": 1.417910447761194e-06,
+      "loss": 0.6909,
+      "step": 19
+    },
+    {
+      "epoch": 0.04484304932735426,
+      "grad_norm": 0.9596092104911804,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 0.7056,
+      "step": 20
+    },
+    {
+      "epoch": 0.04708520179372197,
+      "grad_norm": 0.9225597977638245,
+      "learning_rate": 1.5671641791044779e-06,
+      "loss": 0.6878,
+      "step": 21
+    },
+    {
+      "epoch": 0.04932735426008968,
+      "grad_norm": 0.8473866581916809,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 0.6613,
+      "step": 22
+    },
+    {
+      "epoch": 0.0515695067264574,
+      "grad_norm": 1.1335890293121338,
+      "learning_rate": 1.7164179104477613e-06,
+      "loss": 0.6456,
+      "step": 23
+    },
+    {
+      "epoch": 0.053811659192825115,
+      "grad_norm": 1.018733024597168,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 0.6391,
+      "step": 24
+    },
+    {
+      "epoch": 0.05605381165919283,
+      "grad_norm": 0.8574395775794983,
+      "learning_rate": 1.865671641791045e-06,
+      "loss": 0.627,
+      "step": 25
+    },
+    {
+      "epoch": 0.05829596412556054,
+      "grad_norm": 0.6573004722595215,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 0.6153,
+      "step": 26
+    },
+    {
+      "epoch": 0.06053811659192825,
+      "grad_norm": 0.6763186454772949,
+      "learning_rate": 2.0149253731343284e-06,
+      "loss": 0.6121,
+      "step": 27
+    },
+    {
+      "epoch": 0.06278026905829596,
+      "grad_norm": 0.5623936057090759,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 0.6072,
+      "step": 28
+    },
+    {
+      "epoch": 0.06502242152466367,
+      "grad_norm": 0.4901895523071289,
+      "learning_rate": 2.1641791044776118e-06,
+      "loss": 0.5862,
+      "step": 29
+    },
+    {
+      "epoch": 0.06726457399103139,
+      "grad_norm": 0.565390408039093,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 0.5924,
+      "step": 30
+    },
+    {
+      "epoch": 0.06950672645739911,
+      "grad_norm": 0.6237286329269409,
+      "learning_rate": 2.3134328358208956e-06,
+      "loss": 0.5848,
+      "step": 31
+    },
+    {
+      "epoch": 0.07174887892376682,
+      "grad_norm": 0.6724790930747986,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 0.5748,
+      "step": 32
+    },
+    {
+      "epoch": 0.07399103139013453,
+      "grad_norm": 0.49948763847351074,
+      "learning_rate": 2.4626865671641794e-06,
+      "loss": 0.5846,
+      "step": 33
+    },
+    {
+      "epoch": 0.07623318385650224,
+      "grad_norm": 0.5075132846832275,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 0.5853,
+      "step": 34
+    },
+    {
+      "epoch": 0.07847533632286996,
+      "grad_norm": 0.49447813630104065,
+      "learning_rate": 2.6119402985074627e-06,
+      "loss": 0.5628,
+      "step": 35
+    },
+    {
+      "epoch": 0.08071748878923767,
+      "grad_norm": 0.47228777408599854,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.5649,
+      "step": 36
+    },
+    {
+      "epoch": 0.08295964125560538,
+      "grad_norm": 0.473441481590271,
+      "learning_rate": 2.7611940298507465e-06,
+      "loss": 0.5505,
+      "step": 37
+    },
+    {
+      "epoch": 0.08520179372197309,
+      "grad_norm": 0.4014796018600464,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 0.5562,
+      "step": 38
+    },
+    {
+      "epoch": 0.08744394618834081,
+      "grad_norm": 0.4066062569618225,
+      "learning_rate": 2.9104477611940303e-06,
+      "loss": 0.5618,
+      "step": 39
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 0.4057469069957733,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 0.551,
+      "step": 40
+    },
+    {
+      "epoch": 0.09192825112107623,
+      "grad_norm": 0.4866620898246765,
+      "learning_rate": 3.0597014925373137e-06,
+      "loss": 0.5409,
+      "step": 41
+    },
+    {
+      "epoch": 0.09417040358744394,
+      "grad_norm": 0.37161940336227417,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 0.5374,
+      "step": 42
+    },
+    {
+      "epoch": 0.09641255605381166,
+      "grad_norm": 0.34162813425064087,
+      "learning_rate": 3.208955223880597e-06,
+      "loss": 0.5298,
+      "step": 43
+    },
+    {
+      "epoch": 0.09865470852017937,
+      "grad_norm": 0.3495389223098755,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 0.5303,
+      "step": 44
+    },
+    {
+      "epoch": 0.10089686098654709,
+      "grad_norm": 0.34511131048202515,
+      "learning_rate": 3.3582089552238813e-06,
+      "loss": 0.5333,
+      "step": 45
+    },
+    {
+      "epoch": 0.1031390134529148,
+      "grad_norm": 0.3448837697505951,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 0.5273,
+      "step": 46
+    },
+    {
+      "epoch": 0.10538116591928251,
+      "grad_norm": 0.33499377965927124,
+      "learning_rate": 3.5074626865671646e-06,
+      "loss": 0.5175,
+      "step": 47
+    },
+    {
+      "epoch": 0.10762331838565023,
+      "grad_norm": 0.31002774834632874,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.5185,
+      "step": 48
+    },
+    {
+      "epoch": 0.10986547085201794,
+      "grad_norm": 0.3834333121776581,
+      "learning_rate": 3.656716417910448e-06,
+      "loss": 0.525,
+      "step": 49
+    },
+    {
+      "epoch": 0.11210762331838565,
+      "grad_norm": 0.316884845495224,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 0.5158,
+      "step": 50
+    },
+    {
+      "epoch": 0.11434977578475336,
+      "grad_norm": 0.3580138087272644,
+      "learning_rate": 3.8059701492537314e-06,
+      "loss": 0.5161,
+      "step": 51
+    },
+    {
+      "epoch": 0.11659192825112108,
+      "grad_norm": 0.33753862977027893,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 0.4956,
+      "step": 52
+    },
+    {
+      "epoch": 0.11883408071748879,
+      "grad_norm": 0.2882518768310547,
+      "learning_rate": 3.955223880597015e-06,
+      "loss": 0.502,
+      "step": 53
+    },
+    {
+      "epoch": 0.1210762331838565,
+      "grad_norm": 0.29446494579315186,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 0.5089,
+      "step": 54
+    },
+    {
+      "epoch": 0.12331838565022421,
+      "grad_norm": 0.28530511260032654,
+      "learning_rate": 4.104477611940299e-06,
+      "loss": 0.5189,
+      "step": 55
+    },
+    {
+      "epoch": 0.12556053811659193,
+      "grad_norm": 0.2840350568294525,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 0.5054,
+      "step": 56
+    },
+    {
+      "epoch": 0.12780269058295965,
+      "grad_norm": 0.2816079556941986,
+      "learning_rate": 4.253731343283583e-06,
+      "loss": 0.4888,
+      "step": 57
+    },
+    {
+      "epoch": 0.13004484304932734,
+      "grad_norm": 0.2812719941139221,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 0.4998,
+      "step": 58
+    },
+    {
+      "epoch": 0.13228699551569506,
+      "grad_norm": 0.28168466687202454,
+      "learning_rate": 4.402985074626866e-06,
+      "loss": 0.5051,
+      "step": 59
+    },
+    {
+      "epoch": 0.13452914798206278,
+      "grad_norm": 0.37374943494796753,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 0.5023,
+      "step": 60
+    },
+    {
+      "epoch": 0.1367713004484305,
+      "grad_norm": 0.2893935739994049,
+      "learning_rate": 4.5522388059701495e-06,
+      "loss": 0.4935,
+      "step": 61
+    },
+    {
+      "epoch": 0.13901345291479822,
+      "grad_norm": 0.26644203066825867,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 0.4969,
+      "step": 62
+    },
+    {
+      "epoch": 0.1412556053811659,
+      "grad_norm": 0.4302164316177368,
+      "learning_rate": 4.701492537313434e-06,
+      "loss": 0.4963,
+      "step": 63
+    },
+    {
+      "epoch": 0.14349775784753363,
+      "grad_norm": 0.3002747595310211,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.4908,
+      "step": 64
+    },
+    {
+      "epoch": 0.14573991031390135,
+      "grad_norm": 0.2890976071357727,
+      "learning_rate": 4.850746268656717e-06,
+      "loss": 0.4908,
+      "step": 65
+    },
+    {
+      "epoch": 0.14798206278026907,
+      "grad_norm": 0.41916659474372864,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 0.4814,
+      "step": 66
+    },
+    {
+      "epoch": 0.15022421524663676,
+      "grad_norm": 0.2601127624511719,
+      "learning_rate": 5e-06,
+      "loss": 0.492,
+      "step": 67
+    },
+    {
+      "epoch": 0.15246636771300448,
+      "grad_norm": 0.2654106318950653,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 0.4854,
+      "step": 68
+    },
+    {
+      "epoch": 0.1547085201793722,
+      "grad_norm": 0.28933921456336975,
+      "learning_rate": 5.149253731343285e-06,
+      "loss": 0.4917,
+      "step": 69
+    },
+    {
+      "epoch": 0.15695067264573992,
+      "grad_norm": 0.26526781916618347,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 0.4807,
+      "step": 70
+    },
+    {
+      "epoch": 0.1591928251121076,
+      "grad_norm": 0.2611664831638336,
+      "learning_rate": 5.298507462686567e-06,
+      "loss": 0.4839,
+      "step": 71
+    },
+    {
+      "epoch": 0.16143497757847533,
+      "grad_norm": 0.271530419588089,
+      "learning_rate": 5.37313432835821e-06,
+      "loss": 0.4642,
+      "step": 72
+    },
+    {
+      "epoch": 0.16367713004484305,
+      "grad_norm": 0.2678532898426056,
+      "learning_rate": 5.447761194029851e-06,
+      "loss": 0.4621,
+      "step": 73
+    },
+    {
+      "epoch": 0.16591928251121077,
+      "grad_norm": 0.2560504078865051,
+      "learning_rate": 5.522388059701493e-06,
+      "loss": 0.4812,
+      "step": 74
+    },
+    {
+      "epoch": 0.1681614349775785,
+      "grad_norm": 0.2620652914047241,
+      "learning_rate": 5.597014925373134e-06,
+      "loss": 0.4857,
+      "step": 75
+    },
+    {
+      "epoch": 0.17040358744394618,
+      "grad_norm": 0.25455743074417114,
+      "learning_rate": 5.671641791044776e-06,
+      "loss": 0.4698,
+      "step": 76
+    },
+    {
+      "epoch": 0.1726457399103139,
+      "grad_norm": 0.36607858538627625,
+      "learning_rate": 5.746268656716418e-06,
+      "loss": 0.468,
+      "step": 77
+    },
+    {
+      "epoch": 0.17488789237668162,
+      "grad_norm": 0.4636121094226837,
+      "learning_rate": 5.820895522388061e-06,
+      "loss": 0.4787,
+      "step": 78
+    },
+    {
+      "epoch": 0.17713004484304934,
+      "grad_norm": 0.3027566373348236,
+      "learning_rate": 5.895522388059702e-06,
+      "loss": 0.4756,
+      "step": 79
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 0.5307738780975342,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 0.4651,
+      "step": 80
+    },
+    {
+      "epoch": 0.18161434977578475,
+      "grad_norm": 0.2765562832355499,
+      "learning_rate": 6.044776119402986e-06,
+      "loss": 0.4799,
+      "step": 81
+    },
+    {
+      "epoch": 0.18385650224215247,
+      "grad_norm": 0.27763932943344116,
+      "learning_rate": 6.119402985074627e-06,
+      "loss": 0.4653,
+      "step": 82
+    },
+    {
+      "epoch": 0.1860986547085202,
+      "grad_norm": 0.29424914717674255,
+      "learning_rate": 6.194029850746269e-06,
+      "loss": 0.4644,
+      "step": 83
+    },
+    {
+      "epoch": 0.18834080717488788,
+      "grad_norm": 0.2893185317516327,
+      "learning_rate": 6.2686567164179116e-06,
+      "loss": 0.4647,
+      "step": 84
+    },
+    {
+      "epoch": 0.1905829596412556,
+      "grad_norm": 0.2873184084892273,
+      "learning_rate": 6.343283582089553e-06,
+      "loss": 0.4561,
+      "step": 85
+    },
+    {
+      "epoch": 0.19282511210762332,
+      "grad_norm": 0.3052224814891815,
+      "learning_rate": 6.417910447761194e-06,
+      "loss": 0.4621,
+      "step": 86
+    },
+    {
+      "epoch": 0.19506726457399104,
+      "grad_norm": 0.2775309383869171,
+      "learning_rate": 6.492537313432837e-06,
+      "loss": 0.4633,
+      "step": 87
+    },
+    {
+      "epoch": 0.19730941704035873,
+      "grad_norm": 0.2616846263408661,
+      "learning_rate": 6.567164179104478e-06,
+      "loss": 0.4624,
+      "step": 88
+    },
+    {
+      "epoch": 0.19955156950672645,
+      "grad_norm": 0.29092642664909363,
+      "learning_rate": 6.64179104477612e-06,
+      "loss": 0.4592,
+      "step": 89
+    },
+    {
+      "epoch": 0.20179372197309417,
+      "grad_norm": 0.31288766860961914,
+      "learning_rate": 6.7164179104477625e-06,
+      "loss": 0.4546,
+      "step": 90
+    },
+    {
+      "epoch": 0.2040358744394619,
+      "grad_norm": 0.3467235565185547,
+      "learning_rate": 6.791044776119403e-06,
+      "loss": 0.4644,
+      "step": 91
+    },
+    {
+      "epoch": 0.2062780269058296,
+      "grad_norm": 0.3914223313331604,
+      "learning_rate": 6.865671641791045e-06,
+      "loss": 0.4628,
+      "step": 92
+    },
+    {
+      "epoch": 0.2085201793721973,
+      "grad_norm": 0.3396969437599182,
+      "learning_rate": 6.9402985074626876e-06,
+      "loss": 0.4585,
+      "step": 93
+    },
+    {
+      "epoch": 0.21076233183856502,
+      "grad_norm": 0.4306129217147827,
+      "learning_rate": 7.014925373134329e-06,
+      "loss": 0.458,
+      "step": 94
+    },
+    {
+      "epoch": 0.21300448430493274,
+      "grad_norm": 0.39612337946891785,
+      "learning_rate": 7.089552238805971e-06,
+      "loss": 0.4467,
+      "step": 95
+    },
+    {
+      "epoch": 0.21524663677130046,
+      "grad_norm": 0.3230355381965637,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 0.4621,
+      "step": 96
+    },
+    {
+      "epoch": 0.21748878923766815,
+      "grad_norm": 0.33907926082611084,
+      "learning_rate": 7.238805970149254e-06,
+      "loss": 0.4591,
+      "step": 97
+    },
+    {
+      "epoch": 0.21973094170403587,
+      "grad_norm": 0.3662698566913605,
+      "learning_rate": 7.313432835820896e-06,
+      "loss": 0.4443,
+      "step": 98
+    },
+    {
+      "epoch": 0.2219730941704036,
+      "grad_norm": 0.3996683955192566,
+      "learning_rate": 7.3880597014925385e-06,
+      "loss": 0.4668,
+      "step": 99
+    },
+    {
+      "epoch": 0.2242152466367713,
+      "grad_norm": 0.32585129141807556,
+      "learning_rate": 7.46268656716418e-06,
+      "loss": 0.4596,
+      "step": 100
+    },
+    {
+      "epoch": 0.226457399103139,
+      "grad_norm": 0.33397114276885986,
+      "learning_rate": 7.537313432835821e-06,
+      "loss": 0.4422,
+      "step": 101
+    },
+    {
+      "epoch": 0.22869955156950672,
+      "grad_norm": 0.3338150978088379,
+      "learning_rate": 7.611940298507463e-06,
+      "loss": 0.4565,
+      "step": 102
+    },
+    {
+      "epoch": 0.23094170403587444,
+      "grad_norm": 0.3083316683769226,
+      "learning_rate": 7.686567164179105e-06,
+      "loss": 0.4443,
+      "step": 103
+    },
+    {
+      "epoch": 0.23318385650224216,
+      "grad_norm": 0.32064288854599,
+      "learning_rate": 7.761194029850747e-06,
+      "loss": 0.4559,
+      "step": 104
+    },
+    {
+      "epoch": 0.23542600896860988,
+      "grad_norm": 0.41682082414627075,
+      "learning_rate": 7.835820895522389e-06,
+      "loss": 0.4512,
+      "step": 105
+    },
+    {
+      "epoch": 0.23766816143497757,
+      "grad_norm": 0.33324334025382996,
+      "learning_rate": 7.91044776119403e-06,
+      "loss": 0.456,
+      "step": 106
+    },
+    {
+      "epoch": 0.2399103139013453,
+      "grad_norm": 0.33542048931121826,
+      "learning_rate": 7.985074626865672e-06,
+      "loss": 0.4509,
+      "step": 107
+    },
+    {
+      "epoch": 0.242152466367713,
+      "grad_norm": 0.3272627592086792,
+      "learning_rate": 8.059701492537314e-06,
+      "loss": 0.4544,
+      "step": 108
+    },
+    {
+      "epoch": 0.24439461883408073,
+      "grad_norm": 0.3635421693325043,
+      "learning_rate": 8.134328358208955e-06,
+      "loss": 0.4343,
+      "step": 109
+    },
+    {
+      "epoch": 0.24663677130044842,
+      "grad_norm": 0.4190502464771271,
+      "learning_rate": 8.208955223880599e-06,
+      "loss": 0.446,
+      "step": 110
+    },
+    {
+      "epoch": 0.24887892376681614,
+      "grad_norm": 0.3531770408153534,
+      "learning_rate": 8.283582089552239e-06,
+      "loss": 0.4508,
+      "step": 111
+    },
+    {
+      "epoch": 0.25112107623318386,
+      "grad_norm": 0.39075160026550293,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.4436,
+      "step": 112
+    },
+    {
+      "epoch": 0.2533632286995516,
+      "grad_norm": 0.44263431429862976,
+      "learning_rate": 8.432835820895524e-06,
+      "loss": 0.4462,
+      "step": 113
+    },
+    {
+      "epoch": 0.2556053811659193,
+      "grad_norm": 0.5005448460578918,
+      "learning_rate": 8.507462686567165e-06,
+      "loss": 0.4506,
+      "step": 114
+    },
+    {
+      "epoch": 0.257847533632287,
+      "grad_norm": 0.45026519894599915,
+      "learning_rate": 8.582089552238807e-06,
+      "loss": 0.4459,
+      "step": 115
+    },
+    {
+      "epoch": 0.2600896860986547,
+      "grad_norm": 0.32538262009620667,
+      "learning_rate": 8.656716417910447e-06,
+      "loss": 0.4418,
+      "step": 116
+    },
+    {
+      "epoch": 0.2623318385650224,
+      "grad_norm": 0.33152252435684204,
+      "learning_rate": 8.73134328358209e-06,
+      "loss": 0.4467,
+      "step": 117
+    },
+    {
+      "epoch": 0.2645739910313901,
+      "grad_norm": 0.36994752287864685,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 0.455,
+      "step": 118
+    },
+    {
+      "epoch": 0.26681614349775784,
+      "grad_norm": 0.37849873304367065,
+      "learning_rate": 8.880597014925374e-06,
+      "loss": 0.4389,
+      "step": 119
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 0.36802616715431213,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.442,
+      "step": 120
+    },
+    {
+      "epoch": 0.2713004484304933,
+      "grad_norm": 0.3996352553367615,
+      "learning_rate": 9.029850746268657e-06,
+      "loss": 0.4414,
+      "step": 121
+    },
+    {
+      "epoch": 0.273542600896861,
+      "grad_norm": 0.39996999502182007,
+      "learning_rate": 9.104477611940299e-06,
+      "loss": 0.4323,
+      "step": 122
+    },
+    {
+      "epoch": 0.2757847533632287,
+      "grad_norm": 0.3852199614048004,
+      "learning_rate": 9.17910447761194e-06,
+      "loss": 0.4328,
+      "step": 123
+    },
+    {
+      "epoch": 0.27802690582959644,
+      "grad_norm": 0.3845430910587311,
+      "learning_rate": 9.253731343283582e-06,
+      "loss": 0.4465,
+      "step": 124
+    },
+    {
+      "epoch": 0.2802690582959641,
+      "grad_norm": 0.3522239327430725,
+      "learning_rate": 9.328358208955226e-06,
+      "loss": 0.4497,
+      "step": 125
+    },
+    {
+      "epoch": 0.2825112107623318,
+      "grad_norm": 0.40901991724967957,
+      "learning_rate": 9.402985074626867e-06,
+      "loss": 0.4573,
+      "step": 126
+    },
+    {
+      "epoch": 0.28475336322869954,
+      "grad_norm": 0.3677988350391388,
+      "learning_rate": 9.477611940298507e-06,
+      "loss": 0.4381,
+      "step": 127
+    },
+    {
+      "epoch": 0.28699551569506726,
+      "grad_norm": 0.3076108992099762,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 0.4416,
+      "step": 128
+    },
+    {
+      "epoch": 0.289237668161435,
+      "grad_norm": 0.38808339834213257,
+      "learning_rate": 9.626865671641792e-06,
+      "loss": 0.4504,
+      "step": 129
+    },
+    {
+      "epoch": 0.2914798206278027,
+      "grad_norm": 0.3808380961418152,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 0.43,
+      "step": 130
+    },
+    {
+      "epoch": 0.2937219730941704,
+      "grad_norm": 0.39688679575920105,
+      "learning_rate": 9.776119402985076e-06,
+      "loss": 0.4327,
+      "step": 131
+    },
+    {
+      "epoch": 0.29596412556053814,
+      "grad_norm": 0.36736515164375305,
+      "learning_rate": 9.850746268656717e-06,
+      "loss": 0.4309,
+      "step": 132
+    },
+    {
+      "epoch": 0.2982062780269058,
+      "grad_norm": 0.4459657669067383,
+      "learning_rate": 9.925373134328359e-06,
+      "loss": 0.4553,
+      "step": 133
+    },
+    {
+      "epoch": 0.3004484304932735,
+      "grad_norm": 0.4318934977054596,
+      "learning_rate": 1e-05,
+      "loss": 0.4389,
+      "step": 134
+    },
+    {
+      "epoch": 0.30269058295964124,
+      "grad_norm": 0.44442546367645264,
+      "learning_rate": 9.999982978942698e-06,
+      "loss": 0.443,
+      "step": 135
+    },
+    {
+      "epoch": 0.30493273542600896,
+      "grad_norm": 0.4931248724460602,
+      "learning_rate": 9.999931915886675e-06,
+      "loss": 0.4375,
+      "step": 136
+    },
+    {
+      "epoch": 0.3071748878923767,
+      "grad_norm": 0.45028337836265564,
+      "learning_rate": 9.999846811179592e-06,
+      "loss": 0.4393,
+      "step": 137
+    },
+    {
+      "epoch": 0.3094170403587444,
+      "grad_norm": 0.5610398054122925,
+      "learning_rate": 9.999727665400876e-06,
+      "loss": 0.4437,
+      "step": 138
+    },
+    {
+      "epoch": 0.3116591928251121,
+      "grad_norm": 0.5582858920097351,
+      "learning_rate": 9.999574479361724e-06,
+      "loss": 0.4286,
+      "step": 139
+    },
+    {
+      "epoch": 0.31390134529147984,
+      "grad_norm": 0.6548150777816772,
+      "learning_rate": 9.99938725410509e-06,
+      "loss": 0.4401,
+      "step": 140
+    },
+    {
+      "epoch": 0.31614349775784756,
+      "grad_norm": 0.45949167013168335,
+      "learning_rate": 9.999165990905684e-06,
+      "loss": 0.4343,
+      "step": 141
+    },
+    {
+      "epoch": 0.3183856502242152,
+      "grad_norm": 0.4112301468849182,
+      "learning_rate": 9.998910691269957e-06,
+      "loss": 0.4438,
+      "step": 142
+    },
+    {
+      "epoch": 0.32062780269058294,
+      "grad_norm": 0.3826413154602051,
+      "learning_rate": 9.998621356936098e-06,
+      "loss": 0.4355,
+      "step": 143
+    },
+    {
+      "epoch": 0.32286995515695066,
+      "grad_norm": 0.5311616063117981,
+      "learning_rate": 9.99829798987402e-06,
+      "loss": 0.4601,
+      "step": 144
+    },
+    {
+      "epoch": 0.3251121076233184,
+      "grad_norm": 0.4072250723838806,
+      "learning_rate": 9.997940592285338e-06,
+      "loss": 0.4379,
+      "step": 145
+    },
+    {
+      "epoch": 0.3273542600896861,
+      "grad_norm": 0.41043272614479065,
+      "learning_rate": 9.99754916660337e-06,
+      "loss": 0.4295,
+      "step": 146
+    },
+    {
+      "epoch": 0.3295964125560538,
+      "grad_norm": 0.4894586205482483,
+      "learning_rate": 9.997123715493107e-06,
+      "loss": 0.4414,
+      "step": 147
+    },
+    {
+      "epoch": 0.33183856502242154,
+      "grad_norm": 0.3155098557472229,
+      "learning_rate": 9.996664241851197e-06,
+      "loss": 0.4324,
+      "step": 148
+    },
+    {
+      "epoch": 0.33408071748878926,
+      "grad_norm": 0.4175388813018799,
+      "learning_rate": 9.996170748805935e-06,
+      "loss": 0.4411,
+      "step": 149
+    },
+    {
+      "epoch": 0.336322869955157,
+      "grad_norm": 0.3918183147907257,
+      "learning_rate": 9.995643239717228e-06,
+      "loss": 0.4524,
+      "step": 150
+    },
+    {
+      "epoch": 0.33856502242152464,
+      "grad_norm": 0.3581607937812805,
+      "learning_rate": 9.99508171817658e-06,
+      "loss": 0.4308,
+      "step": 151
+    },
+    {
+      "epoch": 0.34080717488789236,
+      "grad_norm": 0.36461561918258667,
+      "learning_rate": 9.994486188007071e-06,
+      "loss": 0.4294,
+      "step": 152
+    },
+    {
+      "epoch": 0.3430493273542601,
+      "grad_norm": 0.4068436622619629,
+      "learning_rate": 9.99385665326332e-06,
+      "loss": 0.4298,
+      "step": 153
+    },
+    {
+      "epoch": 0.3452914798206278,
+      "grad_norm": 0.40921491384506226,
+      "learning_rate": 9.993193118231463e-06,
+      "loss": 0.441,
+      "step": 154
+    },
+    {
+      "epoch": 0.3475336322869955,
+      "grad_norm": 0.3476938307285309,
+      "learning_rate": 9.99249558742913e-06,
+      "loss": 0.4309,
+      "step": 155
+    },
+    {
+      "epoch": 0.34977578475336324,
+      "grad_norm": 0.3763669729232788,
+      "learning_rate": 9.991764065605407e-06,
+      "loss": 0.4311,
+      "step": 156
+    },
+    {
+      "epoch": 0.35201793721973096,
+      "grad_norm": 0.4030418395996094,
+      "learning_rate": 9.9909985577408e-06,
+      "loss": 0.4409,
+      "step": 157
+    },
+    {
+      "epoch": 0.3542600896860987,
+      "grad_norm": 0.35823819041252136,
+      "learning_rate": 9.990199069047216e-06,
+      "loss": 0.4333,
+      "step": 158
+    },
+    {
+      "epoch": 0.35650224215246634,
+      "grad_norm": 0.3343595862388611,
+      "learning_rate": 9.989365604967906e-06,
+      "loss": 0.4483,
+      "step": 159
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 0.39926877617836,
+      "learning_rate": 9.98849817117745e-06,
+      "loss": 0.4283,
+      "step": 160
+    },
+    {
+      "epoch": 0.3609865470852018,
+      "grad_norm": 0.4145961403846741,
+      "learning_rate": 9.987596773581702e-06,
+      "loss": 0.4442,
+      "step": 161
+    },
+    {
+      "epoch": 0.3632286995515695,
+      "grad_norm": 0.3885423541069031,
+      "learning_rate": 9.986661418317759e-06,
+      "loss": 0.4288,
+      "step": 162
+    },
+    {
+      "epoch": 0.3654708520179372,
+      "grad_norm": 0.4642771780490875,
+      "learning_rate": 9.985692111753915e-06,
+      "loss": 0.4374,
+      "step": 163
+    },
+    {
+      "epoch": 0.36771300448430494,
+      "grad_norm": 0.4342643618583679,
+      "learning_rate": 9.984688860489619e-06,
+      "loss": 0.4516,
+      "step": 164
+    },
+    {
+      "epoch": 0.36995515695067266,
+      "grad_norm": 0.35809198021888733,
+      "learning_rate": 9.983651671355432e-06,
+      "loss": 0.4284,
+      "step": 165
+    },
+    {
+      "epoch": 0.3721973094170404,
+      "grad_norm": 0.382501482963562,
+      "learning_rate": 9.982580551412972e-06,
+      "loss": 0.443,
+      "step": 166
+    },
+    {
+      "epoch": 0.3744394618834081,
+      "grad_norm": 0.36501315236091614,
+      "learning_rate": 9.981475507954882e-06,
+      "loss": 0.4373,
+      "step": 167
+    },
+    {
+      "epoch": 0.37668161434977576,
+      "grad_norm": 0.3442380130290985,
+      "learning_rate": 9.980336548504759e-06,
+      "loss": 0.435,
+      "step": 168
+    },
+    {
+      "epoch": 0.3789237668161435,
+      "grad_norm": 0.46496739983558655,
+      "learning_rate": 9.979163680817124e-06,
+      "loss": 0.4358,
+      "step": 169
+    },
+    {
+      "epoch": 0.3811659192825112,
+      "grad_norm": 0.3578864336013794,
+      "learning_rate": 9.977956912877356e-06,
+      "loss": 0.429,
+      "step": 170
+    },
+    {
+      "epoch": 0.3834080717488789,
+      "grad_norm": 0.34448155760765076,
+      "learning_rate": 9.976716252901643e-06,
+      "loss": 0.4206,
+      "step": 171
+    },
+    {
+      "epoch": 0.38565022421524664,
+      "grad_norm": 0.4001348316669464,
+      "learning_rate": 9.97544170933692e-06,
+      "loss": 0.4328,
+      "step": 172
+    },
+    {
+      "epoch": 0.38789237668161436,
+      "grad_norm": 0.35518965125083923,
+      "learning_rate": 9.974133290860821e-06,
+      "loss": 0.4328,
+      "step": 173
+    },
+    {
+      "epoch": 0.3901345291479821,
+      "grad_norm": 0.3327184319496155,
+      "learning_rate": 9.97279100638161e-06,
+      "loss": 0.4201,
+      "step": 174
+    },
+    {
+      "epoch": 0.3923766816143498,
+      "grad_norm": 0.3277827799320221,
+      "learning_rate": 9.97141486503813e-06,
+      "loss": 0.4226,
+      "step": 175
+    },
+    {
+      "epoch": 0.39461883408071746,
+      "grad_norm": 0.3041229546070099,
+      "learning_rate": 9.970004876199731e-06,
+      "loss": 0.4203,
+      "step": 176
+    },
+    {
+      "epoch": 0.3968609865470852,
+      "grad_norm": 0.30995795130729675,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.4347,
+      "step": 177
+    },
+    {
+      "epoch": 0.3991031390134529,
+      "grad_norm": 0.34650734066963196,
+      "learning_rate": 9.967083394667763e-06,
+      "loss": 0.4309,
+      "step": 178
+    },
+    {
+      "epoch": 0.4013452914798206,
+      "grad_norm": 0.3217341899871826,
+      "learning_rate": 9.965571921864875e-06,
+      "loss": 0.4413,
+      "step": 179
+    },
+    {
+      "epoch": 0.40358744394618834,
+      "grad_norm": 0.38960036635398865,
+      "learning_rate": 9.964026641348298e-06,
+      "loss": 0.4255,
+      "step": 180
+    },
+    {
+      "epoch": 0.40582959641255606,
+      "grad_norm": 0.2990666925907135,
+      "learning_rate": 9.962447563638954e-06,
+      "loss": 0.4172,
+      "step": 181
+    },
+    {
+      "epoch": 0.4080717488789238,
+      "grad_norm": 0.3317890763282776,
+      "learning_rate": 9.960834699487873e-06,
+      "loss": 0.4345,
+      "step": 182
+    },
+    {
+      "epoch": 0.4103139013452915,
+      "grad_norm": 0.2946619391441345,
+      "learning_rate": 9.959188059876115e-06,
+      "loss": 0.4276,
+      "step": 183
+    },
+    {
+      "epoch": 0.4125560538116592,
+      "grad_norm": 0.32814356684684753,
+      "learning_rate": 9.9575076560147e-06,
+      "loss": 0.4241,
+      "step": 184
+    },
+    {
+      "epoch": 0.4147982062780269,
+      "grad_norm": 0.3361572027206421,
+      "learning_rate": 9.955793499344526e-06,
+      "loss": 0.4194,
+      "step": 185
+    },
+    {
+      "epoch": 0.4170403587443946,
+      "grad_norm": 0.3031604588031769,
+      "learning_rate": 9.9540456015363e-06,
+      "loss": 0.4303,
+      "step": 186
+    },
+    {
+      "epoch": 0.4192825112107623,
+      "grad_norm": 0.3185907006263733,
+      "learning_rate": 9.952263974490447e-06,
+      "loss": 0.4335,
+      "step": 187
+    },
+    {
+      "epoch": 0.42152466367713004,
+      "grad_norm": 0.35873353481292725,
+      "learning_rate": 9.950448630337037e-06,
+      "loss": 0.4257,
+      "step": 188
+    },
+    {
+      "epoch": 0.42376681614349776,
+      "grad_norm": 0.36934223771095276,
+      "learning_rate": 9.948599581435704e-06,
+      "loss": 0.4322,
+      "step": 189
+    },
+    {
+      "epoch": 0.4260089686098655,
+      "grad_norm": 0.32821154594421387,
+      "learning_rate": 9.946716840375552e-06,
+      "loss": 0.4323,
+      "step": 190
+    },
+    {
+      "epoch": 0.4282511210762332,
+      "grad_norm": 0.3423391878604889,
+      "learning_rate": 9.944800419975078e-06,
+      "loss": 0.4161,
+      "step": 191
+    },
+    {
+      "epoch": 0.4304932735426009,
+      "grad_norm": 0.4166717827320099,
+      "learning_rate": 9.942850333282085e-06,
+      "loss": 0.4198,
+      "step": 192
+    },
+    {
+      "epoch": 0.4327354260089686,
+      "grad_norm": 0.31524884700775146,
+      "learning_rate": 9.940866593573586e-06,
+      "loss": 0.4288,
+      "step": 193
+    },
+    {
+      "epoch": 0.4349775784753363,
+      "grad_norm": 0.5636321902275085,
+      "learning_rate": 9.938849214355722e-06,
+      "loss": 0.43,
+      "step": 194
+    },
+    {
+      "epoch": 0.437219730941704,
+      "grad_norm": 0.3374340832233429,
+      "learning_rate": 9.93679820936366e-06,
+      "loss": 0.4153,
+      "step": 195
+    },
+    {
+      "epoch": 0.43946188340807174,
+      "grad_norm": 0.36334872245788574,
+      "learning_rate": 9.934713592561513e-06,
+      "loss": 0.433,
+      "step": 196
+    },
+    {
+      "epoch": 0.44170403587443946,
+      "grad_norm": 0.40390753746032715,
+      "learning_rate": 9.932595378142233e-06,
+      "loss": 0.4127,
+      "step": 197
+    },
+    {
+      "epoch": 0.4439461883408072,
+      "grad_norm": 0.36284729838371277,
+      "learning_rate": 9.93044358052752e-06,
+      "loss": 0.4271,
+      "step": 198
+    },
+    {
+      "epoch": 0.4461883408071749,
+      "grad_norm": 0.31764981150627136,
+      "learning_rate": 9.928258214367721e-06,
+      "loss": 0.4138,
+      "step": 199
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 0.35598257184028625,
+      "learning_rate": 9.926039294541733e-06,
+      "loss": 0.4195,
+      "step": 200
+    },
+    {
+      "epoch": 0.45067264573991034,
+      "grad_norm": 0.3484888970851898,
+      "learning_rate": 9.923786836156902e-06,
+      "loss": 0.4207,
+      "step": 201
+    },
+    {
+      "epoch": 0.452914798206278,
+      "grad_norm": 0.3748074769973755,
+      "learning_rate": 9.921500854548916e-06,
+      "loss": 0.4397,
+      "step": 202
+    },
+    {
+      "epoch": 0.4551569506726457,
+      "grad_norm": 0.3167206645011902,
+      "learning_rate": 9.919181365281704e-06,
+      "loss": 0.4302,
+      "step": 203
+    },
+    {
+      "epoch": 0.45739910313901344,
+      "grad_norm": 0.34554246068000793,
+      "learning_rate": 9.91682838414733e-06,
+      "loss": 0.4212,
+      "step": 204
+    },
+    {
+      "epoch": 0.45964125560538116,
+      "grad_norm": 0.36509010195732117,
+      "learning_rate": 9.91444192716589e-06,
+      "loss": 0.4226,
+      "step": 205
+    },
+    {
+      "epoch": 0.4618834080717489,
+      "grad_norm": 0.33615776896476746,
+      "learning_rate": 9.912022010585385e-06,
+      "loss": 0.4323,
+      "step": 206
+    },
+    {
+      "epoch": 0.4641255605381166,
+      "grad_norm": 0.3156930208206177,
+      "learning_rate": 9.909568650881633e-06,
+      "loss": 0.4145,
+      "step": 207
+    },
+    {
+      "epoch": 0.4663677130044843,
+      "grad_norm": 0.324225515127182,
+      "learning_rate": 9.907081864758145e-06,
+      "loss": 0.4147,
+      "step": 208
+    },
+    {
+      "epoch": 0.46860986547085204,
+      "grad_norm": 0.32544881105422974,
+      "learning_rate": 9.904561669146015e-06,
+      "loss": 0.4134,
+      "step": 209
+    },
+    {
+      "epoch": 0.47085201793721976,
+      "grad_norm": 0.3008864223957062,
+      "learning_rate": 9.902008081203796e-06,
+      "loss": 0.4116,
+      "step": 210
+    },
+    {
+      "epoch": 0.4730941704035874,
+      "grad_norm": 0.3278709650039673,
+      "learning_rate": 9.899421118317399e-06,
+      "loss": 0.4313,
+      "step": 211
+    },
+    {
+      "epoch": 0.47533632286995514,
+      "grad_norm": 0.29143965244293213,
+      "learning_rate": 9.896800798099957e-06,
+      "loss": 0.4089,
+      "step": 212
+    },
+    {
+      "epoch": 0.47757847533632286,
+      "grad_norm": 0.3461761474609375,
+      "learning_rate": 9.89414713839172e-06,
+      "loss": 0.4205,
+      "step": 213
+    },
+    {
+      "epoch": 0.4798206278026906,
+      "grad_norm": 0.3265862464904785,
+      "learning_rate": 9.89146015725993e-06,
+      "loss": 0.4115,
+      "step": 214
+    },
+    {
+      "epoch": 0.4820627802690583,
+      "grad_norm": 0.3013107180595398,
+      "learning_rate": 9.888739872998682e-06,
+      "loss": 0.4205,
+      "step": 215
+    },
+    {
+      "epoch": 0.484304932735426,
+      "grad_norm": 0.3210050165653229,
+      "learning_rate": 9.88598630412883e-06,
+      "loss": 0.4251,
+      "step": 216
+    },
+    {
+      "epoch": 0.48654708520179374,
+      "grad_norm": 0.28820371627807617,
+      "learning_rate": 9.883199469397832e-06,
+      "loss": 0.422,
+      "step": 217
+    },
+    {
+      "epoch": 0.48878923766816146,
+      "grad_norm": 0.29091912508010864,
+      "learning_rate": 9.880379387779637e-06,
+      "loss": 0.4211,
+      "step": 218
+    },
+    {
+      "epoch": 0.4910313901345291,
+      "grad_norm": 0.3057043254375458,
+      "learning_rate": 9.877526078474553e-06,
+      "loss": 0.4228,
+      "step": 219
+    },
+    {
+      "epoch": 0.49327354260089684,
+      "grad_norm": 0.4122786819934845,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.4156,
+      "step": 220
+    },
+    {
+      "epoch": 0.49551569506726456,
+      "grad_norm": 0.34951844811439514,
+      "learning_rate": 9.871719854735964e-06,
+      "loss": 0.4199,
+      "step": 221
+    },
+    {
+      "epoch": 0.4977578475336323,
+      "grad_norm": 0.33593615889549255,
+      "learning_rate": 9.868766979833686e-06,
+      "loss": 0.419,
+      "step": 222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3360561728477478,
+      "learning_rate": 9.865780956306703e-06,
+      "loss": 0.4172,
+      "step": 223
+    },
+    {
+      "epoch": 0.5022421524663677,
+      "grad_norm": 0.33042022585868835,
+      "learning_rate": 9.862761804485129e-06,
+      "loss": 0.4144,
+      "step": 224
+    },
+    {
+      "epoch": 0.5044843049327354,
+      "grad_norm": 0.3152481019496918,
+      "learning_rate": 9.859709544924624e-06,
+      "loss": 0.4225,
+      "step": 225
+    },
+    {
+      "epoch": 0.5067264573991032,
+      "grad_norm": 0.33611994981765747,
+      "learning_rate": 9.856624198406262e-06,
+      "loss": 0.4296,
+      "step": 226
+    },
+    {
+      "epoch": 0.5089686098654709,
+      "grad_norm": 0.3388019800186157,
+      "learning_rate": 9.85350578593639e-06,
+      "loss": 0.4307,
+      "step": 227
+    },
+    {
+      "epoch": 0.5112107623318386,
+      "grad_norm": 0.2809235751628876,
+      "learning_rate": 9.850354328746474e-06,
+      "loss": 0.4159,
+      "step": 228
+    },
+    {
+      "epoch": 0.5134529147982063,
+      "grad_norm": 0.30323678255081177,
+      "learning_rate": 9.847169848292971e-06,
+      "loss": 0.4201,
+      "step": 229
+    },
+    {
+      "epoch": 0.515695067264574,
+      "grad_norm": 0.3410962224006653,
+      "learning_rate": 9.84395236625717e-06,
+      "loss": 0.4075,
+      "step": 230
+    },
+    {
+      "epoch": 0.5179372197309418,
+      "grad_norm": 0.3696151673793793,
+      "learning_rate": 9.84070190454505e-06,
+      "loss": 0.4251,
+      "step": 231
+    },
+    {
+      "epoch": 0.5201793721973094,
+      "grad_norm": 0.4791179895401001,
+      "learning_rate": 9.837418485287126e-06,
+      "loss": 0.4299,
+      "step": 232
+    },
+    {
+      "epoch": 0.5224215246636771,
+      "grad_norm": 0.36019060015678406,
+      "learning_rate": 9.83410213083831e-06,
+      "loss": 0.4129,
+      "step": 233
+    },
+    {
+      "epoch": 0.5246636771300448,
+      "grad_norm": 0.34516841173171997,
+      "learning_rate": 9.830752863777741e-06,
+      "loss": 0.4277,
+      "step": 234
+    },
+    {
+      "epoch": 0.5269058295964125,
+      "grad_norm": 0.3418406844139099,
+      "learning_rate": 9.827370706908647e-06,
+      "loss": 0.4152,
+      "step": 235
+    },
+    {
+      "epoch": 0.5291479820627802,
+      "grad_norm": 0.39434462785720825,
+      "learning_rate": 9.823955683258185e-06,
+      "loss": 0.4276,
+      "step": 236
+    },
+    {
+      "epoch": 0.531390134529148,
+      "grad_norm": 0.3335305452346802,
+      "learning_rate": 9.820507816077275e-06,
+      "loss": 0.4169,
+      "step": 237
+    },
+    {
+      "epoch": 0.5336322869955157,
+      "grad_norm": 0.3423804044723511,
+      "learning_rate": 9.817027128840462e-06,
+      "loss": 0.4089,
+      "step": 238
+    },
+    {
+      "epoch": 0.5358744394618834,
+      "grad_norm": 0.3174763321876526,
+      "learning_rate": 9.81351364524573e-06,
+      "loss": 0.4188,
+      "step": 239
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 0.3633676767349243,
+      "learning_rate": 9.809967389214366e-06,
+      "loss": 0.4188,
+      "step": 240
+    },
+    {
+      "epoch": 0.5403587443946188,
+      "grad_norm": 0.3306373357772827,
+      "learning_rate": 9.806388384890777e-06,
+      "loss": 0.4164,
+      "step": 241
+    },
+    {
+      "epoch": 0.5426008968609866,
+      "grad_norm": 0.3439992368221283,
+      "learning_rate": 9.802776656642341e-06,
+      "loss": 0.4266,
+      "step": 242
+    },
+    {
+      "epoch": 0.5448430493273543,
+      "grad_norm": 0.370136022567749,
+      "learning_rate": 9.799132229059229e-06,
+      "loss": 0.4112,
+      "step": 243
+    },
+    {
+      "epoch": 0.547085201793722,
+      "grad_norm": 0.3599247932434082,
+      "learning_rate": 9.795455126954247e-06,
+      "loss": 0.4176,
+      "step": 244
+    },
+    {
+      "epoch": 0.5493273542600897,
+      "grad_norm": 0.34241363406181335,
+      "learning_rate": 9.791745375362662e-06,
+      "loss": 0.407,
+      "step": 245
+    },
+    {
+      "epoch": 0.5515695067264574,
+      "grad_norm": 0.32444271445274353,
+      "learning_rate": 9.78800299954203e-06,
+      "loss": 0.4169,
+      "step": 246
+    },
+    {
+      "epoch": 0.5538116591928252,
+      "grad_norm": 0.40615493059158325,
+      "learning_rate": 9.78422802497203e-06,
+      "loss": 0.4157,
+      "step": 247
+    },
+    {
+      "epoch": 0.5560538116591929,
+      "grad_norm": 0.39495211839675903,
+      "learning_rate": 9.780420477354283e-06,
+      "loss": 0.4263,
+      "step": 248
+    },
+    {
+      "epoch": 0.5582959641255605,
+      "grad_norm": 0.3028656542301178,
+      "learning_rate": 9.776580382612186e-06,
+      "loss": 0.4074,
+      "step": 249
+    },
+    {
+      "epoch": 0.5605381165919282,
+      "grad_norm": 0.36316531896591187,
+      "learning_rate": 9.772707766890726e-06,
+      "loss": 0.427,
+      "step": 250
+    },
+    {
+      "epoch": 0.5627802690582959,
+      "grad_norm": 0.35423365235328674,
+      "learning_rate": 9.768802656556309e-06,
+      "loss": 0.417,
+      "step": 251
+    },
+    {
+      "epoch": 0.5650224215246636,
+      "grad_norm": 0.3939313292503357,
+      "learning_rate": 9.76486507819658e-06,
+      "loss": 0.4097,
+      "step": 252
+    },
+    {
+      "epoch": 0.5672645739910314,
+      "grad_norm": 0.3233250379562378,
+      "learning_rate": 9.760895058620236e-06,
+      "loss": 0.4084,
+      "step": 253
+    },
+    {
+      "epoch": 0.5695067264573991,
+      "grad_norm": 0.351610004901886,
+      "learning_rate": 9.756892624856848e-06,
+      "loss": 0.4187,
+      "step": 254
+    },
+    {
+      "epoch": 0.5717488789237668,
+      "grad_norm": 0.3417075276374817,
+      "learning_rate": 9.752857804156678e-06,
+      "loss": 0.4142,
+      "step": 255
+    },
+    {
+      "epoch": 0.5739910313901345,
+      "grad_norm": 0.36849814653396606,
+      "learning_rate": 9.748790623990497e-06,
+      "loss": 0.4216,
+      "step": 256
+    },
+    {
+      "epoch": 0.5762331838565022,
+      "grad_norm": 0.3799193799495697,
+      "learning_rate": 9.744691112049381e-06,
+      "loss": 0.4191,
+      "step": 257
+    },
+    {
+      "epoch": 0.57847533632287,
+      "grad_norm": 0.3511918783187866,
+      "learning_rate": 9.740559296244543e-06,
+      "loss": 0.4108,
+      "step": 258
+    },
+    {
+      "epoch": 0.5807174887892377,
+      "grad_norm": 0.34906819462776184,
+      "learning_rate": 9.736395204707133e-06,
+      "loss": 0.4065,
+      "step": 259
+    },
+    {
+      "epoch": 0.5829596412556054,
+      "grad_norm": 0.31337782740592957,
+      "learning_rate": 9.732198865788047e-06,
+      "loss": 0.4039,
+      "step": 260
+    },
+    {
+      "epoch": 0.5852017937219731,
+      "grad_norm": 0.38677307963371277,
+      "learning_rate": 9.727970308057736e-06,
+      "loss": 0.4294,
+      "step": 261
+    },
+    {
+      "epoch": 0.5874439461883408,
+      "grad_norm": 0.47673216462135315,
+      "learning_rate": 9.723709560306009e-06,
+      "loss": 0.4124,
+      "step": 262
+    },
+    {
+      "epoch": 0.5896860986547086,
+      "grad_norm": 0.3405997157096863,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.406,
+      "step": 263
+    },
+    {
+      "epoch": 0.5919282511210763,
+      "grad_norm": 0.31773215532302856,
+      "learning_rate": 9.715091610993162e-06,
+      "loss": 0.4019,
+      "step": 264
+    },
+    {
+      "epoch": 0.594170403587444,
+      "grad_norm": 0.40832459926605225,
+      "learning_rate": 9.710734468106686e-06,
+      "loss": 0.4208,
+      "step": 265
+    },
+    {
+      "epoch": 0.5964125560538116,
+      "grad_norm": 0.3773450553417206,
+      "learning_rate": 9.706345252547681e-06,
+      "loss": 0.3997,
+      "step": 266
+    },
+    {
+      "epoch": 0.5986547085201793,
+      "grad_norm": 0.40097302198410034,
+      "learning_rate": 9.701923994199784e-06,
+      "loss": 0.4318,
+      "step": 267
+    },
+    {
+      "epoch": 0.600896860986547,
+      "grad_norm": 0.27954745292663574,
+      "learning_rate": 9.697470723164792e-06,
+      "loss": 0.4088,
+      "step": 268
+    },
+    {
+      "epoch": 0.6031390134529148,
+      "grad_norm": 0.33827874064445496,
+      "learning_rate": 9.692985469762454e-06,
+      "loss": 0.4204,
+      "step": 269
+    },
+    {
+      "epoch": 0.6053811659192825,
+      "grad_norm": 0.34893351793289185,
+      "learning_rate": 9.688468264530278e-06,
+      "loss": 0.4098,
+      "step": 270
+    },
+    {
+      "epoch": 0.6076233183856502,
+      "grad_norm": 0.3191416561603546,
+      "learning_rate": 9.683919138223301e-06,
+      "loss": 0.427,
+      "step": 271
+    },
+    {
+      "epoch": 0.6098654708520179,
+      "grad_norm": 0.41698306798934937,
+      "learning_rate": 9.679338121813903e-06,
+      "loss": 0.4109,
+      "step": 272
+    },
+    {
+      "epoch": 0.6121076233183856,
+      "grad_norm": 0.2933104634284973,
+      "learning_rate": 9.674725246491583e-06,
+      "loss": 0.4081,
+      "step": 273
+    },
+    {
+      "epoch": 0.6143497757847534,
+      "grad_norm": 0.34212368726730347,
+      "learning_rate": 9.670080543662742e-06,
+      "loss": 0.4154,
+      "step": 274
+    },
+    {
+      "epoch": 0.6165919282511211,
+      "grad_norm": 0.33427420258522034,
+      "learning_rate": 9.665404044950482e-06,
+      "loss": 0.4074,
+      "step": 275
+    },
+    {
+      "epoch": 0.6188340807174888,
+      "grad_norm": 0.32226645946502686,
+      "learning_rate": 9.660695782194386e-06,
+      "loss": 0.4093,
+      "step": 276
+    },
+    {
+      "epoch": 0.6210762331838565,
+      "grad_norm": 0.35459184646606445,
+      "learning_rate": 9.655955787450299e-06,
+      "loss": 0.4128,
+      "step": 277
+    },
+    {
+      "epoch": 0.6233183856502242,
+      "grad_norm": 0.3879610598087311,
+      "learning_rate": 9.651184092990109e-06,
+      "loss": 0.4102,
+      "step": 278
+    },
+    {
+      "epoch": 0.625560538116592,
+      "grad_norm": 0.3073469400405884,
+      "learning_rate": 9.64638073130153e-06,
+      "loss": 0.4145,
+      "step": 279
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 0.3368130326271057,
+      "learning_rate": 9.641545735087877e-06,
+      "loss": 0.4004,
+      "step": 280
+    },
+    {
+      "epoch": 0.6300448430493274,
+      "grad_norm": 0.3618389070034027,
+      "learning_rate": 9.636679137267852e-06,
+      "loss": 0.4256,
+      "step": 281
+    },
+    {
+      "epoch": 0.6322869955156951,
+      "grad_norm": 0.3407100439071655,
+      "learning_rate": 9.631780970975311e-06,
+      "loss": 0.4142,
+      "step": 282
+    },
+    {
+      "epoch": 0.6345291479820628,
+      "grad_norm": 0.36761343479156494,
+      "learning_rate": 9.62685126955904e-06,
+      "loss": 0.4076,
+      "step": 283
+    },
+    {
+      "epoch": 0.6367713004484304,
+      "grad_norm": 0.3400237262248993,
+      "learning_rate": 9.621890066582534e-06,
+      "loss": 0.4087,
+      "step": 284
+    },
+    {
+      "epoch": 0.6390134529147982,
+      "grad_norm": 0.3334314227104187,
+      "learning_rate": 9.616897395823758e-06,
+      "loss": 0.4056,
+      "step": 285
+    },
+    {
+      "epoch": 0.6412556053811659,
+      "grad_norm": 0.3367193043231964,
+      "learning_rate": 9.611873291274927e-06,
+      "loss": 0.4028,
+      "step": 286
+    },
+    {
+      "epoch": 0.6434977578475336,
+      "grad_norm": 0.3631322383880615,
+      "learning_rate": 9.60681778714227e-06,
+      "loss": 0.4019,
+      "step": 287
+    },
+    {
+      "epoch": 0.6457399103139013,
+      "grad_norm": 0.29291173815727234,
+      "learning_rate": 9.601730917845798e-06,
+      "loss": 0.4126,
+      "step": 288
+    },
+    {
+      "epoch": 0.647982062780269,
+      "grad_norm": 0.32067742943763733,
+      "learning_rate": 9.596612718019067e-06,
+      "loss": 0.4193,
+      "step": 289
+    },
+    {
+      "epoch": 0.6502242152466368,
+      "grad_norm": 0.33895620703697205,
+      "learning_rate": 9.591463222508947e-06,
+      "loss": 0.4162,
+      "step": 290
+    },
+    {
+      "epoch": 0.6524663677130045,
+      "grad_norm": 0.332032710313797,
+      "learning_rate": 9.586282466375381e-06,
+      "loss": 0.4132,
+      "step": 291
+    },
+    {
+      "epoch": 0.6547085201793722,
+      "grad_norm": 0.3511500656604767,
+      "learning_rate": 9.581070484891146e-06,
+      "loss": 0.4089,
+      "step": 292
+    },
+    {
+      "epoch": 0.6569506726457399,
+      "grad_norm": 0.36041462421417236,
+      "learning_rate": 9.575827313541618e-06,
+      "loss": 0.3988,
+      "step": 293
+    },
+    {
+      "epoch": 0.6591928251121076,
+      "grad_norm": 0.34588080644607544,
+      "learning_rate": 9.570552988024527e-06,
+      "loss": 0.4099,
+      "step": 294
+    },
+    {
+      "epoch": 0.6614349775784754,
+      "grad_norm": 0.2956308126449585,
+      "learning_rate": 9.565247544249709e-06,
+      "loss": 0.4109,
+      "step": 295
+    },
+    {
+      "epoch": 0.6636771300448431,
+      "grad_norm": 0.31860873103141785,
+      "learning_rate": 9.55991101833887e-06,
+      "loss": 0.4119,
+      "step": 296
+    },
+    {
+      "epoch": 0.6659192825112108,
+      "grad_norm": 0.3061535656452179,
+      "learning_rate": 9.554543446625332e-06,
+      "loss": 0.4168,
+      "step": 297
+    },
+    {
+      "epoch": 0.6681614349775785,
+      "grad_norm": 0.2784154713153839,
+      "learning_rate": 9.5491448656538e-06,
+      "loss": 0.4055,
+      "step": 298
+    },
+    {
+      "epoch": 0.6704035874439462,
+      "grad_norm": 0.28013232350349426,
+      "learning_rate": 9.54371531218009e-06,
+      "loss": 0.4067,
+      "step": 299
+    },
+    {
+      "epoch": 0.672645739910314,
+      "grad_norm": 0.2972075343132019,
+      "learning_rate": 9.538254823170903e-06,
+      "loss": 0.4071,
+      "step": 300
+    },
+    {
+      "epoch": 0.6748878923766816,
+      "grad_norm": 0.3000189960002899,
+      "learning_rate": 9.532763435803556e-06,
+      "loss": 0.4318,
+      "step": 301
+    },
+    {
+      "epoch": 0.6771300448430493,
+      "grad_norm": 0.2776426076889038,
+      "learning_rate": 9.527241187465735e-06,
+      "loss": 0.4155,
+      "step": 302
+    },
+    {
+      "epoch": 0.679372197309417,
+      "grad_norm": 0.28904399275779724,
+      "learning_rate": 9.521688115755243e-06,
+      "loss": 0.4277,
+      "step": 303
+    },
+    {
+      "epoch": 0.6816143497757847,
+      "grad_norm": 0.2931148111820221,
+      "learning_rate": 9.51610425847974e-06,
+      "loss": 0.4173,
+      "step": 304
+    },
+    {
+      "epoch": 0.6838565022421524,
+      "grad_norm": 0.30723485350608826,
+      "learning_rate": 9.51048965365649e-06,
+      "loss": 0.4139,
+      "step": 305
+    },
+    {
+      "epoch": 0.6860986547085202,
+      "grad_norm": 0.27698543667793274,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.4247,
+      "step": 306
+    },
+    {
+      "epoch": 0.6883408071748879,
+      "grad_norm": 0.2815530002117157,
+      "learning_rate": 9.499168354482244e-06,
+      "loss": 0.4062,
+      "step": 307
+    },
+    {
+      "epoch": 0.6905829596412556,
+      "grad_norm": 0.2806107699871063,
+      "learning_rate": 9.493461737211443e-06,
+      "loss": 0.4248,
+      "step": 308
+    },
+    {
+      "epoch": 0.6928251121076233,
+      "grad_norm": 0.2786705791950226,
+      "learning_rate": 9.487724526552753e-06,
+      "loss": 0.4282,
+      "step": 309
+    },
+    {
+      "epoch": 0.695067264573991,
+      "grad_norm": 0.28470921516418457,
+      "learning_rate": 9.481956761567531e-06,
+      "loss": 0.4019,
+      "step": 310
+    },
+    {
+      "epoch": 0.6973094170403588,
+      "grad_norm": 0.3217407166957855,
+      "learning_rate": 9.476158481525163e-06,
+      "loss": 0.4128,
+      "step": 311
+    },
+    {
+      "epoch": 0.6995515695067265,
+      "grad_norm": 0.32459142804145813,
+      "learning_rate": 9.47032972590279e-06,
+      "loss": 0.3983,
+      "step": 312
+    },
+    {
+      "epoch": 0.7017937219730942,
+      "grad_norm": 0.273169606924057,
+      "learning_rate": 9.464470534385047e-06,
+      "loss": 0.4102,
+      "step": 313
+    },
+    {
+      "epoch": 0.7040358744394619,
+      "grad_norm": 0.3542450964450836,
+      "learning_rate": 9.458580946863784e-06,
+      "loss": 0.4039,
+      "step": 314
+    },
+    {
+      "epoch": 0.7062780269058296,
+      "grad_norm": 0.3514912724494934,
+      "learning_rate": 9.452661003437809e-06,
+      "loss": 0.4067,
+      "step": 315
+    },
+    {
+      "epoch": 0.7085201793721974,
+      "grad_norm": 0.335882306098938,
+      "learning_rate": 9.446710744412595e-06,
+      "loss": 0.4181,
+      "step": 316
+    },
+    {
+      "epoch": 0.7107623318385651,
+      "grad_norm": 0.2935275137424469,
+      "learning_rate": 9.440730210300027e-06,
+      "loss": 0.4078,
+      "step": 317
+    },
+    {
+      "epoch": 0.7130044843049327,
+      "grad_norm": 0.28140637278556824,
+      "learning_rate": 9.434719441818106e-06,
+      "loss": 0.419,
+      "step": 318
+    },
+    {
+      "epoch": 0.7152466367713004,
+      "grad_norm": 0.34279656410217285,
+      "learning_rate": 9.428678479890689e-06,
+      "loss": 0.4194,
+      "step": 319
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 0.28014835715293884,
+      "learning_rate": 9.422607365647199e-06,
+      "loss": 0.4126,
+      "step": 320
+    },
+    {
+      "epoch": 0.7197309417040358,
+      "grad_norm": 0.3289497196674347,
+      "learning_rate": 9.41650614042235e-06,
+      "loss": 0.4054,
+      "step": 321
+    },
+    {
+      "epoch": 0.7219730941704036,
+      "grad_norm": 0.2863677144050598,
+      "learning_rate": 9.410374845755862e-06,
+      "loss": 0.4055,
+      "step": 322
+    },
+    {
+      "epoch": 0.7242152466367713,
+      "grad_norm": 0.29787036776542664,
+      "learning_rate": 9.404213523392183e-06,
+      "loss": 0.4105,
+      "step": 323
+    },
+    {
+      "epoch": 0.726457399103139,
+      "grad_norm": 0.28383153676986694,
+      "learning_rate": 9.3980222152802e-06,
+      "loss": 0.4069,
+      "step": 324
+    },
+    {
+      "epoch": 0.7286995515695067,
+      "grad_norm": 0.3196985423564911,
+      "learning_rate": 9.39180096357296e-06,
+      "loss": 0.3957,
+      "step": 325
+    },
+    {
+      "epoch": 0.7309417040358744,
+      "grad_norm": 0.3063465356826782,
+      "learning_rate": 9.385549810627374e-06,
+      "loss": 0.4037,
+      "step": 326
+    },
+    {
+      "epoch": 0.7331838565022422,
+      "grad_norm": 0.3238948583602905,
+      "learning_rate": 9.379268799003934e-06,
+      "loss": 0.4032,
+      "step": 327
+    },
+    {
+      "epoch": 0.7354260089686099,
+      "grad_norm": 0.31256479024887085,
+      "learning_rate": 9.372957971466425e-06,
+      "loss": 0.4012,
+      "step": 328
+    },
+    {
+      "epoch": 0.7376681614349776,
+      "grad_norm": 0.31509679555892944,
+      "learning_rate": 9.366617370981631e-06,
+      "loss": 0.4075,
+      "step": 329
+    },
+    {
+      "epoch": 0.7399103139013453,
+      "grad_norm": 0.3181549608707428,
+      "learning_rate": 9.36024704071904e-06,
+      "loss": 0.4116,
+      "step": 330
+    },
+    {
+      "epoch": 0.742152466367713,
+      "grad_norm": 0.3025352954864502,
+      "learning_rate": 9.353847024050552e-06,
+      "loss": 0.4071,
+      "step": 331
+    },
+    {
+      "epoch": 0.7443946188340808,
+      "grad_norm": 0.2810025215148926,
+      "learning_rate": 9.347417364550193e-06,
+      "loss": 0.4181,
+      "step": 332
+    },
+    {
+      "epoch": 0.7466367713004485,
+      "grad_norm": 0.27573129534721375,
+      "learning_rate": 9.3409581059938e-06,
+      "loss": 0.4113,
+      "step": 333
+    },
+    {
+      "epoch": 0.7488789237668162,
+      "grad_norm": 0.5964189767837524,
+      "learning_rate": 9.334469292358736e-06,
+      "loss": 0.4099,
+      "step": 334
+    },
+    {
+      "epoch": 0.7511210762331838,
+      "grad_norm": 0.2760773301124573,
+      "learning_rate": 9.327950967823594e-06,
+      "loss": 0.4133,
+      "step": 335
+    },
+    {
+      "epoch": 0.7533632286995515,
+      "grad_norm": 0.3778616786003113,
+      "learning_rate": 9.321403176767878e-06,
+      "loss": 0.4017,
+      "step": 336
+    },
+    {
+      "epoch": 0.7556053811659192,
+      "grad_norm": 0.31446725130081177,
+      "learning_rate": 9.314825963771724e-06,
+      "loss": 0.3996,
+      "step": 337
+    },
+    {
+      "epoch": 0.757847533632287,
+      "grad_norm": 0.3497312664985657,
+      "learning_rate": 9.308219373615574e-06,
+      "loss": 0.4078,
+      "step": 338
+    },
+    {
+      "epoch": 0.7600896860986547,
+      "grad_norm": 0.3311575651168823,
+      "learning_rate": 9.301583451279892e-06,
+      "loss": 0.4084,
+      "step": 339
+    },
+    {
+      "epoch": 0.7623318385650224,
+      "grad_norm": 0.3345532715320587,
+      "learning_rate": 9.294918241944843e-06,
+      "loss": 0.404,
+      "step": 340
+    },
+    {
+      "epoch": 0.7645739910313901,
+      "grad_norm": 0.349703311920166,
+      "learning_rate": 9.288223790989991e-06,
+      "loss": 0.414,
+      "step": 341
+    },
+    {
+      "epoch": 0.7668161434977578,
+      "grad_norm": 0.3346048891544342,
+      "learning_rate": 9.28150014399399e-06,
+      "loss": 0.4063,
+      "step": 342
+    },
+    {
+      "epoch": 0.7690582959641256,
+      "grad_norm": 0.27873894572257996,
+      "learning_rate": 9.27474734673427e-06,
+      "loss": 0.4038,
+      "step": 343
+    },
+    {
+      "epoch": 0.7713004484304933,
+      "grad_norm": 0.3184773623943329,
+      "learning_rate": 9.267965445186733e-06,
+      "loss": 0.4181,
+      "step": 344
+    },
+    {
+      "epoch": 0.773542600896861,
+      "grad_norm": 0.32033205032348633,
+      "learning_rate": 9.261154485525433e-06,
+      "loss": 0.3895,
+      "step": 345
+    },
+    {
+      "epoch": 0.7757847533632287,
+      "grad_norm": 0.33291661739349365,
+      "learning_rate": 9.25431451412226e-06,
+      "loss": 0.4104,
+      "step": 346
+    },
+    {
+      "epoch": 0.7780269058295964,
+      "grad_norm": 0.33981722593307495,
+      "learning_rate": 9.247445577546637e-06,
+      "loss": 0.4099,
+      "step": 347
+    },
+    {
+      "epoch": 0.7802690582959642,
+      "grad_norm": 0.35642552375793457,
+      "learning_rate": 9.240547722565185e-06,
+      "loss": 0.4053,
+      "step": 348
+    },
+    {
+      "epoch": 0.7825112107623319,
+      "grad_norm": 0.2905188798904419,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.4067,
+      "step": 349
+    },
+    {
+      "epoch": 0.7847533632286996,
+      "grad_norm": 0.7361780405044556,
+      "learning_rate": 9.226665445435428e-06,
+      "loss": 0.4183,
+      "step": 350
+    },
+    {
+      "epoch": 0.7869955156950673,
+      "grad_norm": 0.325268417596817,
+      "learning_rate": 9.219681117803537e-06,
+      "loss": 0.4117,
+      "step": 351
+    },
+    {
+      "epoch": 0.7892376681614349,
+      "grad_norm": 0.43043798208236694,
+      "learning_rate": 9.212668060798e-06,
+      "loss": 0.407,
+      "step": 352
+    },
+    {
+      "epoch": 0.7914798206278026,
+      "grad_norm": 0.30372267961502075,
+      "learning_rate": 9.20562632216668e-06,
+      "loss": 0.3869,
+      "step": 353
+    },
+    {
+      "epoch": 0.7937219730941704,
+      "grad_norm": 0.29308637976646423,
+      "learning_rate": 9.19855594985271e-06,
+      "loss": 0.408,
+      "step": 354
+    },
+    {
+      "epoch": 0.7959641255605381,
+      "grad_norm": 0.33781367540359497,
+      "learning_rate": 9.191456991994174e-06,
+      "loss": 0.4113,
+      "step": 355
+    },
+    {
+      "epoch": 0.7982062780269058,
+      "grad_norm": 0.3288229703903198,
+      "learning_rate": 9.184329496923783e-06,
+      "loss": 0.3934,
+      "step": 356
+    },
+    {
+      "epoch": 0.8004484304932735,
+      "grad_norm": 0.31993719935417175,
+      "learning_rate": 9.177173513168533e-06,
+      "loss": 0.4063,
+      "step": 357
+    },
+    {
+      "epoch": 0.8026905829596412,
+      "grad_norm": 0.6845749020576477,
+      "learning_rate": 9.16998908944939e-06,
+      "loss": 0.3998,
+      "step": 358
+    },
+    {
+      "epoch": 0.804932735426009,
+      "grad_norm": 0.31695330142974854,
+      "learning_rate": 9.16277627468095e-06,
+      "loss": 0.4125,
+      "step": 359
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 0.3115200698375702,
+      "learning_rate": 9.155535117971105e-06,
+      "loss": 0.4079,
+      "step": 360
+    },
+    {
+      "epoch": 0.8094170403587444,
+      "grad_norm": 0.3077631890773773,
+      "learning_rate": 9.148265668620713e-06,
+      "loss": 0.4183,
+      "step": 361
+    },
+    {
+      "epoch": 0.8116591928251121,
+      "grad_norm": 0.32053032517433167,
+      "learning_rate": 9.14096797612326e-06,
+      "loss": 0.4156,
+      "step": 362
+    },
+    {
+      "epoch": 0.8139013452914798,
+      "grad_norm": 0.484241247177124,
+      "learning_rate": 9.133642090164521e-06,
+      "loss": 0.4039,
+      "step": 363
+    },
+    {
+      "epoch": 0.8161434977578476,
+      "grad_norm": 0.3111223578453064,
+      "learning_rate": 9.126288060622228e-06,
+      "loss": 0.4143,
+      "step": 364
+    },
+    {
+      "epoch": 0.8183856502242153,
+      "grad_norm": 0.29185277223587036,
+      "learning_rate": 9.118905937565723e-06,
+      "loss": 0.4144,
+      "step": 365
+    },
+    {
+      "epoch": 0.820627802690583,
+      "grad_norm": 0.287458211183548,
+      "learning_rate": 9.111495771255623e-06,
+      "loss": 0.4062,
+      "step": 366
+    },
+    {
+      "epoch": 0.8228699551569507,
+      "grad_norm": 0.33976373076438904,
+      "learning_rate": 9.104057612143473e-06,
+      "loss": 0.4053,
+      "step": 367
+    },
+    {
+      "epoch": 0.8251121076233184,
+      "grad_norm": 0.3475874066352844,
+      "learning_rate": 9.096591510871406e-06,
+      "loss": 0.4092,
+      "step": 368
+    },
+    {
+      "epoch": 0.827354260089686,
+      "grad_norm": 0.42061370611190796,
+      "learning_rate": 9.089097518271797e-06,
+      "loss": 0.4052,
+      "step": 369
+    },
+    {
+      "epoch": 0.8295964125560538,
+      "grad_norm": 0.29814794659614563,
+      "learning_rate": 9.081575685366919e-06,
+      "loss": 0.4063,
+      "step": 370
+    },
+    {
+      "epoch": 0.8318385650224215,
+      "grad_norm": 0.33191218972206116,
+      "learning_rate": 9.074026063368589e-06,
+      "loss": 0.4017,
+      "step": 371
+    },
+    {
+      "epoch": 0.8340807174887892,
+      "grad_norm": 0.29336613416671753,
+      "learning_rate": 9.066448703677828e-06,
+      "loss": 0.4151,
+      "step": 372
+    },
+    {
+      "epoch": 0.8363228699551569,
+      "grad_norm": 0.3203662037849426,
+      "learning_rate": 9.058843657884505e-06,
+      "loss": 0.4117,
+      "step": 373
+    },
+    {
+      "epoch": 0.8385650224215246,
+      "grad_norm": 0.325100302696228,
+      "learning_rate": 9.051210977766987e-06,
+      "loss": 0.402,
+      "step": 374
+    },
+    {
+      "epoch": 0.8408071748878924,
+      "grad_norm": 0.3161066770553589,
+      "learning_rate": 9.04355071529179e-06,
+      "loss": 0.4075,
+      "step": 375
+    },
+    {
+      "epoch": 0.8430493273542601,
+      "grad_norm": 0.4152739346027374,
+      "learning_rate": 9.035862922613218e-06,
+      "loss": 0.4039,
+      "step": 376
+    },
+    {
+      "epoch": 0.8452914798206278,
+      "grad_norm": 0.3031544089317322,
+      "learning_rate": 9.02814765207302e-06,
+      "loss": 0.4014,
+      "step": 377
+    },
+    {
+      "epoch": 0.8475336322869955,
+      "grad_norm": 0.36650827527046204,
+      "learning_rate": 9.020404956200016e-06,
+      "loss": 0.4064,
+      "step": 378
+    },
+    {
+      "epoch": 0.8497757847533632,
+      "grad_norm": 0.3329921066761017,
+      "learning_rate": 9.012634887709755e-06,
+      "loss": 0.4221,
+      "step": 379
+    },
+    {
+      "epoch": 0.852017937219731,
+      "grad_norm": 0.32171863317489624,
+      "learning_rate": 9.004837499504152e-06,
+      "loss": 0.4101,
+      "step": 380
+    },
+    {
+      "epoch": 0.8542600896860987,
+      "grad_norm": 0.3472563624382019,
+      "learning_rate": 8.997012844671118e-06,
+      "loss": 0.3965,
+      "step": 381
+    },
+    {
+      "epoch": 0.8565022421524664,
+      "grad_norm": 0.2860943078994751,
+      "learning_rate": 8.989160976484218e-06,
+      "loss": 0.4261,
+      "step": 382
+    },
+    {
+      "epoch": 0.8587443946188341,
+      "grad_norm": 0.37671801447868347,
+      "learning_rate": 8.981281948402289e-06,
+      "loss": 0.4173,
+      "step": 383
+    },
+    {
+      "epoch": 0.8609865470852018,
+      "grad_norm": 0.4690786600112915,
+      "learning_rate": 8.973375814069087e-06,
+      "loss": 0.4028,
+      "step": 384
+    },
+    {
+      "epoch": 0.8632286995515696,
+      "grad_norm": 0.27456754446029663,
+      "learning_rate": 8.965442627312916e-06,
+      "loss": 0.3988,
+      "step": 385
+    },
+    {
+      "epoch": 0.8654708520179372,
+      "grad_norm": 0.315211683511734,
+      "learning_rate": 8.957482442146271e-06,
+      "loss": 0.4156,
+      "step": 386
+    },
+    {
+      "epoch": 0.8677130044843049,
+      "grad_norm": 0.312274694442749,
+      "learning_rate": 8.949495312765456e-06,
+      "loss": 0.4031,
+      "step": 387
+    },
+    {
+      "epoch": 0.8699551569506726,
+      "grad_norm": 0.293405145406723,
+      "learning_rate": 8.941481293550225e-06,
+      "loss": 0.416,
+      "step": 388
+    },
+    {
+      "epoch": 0.8721973094170403,
+      "grad_norm": 0.30306535959243774,
+      "learning_rate": 8.933440439063412e-06,
+      "loss": 0.3966,
+      "step": 389
+    },
+    {
+      "epoch": 0.874439461883408,
+      "grad_norm": 0.3342600166797638,
+      "learning_rate": 8.925372804050554e-06,
+      "loss": 0.3901,
+      "step": 390
+    },
+    {
+      "epoch": 0.8766816143497758,
+      "grad_norm": 0.31084662675857544,
+      "learning_rate": 8.917278443439524e-06,
+      "loss": 0.3984,
+      "step": 391
+    },
+    {
+      "epoch": 0.8789237668161435,
+      "grad_norm": 0.32946890592575073,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.4014,
+      "step": 392
+    },
+    {
+      "epoch": 0.8811659192825112,
+      "grad_norm": 0.27684393525123596,
+      "learning_rate": 8.901009766043846e-06,
+      "loss": 0.3982,
+      "step": 393
+    },
+    {
+      "epoch": 0.8834080717488789,
+      "grad_norm": 1.5456547737121582,
+      "learning_rate": 8.892835560023236e-06,
+      "loss": 0.4046,
+      "step": 394
+    },
+    {
+      "epoch": 0.8856502242152466,
+      "grad_norm": 0.3096829652786255,
+      "learning_rate": 8.884634849931771e-06,
+      "loss": 0.4026,
+      "step": 395
+    },
+    {
+      "epoch": 0.8878923766816144,
+      "grad_norm": 0.29568785429000854,
+      "learning_rate": 8.876407691603353e-06,
+      "loss": 0.4288,
+      "step": 396
+    },
+    {
+      "epoch": 0.8901345291479821,
+      "grad_norm": 0.30268481373786926,
+      "learning_rate": 8.868154141051955e-06,
+      "loss": 0.3937,
+      "step": 397
+    },
+    {
+      "epoch": 0.8923766816143498,
+      "grad_norm": 0.2969793677330017,
+      "learning_rate": 8.85987425447124e-06,
+      "loss": 0.4123,
+      "step": 398
+    },
+    {
+      "epoch": 0.8946188340807175,
+      "grad_norm": 0.3506118953227997,
+      "learning_rate": 8.851568088234178e-06,
+      "loss": 0.4029,
+      "step": 399
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 0.3449394702911377,
+      "learning_rate": 8.843235698892661e-06,
+      "loss": 0.3985,
+      "step": 400
+    },
+    {
+      "epoch": 0.899103139013453,
+      "grad_norm": 0.29561424255371094,
+      "learning_rate": 8.834877143177121e-06,
+      "loss": 0.4012,
+      "step": 401
+    },
+    {
+      "epoch": 0.9013452914798207,
+      "grad_norm": 0.3387802839279175,
+      "learning_rate": 8.826492477996138e-06,
+      "loss": 0.4067,
+      "step": 402
+    },
+    {
+      "epoch": 0.9035874439461884,
+      "grad_norm": 0.390193909406662,
+      "learning_rate": 8.818081760436062e-06,
+      "loss": 0.4191,
+      "step": 403
+    },
+    {
+      "epoch": 0.905829596412556,
+      "grad_norm": 0.3167582154273987,
+      "learning_rate": 8.80964504776061e-06,
+      "loss": 0.4059,
+      "step": 404
+    },
+    {
+      "epoch": 0.9080717488789237,
+      "grad_norm": 0.2925993800163269,
+      "learning_rate": 8.801182397410495e-06,
+      "loss": 0.4044,
+      "step": 405
+    },
+    {
+      "epoch": 0.9103139013452914,
+      "grad_norm": 0.5227466821670532,
+      "learning_rate": 8.792693867003017e-06,
+      "loss": 0.3956,
+      "step": 406
+    },
+    {
+      "epoch": 0.9125560538116592,
+      "grad_norm": 0.28477007150650024,
+      "learning_rate": 8.784179514331683e-06,
+      "loss": 0.4042,
+      "step": 407
+    },
+    {
+      "epoch": 0.9147982062780269,
+      "grad_norm": 0.3327925205230713,
+      "learning_rate": 8.775639397365803e-06,
+      "loss": 0.3999,
+      "step": 408
+    },
+    {
+      "epoch": 0.9170403587443946,
+      "grad_norm": 0.2806262671947479,
+      "learning_rate": 8.76707357425011e-06,
+      "loss": 0.3989,
+      "step": 409
+    },
+    {
+      "epoch": 0.9192825112107623,
+      "grad_norm": 0.31155821681022644,
+      "learning_rate": 8.758482103304348e-06,
+      "loss": 0.4187,
+      "step": 410
+    },
+    {
+      "epoch": 0.92152466367713,
+      "grad_norm": 0.2895669639110565,
+      "learning_rate": 8.749865043022884e-06,
+      "loss": 0.4134,
+      "step": 411
+    },
+    {
+      "epoch": 0.9237668161434978,
+      "grad_norm": 0.333717405796051,
+      "learning_rate": 8.74122245207431e-06,
+      "loss": 0.4078,
+      "step": 412
+    },
+    {
+      "epoch": 0.9260089686098655,
+      "grad_norm": 0.29114285111427307,
+      "learning_rate": 8.73255438930104e-06,
+      "loss": 0.3975,
+      "step": 413
+    },
+    {
+      "epoch": 0.9282511210762332,
+      "grad_norm": 0.28838276863098145,
+      "learning_rate": 8.72386091371891e-06,
+      "loss": 0.3964,
+      "step": 414
+    },
+    {
+      "epoch": 0.9304932735426009,
+      "grad_norm": 0.3286455273628235,
+      "learning_rate": 8.71514208451678e-06,
+      "loss": 0.4003,
+      "step": 415
+    },
+    {
+      "epoch": 0.9327354260089686,
+      "grad_norm": 0.2625728249549866,
+      "learning_rate": 8.706397961056127e-06,
+      "loss": 0.4008,
+      "step": 416
+    },
+    {
+      "epoch": 0.9349775784753364,
+      "grad_norm": 0.31902843713760376,
+      "learning_rate": 8.69762860287064e-06,
+      "loss": 0.4161,
+      "step": 417
+    },
+    {
+      "epoch": 0.9372197309417041,
+      "grad_norm": 0.26823586225509644,
+      "learning_rate": 8.688834069665819e-06,
+      "loss": 0.394,
+      "step": 418
+    },
+    {
+      "epoch": 0.9394618834080718,
+      "grad_norm": 0.26950374245643616,
+      "learning_rate": 8.680014421318565e-06,
+      "loss": 0.4176,
+      "step": 419
+    },
+    {
+      "epoch": 0.9417040358744395,
+      "grad_norm": 0.2729303240776062,
+      "learning_rate": 8.671169717876777e-06,
+      "loss": 0.397,
+      "step": 420
+    },
+    {
+      "epoch": 0.9439461883408071,
+      "grad_norm": 0.30254143476486206,
+      "learning_rate": 8.662300019558931e-06,
+      "loss": 0.4003,
+      "step": 421
+    },
+    {
+      "epoch": 0.9461883408071748,
+      "grad_norm": 0.27907007932662964,
+      "learning_rate": 8.653405386753688e-06,
+      "loss": 0.4015,
+      "step": 422
+    },
+    {
+      "epoch": 0.9484304932735426,
+      "grad_norm": 0.2718103528022766,
+      "learning_rate": 8.64448588001947e-06,
+      "loss": 0.4077,
+      "step": 423
+    },
+    {
+      "epoch": 0.9506726457399103,
+      "grad_norm": 0.29295119643211365,
+      "learning_rate": 8.63554156008405e-06,
+      "loss": 0.3959,
+      "step": 424
+    },
+    {
+      "epoch": 0.952914798206278,
+      "grad_norm": 0.29576820135116577,
+      "learning_rate": 8.626572487844143e-06,
+      "loss": 0.3982,
+      "step": 425
+    },
+    {
+      "epoch": 0.9551569506726457,
+      "grad_norm": 0.27937939763069153,
+      "learning_rate": 8.617578724364984e-06,
+      "loss": 0.3953,
+      "step": 426
+    },
+    {
+      "epoch": 0.9573991031390134,
+      "grad_norm": 0.3426397740840912,
+      "learning_rate": 8.608560330879916e-06,
+      "loss": 0.4035,
+      "step": 427
+    },
+    {
+      "epoch": 0.9596412556053812,
+      "grad_norm": 0.2796214520931244,
+      "learning_rate": 8.599517368789981e-06,
+      "loss": 0.4066,
+      "step": 428
+    },
+    {
+      "epoch": 0.9618834080717489,
+      "grad_norm": 0.29795750975608826,
+      "learning_rate": 8.590449899663484e-06,
+      "loss": 0.3948,
+      "step": 429
+    },
+    {
+      "epoch": 0.9641255605381166,
+      "grad_norm": 0.2671406865119934,
+      "learning_rate": 8.581357985235595e-06,
+      "loss": 0.3889,
+      "step": 430
+    },
+    {
+      "epoch": 0.9663677130044843,
+      "grad_norm": 0.259992778301239,
+      "learning_rate": 8.57224168740791e-06,
+      "loss": 0.401,
+      "step": 431
+    },
+    {
+      "epoch": 0.968609865470852,
+      "grad_norm": 0.2756447494029999,
+      "learning_rate": 8.56310106824804e-06,
+      "loss": 0.3923,
+      "step": 432
+    },
+    {
+      "epoch": 0.9708520179372198,
+      "grad_norm": 0.2677665948867798,
+      "learning_rate": 8.553936189989187e-06,
+      "loss": 0.3879,
+      "step": 433
+    },
+    {
+      "epoch": 0.9730941704035875,
+      "grad_norm": 0.27413448691368103,
+      "learning_rate": 8.544747115029717e-06,
+      "loss": 0.4114,
+      "step": 434
+    },
+    {
+      "epoch": 0.9753363228699552,
+      "grad_norm": 0.31824958324432373,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.4074,
+      "step": 435
+    },
+    {
+      "epoch": 0.9775784753363229,
+      "grad_norm": 0.5684327483177185,
+      "learning_rate": 8.526296625425676e-06,
+      "loss": 0.4066,
+      "step": 436
+    },
+    {
+      "epoch": 0.9798206278026906,
+      "grad_norm": 0.31156355142593384,
+      "learning_rate": 8.517035336399842e-06,
+      "loss": 0.4005,
+      "step": 437
+    },
+    {
+      "epoch": 0.9820627802690582,
+      "grad_norm": 0.3012964129447937,
+      "learning_rate": 8.50775010191001e-06,
+      "loss": 0.398,
+      "step": 438
+    },
+    {
+      "epoch": 0.984304932735426,
+      "grad_norm": 0.28542065620422363,
+      "learning_rate": 8.498440985173981e-06,
+      "loss": 0.4017,
+      "step": 439
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 0.33007916808128357,
+      "learning_rate": 8.489108049572161e-06,
+      "loss": 0.4078,
+      "step": 440
+    },
+    {
+      "epoch": 0.9887892376681614,
+      "grad_norm": 0.2822427749633789,
+      "learning_rate": 8.479751358647122e-06,
+      "loss": 0.4002,
+      "step": 441
+    },
+    {
+      "epoch": 0.9910313901345291,
+      "grad_norm": 0.31078580021858215,
+      "learning_rate": 8.470370976103171e-06,
+      "loss": 0.4065,
+      "step": 442
+    },
+    {
+      "epoch": 0.9932735426008968,
+      "grad_norm": 0.3202306926250458,
+      "learning_rate": 8.46096696580592e-06,
+      "loss": 0.3944,
+      "step": 443
+    },
+    {
+      "epoch": 0.9955156950672646,
+      "grad_norm": 0.2832654118537903,
+      "learning_rate": 8.451539391781851e-06,
+      "loss": 0.4015,
+      "step": 444
+    },
+    {
+      "epoch": 0.9977578475336323,
+      "grad_norm": 0.3062937557697296,
+      "learning_rate": 8.442088318217875e-06,
+      "loss": 0.4016,
+      "step": 445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2647171914577484,
+      "learning_rate": 8.432613809460895e-06,
+      "loss": 0.3765,
+      "step": 446
+    },
+    {
+      "epoch": 1.0022421524663676,
+      "grad_norm": 0.34632837772369385,
+      "learning_rate": 8.423115930017373e-06,
+      "loss": 0.3602,
+      "step": 447
+    },
+    {
+      "epoch": 1.0044843049327354,
+      "grad_norm": 0.274224191904068,
+      "learning_rate": 8.413594744552895e-06,
+      "loss": 0.3599,
+      "step": 448
+    },
+    {
+      "epoch": 1.006726457399103,
+      "grad_norm": 0.35535210371017456,
+      "learning_rate": 8.40405031789171e-06,
+      "loss": 0.3638,
+      "step": 449
+    },
+    {
+      "epoch": 1.0089686098654709,
+      "grad_norm": 0.29200729727745056,
+      "learning_rate": 8.394482715016318e-06,
+      "loss": 0.351,
+      "step": 450
+    },
+    {
+      "epoch": 1.0112107623318385,
+      "grad_norm": 0.3285231590270996,
+      "learning_rate": 8.384892001067004e-06,
+      "loss": 0.3546,
+      "step": 451
+    },
+    {
+      "epoch": 1.0134529147982063,
+      "grad_norm": 0.46168652176856995,
+      "learning_rate": 8.375278241341404e-06,
+      "loss": 0.3641,
+      "step": 452
+    },
+    {
+      "epoch": 1.015695067264574,
+      "grad_norm": 0.28547239303588867,
+      "learning_rate": 8.36564150129406e-06,
+      "loss": 0.3421,
+      "step": 453
+    },
+    {
+      "epoch": 1.0179372197309418,
+      "grad_norm": 0.3175661563873291,
+      "learning_rate": 8.355981846535972e-06,
+      "loss": 0.3463,
+      "step": 454
+    },
+    {
+      "epoch": 1.0201793721973094,
+      "grad_norm": 0.35064899921417236,
+      "learning_rate": 8.346299342834158e-06,
+      "loss": 0.3532,
+      "step": 455
+    },
+    {
+      "epoch": 1.0224215246636772,
+      "grad_norm": 0.3195887506008148,
+      "learning_rate": 8.336594056111197e-06,
+      "loss": 0.3523,
+      "step": 456
+    },
+    {
+      "epoch": 1.0246636771300448,
+      "grad_norm": 0.2751784026622772,
+      "learning_rate": 8.326866052444787e-06,
+      "loss": 0.3625,
+      "step": 457
+    },
+    {
+      "epoch": 1.0269058295964126,
+      "grad_norm": 0.28595998883247375,
+      "learning_rate": 8.317115398067289e-06,
+      "loss": 0.35,
+      "step": 458
+    },
+    {
+      "epoch": 1.0291479820627802,
+      "grad_norm": 0.34845221042633057,
+      "learning_rate": 8.30734215936528e-06,
+      "loss": 0.354,
+      "step": 459
+    },
+    {
+      "epoch": 1.031390134529148,
+      "grad_norm": 0.2812776267528534,
+      "learning_rate": 8.297546402879106e-06,
+      "loss": 0.3602,
+      "step": 460
+    },
+    {
+      "epoch": 1.0336322869955157,
+      "grad_norm": 0.3051612675189972,
+      "learning_rate": 8.28772819530242e-06,
+      "loss": 0.3543,
+      "step": 461
+    },
+    {
+      "epoch": 1.0358744394618835,
+      "grad_norm": 0.31309396028518677,
+      "learning_rate": 8.27788760348173e-06,
+      "loss": 0.3601,
+      "step": 462
+    },
+    {
+      "epoch": 1.0381165919282511,
+      "grad_norm": 0.28610166907310486,
+      "learning_rate": 8.268024694415949e-06,
+      "loss": 0.3566,
+      "step": 463
+    },
+    {
+      "epoch": 1.0403587443946187,
+      "grad_norm": 0.34127941727638245,
+      "learning_rate": 8.258139535255926e-06,
+      "loss": 0.3563,
+      "step": 464
+    },
+    {
+      "epoch": 1.0426008968609866,
+      "grad_norm": 0.27622556686401367,
+      "learning_rate": 8.248232193304016e-06,
+      "loss": 0.3463,
+      "step": 465
+    },
+    {
+      "epoch": 1.0448430493273542,
+      "grad_norm": 0.3454950451850891,
+      "learning_rate": 8.238302736013587e-06,
+      "loss": 0.358,
+      "step": 466
+    },
+    {
+      "epoch": 1.047085201793722,
+      "grad_norm": 0.315620481967926,
+      "learning_rate": 8.22835123098858e-06,
+      "loss": 0.3408,
+      "step": 467
+    },
+    {
+      "epoch": 1.0493273542600896,
+      "grad_norm": 0.30485957860946655,
+      "learning_rate": 8.218377745983058e-06,
+      "loss": 0.3583,
+      "step": 468
+    },
+    {
+      "epoch": 1.0515695067264574,
+      "grad_norm": 0.275348961353302,
+      "learning_rate": 8.208382348900718e-06,
+      "loss": 0.3631,
+      "step": 469
+    },
+    {
+      "epoch": 1.053811659192825,
+      "grad_norm": 0.2757943272590637,
+      "learning_rate": 8.198365107794457e-06,
+      "loss": 0.3571,
+      "step": 470
+    },
+    {
+      "epoch": 1.0560538116591929,
+      "grad_norm": 0.2984645962715149,
+      "learning_rate": 8.188326090865885e-06,
+      "loss": 0.3451,
+      "step": 471
+    },
+    {
+      "epoch": 1.0582959641255605,
+      "grad_norm": 0.3153923451900482,
+      "learning_rate": 8.178265366464875e-06,
+      "loss": 0.3596,
+      "step": 472
+    },
+    {
+      "epoch": 1.0605381165919283,
+      "grad_norm": 0.296587735414505,
+      "learning_rate": 8.168183003089095e-06,
+      "loss": 0.359,
+      "step": 473
+    },
+    {
+      "epoch": 1.062780269058296,
+      "grad_norm": 0.2634204030036926,
+      "learning_rate": 8.158079069383535e-06,
+      "loss": 0.3492,
+      "step": 474
+    },
+    {
+      "epoch": 1.0650224215246638,
+      "grad_norm": 0.29771867394447327,
+      "learning_rate": 8.147953634140054e-06,
+      "loss": 0.3475,
+      "step": 475
+    },
+    {
+      "epoch": 1.0672645739910314,
+      "grad_norm": 0.2846080958843231,
+      "learning_rate": 8.137806766296896e-06,
+      "loss": 0.3526,
+      "step": 476
+    },
+    {
+      "epoch": 1.0695067264573992,
+      "grad_norm": 0.2761863172054291,
+      "learning_rate": 8.127638534938227e-06,
+      "loss": 0.3542,
+      "step": 477
+    },
+    {
+      "epoch": 1.0717488789237668,
+      "grad_norm": 0.27391237020492554,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.3486,
+      "step": 478
+    },
+    {
+      "epoch": 1.0739910313901346,
+      "grad_norm": 0.332175076007843,
+      "learning_rate": 8.107238258737818e-06,
+      "loss": 0.3551,
+      "step": 479
+    },
+    {
+      "epoch": 1.0762331838565022,
+      "grad_norm": 0.2657061815261841,
+      "learning_rate": 8.097006352789787e-06,
+      "loss": 0.3565,
+      "step": 480
+    },
+    {
+      "epoch": 1.07847533632287,
+      "grad_norm": 0.2759336531162262,
+      "learning_rate": 8.086753361112715e-06,
+      "loss": 0.35,
+      "step": 481
+    },
+    {
+      "epoch": 1.0807174887892377,
+      "grad_norm": 0.3019334077835083,
+      "learning_rate": 8.076479353513308e-06,
+      "loss": 0.3509,
+      "step": 482
+    },
+    {
+      "epoch": 1.0829596412556053,
+      "grad_norm": 0.2541658878326416,
+      "learning_rate": 8.066184399941353e-06,
+      "loss": 0.354,
+      "step": 483
+    },
+    {
+      "epoch": 1.0852017937219731,
+      "grad_norm": 0.35716456174850464,
+      "learning_rate": 8.055868570489247e-06,
+      "loss": 0.3552,
+      "step": 484
+    },
+    {
+      "epoch": 1.0874439461883407,
+      "grad_norm": 0.24950361251831055,
+      "learning_rate": 8.045531935391524e-06,
+      "loss": 0.3585,
+      "step": 485
+    },
+    {
+      "epoch": 1.0896860986547086,
+      "grad_norm": 0.27788081765174866,
+      "learning_rate": 8.035174565024362e-06,
+      "loss": 0.3494,
+      "step": 486
+    },
+    {
+      "epoch": 1.0919282511210762,
+      "grad_norm": 0.29102277755737305,
+      "learning_rate": 8.024796529905125e-06,
+      "loss": 0.3501,
+      "step": 487
+    },
+    {
+      "epoch": 1.094170403587444,
+      "grad_norm": 0.25892260670661926,
+      "learning_rate": 8.014397900691859e-06,
+      "loss": 0.3544,
+      "step": 488
+    },
+    {
+      "epoch": 1.0964125560538116,
+      "grad_norm": 0.29088884592056274,
+      "learning_rate": 8.003978748182833e-06,
+      "loss": 0.3496,
+      "step": 489
+    },
+    {
+      "epoch": 1.0986547085201794,
+      "grad_norm": 0.2847244441509247,
+      "learning_rate": 7.993539143316044e-06,
+      "loss": 0.351,
+      "step": 490
+    },
+    {
+      "epoch": 1.100896860986547,
+      "grad_norm": 0.2778773605823517,
+      "learning_rate": 7.983079157168736e-06,
+      "loss": 0.3537,
+      "step": 491
+    },
+    {
+      "epoch": 1.1031390134529149,
+      "grad_norm": 0.25572022795677185,
+      "learning_rate": 7.972598860956917e-06,
+      "loss": 0.3624,
+      "step": 492
+    },
+    {
+      "epoch": 1.1053811659192825,
+      "grad_norm": 0.3382296562194824,
+      "learning_rate": 7.96209832603488e-06,
+      "loss": 0.3613,
+      "step": 493
+    },
+    {
+      "epoch": 1.1076233183856503,
+      "grad_norm": 0.2746177911758423,
+      "learning_rate": 7.951577623894701e-06,
+      "loss": 0.3551,
+      "step": 494
+    },
+    {
+      "epoch": 1.109865470852018,
+      "grad_norm": 0.2753693461418152,
+      "learning_rate": 7.941036826165778e-06,
+      "loss": 0.3388,
+      "step": 495
+    },
+    {
+      "epoch": 1.1121076233183858,
+      "grad_norm": 0.3050689399242401,
+      "learning_rate": 7.930476004614313e-06,
+      "loss": 0.3485,
+      "step": 496
+    },
+    {
+      "epoch": 1.1143497757847534,
+      "grad_norm": 0.29461848735809326,
+      "learning_rate": 7.919895231142848e-06,
+      "loss": 0.3595,
+      "step": 497
+    },
+    {
+      "epoch": 1.116591928251121,
+      "grad_norm": 0.26750698685646057,
+      "learning_rate": 7.909294577789765e-06,
+      "loss": 0.3574,
+      "step": 498
+    },
+    {
+      "epoch": 1.1188340807174888,
+      "grad_norm": 0.28324681520462036,
+      "learning_rate": 7.898674116728792e-06,
+      "loss": 0.3585,
+      "step": 499
+    },
+    {
+      "epoch": 1.1210762331838564,
+      "grad_norm": 0.2557442784309387,
+      "learning_rate": 7.888033920268524e-06,
+      "loss": 0.3497,
+      "step": 500
+    },
+    {
+      "epoch": 1.1233183856502242,
+      "grad_norm": 0.28862401843070984,
+      "learning_rate": 7.877374060851915e-06,
+      "loss": 0.36,
+      "step": 501
+    },
+    {
+      "epoch": 1.1255605381165918,
+      "grad_norm": 0.28569144010543823,
+      "learning_rate": 7.866694611055796e-06,
+      "loss": 0.3485,
+      "step": 502
+    },
+    {
+      "epoch": 1.1278026905829597,
+      "grad_norm": 0.2830195128917694,
+      "learning_rate": 7.85599564359038e-06,
+      "loss": 0.3567,
+      "step": 503
+    },
+    {
+      "epoch": 1.1300448430493273,
+      "grad_norm": 0.3148826062679291,
+      "learning_rate": 7.845277231298761e-06,
+      "loss": 0.3562,
+      "step": 504
+    },
+    {
+      "epoch": 1.1322869955156951,
+      "grad_norm": 0.31142207980155945,
+      "learning_rate": 7.834539447156424e-06,
+      "loss": 0.3455,
+      "step": 505
+    },
+    {
+      "epoch": 1.1345291479820627,
+      "grad_norm": 0.2780856490135193,
+      "learning_rate": 7.823782364270743e-06,
+      "loss": 0.3706,
+      "step": 506
+    },
+    {
+      "epoch": 1.1367713004484306,
+      "grad_norm": 0.3270454406738281,
+      "learning_rate": 7.813006055880487e-06,
+      "loss": 0.3687,
+      "step": 507
+    },
+    {
+      "epoch": 1.1390134529147982,
+      "grad_norm": 0.3276538550853729,
+      "learning_rate": 7.802210595355327e-06,
+      "loss": 0.3603,
+      "step": 508
+    },
+    {
+      "epoch": 1.141255605381166,
+      "grad_norm": 0.26958054304122925,
+      "learning_rate": 7.791396056195315e-06,
+      "loss": 0.3503,
+      "step": 509
+    },
+    {
+      "epoch": 1.1434977578475336,
+      "grad_norm": 0.3360564112663269,
+      "learning_rate": 7.780562512030414e-06,
+      "loss": 0.3435,
+      "step": 510
+    },
+    {
+      "epoch": 1.1457399103139014,
+      "grad_norm": 0.3261363208293915,
+      "learning_rate": 7.76971003661997e-06,
+      "loss": 0.3513,
+      "step": 511
+    },
+    {
+      "epoch": 1.147982062780269,
+      "grad_norm": 0.27897167205810547,
+      "learning_rate": 7.75883870385223e-06,
+      "loss": 0.3606,
+      "step": 512
+    },
+    {
+      "epoch": 1.1502242152466366,
+      "grad_norm": 0.3128066956996918,
+      "learning_rate": 7.747948587743823e-06,
+      "loss": 0.3494,
+      "step": 513
+    },
+    {
+      "epoch": 1.1524663677130045,
+      "grad_norm": 0.33549627661705017,
+      "learning_rate": 7.737039762439263e-06,
+      "loss": 0.3616,
+      "step": 514
+    },
+    {
+      "epoch": 1.1547085201793723,
+      "grad_norm": 0.2662554681301117,
+      "learning_rate": 7.726112302210449e-06,
+      "loss": 0.3554,
+      "step": 515
+    },
+    {
+      "epoch": 1.15695067264574,
+      "grad_norm": 0.3189751207828522,
+      "learning_rate": 7.71516628145615e-06,
+      "loss": 0.3523,
+      "step": 516
+    },
+    {
+      "epoch": 1.1591928251121075,
+      "grad_norm": 0.2950923442840576,
+      "learning_rate": 7.704201774701505e-06,
+      "loss": 0.3573,
+      "step": 517
+    },
+    {
+      "epoch": 1.1614349775784754,
+      "grad_norm": 0.3824740946292877,
+      "learning_rate": 7.693218856597515e-06,
+      "loss": 0.3656,
+      "step": 518
+    },
+    {
+      "epoch": 1.163677130044843,
+      "grad_norm": 0.29444941878318787,
+      "learning_rate": 7.682217601920529e-06,
+      "loss": 0.3554,
+      "step": 519
+    },
+    {
+      "epoch": 1.1659192825112108,
+      "grad_norm": 0.2922634780406952,
+      "learning_rate": 7.671198085571744e-06,
+      "loss": 0.3537,
+      "step": 520
+    },
+    {
+      "epoch": 1.1681614349775784,
+      "grad_norm": 0.32547417283058167,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.3562,
+      "step": 521
+    },
+    {
+      "epoch": 1.1704035874439462,
+      "grad_norm": 0.30854475498199463,
+      "learning_rate": 7.649104568084701e-06,
+      "loss": 0.3409,
+      "step": 522
+    },
+    {
+      "epoch": 1.1726457399103138,
+      "grad_norm": 0.3006819784641266,
+      "learning_rate": 7.638030717368458e-06,
+      "loss": 0.3522,
+      "step": 523
+    },
+    {
+      "epoch": 1.1748878923766817,
+      "grad_norm": 0.30922257900238037,
+      "learning_rate": 7.626938905823409e-06,
+      "loss": 0.3606,
+      "step": 524
+    },
+    {
+      "epoch": 1.1771300448430493,
+      "grad_norm": 0.480508029460907,
+      "learning_rate": 7.615829208967303e-06,
+      "loss": 0.3634,
+      "step": 525
+    },
+    {
+      "epoch": 1.1793721973094171,
+      "grad_norm": 0.3412018120288849,
+      "learning_rate": 7.604701702439652e-06,
+      "loss": 0.3645,
+      "step": 526
+    },
+    {
+      "epoch": 1.1816143497757847,
+      "grad_norm": 0.2719302177429199,
+      "learning_rate": 7.5935564620012285e-06,
+      "loss": 0.3575,
+      "step": 527
+    },
+    {
+      "epoch": 1.1838565022421526,
+      "grad_norm": 0.3078562319278717,
+      "learning_rate": 7.58239356353354e-06,
+      "loss": 0.3638,
+      "step": 528
+    },
+    {
+      "epoch": 1.1860986547085202,
+      "grad_norm": 0.26822179555892944,
+      "learning_rate": 7.571213083038322e-06,
+      "loss": 0.3458,
+      "step": 529
+    },
+    {
+      "epoch": 1.188340807174888,
+      "grad_norm": 0.29158663749694824,
+      "learning_rate": 7.560015096637015e-06,
+      "loss": 0.3605,
+      "step": 530
+    },
+    {
+      "epoch": 1.1905829596412556,
+      "grad_norm": 0.2806437313556671,
+      "learning_rate": 7.548799680570243e-06,
+      "loss": 0.3583,
+      "step": 531
+    },
+    {
+      "epoch": 1.1928251121076232,
+      "grad_norm": 0.40184906125068665,
+      "learning_rate": 7.537566911197307e-06,
+      "loss": 0.3623,
+      "step": 532
+    },
+    {
+      "epoch": 1.195067264573991,
+      "grad_norm": 0.2827012836933136,
+      "learning_rate": 7.526316864995648e-06,
+      "loss": 0.3524,
+      "step": 533
+    },
+    {
+      "epoch": 1.1973094170403586,
+      "grad_norm": 0.289886474609375,
+      "learning_rate": 7.515049618560337e-06,
+      "loss": 0.3537,
+      "step": 534
+    },
+    {
+      "epoch": 1.1995515695067265,
+      "grad_norm": 0.3068574368953705,
+      "learning_rate": 7.503765248603559e-06,
+      "loss": 0.3509,
+      "step": 535
+    },
+    {
+      "epoch": 1.201793721973094,
+      "grad_norm": 0.3583664894104004,
+      "learning_rate": 7.492463831954069e-06,
+      "loss": 0.3558,
+      "step": 536
+    },
+    {
+      "epoch": 1.204035874439462,
+      "grad_norm": 0.29033374786376953,
+      "learning_rate": 7.481145445556697e-06,
+      "loss": 0.3419,
+      "step": 537
+    },
+    {
+      "epoch": 1.2062780269058295,
+      "grad_norm": 0.27806687355041504,
+      "learning_rate": 7.469810166471802e-06,
+      "loss": 0.3668,
+      "step": 538
+    },
+    {
+      "epoch": 1.2085201793721974,
+      "grad_norm": 0.27455905079841614,
+      "learning_rate": 7.458458071874759e-06,
+      "loss": 0.3572,
+      "step": 539
+    },
+    {
+      "epoch": 1.210762331838565,
+      "grad_norm": 0.2968118488788605,
+      "learning_rate": 7.447089239055428e-06,
+      "loss": 0.3515,
+      "step": 540
+    },
+    {
+      "epoch": 1.2130044843049328,
+      "grad_norm": 0.27387139201164246,
+      "learning_rate": 7.4357037454176314e-06,
+      "loss": 0.3541,
+      "step": 541
+    },
+    {
+      "epoch": 1.2152466367713004,
+      "grad_norm": 0.28574517369270325,
+      "learning_rate": 7.424301668478626e-06,
+      "loss": 0.3496,
+      "step": 542
+    },
+    {
+      "epoch": 1.2174887892376682,
+      "grad_norm": 0.28276965022087097,
+      "learning_rate": 7.412883085868573e-06,
+      "loss": 0.36,
+      "step": 543
+    },
+    {
+      "epoch": 1.2197309417040358,
+      "grad_norm": 0.272532194852829,
+      "learning_rate": 7.401448075330011e-06,
+      "loss": 0.3436,
+      "step": 544
+    },
+    {
+      "epoch": 1.2219730941704037,
+      "grad_norm": 0.3622370660305023,
+      "learning_rate": 7.38999671471733e-06,
+      "loss": 0.3609,
+      "step": 545
+    },
+    {
+      "epoch": 1.2242152466367713,
+      "grad_norm": 0.3498156666755676,
+      "learning_rate": 7.378529081996233e-06,
+      "loss": 0.349,
+      "step": 546
+    },
+    {
+      "epoch": 1.226457399103139,
+      "grad_norm": 0.26131412386894226,
+      "learning_rate": 7.367045255243217e-06,
+      "loss": 0.3479,
+      "step": 547
+    },
+    {
+      "epoch": 1.2286995515695067,
+      "grad_norm": 0.289299875497818,
+      "learning_rate": 7.355545312645029e-06,
+      "loss": 0.3599,
+      "step": 548
+    },
+    {
+      "epoch": 1.2309417040358746,
+      "grad_norm": 0.2824910879135132,
+      "learning_rate": 7.344029332498143e-06,
+      "loss": 0.3596,
+      "step": 549
+    },
+    {
+      "epoch": 1.2331838565022422,
+      "grad_norm": 0.2902211844921112,
+      "learning_rate": 7.332497393208221e-06,
+      "loss": 0.345,
+      "step": 550
+    },
+    {
+      "epoch": 1.2354260089686098,
+      "grad_norm": 0.2738643288612366,
+      "learning_rate": 7.320949573289584e-06,
+      "loss": 0.3614,
+      "step": 551
+    },
+    {
+      "epoch": 1.2376681614349776,
+      "grad_norm": 0.3469460606575012,
+      "learning_rate": 7.309385951364672e-06,
+      "loss": 0.3633,
+      "step": 552
+    },
+    {
+      "epoch": 1.2399103139013452,
+      "grad_norm": 0.278457909822464,
+      "learning_rate": 7.297806606163516e-06,
+      "loss": 0.3476,
+      "step": 553
+    },
+    {
+      "epoch": 1.242152466367713,
+      "grad_norm": 0.2693907618522644,
+      "learning_rate": 7.286211616523193e-06,
+      "loss": 0.3509,
+      "step": 554
+    },
+    {
+      "epoch": 1.2443946188340806,
+      "grad_norm": 0.31849440932273865,
+      "learning_rate": 7.274601061387298e-06,
+      "loss": 0.3464,
+      "step": 555
+    },
+    {
+      "epoch": 1.2466367713004485,
+      "grad_norm": 0.33173537254333496,
+      "learning_rate": 7.2629750198054e-06,
+      "loss": 0.3629,
+      "step": 556
+    },
+    {
+      "epoch": 1.248878923766816,
+      "grad_norm": 0.3126336634159088,
+      "learning_rate": 7.251333570932507e-06,
+      "loss": 0.348,
+      "step": 557
+    },
+    {
+      "epoch": 1.251121076233184,
+      "grad_norm": 0.3139651417732239,
+      "learning_rate": 7.239676794028526e-06,
+      "loss": 0.3591,
+      "step": 558
+    },
+    {
+      "epoch": 1.2533632286995515,
+      "grad_norm": 0.3590622842311859,
+      "learning_rate": 7.2280047684577235e-06,
+      "loss": 0.3483,
+      "step": 559
+    },
+    {
+      "epoch": 1.2556053811659194,
+      "grad_norm": 0.2720869779586792,
+      "learning_rate": 7.216317573688189e-06,
+      "loss": 0.3553,
+      "step": 560
+    },
+    {
+      "epoch": 1.257847533632287,
+      "grad_norm": 0.28638508915901184,
+      "learning_rate": 7.204615289291283e-06,
+      "loss": 0.3565,
+      "step": 561
+    },
+    {
+      "epoch": 1.2600896860986546,
+      "grad_norm": 0.37933826446533203,
+      "learning_rate": 7.192897994941111e-06,
+      "loss": 0.3499,
+      "step": 562
+    },
+    {
+      "epoch": 1.2623318385650224,
+      "grad_norm": 0.2804115116596222,
+      "learning_rate": 7.1811657704139646e-06,
+      "loss": 0.3567,
+      "step": 563
+    },
+    {
+      "epoch": 1.2645739910313902,
+      "grad_norm": 0.3081219494342804,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.3513,
+      "step": 564
+    },
+    {
+      "epoch": 1.2668161434977578,
+      "grad_norm": 0.30695152282714844,
+      "learning_rate": 7.1576568504416464e-06,
+      "loss": 0.3631,
+      "step": 565
+    },
+    {
+      "epoch": 1.2690582959641254,
+      "grad_norm": 0.27527016401290894,
+      "learning_rate": 7.145880315055145e-06,
+      "loss": 0.3456,
+      "step": 566
+    },
+    {
+      "epoch": 1.2713004484304933,
+      "grad_norm": 0.3253130316734314,
+      "learning_rate": 7.13408916960792e-06,
+      "loss": 0.3615,
+      "step": 567
+    },
+    {
+      "epoch": 1.273542600896861,
+      "grad_norm": 0.25831273198127747,
+      "learning_rate": 7.122283494379076e-06,
+      "loss": 0.3584,
+      "step": 568
+    },
+    {
+      "epoch": 1.2757847533632287,
+      "grad_norm": 0.3219277858734131,
+      "learning_rate": 7.110463369746644e-06,
+      "loss": 0.3597,
+      "step": 569
+    },
+    {
+      "epoch": 1.2780269058295963,
+      "grad_norm": 0.3154788017272949,
+      "learning_rate": 7.098628876187031e-06,
+      "loss": 0.3472,
+      "step": 570
+    },
+    {
+      "epoch": 1.2802690582959642,
+      "grad_norm": 0.28053873777389526,
+      "learning_rate": 7.086780094274475e-06,
+      "loss": 0.3556,
+      "step": 571
+    },
+    {
+      "epoch": 1.2825112107623318,
+      "grad_norm": 0.41480255126953125,
+      "learning_rate": 7.074917104680494e-06,
+      "loss": 0.3638,
+      "step": 572
+    },
+    {
+      "epoch": 1.2847533632286996,
+      "grad_norm": 0.36695724725723267,
+      "learning_rate": 7.063039988173336e-06,
+      "loss": 0.3597,
+      "step": 573
+    },
+    {
+      "epoch": 1.2869955156950672,
+      "grad_norm": 0.2615639865398407,
+      "learning_rate": 7.051148825617435e-06,
+      "loss": 0.3583,
+      "step": 574
+    },
+    {
+      "epoch": 1.289237668161435,
+      "grad_norm": 0.2954905927181244,
+      "learning_rate": 7.039243697972856e-06,
+      "loss": 0.3514,
+      "step": 575
+    },
+    {
+      "epoch": 1.2914798206278026,
+      "grad_norm": 0.33068108558654785,
+      "learning_rate": 7.027324686294741e-06,
+      "loss": 0.3519,
+      "step": 576
+    },
+    {
+      "epoch": 1.2937219730941705,
+      "grad_norm": 0.7653133869171143,
+      "learning_rate": 7.0153918717327644e-06,
+      "loss": 0.3564,
+      "step": 577
+    },
+    {
+      "epoch": 1.295964125560538,
+      "grad_norm": 0.274482786655426,
+      "learning_rate": 7.003445335530572e-06,
+      "loss": 0.3444,
+      "step": 578
+    },
+    {
+      "epoch": 1.298206278026906,
+      "grad_norm": 0.3408889174461365,
+      "learning_rate": 6.991485159025236e-06,
+      "loss": 0.3504,
+      "step": 579
+    },
+    {
+      "epoch": 1.3004484304932735,
+      "grad_norm": 0.30217838287353516,
+      "learning_rate": 6.979511423646696e-06,
+      "loss": 0.3475,
+      "step": 580
+    },
+    {
+      "epoch": 1.3026905829596411,
+      "grad_norm": 0.3139374256134033,
+      "learning_rate": 6.9675242109172055e-06,
+      "loss": 0.3621,
+      "step": 581
+    },
+    {
+      "epoch": 1.304932735426009,
+      "grad_norm": 0.317721962928772,
+      "learning_rate": 6.95552360245078e-06,
+      "loss": 0.3542,
+      "step": 582
+    },
+    {
+      "epoch": 1.3071748878923768,
+      "grad_norm": 0.27298814058303833,
+      "learning_rate": 6.943509679952638e-06,
+      "loss": 0.3513,
+      "step": 583
+    },
+    {
+      "epoch": 1.3094170403587444,
+      "grad_norm": 0.37444284558296204,
+      "learning_rate": 6.931482525218641e-06,
+      "loss": 0.3564,
+      "step": 584
+    },
+    {
+      "epoch": 1.311659192825112,
+      "grad_norm": 0.31301143765449524,
+      "learning_rate": 6.919442220134749e-06,
+      "loss": 0.3457,
+      "step": 585
+    },
+    {
+      "epoch": 1.3139013452914798,
+      "grad_norm": 0.276804655790329,
+      "learning_rate": 6.9073888466764495e-06,
+      "loss": 0.3505,
+      "step": 586
+    },
+    {
+      "epoch": 1.3161434977578477,
+      "grad_norm": 0.2864084243774414,
+      "learning_rate": 6.895322486908206e-06,
+      "loss": 0.3503,
+      "step": 587
+    },
+    {
+      "epoch": 1.3183856502242153,
+      "grad_norm": 0.2761646509170532,
+      "learning_rate": 6.883243222982899e-06,
+      "loss": 0.3559,
+      "step": 588
+    },
+    {
+      "epoch": 1.3206278026905829,
+      "grad_norm": 0.3123345375061035,
+      "learning_rate": 6.871151137141266e-06,
+      "loss": 0.3491,
+      "step": 589
+    },
+    {
+      "epoch": 1.3228699551569507,
+      "grad_norm": 0.3964303731918335,
+      "learning_rate": 6.859046311711344e-06,
+      "loss": 0.3628,
+      "step": 590
+    },
+    {
+      "epoch": 1.3251121076233183,
+      "grad_norm": 0.39678052067756653,
+      "learning_rate": 6.846928829107899e-06,
+      "loss": 0.3676,
+      "step": 591
+    },
+    {
+      "epoch": 1.3273542600896862,
+      "grad_norm": 0.2885213792324066,
+      "learning_rate": 6.834798771831882e-06,
+      "loss": 0.3609,
+      "step": 592
+    },
+    {
+      "epoch": 1.3295964125560538,
+      "grad_norm": 0.2681860327720642,
+      "learning_rate": 6.82265622246985e-06,
+      "loss": 0.3558,
+      "step": 593
+    },
+    {
+      "epoch": 1.3318385650224216,
+      "grad_norm": 0.36434921622276306,
+      "learning_rate": 6.810501263693416e-06,
+      "loss": 0.3475,
+      "step": 594
+    },
+    {
+      "epoch": 1.3340807174887892,
+      "grad_norm": 0.3095291256904602,
+      "learning_rate": 6.79833397825868e-06,
+      "loss": 0.3642,
+      "step": 595
+    },
+    {
+      "epoch": 1.336322869955157,
+      "grad_norm": 0.2936202883720398,
+      "learning_rate": 6.786154449005664e-06,
+      "loss": 0.3496,
+      "step": 596
+    },
+    {
+      "epoch": 1.3385650224215246,
+      "grad_norm": 0.3011555075645447,
+      "learning_rate": 6.7739627588577594e-06,
+      "loss": 0.352,
+      "step": 597
+    },
+    {
+      "epoch": 1.3408071748878925,
+      "grad_norm": 1.0378189086914062,
+      "learning_rate": 6.761758990821143e-06,
+      "loss": 0.3509,
+      "step": 598
+    },
+    {
+      "epoch": 1.34304932735426,
+      "grad_norm": 0.3466256856918335,
+      "learning_rate": 6.749543227984234e-06,
+      "loss": 0.3523,
+      "step": 599
+    },
+    {
+      "epoch": 1.3452914798206277,
+      "grad_norm": 0.2917185425758362,
+      "learning_rate": 6.73731555351711e-06,
+      "loss": 0.3547,
+      "step": 600
+    },
+    {
+      "epoch": 1.3475336322869955,
+      "grad_norm": 0.3015524446964264,
+      "learning_rate": 6.725076050670949e-06,
+      "loss": 0.3493,
+      "step": 601
+    },
+    {
+      "epoch": 1.3497757847533634,
+      "grad_norm": 0.35212278366088867,
+      "learning_rate": 6.712824802777465e-06,
+      "loss": 0.3559,
+      "step": 602
+    },
+    {
+      "epoch": 1.352017937219731,
+      "grad_norm": 0.28707870841026306,
+      "learning_rate": 6.700561893248332e-06,
+      "loss": 0.369,
+      "step": 603
+    },
+    {
+      "epoch": 1.3542600896860986,
+      "grad_norm": 0.30742138624191284,
+      "learning_rate": 6.688287405574627e-06,
+      "loss": 0.3427,
+      "step": 604
+    },
+    {
+      "epoch": 1.3565022421524664,
+      "grad_norm": 0.30532118678092957,
+      "learning_rate": 6.676001423326252e-06,
+      "loss": 0.3524,
+      "step": 605
+    },
+    {
+      "epoch": 1.358744394618834,
+      "grad_norm": 0.2950114905834198,
+      "learning_rate": 6.66370403015137e-06,
+      "loss": 0.3477,
+      "step": 606
+    },
+    {
+      "epoch": 1.3609865470852018,
+      "grad_norm": 0.34040990471839905,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.3468,
+      "step": 607
+    },
+    {
+      "epoch": 1.3632286995515694,
+      "grad_norm": 0.29443567991256714,
+      "learning_rate": 6.639075346002621e-06,
+      "loss": 0.3563,
+      "step": 608
+    },
+    {
+      "epoch": 1.3654708520179373,
+      "grad_norm": 0.2683103680610657,
+      "learning_rate": 6.626744222711254e-06,
+      "loss": 0.3493,
+      "step": 609
+    },
+    {
+      "epoch": 1.3677130044843049,
+      "grad_norm": 0.26770246028900146,
+      "learning_rate": 6.614402023857231e-06,
+      "loss": 0.3388,
+      "step": 610
+    },
+    {
+      "epoch": 1.3699551569506727,
+      "grad_norm": 0.26347848773002625,
+      "learning_rate": 6.602048833471466e-06,
+      "loss": 0.3491,
+      "step": 611
+    },
+    {
+      "epoch": 1.3721973094170403,
+      "grad_norm": 0.25639399886131287,
+      "learning_rate": 6.589684735659702e-06,
+      "loss": 0.3497,
+      "step": 612
+    },
+    {
+      "epoch": 1.3744394618834082,
+      "grad_norm": 0.2923979163169861,
+      "learning_rate": 6.5773098146019464e-06,
+      "loss": 0.3444,
+      "step": 613
+    },
+    {
+      "epoch": 1.3766816143497758,
+      "grad_norm": 0.2778940796852112,
+      "learning_rate": 6.564924154551895e-06,
+      "loss": 0.3447,
+      "step": 614
+    },
+    {
+      "epoch": 1.3789237668161434,
+      "grad_norm": 0.259247750043869,
+      "learning_rate": 6.55252783983636e-06,
+      "loss": 0.3492,
+      "step": 615
+    },
+    {
+      "epoch": 1.3811659192825112,
+      "grad_norm": 0.2662297189235687,
+      "learning_rate": 6.540120954854694e-06,
+      "loss": 0.3563,
+      "step": 616
+    },
+    {
+      "epoch": 1.383408071748879,
+      "grad_norm": 0.25957462191581726,
+      "learning_rate": 6.527703584078219e-06,
+      "loss": 0.3414,
+      "step": 617
+    },
+    {
+      "epoch": 1.3856502242152466,
+      "grad_norm": 0.27066195011138916,
+      "learning_rate": 6.515275812049644e-06,
+      "loss": 0.3701,
+      "step": 618
+    },
+    {
+      "epoch": 1.3878923766816142,
+      "grad_norm": 0.28484660387039185,
+      "learning_rate": 6.502837723382499e-06,
+      "loss": 0.3447,
+      "step": 619
+    },
+    {
+      "epoch": 1.390134529147982,
+      "grad_norm": 0.3598700761795044,
+      "learning_rate": 6.490389402760549e-06,
+      "loss": 0.3377,
+      "step": 620
+    },
+    {
+      "epoch": 1.39237668161435,
+      "grad_norm": 0.2666773796081543,
+      "learning_rate": 6.477930934937231e-06,
+      "loss": 0.3496,
+      "step": 621
+    },
+    {
+      "epoch": 1.3946188340807175,
+      "grad_norm": 0.32423263788223267,
+      "learning_rate": 6.4654624047350575e-06,
+      "loss": 0.3505,
+      "step": 622
+    },
+    {
+      "epoch": 1.3968609865470851,
+      "grad_norm": 0.24971212446689606,
+      "learning_rate": 6.452983897045057e-06,
+      "loss": 0.3434,
+      "step": 623
+    },
+    {
+      "epoch": 1.399103139013453,
+      "grad_norm": 0.27574530243873596,
+      "learning_rate": 6.440495496826189e-06,
+      "loss": 0.3526,
+      "step": 624
+    },
+    {
+      "epoch": 1.4013452914798206,
+      "grad_norm": 0.3362599313259125,
+      "learning_rate": 6.427997289104763e-06,
+      "loss": 0.349,
+      "step": 625
+    },
+    {
+      "epoch": 1.4035874439461884,
+      "grad_norm": 0.34073925018310547,
+      "learning_rate": 6.41548935897386e-06,
+      "loss": 0.3554,
+      "step": 626
+    },
+    {
+      "epoch": 1.405829596412556,
+      "grad_norm": 0.3609859347343445,
+      "learning_rate": 6.402971791592762e-06,
+      "loss": 0.3467,
+      "step": 627
+    },
+    {
+      "epoch": 1.4080717488789238,
+      "grad_norm": 0.29569607973098755,
+      "learning_rate": 6.39044467218636e-06,
+      "loss": 0.3523,
+      "step": 628
+    },
+    {
+      "epoch": 1.4103139013452914,
+      "grad_norm": 0.2835816442966461,
+      "learning_rate": 6.377908086044582e-06,
+      "loss": 0.3622,
+      "step": 629
+    },
+    {
+      "epoch": 1.4125560538116593,
+      "grad_norm": 0.27152466773986816,
+      "learning_rate": 6.365362118521807e-06,
+      "loss": 0.352,
+      "step": 630
+    },
+    {
+      "epoch": 1.4147982062780269,
+      "grad_norm": 0.2971261739730835,
+      "learning_rate": 6.352806855036287e-06,
+      "loss": 0.3408,
+      "step": 631
+    },
+    {
+      "epoch": 1.4170403587443947,
+      "grad_norm": 0.24410578608512878,
+      "learning_rate": 6.34024238106957e-06,
+      "loss": 0.3455,
+      "step": 632
+    },
+    {
+      "epoch": 1.4192825112107623,
+      "grad_norm": 0.2897018790245056,
+      "learning_rate": 6.327668782165903e-06,
+      "loss": 0.3508,
+      "step": 633
+    },
+    {
+      "epoch": 1.42152466367713,
+      "grad_norm": 0.29007428884506226,
+      "learning_rate": 6.31508614393167e-06,
+      "loss": 0.3538,
+      "step": 634
+    },
+    {
+      "epoch": 1.4237668161434978,
+      "grad_norm": 0.26024141907691956,
+      "learning_rate": 6.30249455203479e-06,
+      "loss": 0.3548,
+      "step": 635
+    },
+    {
+      "epoch": 1.4260089686098656,
+      "grad_norm": 0.2545620799064636,
+      "learning_rate": 6.2898940922041475e-06,
+      "loss": 0.3551,
+      "step": 636
+    },
+    {
+      "epoch": 1.4282511210762332,
+      "grad_norm": 0.2681948244571686,
+      "learning_rate": 6.277284850229002e-06,
+      "loss": 0.3544,
+      "step": 637
+    },
+    {
+      "epoch": 1.4304932735426008,
+      "grad_norm": 0.261748731136322,
+      "learning_rate": 6.264666911958404e-06,
+      "loss": 0.3409,
+      "step": 638
+    },
+    {
+      "epoch": 1.4327354260089686,
+      "grad_norm": 0.2525177299976349,
+      "learning_rate": 6.252040363300617e-06,
+      "loss": 0.3597,
+      "step": 639
+    },
+    {
+      "epoch": 1.4349775784753362,
+      "grad_norm": 0.28739526867866516,
+      "learning_rate": 6.23940529022252e-06,
+      "loss": 0.3581,
+      "step": 640
+    },
+    {
+      "epoch": 1.437219730941704,
+      "grad_norm": 0.33014583587646484,
+      "learning_rate": 6.2267617787490375e-06,
+      "loss": 0.3547,
+      "step": 641
+    },
+    {
+      "epoch": 1.4394618834080717,
+      "grad_norm": 0.46392080187797546,
+      "learning_rate": 6.214109914962542e-06,
+      "loss": 0.3412,
+      "step": 642
+    },
+    {
+      "epoch": 1.4417040358744395,
+      "grad_norm": 0.2541409134864807,
+      "learning_rate": 6.201449785002272e-06,
+      "loss": 0.3509,
+      "step": 643
+    },
+    {
+      "epoch": 1.4439461883408071,
+      "grad_norm": 0.3008180558681488,
+      "learning_rate": 6.188781475063747e-06,
+      "loss": 0.3511,
+      "step": 644
+    },
+    {
+      "epoch": 1.446188340807175,
+      "grad_norm": 0.2632216811180115,
+      "learning_rate": 6.1761050713981795e-06,
+      "loss": 0.3542,
+      "step": 645
+    },
+    {
+      "epoch": 1.4484304932735426,
+      "grad_norm": 0.37776893377304077,
+      "learning_rate": 6.1634206603118844e-06,
+      "loss": 0.3639,
+      "step": 646
+    },
+    {
+      "epoch": 1.4506726457399104,
+      "grad_norm": 0.30006882548332214,
+      "learning_rate": 6.1507283281657005e-06,
+      "loss": 0.3626,
+      "step": 647
+    },
+    {
+      "epoch": 1.452914798206278,
+      "grad_norm": 0.29537200927734375,
+      "learning_rate": 6.13802816137439e-06,
+      "loss": 0.3538,
+      "step": 648
+    },
+    {
+      "epoch": 1.4551569506726456,
+      "grad_norm": 0.3013269901275635,
+      "learning_rate": 6.125320246406061e-06,
+      "loss": 0.3555,
+      "step": 649
+    },
+    {
+      "epoch": 1.4573991031390134,
+      "grad_norm": 0.2460028976202011,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.3556,
+      "step": 650
+    },
+    {
+      "epoch": 1.4596412556053813,
+      "grad_norm": 0.2585391104221344,
+      "learning_rate": 6.099881518073948e-06,
+      "loss": 0.3494,
+      "step": 651
+    },
+    {
+      "epoch": 1.4618834080717489,
+      "grad_norm": 0.28896862268447876,
+      "learning_rate": 6.087150877907786e-06,
+      "loss": 0.35,
+      "step": 652
+    },
+    {
+      "epoch": 1.4641255605381165,
+      "grad_norm": 0.2583191692829132,
+      "learning_rate": 6.0744128359586675e-06,
+      "loss": 0.355,
+      "step": 653
+    },
+    {
+      "epoch": 1.4663677130044843,
+      "grad_norm": 0.2757875621318817,
+      "learning_rate": 6.06166747895257e-06,
+      "loss": 0.3512,
+      "step": 654
+    },
+    {
+      "epoch": 1.4686098654708521,
+      "grad_norm": 0.24874940514564514,
+      "learning_rate": 6.048914893665272e-06,
+      "loss": 0.3449,
+      "step": 655
+    },
+    {
+      "epoch": 1.4708520179372198,
+      "grad_norm": 0.27993834018707275,
+      "learning_rate": 6.036155166921772e-06,
+      "loss": 0.3541,
+      "step": 656
+    },
+    {
+      "epoch": 1.4730941704035874,
+      "grad_norm": 0.2902776598930359,
+      "learning_rate": 6.023388385595683e-06,
+      "loss": 0.3736,
+      "step": 657
+    },
+    {
+      "epoch": 1.4753363228699552,
+      "grad_norm": 0.2919728457927704,
+      "learning_rate": 6.0106146366086514e-06,
+      "loss": 0.3563,
+      "step": 658
+    },
+    {
+      "epoch": 1.4775784753363228,
+      "grad_norm": 0.31043824553489685,
+      "learning_rate": 5.997834006929765e-06,
+      "loss": 0.3444,
+      "step": 659
+    },
+    {
+      "epoch": 1.4798206278026906,
+      "grad_norm": 0.28806355595588684,
+      "learning_rate": 5.985046583574953e-06,
+      "loss": 0.3498,
+      "step": 660
+    },
+    {
+      "epoch": 1.4820627802690582,
+      "grad_norm": 0.2964230477809906,
+      "learning_rate": 5.972252453606402e-06,
+      "loss": 0.3447,
+      "step": 661
+    },
+    {
+      "epoch": 1.484304932735426,
+      "grad_norm": 0.4060839116573334,
+      "learning_rate": 5.959451704131962e-06,
+      "loss": 0.3593,
+      "step": 662
+    },
+    {
+      "epoch": 1.4865470852017937,
+      "grad_norm": 0.2563171088695526,
+      "learning_rate": 5.946644422304547e-06,
+      "loss": 0.351,
+      "step": 663
+    },
+    {
+      "epoch": 1.4887892376681615,
+      "grad_norm": 0.27978602051734924,
+      "learning_rate": 5.93383069532155e-06,
+      "loss": 0.3455,
+      "step": 664
+    },
+    {
+      "epoch": 1.4910313901345291,
+      "grad_norm": 0.26793238520622253,
+      "learning_rate": 5.921010610424242e-06,
+      "loss": 0.3479,
+      "step": 665
+    },
+    {
+      "epoch": 1.493273542600897,
+      "grad_norm": 0.3125176429748535,
+      "learning_rate": 5.908184254897183e-06,
+      "loss": 0.3428,
+      "step": 666
+    },
+    {
+      "epoch": 1.4955156950672646,
+      "grad_norm": 0.25852832198143005,
+      "learning_rate": 5.895351716067626e-06,
+      "loss": 0.3542,
+      "step": 667
+    },
+    {
+      "epoch": 1.4977578475336322,
+      "grad_norm": 0.2737487554550171,
+      "learning_rate": 5.882513081304924e-06,
+      "loss": 0.3553,
+      "step": 668
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.25031042098999023,
+      "learning_rate": 5.86966843801993e-06,
+      "loss": 0.3572,
+      "step": 669
+    },
+    {
+      "epoch": 1.5022421524663678,
+      "grad_norm": 0.27862465381622314,
+      "learning_rate": 5.856817873664409e-06,
+      "loss": 0.3511,
+      "step": 670
+    },
+    {
+      "epoch": 1.5044843049327354,
+      "grad_norm": 0.26174575090408325,
+      "learning_rate": 5.843961475730437e-06,
+      "loss": 0.3434,
+      "step": 671
+    },
+    {
+      "epoch": 1.506726457399103,
+      "grad_norm": 0.26741865277290344,
+      "learning_rate": 5.83109933174981e-06,
+      "loss": 0.3572,
+      "step": 672
+    },
+    {
+      "epoch": 1.5089686098654709,
+      "grad_norm": 1.1492397785186768,
+      "learning_rate": 5.818231529293441e-06,
+      "loss": 0.3554,
+      "step": 673
+    },
+    {
+      "epoch": 1.5112107623318387,
+      "grad_norm": 0.2663498520851135,
+      "learning_rate": 5.8053581559707754e-06,
+      "loss": 0.3438,
+      "step": 674
+    },
+    {
+      "epoch": 1.5134529147982063,
+      "grad_norm": 0.24593260884284973,
+      "learning_rate": 5.79247929942918e-06,
+      "loss": 0.353,
+      "step": 675
+    },
+    {
+      "epoch": 1.515695067264574,
+      "grad_norm": 0.29982686042785645,
+      "learning_rate": 5.779595047353357e-06,
+      "loss": 0.3396,
+      "step": 676
+    },
+    {
+      "epoch": 1.5179372197309418,
+      "grad_norm": 0.24093005061149597,
+      "learning_rate": 5.766705487464745e-06,
+      "loss": 0.3532,
+      "step": 677
+    },
+    {
+      "epoch": 1.5201793721973094,
+      "grad_norm": 0.2645823359489441,
+      "learning_rate": 5.753810707520918e-06,
+      "loss": 0.3576,
+      "step": 678
+    },
+    {
+      "epoch": 1.522421524663677,
+      "grad_norm": 0.2721596658229828,
+      "learning_rate": 5.740910795314992e-06,
+      "loss": 0.3584,
+      "step": 679
+    },
+    {
+      "epoch": 1.5246636771300448,
+      "grad_norm": 0.5878199934959412,
+      "learning_rate": 5.728005838675026e-06,
+      "loss": 0.3524,
+      "step": 680
+    },
+    {
+      "epoch": 1.5269058295964126,
+      "grad_norm": 0.26442667841911316,
+      "learning_rate": 5.7150959254634186e-06,
+      "loss": 0.3476,
+      "step": 681
+    },
+    {
+      "epoch": 1.5291479820627802,
+      "grad_norm": 0.2946506142616272,
+      "learning_rate": 5.702181143576323e-06,
+      "loss": 0.3529,
+      "step": 682
+    },
+    {
+      "epoch": 1.5313901345291479,
+      "grad_norm": 0.2519643306732178,
+      "learning_rate": 5.689261580943033e-06,
+      "loss": 0.3656,
+      "step": 683
+    },
+    {
+      "epoch": 1.5336322869955157,
+      "grad_norm": 0.26235291361808777,
+      "learning_rate": 5.676337325525398e-06,
+      "loss": 0.3544,
+      "step": 684
+    },
+    {
+      "epoch": 1.5358744394618835,
+      "grad_norm": 0.23943795263767242,
+      "learning_rate": 5.663408465317214e-06,
+      "loss": 0.3483,
+      "step": 685
+    },
+    {
+      "epoch": 1.5381165919282511,
+      "grad_norm": 0.25865891575813293,
+      "learning_rate": 5.6504750883436275e-06,
+      "loss": 0.3489,
+      "step": 686
+    },
+    {
+      "epoch": 1.5403587443946187,
+      "grad_norm": 0.2687196433544159,
+      "learning_rate": 5.63753728266054e-06,
+      "loss": 0.3525,
+      "step": 687
+    },
+    {
+      "epoch": 1.5426008968609866,
+      "grad_norm": 0.2684645354747772,
+      "learning_rate": 5.624595136354004e-06,
+      "loss": 0.3578,
+      "step": 688
+    },
+    {
+      "epoch": 1.5448430493273544,
+      "grad_norm": 0.28312748670578003,
+      "learning_rate": 5.611648737539626e-06,
+      "loss": 0.3568,
+      "step": 689
+    },
+    {
+      "epoch": 1.547085201793722,
+      "grad_norm": 0.26138702034950256,
+      "learning_rate": 5.5986981743619615e-06,
+      "loss": 0.3669,
+      "step": 690
+    },
+    {
+      "epoch": 1.5493273542600896,
+      "grad_norm": 0.2802727222442627,
+      "learning_rate": 5.585743534993926e-06,
+      "loss": 0.3546,
+      "step": 691
+    },
+    {
+      "epoch": 1.5515695067264574,
+      "grad_norm": 0.24105150997638702,
+      "learning_rate": 5.572784907636181e-06,
+      "loss": 0.3543,
+      "step": 692
+    },
+    {
+      "epoch": 1.5538116591928253,
+      "grad_norm": 0.3083956241607666,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 0.349,
+      "step": 693
+    },
+    {
+      "epoch": 1.5560538116591929,
+      "grad_norm": 0.24600344896316528,
+      "learning_rate": 5.546856041889374e-06,
+      "loss": 0.3513,
+      "step": 694
+    },
+    {
+      "epoch": 1.5582959641255605,
+      "grad_norm": 0.26413145661354065,
+      "learning_rate": 5.533885980034995e-06,
+      "loss": 0.35,
+      "step": 695
+    },
+    {
+      "epoch": 1.5605381165919283,
+      "grad_norm": 0.28658759593963623,
+      "learning_rate": 5.520912283259073e-06,
+      "loss": 0.3558,
+      "step": 696
+    },
+    {
+      "epoch": 1.562780269058296,
+      "grad_norm": 0.25767186284065247,
+      "learning_rate": 5.507935039892024e-06,
+      "loss": 0.3396,
+      "step": 697
+    },
+    {
+      "epoch": 1.5650224215246635,
+      "grad_norm": 0.24957624077796936,
+      "learning_rate": 5.494954338288404e-06,
+      "loss": 0.356,
+      "step": 698
+    },
+    {
+      "epoch": 1.5672645739910314,
+      "grad_norm": 0.2554234564304352,
+      "learning_rate": 5.481970266826324e-06,
+      "loss": 0.356,
+      "step": 699
+    },
+    {
+      "epoch": 1.5695067264573992,
+      "grad_norm": 0.24980078637599945,
+      "learning_rate": 5.468982913906831e-06,
+      "loss": 0.3542,
+      "step": 700
+    },
+    {
+      "epoch": 1.5717488789237668,
+      "grad_norm": 0.2521933615207672,
+      "learning_rate": 5.455992367953318e-06,
+      "loss": 0.3502,
+      "step": 701
+    },
+    {
+      "epoch": 1.5739910313901344,
+      "grad_norm": 0.2527569830417633,
+      "learning_rate": 5.442998717410916e-06,
+      "loss": 0.3518,
+      "step": 702
+    },
+    {
+      "epoch": 1.5762331838565022,
+      "grad_norm": 0.24163737893104553,
+      "learning_rate": 5.43000205074589e-06,
+      "loss": 0.3538,
+      "step": 703
+    },
+    {
+      "epoch": 1.57847533632287,
+      "grad_norm": 0.25389352440834045,
+      "learning_rate": 5.417002456445046e-06,
+      "loss": 0.3563,
+      "step": 704
+    },
+    {
+      "epoch": 1.5807174887892377,
+      "grad_norm": 0.2582768499851227,
+      "learning_rate": 5.40400002301512e-06,
+      "loss": 0.3479,
+      "step": 705
+    },
+    {
+      "epoch": 1.5829596412556053,
+      "grad_norm": 0.2500203251838684,
+      "learning_rate": 5.390994838982178e-06,
+      "loss": 0.3512,
+      "step": 706
+    },
+    {
+      "epoch": 1.5852017937219731,
+      "grad_norm": 0.24961726367473602,
+      "learning_rate": 5.377986992891011e-06,
+      "loss": 0.3382,
+      "step": 707
+    },
+    {
+      "epoch": 1.587443946188341,
+      "grad_norm": 0.3987758755683899,
+      "learning_rate": 5.364976573304538e-06,
+      "loss": 0.3554,
+      "step": 708
+    },
+    {
+      "epoch": 1.5896860986547086,
+      "grad_norm": 0.23992732167243958,
+      "learning_rate": 5.351963668803199e-06,
+      "loss": 0.3536,
+      "step": 709
+    },
+    {
+      "epoch": 1.5919282511210762,
+      "grad_norm": 0.24111761152744293,
+      "learning_rate": 5.338948367984347e-06,
+      "loss": 0.3521,
+      "step": 710
+    },
+    {
+      "epoch": 1.594170403587444,
+      "grad_norm": 0.31662869453430176,
+      "learning_rate": 5.32593075946166e-06,
+      "loss": 0.3491,
+      "step": 711
+    },
+    {
+      "epoch": 1.5964125560538116,
+      "grad_norm": 0.2557629644870758,
+      "learning_rate": 5.312910931864518e-06,
+      "loss": 0.3485,
+      "step": 712
+    },
+    {
+      "epoch": 1.5986547085201792,
+      "grad_norm": 0.252812922000885,
+      "learning_rate": 5.2998889738374145e-06,
+      "loss": 0.3429,
+      "step": 713
+    },
+    {
+      "epoch": 1.600896860986547,
+      "grad_norm": 0.2637932598590851,
+      "learning_rate": 5.286864974039349e-06,
+      "loss": 0.3438,
+      "step": 714
+    },
+    {
+      "epoch": 1.6031390134529149,
+      "grad_norm": 0.24221956729888916,
+      "learning_rate": 5.273839021143217e-06,
+      "loss": 0.3505,
+      "step": 715
+    },
+    {
+      "epoch": 1.6053811659192825,
+      "grad_norm": 0.2925490736961365,
+      "learning_rate": 5.26081120383522e-06,
+      "loss": 0.3531,
+      "step": 716
+    },
+    {
+      "epoch": 1.60762331838565,
+      "grad_norm": 0.2337847203016281,
+      "learning_rate": 5.247781610814243e-06,
+      "loss": 0.3627,
+      "step": 717
+    },
+    {
+      "epoch": 1.609865470852018,
+      "grad_norm": 0.23567251861095428,
+      "learning_rate": 5.234750330791268e-06,
+      "loss": 0.3518,
+      "step": 718
+    },
+    {
+      "epoch": 1.6121076233183858,
+      "grad_norm": 0.2610297203063965,
+      "learning_rate": 5.221717452488759e-06,
+      "loss": 0.354,
+      "step": 719
+    },
+    {
+      "epoch": 1.6143497757847534,
+      "grad_norm": 0.2301870733499527,
+      "learning_rate": 5.2086830646400666e-06,
+      "loss": 0.3442,
+      "step": 720
+    },
+    {
+      "epoch": 1.616591928251121,
+      "grad_norm": 0.23972411453723907,
+      "learning_rate": 5.195647255988812e-06,
+      "loss": 0.3451,
+      "step": 721
+    },
+    {
+      "epoch": 1.6188340807174888,
+      "grad_norm": 0.2249157428741455,
+      "learning_rate": 5.182610115288296e-06,
+      "loss": 0.3574,
+      "step": 722
+    },
+    {
+      "epoch": 1.6210762331838566,
+      "grad_norm": 0.2359374761581421,
+      "learning_rate": 5.169571731300884e-06,
+      "loss": 0.3475,
+      "step": 723
+    },
+    {
+      "epoch": 1.6233183856502242,
+      "grad_norm": 0.2330854833126068,
+      "learning_rate": 5.156532192797411e-06,
+      "loss": 0.3413,
+      "step": 724
+    },
+    {
+      "epoch": 1.6255605381165918,
+      "grad_norm": 0.2326117902994156,
+      "learning_rate": 5.143491588556567e-06,
+      "loss": 0.3494,
+      "step": 725
+    },
+    {
+      "epoch": 1.6278026905829597,
+      "grad_norm": 0.2461560219526291,
+      "learning_rate": 5.1304500073643045e-06,
+      "loss": 0.3538,
+      "step": 726
+    },
+    {
+      "epoch": 1.6300448430493275,
+      "grad_norm": 0.2723434865474701,
+      "learning_rate": 5.117407538013222e-06,
+      "loss": 0.3368,
+      "step": 727
+    },
+    {
+      "epoch": 1.6322869955156951,
+      "grad_norm": 0.225622296333313,
+      "learning_rate": 5.104364269301964e-06,
+      "loss": 0.3466,
+      "step": 728
+    },
+    {
+      "epoch": 1.6345291479820627,
+      "grad_norm": 0.25469058752059937,
+      "learning_rate": 5.0913202900346246e-06,
+      "loss": 0.3507,
+      "step": 729
+    },
+    {
+      "epoch": 1.6367713004484306,
+      "grad_norm": 0.25067952275276184,
+      "learning_rate": 5.078275689020129e-06,
+      "loss": 0.3442,
+      "step": 730
+    },
+    {
+      "epoch": 1.6390134529147982,
+      "grad_norm": 0.2440514713525772,
+      "learning_rate": 5.065230555071639e-06,
+      "loss": 0.3469,
+      "step": 731
+    },
+    {
+      "epoch": 1.6412556053811658,
+      "grad_norm": 0.2410144954919815,
+      "learning_rate": 5.052184977005942e-06,
+      "loss": 0.3497,
+      "step": 732
+    },
+    {
+      "epoch": 1.6434977578475336,
+      "grad_norm": 0.6946483254432678,
+      "learning_rate": 5.03913904364285e-06,
+      "loss": 0.3434,
+      "step": 733
+    },
+    {
+      "epoch": 1.6457399103139014,
+      "grad_norm": 0.2710442841053009,
+      "learning_rate": 5.026092843804599e-06,
+      "loss": 0.3323,
+      "step": 734
+    },
+    {
+      "epoch": 1.647982062780269,
+      "grad_norm": 0.24127882719039917,
+      "learning_rate": 5.013046466315229e-06,
+      "loss": 0.3495,
+      "step": 735
+    },
+    {
+      "epoch": 1.6502242152466366,
+      "grad_norm": 0.27001267671585083,
+      "learning_rate": 5e-06,
+      "loss": 0.3552,
+      "step": 736
+    },
+    {
+      "epoch": 1.6524663677130045,
+      "grad_norm": 0.23515008389949799,
+      "learning_rate": 4.986953533684772e-06,
+      "loss": 0.3442,
+      "step": 737
+    },
+    {
+      "epoch": 1.6547085201793723,
+      "grad_norm": 0.24893882870674133,
+      "learning_rate": 4.973907156195405e-06,
+      "loss": 0.3451,
+      "step": 738
+    },
+    {
+      "epoch": 1.65695067264574,
+      "grad_norm": 0.23909777402877808,
+      "learning_rate": 4.960860956357151e-06,
+      "loss": 0.3503,
+      "step": 739
+    },
+    {
+      "epoch": 1.6591928251121075,
+      "grad_norm": 0.31813114881515503,
+      "learning_rate": 4.94781502299406e-06,
+      "loss": 0.3425,
+      "step": 740
+    },
+    {
+      "epoch": 1.6614349775784754,
+      "grad_norm": 0.2263074517250061,
+      "learning_rate": 4.934769444928362e-06,
+      "loss": 0.3434,
+      "step": 741
+    },
+    {
+      "epoch": 1.6636771300448432,
+      "grad_norm": 0.23301279544830322,
+      "learning_rate": 4.921724310979872e-06,
+      "loss": 0.3444,
+      "step": 742
+    },
+    {
+      "epoch": 1.6659192825112108,
+      "grad_norm": 0.23419252038002014,
+      "learning_rate": 4.908679709965376e-06,
+      "loss": 0.3366,
+      "step": 743
+    },
+    {
+      "epoch": 1.6681614349775784,
+      "grad_norm": 0.25969913601875305,
+      "learning_rate": 4.895635730698036e-06,
+      "loss": 0.3564,
+      "step": 744
+    },
+    {
+      "epoch": 1.6704035874439462,
+      "grad_norm": 0.2215176820755005,
+      "learning_rate": 4.882592461986781e-06,
+      "loss": 0.3503,
+      "step": 745
+    },
+    {
+      "epoch": 1.672645739910314,
+      "grad_norm": 0.2533276379108429,
+      "learning_rate": 4.869549992635697e-06,
+      "loss": 0.359,
+      "step": 746
+    },
+    {
+      "epoch": 1.6748878923766815,
+      "grad_norm": 0.28641271591186523,
+      "learning_rate": 4.856508411443433e-06,
+      "loss": 0.3519,
+      "step": 747
+    },
+    {
+      "epoch": 1.6771300448430493,
+      "grad_norm": 0.2498161345720291,
+      "learning_rate": 4.843467807202592e-06,
+      "loss": 0.3424,
+      "step": 748
+    },
+    {
+      "epoch": 1.6793721973094171,
+      "grad_norm": 0.3558255732059479,
+      "learning_rate": 4.830428268699118e-06,
+      "loss": 0.3419,
+      "step": 749
+    },
+    {
+      "epoch": 1.6816143497757847,
+      "grad_norm": 0.24943213164806366,
+      "learning_rate": 4.817389884711706e-06,
+      "loss": 0.3447,
+      "step": 750
+    },
+    {
+      "epoch": 1.6838565022421523,
+      "grad_norm": 0.2571375370025635,
+      "learning_rate": 4.804352744011189e-06,
+      "loss": 0.3432,
+      "step": 751
+    },
+    {
+      "epoch": 1.6860986547085202,
+      "grad_norm": 0.2723126709461212,
+      "learning_rate": 4.791316935359935e-06,
+      "loss": 0.3466,
+      "step": 752
+    },
+    {
+      "epoch": 1.688340807174888,
+      "grad_norm": 0.36954352259635925,
+      "learning_rate": 4.778282547511242e-06,
+      "loss": 0.3585,
+      "step": 753
+    },
+    {
+      "epoch": 1.6905829596412556,
+      "grad_norm": 0.35934436321258545,
+      "learning_rate": 4.765249669208733e-06,
+      "loss": 0.342,
+      "step": 754
+    },
+    {
+      "epoch": 1.6928251121076232,
+      "grad_norm": 0.23793573677539825,
+      "learning_rate": 4.752218389185757e-06,
+      "loss": 0.3537,
+      "step": 755
+    },
+    {
+      "epoch": 1.695067264573991,
+      "grad_norm": 0.2712377607822418,
+      "learning_rate": 4.739188796164782e-06,
+      "loss": 0.3521,
+      "step": 756
+    },
+    {
+      "epoch": 1.6973094170403589,
+      "grad_norm": 0.22843128442764282,
+      "learning_rate": 4.726160978856782e-06,
+      "loss": 0.3488,
+      "step": 757
+    },
+    {
+      "epoch": 1.6995515695067265,
+      "grad_norm": 0.3360415995121002,
+      "learning_rate": 4.713135025960652e-06,
+      "loss": 0.3485,
+      "step": 758
+    },
+    {
+      "epoch": 1.701793721973094,
+      "grad_norm": 0.24069413542747498,
+      "learning_rate": 4.700111026162587e-06,
+      "loss": 0.3356,
+      "step": 759
+    },
+    {
+      "epoch": 1.704035874439462,
+      "grad_norm": 0.24138867855072021,
+      "learning_rate": 4.687089068135484e-06,
+      "loss": 0.3582,
+      "step": 760
+    },
+    {
+      "epoch": 1.7062780269058297,
+      "grad_norm": 0.3291063904762268,
+      "learning_rate": 4.674069240538341e-06,
+      "loss": 0.3454,
+      "step": 761
+    },
+    {
+      "epoch": 1.7085201793721974,
+      "grad_norm": 0.2415570765733719,
+      "learning_rate": 4.661051632015655e-06,
+      "loss": 0.3407,
+      "step": 762
+    },
+    {
+      "epoch": 1.710762331838565,
+      "grad_norm": 0.2438618242740631,
+      "learning_rate": 4.648036331196804e-06,
+      "loss": 0.3528,
+      "step": 763
+    },
+    {
+      "epoch": 1.7130044843049328,
+      "grad_norm": 0.2383349984884262,
+      "learning_rate": 4.635023426695462e-06,
+      "loss": 0.3562,
+      "step": 764
+    },
+    {
+      "epoch": 1.7152466367713004,
+      "grad_norm": 0.2593696713447571,
+      "learning_rate": 4.6220130071089905e-06,
+      "loss": 0.3476,
+      "step": 765
+    },
+    {
+      "epoch": 1.717488789237668,
+      "grad_norm": 0.25911515951156616,
+      "learning_rate": 4.609005161017824e-06,
+      "loss": 0.3547,
+      "step": 766
+    },
+    {
+      "epoch": 1.7197309417040358,
+      "grad_norm": 0.28044837713241577,
+      "learning_rate": 4.59599997698488e-06,
+      "loss": 0.359,
+      "step": 767
+    },
+    {
+      "epoch": 1.7219730941704037,
+      "grad_norm": 0.23994892835617065,
+      "learning_rate": 4.582997543554956e-06,
+      "loss": 0.3552,
+      "step": 768
+    },
+    {
+      "epoch": 1.7242152466367713,
+      "grad_norm": 0.23979540169239044,
+      "learning_rate": 4.569997949254112e-06,
+      "loss": 0.3502,
+      "step": 769
+    },
+    {
+      "epoch": 1.726457399103139,
+      "grad_norm": 0.23544222116470337,
+      "learning_rate": 4.557001282589086e-06,
+      "loss": 0.3455,
+      "step": 770
+    },
+    {
+      "epoch": 1.7286995515695067,
+      "grad_norm": 0.22635391354560852,
+      "learning_rate": 4.544007632046682e-06,
+      "loss": 0.345,
+      "step": 771
+    },
+    {
+      "epoch": 1.7309417040358746,
+      "grad_norm": 0.2678516209125519,
+      "learning_rate": 4.531017086093171e-06,
+      "loss": 0.356,
+      "step": 772
+    },
+    {
+      "epoch": 1.7331838565022422,
+      "grad_norm": 0.23230041563510895,
+      "learning_rate": 4.518029733173677e-06,
+      "loss": 0.3583,
+      "step": 773
+    },
+    {
+      "epoch": 1.7354260089686098,
+      "grad_norm": 0.2455994188785553,
+      "learning_rate": 4.505045661711596e-06,
+      "loss": 0.3509,
+      "step": 774
+    },
+    {
+      "epoch": 1.7376681614349776,
+      "grad_norm": 0.24151350557804108,
+      "learning_rate": 4.4920649601079794e-06,
+      "loss": 0.3414,
+      "step": 775
+    },
+    {
+      "epoch": 1.7399103139013454,
+      "grad_norm": 0.28885453939437866,
+      "learning_rate": 4.479087716740928e-06,
+      "loss": 0.3475,
+      "step": 776
+    },
+    {
+      "epoch": 1.742152466367713,
+      "grad_norm": 0.27120155096054077,
+      "learning_rate": 4.466114019965006e-06,
+      "loss": 0.3543,
+      "step": 777
+    },
+    {
+      "epoch": 1.7443946188340806,
+      "grad_norm": 0.22735843062400818,
+      "learning_rate": 4.4531439581106295e-06,
+      "loss": 0.35,
+      "step": 778
+    },
+    {
+      "epoch": 1.7466367713004485,
+      "grad_norm": 0.22621771693229675,
+      "learning_rate": 4.4401776194834615e-06,
+      "loss": 0.3618,
+      "step": 779
+    },
+    {
+      "epoch": 1.7488789237668163,
+      "grad_norm": 0.24351833760738373,
+      "learning_rate": 4.427215092363821e-06,
+      "loss": 0.3422,
+      "step": 780
+    },
+    {
+      "epoch": 1.7511210762331837,
+      "grad_norm": 0.2636108100414276,
+      "learning_rate": 4.414256465006075e-06,
+      "loss": 0.3441,
+      "step": 781
+    },
+    {
+      "epoch": 1.7533632286995515,
+      "grad_norm": 0.25717464089393616,
+      "learning_rate": 4.401301825638039e-06,
+      "loss": 0.3561,
+      "step": 782
+    },
+    {
+      "epoch": 1.7556053811659194,
+      "grad_norm": 0.2520971894264221,
+      "learning_rate": 4.388351262460375e-06,
+      "loss": 0.3679,
+      "step": 783
+    },
+    {
+      "epoch": 1.757847533632287,
+      "grad_norm": 0.2254844307899475,
+      "learning_rate": 4.375404863645997e-06,
+      "loss": 0.3438,
+      "step": 784
+    },
+    {
+      "epoch": 1.7600896860986546,
+      "grad_norm": 0.2287048101425171,
+      "learning_rate": 4.3624627173394615e-06,
+      "loss": 0.3491,
+      "step": 785
+    },
+    {
+      "epoch": 1.7623318385650224,
+      "grad_norm": 0.2447231411933899,
+      "learning_rate": 4.349524911656373e-06,
+      "loss": 0.3483,
+      "step": 786
+    },
+    {
+      "epoch": 1.7645739910313902,
+      "grad_norm": 0.301116943359375,
+      "learning_rate": 4.3365915346827874e-06,
+      "loss": 0.3566,
+      "step": 787
+    },
+    {
+      "epoch": 1.7668161434977578,
+      "grad_norm": 0.24222904443740845,
+      "learning_rate": 4.323662674474603e-06,
+      "loss": 0.3468,
+      "step": 788
+    },
+    {
+      "epoch": 1.7690582959641254,
+      "grad_norm": 0.34491944313049316,
+      "learning_rate": 4.310738419056969e-06,
+      "loss": 0.3342,
+      "step": 789
+    },
+    {
+      "epoch": 1.7713004484304933,
+      "grad_norm": 0.23477140069007874,
+      "learning_rate": 4.297818856423679e-06,
+      "loss": 0.3338,
+      "step": 790
+    },
+    {
+      "epoch": 1.773542600896861,
+      "grad_norm": 0.23556838929653168,
+      "learning_rate": 4.284904074536582e-06,
+      "loss": 0.3442,
+      "step": 791
+    },
+    {
+      "epoch": 1.7757847533632287,
+      "grad_norm": 0.33268797397613525,
+      "learning_rate": 4.271994161324977e-06,
+      "loss": 0.3471,
+      "step": 792
+    },
+    {
+      "epoch": 1.7780269058295963,
+      "grad_norm": 0.22434638440608978,
+      "learning_rate": 4.259089204685009e-06,
+      "loss": 0.339,
+      "step": 793
+    },
+    {
+      "epoch": 1.7802690582959642,
+      "grad_norm": 0.28118422627449036,
+      "learning_rate": 4.2461892924790825e-06,
+      "loss": 0.3429,
+      "step": 794
+    },
+    {
+      "epoch": 1.782511210762332,
+      "grad_norm": 0.2324005663394928,
+      "learning_rate": 4.233294512535257e-06,
+      "loss": 0.3518,
+      "step": 795
+    },
+    {
+      "epoch": 1.7847533632286996,
+      "grad_norm": 0.2337992936372757,
+      "learning_rate": 4.220404952646645e-06,
+      "loss": 0.3483,
+      "step": 796
+    },
+    {
+      "epoch": 1.7869955156950672,
+      "grad_norm": 0.2364567518234253,
+      "learning_rate": 4.2075207005708215e-06,
+      "loss": 0.3439,
+      "step": 797
+    },
+    {
+      "epoch": 1.789237668161435,
+      "grad_norm": 0.27401337027549744,
+      "learning_rate": 4.194641844029227e-06,
+      "loss": 0.3553,
+      "step": 798
+    },
+    {
+      "epoch": 1.7914798206278026,
+      "grad_norm": 0.26304757595062256,
+      "learning_rate": 4.181768470706561e-06,
+      "loss": 0.3378,
+      "step": 799
+    },
+    {
+      "epoch": 1.7937219730941703,
+      "grad_norm": 0.22789053618907928,
+      "learning_rate": 4.168900668250192e-06,
+      "loss": 0.3563,
+      "step": 800
+    },
+    {
+      "epoch": 1.795964125560538,
+      "grad_norm": 0.2613663971424103,
+      "learning_rate": 4.156038524269563e-06,
+      "loss": 0.3476,
+      "step": 801
+    },
+    {
+      "epoch": 1.798206278026906,
+      "grad_norm": 0.24122577905654907,
+      "learning_rate": 4.143182126335594e-06,
+      "loss": 0.3534,
+      "step": 802
+    },
+    {
+      "epoch": 1.8004484304932735,
+      "grad_norm": 0.24867278337478638,
+      "learning_rate": 4.130331561980072e-06,
+      "loss": 0.3474,
+      "step": 803
+    },
+    {
+      "epoch": 1.8026905829596411,
+      "grad_norm": 0.22728633880615234,
+      "learning_rate": 4.117486918695077e-06,
+      "loss": 0.3533,
+      "step": 804
+    },
+    {
+      "epoch": 1.804932735426009,
+      "grad_norm": 0.28948530554771423,
+      "learning_rate": 4.104648283932376e-06,
+      "loss": 0.344,
+      "step": 805
+    },
+    {
+      "epoch": 1.8071748878923768,
+      "grad_norm": 0.23948384821414948,
+      "learning_rate": 4.091815745102818e-06,
+      "loss": 0.3486,
+      "step": 806
+    },
+    {
+      "epoch": 1.8094170403587444,
+      "grad_norm": 0.23691882193088531,
+      "learning_rate": 4.078989389575759e-06,
+      "loss": 0.3391,
+      "step": 807
+    },
+    {
+      "epoch": 1.811659192825112,
+      "grad_norm": 0.25575342774391174,
+      "learning_rate": 4.066169304678452e-06,
+      "loss": 0.3639,
+      "step": 808
+    },
+    {
+      "epoch": 1.8139013452914798,
+      "grad_norm": 0.26367029547691345,
+      "learning_rate": 4.053355577695454e-06,
+      "loss": 0.347,
+      "step": 809
+    },
+    {
+      "epoch": 1.8161434977578477,
+      "grad_norm": 0.34364739060401917,
+      "learning_rate": 4.040548295868039e-06,
+      "loss": 0.3439,
+      "step": 810
+    },
+    {
+      "epoch": 1.8183856502242153,
+      "grad_norm": 0.44863903522491455,
+      "learning_rate": 4.0277475463935975e-06,
+      "loss": 0.3551,
+      "step": 811
+    },
+    {
+      "epoch": 1.8206278026905829,
+      "grad_norm": 0.2285727858543396,
+      "learning_rate": 4.014953416425049e-06,
+      "loss": 0.3409,
+      "step": 812
+    },
+    {
+      "epoch": 1.8228699551569507,
+      "grad_norm": 0.2608487606048584,
+      "learning_rate": 4.002165993070237e-06,
+      "loss": 0.348,
+      "step": 813
+    },
+    {
+      "epoch": 1.8251121076233185,
+      "grad_norm": 0.24546630680561066,
+      "learning_rate": 3.9893853633913485e-06,
+      "loss": 0.3371,
+      "step": 814
+    },
+    {
+      "epoch": 1.827354260089686,
+      "grad_norm": 0.25345146656036377,
+      "learning_rate": 3.9766116144043195e-06,
+      "loss": 0.3475,
+      "step": 815
+    },
+    {
+      "epoch": 1.8295964125560538,
+      "grad_norm": 0.24488048255443573,
+      "learning_rate": 3.963844833078229e-06,
+      "loss": 0.3477,
+      "step": 816
+    },
+    {
+      "epoch": 1.8318385650224216,
+      "grad_norm": 0.24483051896095276,
+      "learning_rate": 3.951085106334729e-06,
+      "loss": 0.3345,
+      "step": 817
+    },
+    {
+      "epoch": 1.8340807174887892,
+      "grad_norm": 0.3017670810222626,
+      "learning_rate": 3.938332521047434e-06,
+      "loss": 0.3417,
+      "step": 818
+    },
+    {
+      "epoch": 1.8363228699551568,
+      "grad_norm": 0.2701224684715271,
+      "learning_rate": 3.925587164041335e-06,
+      "loss": 0.3484,
+      "step": 819
+    },
+    {
+      "epoch": 1.8385650224215246,
+      "grad_norm": 0.25930464267730713,
+      "learning_rate": 3.912849122092216e-06,
+      "loss": 0.3533,
+      "step": 820
+    },
+    {
+      "epoch": 1.8408071748878925,
+      "grad_norm": 0.23448678851127625,
+      "learning_rate": 3.900118481926053e-06,
+      "loss": 0.3607,
+      "step": 821
+    },
+    {
+      "epoch": 1.84304932735426,
+      "grad_norm": 0.24666792154312134,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.3471,
+      "step": 822
+    },
+    {
+      "epoch": 1.8452914798206277,
+      "grad_norm": 0.2466127723455429,
+      "learning_rate": 3.8746797535939404e-06,
+      "loss": 0.3497,
+      "step": 823
+    },
+    {
+      "epoch": 1.8475336322869955,
+      "grad_norm": 0.2509455382823944,
+      "learning_rate": 3.861971838625611e-06,
+      "loss": 0.341,
+      "step": 824
+    },
+    {
+      "epoch": 1.8497757847533634,
+      "grad_norm": 0.2314426451921463,
+      "learning_rate": 3.849271671834301e-06,
+      "loss": 0.3457,
+      "step": 825
+    },
+    {
+      "epoch": 1.852017937219731,
+      "grad_norm": 0.2355356514453888,
+      "learning_rate": 3.836579339688116e-06,
+      "loss": 0.3349,
+      "step": 826
+    },
+    {
+      "epoch": 1.8542600896860986,
+      "grad_norm": 0.2405450940132141,
+      "learning_rate": 3.823894928601822e-06,
+      "loss": 0.3471,
+      "step": 827
+    },
+    {
+      "epoch": 1.8565022421524664,
+      "grad_norm": 0.24157169461250305,
+      "learning_rate": 3.811218524936256e-06,
+      "loss": 0.3462,
+      "step": 828
+    },
+    {
+      "epoch": 1.8587443946188342,
+      "grad_norm": 0.2270277589559555,
+      "learning_rate": 3.7985502149977307e-06,
+      "loss": 0.3367,
+      "step": 829
+    },
+    {
+      "epoch": 1.8609865470852018,
+      "grad_norm": 0.2341500073671341,
+      "learning_rate": 3.7858900850374596e-06,
+      "loss": 0.3556,
+      "step": 830
+    },
+    {
+      "epoch": 1.8632286995515694,
+      "grad_norm": 0.22112977504730225,
+      "learning_rate": 3.7732382212509634e-06,
+      "loss": 0.3359,
+      "step": 831
+    },
+    {
+      "epoch": 1.8654708520179373,
+      "grad_norm": 0.24002204835414886,
+      "learning_rate": 3.760594709777482e-06,
+      "loss": 0.3584,
+      "step": 832
+    },
+    {
+      "epoch": 1.8677130044843049,
+      "grad_norm": 0.2318405956029892,
+      "learning_rate": 3.7479596366993854e-06,
+      "loss": 0.3461,
+      "step": 833
+    },
+    {
+      "epoch": 1.8699551569506725,
+      "grad_norm": 0.3338366448879242,
+      "learning_rate": 3.7353330880415963e-06,
+      "loss": 0.3531,
+      "step": 834
+    },
+    {
+      "epoch": 1.8721973094170403,
+      "grad_norm": 0.22657963633537292,
+      "learning_rate": 3.722715149771001e-06,
+      "loss": 0.352,
+      "step": 835
+    },
+    {
+      "epoch": 1.8744394618834082,
+      "grad_norm": 0.24927470088005066,
+      "learning_rate": 3.7101059077958546e-06,
+      "loss": 0.336,
+      "step": 836
+    },
+    {
+      "epoch": 1.8766816143497758,
+      "grad_norm": 0.24158969521522522,
+      "learning_rate": 3.6975054479652107e-06,
+      "loss": 0.3451,
+      "step": 837
+    },
+    {
+      "epoch": 1.8789237668161434,
+      "grad_norm": 0.23047369718551636,
+      "learning_rate": 3.6849138560683305e-06,
+      "loss": 0.3475,
+      "step": 838
+    },
+    {
+      "epoch": 1.8811659192825112,
+      "grad_norm": 0.2358091175556183,
+      "learning_rate": 3.672331217834098e-06,
+      "loss": 0.3544,
+      "step": 839
+    },
+    {
+      "epoch": 1.883408071748879,
+      "grad_norm": 0.23862481117248535,
+      "learning_rate": 3.6597576189304317e-06,
+      "loss": 0.3352,
+      "step": 840
+    },
+    {
+      "epoch": 1.8856502242152466,
+      "grad_norm": 0.25188905000686646,
+      "learning_rate": 3.6471931449637127e-06,
+      "loss": 0.3547,
+      "step": 841
+    },
+    {
+      "epoch": 1.8878923766816142,
+      "grad_norm": 0.2471955120563507,
+      "learning_rate": 3.634637881478196e-06,
+      "loss": 0.3409,
+      "step": 842
+    },
+    {
+      "epoch": 1.890134529147982,
+      "grad_norm": 0.23964948952198029,
+      "learning_rate": 3.62209191395542e-06,
+      "loss": 0.345,
+      "step": 843
+    },
+    {
+      "epoch": 1.89237668161435,
+      "grad_norm": 0.24352289736270905,
+      "learning_rate": 3.6095553278136408e-06,
+      "loss": 0.3502,
+      "step": 844
+    },
+    {
+      "epoch": 1.8946188340807175,
+      "grad_norm": 0.26733526587486267,
+      "learning_rate": 3.5970282084072406e-06,
+      "loss": 0.3456,
+      "step": 845
+    },
+    {
+      "epoch": 1.8968609865470851,
+      "grad_norm": 0.2238570749759674,
+      "learning_rate": 3.5845106410261417e-06,
+      "loss": 0.3452,
+      "step": 846
+    },
+    {
+      "epoch": 1.899103139013453,
+      "grad_norm": 0.24052174389362335,
+      "learning_rate": 3.5720027108952394e-06,
+      "loss": 0.3504,
+      "step": 847
+    },
+    {
+      "epoch": 1.9013452914798208,
+      "grad_norm": 0.23180270195007324,
+      "learning_rate": 3.5595045031738123e-06,
+      "loss": 0.3424,
+      "step": 848
+    },
+    {
+      "epoch": 1.9035874439461884,
+      "grad_norm": 0.24206875264644623,
+      "learning_rate": 3.5470161029549444e-06,
+      "loss": 0.3462,
+      "step": 849
+    },
+    {
+      "epoch": 1.905829596412556,
+      "grad_norm": 0.26114320755004883,
+      "learning_rate": 3.534537595264944e-06,
+      "loss": 0.3414,
+      "step": 850
+    },
+    {
+      "epoch": 1.9080717488789238,
+      "grad_norm": 0.23896445333957672,
+      "learning_rate": 3.522069065062771e-06,
+      "loss": 0.3482,
+      "step": 851
+    },
+    {
+      "epoch": 1.9103139013452914,
+      "grad_norm": 0.33738887310028076,
+      "learning_rate": 3.5096105972394516e-06,
+      "loss": 0.3424,
+      "step": 852
+    },
+    {
+      "epoch": 1.912556053811659,
+      "grad_norm": 0.21731296181678772,
+      "learning_rate": 3.497162276617503e-06,
+      "loss": 0.3415,
+      "step": 853
+    },
+    {
+      "epoch": 1.9147982062780269,
+      "grad_norm": 0.23101060092449188,
+      "learning_rate": 3.4847241879503574e-06,
+      "loss": 0.3513,
+      "step": 854
+    },
+    {
+      "epoch": 1.9170403587443947,
+      "grad_norm": 0.73750239610672,
+      "learning_rate": 3.472296415921783e-06,
+      "loss": 0.3459,
+      "step": 855
+    },
+    {
+      "epoch": 1.9192825112107623,
+      "grad_norm": 0.31533709168434143,
+      "learning_rate": 3.4598790451453066e-06,
+      "loss": 0.3502,
+      "step": 856
+    },
+    {
+      "epoch": 1.92152466367713,
+      "grad_norm": 0.23813597857952118,
+      "learning_rate": 3.4474721601636407e-06,
+      "loss": 0.3481,
+      "step": 857
+    },
+    {
+      "epoch": 1.9237668161434978,
+      "grad_norm": 0.2526616156101227,
+      "learning_rate": 3.435075845448105e-06,
+      "loss": 0.3469,
+      "step": 858
+    },
+    {
+      "epoch": 1.9260089686098656,
+      "grad_norm": 0.22211864590644836,
+      "learning_rate": 3.4226901853980552e-06,
+      "loss": 0.3547,
+      "step": 859
+    },
+    {
+      "epoch": 1.9282511210762332,
+      "grad_norm": 0.22354310750961304,
+      "learning_rate": 3.4103152643402985e-06,
+      "loss": 0.3506,
+      "step": 860
+    },
+    {
+      "epoch": 1.9304932735426008,
+      "grad_norm": 0.28579089045524597,
+      "learning_rate": 3.397951166528534e-06,
+      "loss": 0.3357,
+      "step": 861
+    },
+    {
+      "epoch": 1.9327354260089686,
+      "grad_norm": 0.23748673498630524,
+      "learning_rate": 3.3855979761427705e-06,
+      "loss": 0.3495,
+      "step": 862
+    },
+    {
+      "epoch": 1.9349775784753365,
+      "grad_norm": 0.22936514019966125,
+      "learning_rate": 3.373255777288748e-06,
+      "loss": 0.3524,
+      "step": 863
+    },
+    {
+      "epoch": 1.937219730941704,
+      "grad_norm": 0.2715314030647278,
+      "learning_rate": 3.360924653997378e-06,
+      "loss": 0.3427,
+      "step": 864
+    },
+    {
+      "epoch": 1.9394618834080717,
+      "grad_norm": 0.23739570379257202,
+      "learning_rate": 3.3486046902241663e-06,
+      "loss": 0.3531,
+      "step": 865
+    },
+    {
+      "epoch": 1.9417040358744395,
+      "grad_norm": 0.30479979515075684,
+      "learning_rate": 3.3362959698486307e-06,
+      "loss": 0.3464,
+      "step": 866
+    },
+    {
+      "epoch": 1.9439461883408071,
+      "grad_norm": 0.24834442138671875,
+      "learning_rate": 3.3239985766737493e-06,
+      "loss": 0.332,
+      "step": 867
+    },
+    {
+      "epoch": 1.9461883408071747,
+      "grad_norm": 0.2837355434894562,
+      "learning_rate": 3.3117125944253737e-06,
+      "loss": 0.3468,
+      "step": 868
+    },
+    {
+      "epoch": 1.9484304932735426,
+      "grad_norm": 0.22623519599437714,
+      "learning_rate": 3.2994381067516702e-06,
+      "loss": 0.3507,
+      "step": 869
+    },
+    {
+      "epoch": 1.9506726457399104,
+      "grad_norm": 0.22503317892551422,
+      "learning_rate": 3.287175197222537e-06,
+      "loss": 0.3446,
+      "step": 870
+    },
+    {
+      "epoch": 1.952914798206278,
+      "grad_norm": 0.24953000247478485,
+      "learning_rate": 3.2749239493290513e-06,
+      "loss": 0.3471,
+      "step": 871
+    },
+    {
+      "epoch": 1.9551569506726456,
+      "grad_norm": 0.2418135702610016,
+      "learning_rate": 3.262684446482892e-06,
+      "loss": 0.3423,
+      "step": 872
+    },
+    {
+      "epoch": 1.9573991031390134,
+      "grad_norm": 0.2755568325519562,
+      "learning_rate": 3.250456772015767e-06,
+      "loss": 0.3488,
+      "step": 873
+    },
+    {
+      "epoch": 1.9596412556053813,
+      "grad_norm": 0.24255864322185516,
+      "learning_rate": 3.2382410091788567e-06,
+      "loss": 0.3456,
+      "step": 874
+    },
+    {
+      "epoch": 1.9618834080717489,
+      "grad_norm": 0.25122323632240295,
+      "learning_rate": 3.2260372411422435e-06,
+      "loss": 0.3467,
+      "step": 875
+    },
+    {
+      "epoch": 1.9641255605381165,
+      "grad_norm": 0.2649831473827362,
+      "learning_rate": 3.2138455509943365e-06,
+      "loss": 0.3532,
+      "step": 876
+    },
+    {
+      "epoch": 1.9663677130044843,
+      "grad_norm": 0.8044440150260925,
+      "learning_rate": 3.201666021741322e-06,
+      "loss": 0.3469,
+      "step": 877
+    },
+    {
+      "epoch": 1.9686098654708521,
+      "grad_norm": 0.26725465059280396,
+      "learning_rate": 3.189498736306584e-06,
+      "loss": 0.34,
+      "step": 878
+    },
+    {
+      "epoch": 1.9708520179372198,
+      "grad_norm": 0.24081368744373322,
+      "learning_rate": 3.1773437775301515e-06,
+      "loss": 0.3363,
+      "step": 879
+    },
+    {
+      "epoch": 1.9730941704035874,
+      "grad_norm": 0.23061859607696533,
+      "learning_rate": 3.1652012281681187e-06,
+      "loss": 0.3423,
+      "step": 880
+    },
+    {
+      "epoch": 1.9753363228699552,
+      "grad_norm": 0.25481006503105164,
+      "learning_rate": 3.1530711708921008e-06,
+      "loss": 0.3506,
+      "step": 881
+    },
+    {
+      "epoch": 1.977578475336323,
+      "grad_norm": 0.2887466549873352,
+      "learning_rate": 3.140953688288658e-06,
+      "loss": 0.3522,
+      "step": 882
+    },
+    {
+      "epoch": 1.9798206278026906,
+      "grad_norm": 0.24619154632091522,
+      "learning_rate": 3.1288488628587343e-06,
+      "loss": 0.3464,
+      "step": 883
+    },
+    {
+      "epoch": 1.9820627802690582,
+      "grad_norm": 0.3197013735771179,
+      "learning_rate": 3.1167567770171026e-06,
+      "loss": 0.3473,
+      "step": 884
+    },
+    {
+      "epoch": 1.984304932735426,
+      "grad_norm": 0.2541103661060333,
+      "learning_rate": 3.1046775130917965e-06,
+      "loss": 0.3427,
+      "step": 885
+    },
+    {
+      "epoch": 1.9865470852017937,
+      "grad_norm": 0.2391168177127838,
+      "learning_rate": 3.0926111533235526e-06,
+      "loss": 0.3447,
+      "step": 886
+    },
+    {
+      "epoch": 1.9887892376681613,
+      "grad_norm": 0.24867595732212067,
+      "learning_rate": 3.0805577798652527e-06,
+      "loss": 0.3519,
+      "step": 887
+    },
+    {
+      "epoch": 1.9910313901345291,
+      "grad_norm": 0.25274911522865295,
+      "learning_rate": 3.0685174747813595e-06,
+      "loss": 0.3516,
+      "step": 888
+    },
+    {
+      "epoch": 1.993273542600897,
+      "grad_norm": 0.2440386265516281,
+      "learning_rate": 3.056490320047364e-06,
+      "loss": 0.3409,
+      "step": 889
+    },
+    {
+      "epoch": 1.9955156950672646,
+      "grad_norm": 0.24986571073532104,
+      "learning_rate": 3.044476397549221e-06,
+      "loss": 0.3525,
+      "step": 890
+    },
+    {
+      "epoch": 1.9977578475336322,
+      "grad_norm": 0.2795746922492981,
+      "learning_rate": 3.032475789082795e-06,
+      "loss": 0.3471,
+      "step": 891
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27022066712379456,
+      "learning_rate": 3.0204885763533065e-06,
+      "loss": 0.3106,
+      "step": 892
+    },
+    {
+      "epoch": 2.002242152466368,
+      "grad_norm": 0.29273587465286255,
+      "learning_rate": 3.0085148409747656e-06,
+      "loss": 0.3017,
+      "step": 893
+    },
+    {
+      "epoch": 2.004484304932735,
+      "grad_norm": 0.2777996063232422,
+      "learning_rate": 2.9965546644694287e-06,
+      "loss": 0.304,
+      "step": 894
+    },
+    {
+      "epoch": 2.006726457399103,
+      "grad_norm": 0.2917490303516388,
+      "learning_rate": 2.984608128267238e-06,
+      "loss": 0.304,
+      "step": 895
+    },
+    {
+      "epoch": 2.008968609865471,
+      "grad_norm": 0.3169640302658081,
+      "learning_rate": 2.97267531370526e-06,
+      "loss": 0.2961,
+      "step": 896
+    },
+    {
+      "epoch": 2.0112107623318387,
+      "grad_norm": 0.3361988663673401,
+      "learning_rate": 2.9607563020271446e-06,
+      "loss": 0.3049,
+      "step": 897
+    },
+    {
+      "epoch": 2.013452914798206,
+      "grad_norm": 0.32020026445388794,
+      "learning_rate": 2.948851174382565e-06,
+      "loss": 0.2865,
+      "step": 898
+    },
+    {
+      "epoch": 2.015695067264574,
+      "grad_norm": 0.2716546952724457,
+      "learning_rate": 2.936960011826666e-06,
+      "loss": 0.3049,
+      "step": 899
+    },
+    {
+      "epoch": 2.0179372197309418,
+      "grad_norm": 0.28449270129203796,
+      "learning_rate": 2.925082895319509e-06,
+      "loss": 0.3085,
+      "step": 900
+    },
+    {
+      "epoch": 2.0201793721973096,
+      "grad_norm": 0.2632957398891449,
+      "learning_rate": 2.913219905725526e-06,
+      "loss": 0.2944,
+      "step": 901
+    },
+    {
+      "epoch": 2.022421524663677,
+      "grad_norm": 0.2717359960079193,
+      "learning_rate": 2.9013711238129693e-06,
+      "loss": 0.3053,
+      "step": 902
+    },
+    {
+      "epoch": 2.024663677130045,
+      "grad_norm": 0.2436523735523224,
+      "learning_rate": 2.8895366302533566e-06,
+      "loss": 0.2901,
+      "step": 903
+    },
+    {
+      "epoch": 2.0269058295964126,
+      "grad_norm": 0.28306150436401367,
+      "learning_rate": 2.8777165056209256e-06,
+      "loss": 0.29,
+      "step": 904
+    },
+    {
+      "epoch": 2.0291479820627805,
+      "grad_norm": 0.270733505487442,
+      "learning_rate": 2.8659108303920823e-06,
+      "loss": 0.3188,
+      "step": 905
+    },
+    {
+      "epoch": 2.031390134529148,
+      "grad_norm": 0.26432088017463684,
+      "learning_rate": 2.8541196849448582e-06,
+      "loss": 0.2906,
+      "step": 906
+    },
+    {
+      "epoch": 2.0336322869955157,
+      "grad_norm": 0.2751922309398651,
+      "learning_rate": 2.842343149558355e-06,
+      "loss": 0.3204,
+      "step": 907
+    },
+    {
+      "epoch": 2.0358744394618835,
+      "grad_norm": 0.2714661657810211,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.302,
+      "step": 908
+    },
+    {
+      "epoch": 2.038116591928251,
+      "grad_norm": 0.2727791965007782,
+      "learning_rate": 2.8188342295860384e-06,
+      "loss": 0.2933,
+      "step": 909
+    },
+    {
+      "epoch": 2.0403587443946187,
+      "grad_norm": 0.24390915036201477,
+      "learning_rate": 2.8071020050588927e-06,
+      "loss": 0.3094,
+      "step": 910
+    },
+    {
+      "epoch": 2.0426008968609866,
+      "grad_norm": 0.2603088319301605,
+      "learning_rate": 2.7953847107087173e-06,
+      "loss": 0.2872,
+      "step": 911
+    },
+    {
+      "epoch": 2.0448430493273544,
+      "grad_norm": 0.24994972348213196,
+      "learning_rate": 2.7836824263118145e-06,
+      "loss": 0.2881,
+      "step": 912
+    },
+    {
+      "epoch": 2.0470852017937218,
+      "grad_norm": 0.3992326557636261,
+      "learning_rate": 2.7719952315422773e-06,
+      "loss": 0.3155,
+      "step": 913
+    },
+    {
+      "epoch": 2.0493273542600896,
+      "grad_norm": 0.2876724600791931,
+      "learning_rate": 2.760323205971476e-06,
+      "loss": 0.2962,
+      "step": 914
+    },
+    {
+      "epoch": 2.0515695067264574,
+      "grad_norm": 0.24341653287410736,
+      "learning_rate": 2.748666429067496e-06,
+      "loss": 0.3012,
+      "step": 915
+    },
+    {
+      "epoch": 2.0538116591928253,
+      "grad_norm": 0.25780004262924194,
+      "learning_rate": 2.7370249801946026e-06,
+      "loss": 0.3029,
+      "step": 916
+    },
+    {
+      "epoch": 2.0560538116591927,
+      "grad_norm": 0.2519890069961548,
+      "learning_rate": 2.725398938612702e-06,
+      "loss": 0.296,
+      "step": 917
+    },
+    {
+      "epoch": 2.0582959641255605,
+      "grad_norm": 0.24370181560516357,
+      "learning_rate": 2.7137883834768076e-06,
+      "loss": 0.3085,
+      "step": 918
+    },
+    {
+      "epoch": 2.0605381165919283,
+      "grad_norm": 0.23391525447368622,
+      "learning_rate": 2.702193393836485e-06,
+      "loss": 0.2866,
+      "step": 919
+    },
+    {
+      "epoch": 2.062780269058296,
+      "grad_norm": 0.2673562467098236,
+      "learning_rate": 2.690614048635329e-06,
+      "loss": 0.3038,
+      "step": 920
+    },
+    {
+      "epoch": 2.0650224215246635,
+      "grad_norm": 0.2637220323085785,
+      "learning_rate": 2.6790504267104165e-06,
+      "loss": 0.3055,
+      "step": 921
+    },
+    {
+      "epoch": 2.0672645739910314,
+      "grad_norm": 0.25950926542282104,
+      "learning_rate": 2.6675026067917808e-06,
+      "loss": 0.293,
+      "step": 922
+    },
+    {
+      "epoch": 2.069506726457399,
+      "grad_norm": 0.2526501715183258,
+      "learning_rate": 2.6559706675018578e-06,
+      "loss": 0.2907,
+      "step": 923
+    },
+    {
+      "epoch": 2.071748878923767,
+      "grad_norm": 0.3643045127391815,
+      "learning_rate": 2.6444546873549715e-06,
+      "loss": 0.2935,
+      "step": 924
+    },
+    {
+      "epoch": 2.0739910313901344,
+      "grad_norm": 0.27596309781074524,
+      "learning_rate": 2.632954744756784e-06,
+      "loss": 0.3032,
+      "step": 925
+    },
+    {
+      "epoch": 2.0762331838565022,
+      "grad_norm": 0.2736187279224396,
+      "learning_rate": 2.621470918003768e-06,
+      "loss": 0.2916,
+      "step": 926
+    },
+    {
+      "epoch": 2.07847533632287,
+      "grad_norm": 0.2611675560474396,
+      "learning_rate": 2.6100032852826705e-06,
+      "loss": 0.2741,
+      "step": 927
+    },
+    {
+      "epoch": 2.0807174887892375,
+      "grad_norm": 0.25763311982154846,
+      "learning_rate": 2.598551924669989e-06,
+      "loss": 0.3007,
+      "step": 928
+    },
+    {
+      "epoch": 2.0829596412556053,
+      "grad_norm": 0.245026096701622,
+      "learning_rate": 2.5871169141314277e-06,
+      "loss": 0.2911,
+      "step": 929
+    },
+    {
+      "epoch": 2.085201793721973,
+      "grad_norm": 0.26796016097068787,
+      "learning_rate": 2.5756983315213748e-06,
+      "loss": 0.2908,
+      "step": 930
+    },
+    {
+      "epoch": 2.087443946188341,
+      "grad_norm": 0.25604137778282166,
+      "learning_rate": 2.564296254582368e-06,
+      "loss": 0.2946,
+      "step": 931
+    },
+    {
+      "epoch": 2.0896860986547083,
+      "grad_norm": 0.25281235575675964,
+      "learning_rate": 2.5529107609445737e-06,
+      "loss": 0.2868,
+      "step": 932
+    },
+    {
+      "epoch": 2.091928251121076,
+      "grad_norm": 0.25859662890434265,
+      "learning_rate": 2.5415419281252416e-06,
+      "loss": 0.2968,
+      "step": 933
+    },
+    {
+      "epoch": 2.094170403587444,
+      "grad_norm": 0.2615140676498413,
+      "learning_rate": 2.5301898335281994e-06,
+      "loss": 0.292,
+      "step": 934
+    },
+    {
+      "epoch": 2.096412556053812,
+      "grad_norm": 0.26426762342453003,
+      "learning_rate": 2.5188545544433047e-06,
+      "loss": 0.2887,
+      "step": 935
+    },
+    {
+      "epoch": 2.098654708520179,
+      "grad_norm": 0.29007968306541443,
+      "learning_rate": 2.507536168045933e-06,
+      "loss": 0.3063,
+      "step": 936
+    },
+    {
+      "epoch": 2.100896860986547,
+      "grad_norm": 0.25335314869880676,
+      "learning_rate": 2.4962347513964448e-06,
+      "loss": 0.2992,
+      "step": 937
+    },
+    {
+      "epoch": 2.103139013452915,
+      "grad_norm": 0.25333380699157715,
+      "learning_rate": 2.4849503814396624e-06,
+      "loss": 0.2931,
+      "step": 938
+    },
+    {
+      "epoch": 2.1053811659192827,
+      "grad_norm": 0.2501998543739319,
+      "learning_rate": 2.473683135004354e-06,
+      "loss": 0.3045,
+      "step": 939
+    },
+    {
+      "epoch": 2.10762331838565,
+      "grad_norm": 0.26219046115875244,
+      "learning_rate": 2.4624330888026947e-06,
+      "loss": 0.2932,
+      "step": 940
+    },
+    {
+      "epoch": 2.109865470852018,
+      "grad_norm": 0.25986695289611816,
+      "learning_rate": 2.4512003194297578e-06,
+      "loss": 0.3027,
+      "step": 941
+    },
+    {
+      "epoch": 2.1121076233183858,
+      "grad_norm": 0.25550344586372375,
+      "learning_rate": 2.439984903362988e-06,
+      "loss": 0.3007,
+      "step": 942
+    },
+    {
+      "epoch": 2.1143497757847536,
+      "grad_norm": 0.2515295147895813,
+      "learning_rate": 2.428786916961681e-06,
+      "loss": 0.3013,
+      "step": 943
+    },
+    {
+      "epoch": 2.116591928251121,
+      "grad_norm": 0.2937853932380676,
+      "learning_rate": 2.4176064364664615e-06,
+      "loss": 0.3004,
+      "step": 944
+    },
+    {
+      "epoch": 2.118834080717489,
+      "grad_norm": 0.26110005378723145,
+      "learning_rate": 2.406443537998774e-06,
+      "loss": 0.3048,
+      "step": 945
+    },
+    {
+      "epoch": 2.1210762331838566,
+      "grad_norm": 0.25676199793815613,
+      "learning_rate": 2.3952982975603494e-06,
+      "loss": 0.3026,
+      "step": 946
+    },
+    {
+      "epoch": 2.123318385650224,
+      "grad_norm": 0.25228941440582275,
+      "learning_rate": 2.384170791032699e-06,
+      "loss": 0.2995,
+      "step": 947
+    },
+    {
+      "epoch": 2.125560538116592,
+      "grad_norm": 0.3262612521648407,
+      "learning_rate": 2.373061094176591e-06,
+      "loss": 0.3066,
+      "step": 948
+    },
+    {
+      "epoch": 2.1278026905829597,
+      "grad_norm": 0.26757150888442993,
+      "learning_rate": 2.3619692826315454e-06,
+      "loss": 0.2884,
+      "step": 949
+    },
+    {
+      "epoch": 2.1300448430493275,
+      "grad_norm": 0.24544380605220795,
+      "learning_rate": 2.3508954319153e-06,
+      "loss": 0.2971,
+      "step": 950
+    },
+    {
+      "epoch": 2.132286995515695,
+      "grad_norm": 0.25067922472953796,
+      "learning_rate": 2.339839617423318e-06,
+      "loss": 0.3046,
+      "step": 951
+    },
+    {
+      "epoch": 2.1345291479820627,
+      "grad_norm": 0.27066895365715027,
+      "learning_rate": 2.3288019144282586e-06,
+      "loss": 0.2938,
+      "step": 952
+    },
+    {
+      "epoch": 2.1367713004484306,
+      "grad_norm": 0.2884020507335663,
+      "learning_rate": 2.317782398079473e-06,
+      "loss": 0.2953,
+      "step": 953
+    },
+    {
+      "epoch": 2.1390134529147984,
+      "grad_norm": 0.29851874709129333,
+      "learning_rate": 2.306781143402485e-06,
+      "loss": 0.2862,
+      "step": 954
+    },
+    {
+      "epoch": 2.1412556053811658,
+      "grad_norm": 0.2597530484199524,
+      "learning_rate": 2.295798225298495e-06,
+      "loss": 0.3019,
+      "step": 955
+    },
+    {
+      "epoch": 2.1434977578475336,
+      "grad_norm": 0.278565376996994,
+      "learning_rate": 2.2848337185438513e-06,
+      "loss": 0.2986,
+      "step": 956
+    },
+    {
+      "epoch": 2.1457399103139014,
+      "grad_norm": 0.2631435990333557,
+      "learning_rate": 2.273887697789553e-06,
+      "loss": 0.2992,
+      "step": 957
+    },
+    {
+      "epoch": 2.1479820627802693,
+      "grad_norm": 0.26232632994651794,
+      "learning_rate": 2.2629602375607373e-06,
+      "loss": 0.2959,
+      "step": 958
+    },
+    {
+      "epoch": 2.1502242152466366,
+      "grad_norm": 0.2654859125614166,
+      "learning_rate": 2.25205141225618e-06,
+      "loss": 0.2887,
+      "step": 959
+    },
+    {
+      "epoch": 2.1524663677130045,
+      "grad_norm": 0.2495771199464798,
+      "learning_rate": 2.2411612961477704e-06,
+      "loss": 0.3013,
+      "step": 960
+    },
+    {
+      "epoch": 2.1547085201793723,
+      "grad_norm": 0.28251218795776367,
+      "learning_rate": 2.2302899633800296e-06,
+      "loss": 0.2969,
+      "step": 961
+    },
+    {
+      "epoch": 2.15695067264574,
+      "grad_norm": 0.2754749357700348,
+      "learning_rate": 2.219437487969588e-06,
+      "loss": 0.2959,
+      "step": 962
+    },
+    {
+      "epoch": 2.1591928251121075,
+      "grad_norm": 0.25928547978401184,
+      "learning_rate": 2.2086039438046863e-06,
+      "loss": 0.2972,
+      "step": 963
+    },
+    {
+      "epoch": 2.1614349775784754,
+      "grad_norm": 0.24755533039569855,
+      "learning_rate": 2.1977894046446747e-06,
+      "loss": 0.305,
+      "step": 964
+    },
+    {
+      "epoch": 2.163677130044843,
+      "grad_norm": 0.25068092346191406,
+      "learning_rate": 2.186993944119512e-06,
+      "loss": 0.2947,
+      "step": 965
+    },
+    {
+      "epoch": 2.1659192825112106,
+      "grad_norm": 0.2611483037471771,
+      "learning_rate": 2.1762176357292582e-06,
+      "loss": 0.2903,
+      "step": 966
+    },
+    {
+      "epoch": 2.1681614349775784,
+      "grad_norm": 0.27977579832077026,
+      "learning_rate": 2.1654605528435774e-06,
+      "loss": 0.2845,
+      "step": 967
+    },
+    {
+      "epoch": 2.1704035874439462,
+      "grad_norm": 0.25573110580444336,
+      "learning_rate": 2.1547227687012403e-06,
+      "loss": 0.3125,
+      "step": 968
+    },
+    {
+      "epoch": 2.172645739910314,
+      "grad_norm": 0.27416783571243286,
+      "learning_rate": 2.144004356409621e-06,
+      "loss": 0.3014,
+      "step": 969
+    },
+    {
+      "epoch": 2.1748878923766815,
+      "grad_norm": 0.2737100124359131,
+      "learning_rate": 2.1333053889442033e-06,
+      "loss": 0.3029,
+      "step": 970
+    },
+    {
+      "epoch": 2.1771300448430493,
+      "grad_norm": 0.276579886674881,
+      "learning_rate": 2.122625939148086e-06,
+      "loss": 0.3004,
+      "step": 971
+    },
+    {
+      "epoch": 2.179372197309417,
+      "grad_norm": 0.2575071156024933,
+      "learning_rate": 2.1119660797314775e-06,
+      "loss": 0.3105,
+      "step": 972
+    },
+    {
+      "epoch": 2.181614349775785,
+      "grad_norm": 0.2634425163269043,
+      "learning_rate": 2.101325883271208e-06,
+      "loss": 0.2931,
+      "step": 973
+    },
+    {
+      "epoch": 2.1838565022421523,
+      "grad_norm": 0.26021769642829895,
+      "learning_rate": 2.0907054222102367e-06,
+      "loss": 0.2953,
+      "step": 974
+    },
+    {
+      "epoch": 2.18609865470852,
+      "grad_norm": 0.30425941944122314,
+      "learning_rate": 2.0801047688571517e-06,
+      "loss": 0.2959,
+      "step": 975
+    },
+    {
+      "epoch": 2.188340807174888,
+      "grad_norm": 0.2848120629787445,
+      "learning_rate": 2.0695239953856894e-06,
+      "loss": 0.2892,
+      "step": 976
+    },
+    {
+      "epoch": 2.1905829596412554,
+      "grad_norm": 0.25678330659866333,
+      "learning_rate": 2.0589631738342235e-06,
+      "loss": 0.2973,
+      "step": 977
+    },
+    {
+      "epoch": 2.192825112107623,
+      "grad_norm": 0.2615320086479187,
+      "learning_rate": 2.048422376105299e-06,
+      "loss": 0.2955,
+      "step": 978
+    },
+    {
+      "epoch": 2.195067264573991,
+      "grad_norm": 0.2417897880077362,
+      "learning_rate": 2.0379016739651225e-06,
+      "loss": 0.3205,
+      "step": 979
+    },
+    {
+      "epoch": 2.197309417040359,
+      "grad_norm": 0.3421541750431061,
+      "learning_rate": 2.027401139043084e-06,
+      "loss": 0.3016,
+      "step": 980
+    },
+    {
+      "epoch": 2.1995515695067263,
+      "grad_norm": 0.2576618492603302,
+      "learning_rate": 2.0169208428312647e-06,
+      "loss": 0.2896,
+      "step": 981
+    },
+    {
+      "epoch": 2.201793721973094,
+      "grad_norm": 0.2997666895389557,
+      "learning_rate": 2.0064608566839584e-06,
+      "loss": 0.2949,
+      "step": 982
+    },
+    {
+      "epoch": 2.204035874439462,
+      "grad_norm": 0.276877224445343,
+      "learning_rate": 1.996021251817167e-06,
+      "loss": 0.2934,
+      "step": 983
+    },
+    {
+      "epoch": 2.2062780269058297,
+      "grad_norm": 0.2796614170074463,
+      "learning_rate": 1.985602099308142e-06,
+      "loss": 0.2921,
+      "step": 984
+    },
+    {
+      "epoch": 2.208520179372197,
+      "grad_norm": 0.2383633852005005,
+      "learning_rate": 1.975203470094875e-06,
+      "loss": 0.297,
+      "step": 985
+    },
+    {
+      "epoch": 2.210762331838565,
+      "grad_norm": 0.26453033089637756,
+      "learning_rate": 1.964825434975639e-06,
+      "loss": 0.2904,
+      "step": 986
+    },
+    {
+      "epoch": 2.213004484304933,
+      "grad_norm": 0.2566130459308624,
+      "learning_rate": 1.9544680646084774e-06,
+      "loss": 0.2965,
+      "step": 987
+    },
+    {
+      "epoch": 2.2152466367713006,
+      "grad_norm": 0.2635822594165802,
+      "learning_rate": 1.944131429510754e-06,
+      "loss": 0.3011,
+      "step": 988
+    },
+    {
+      "epoch": 2.217488789237668,
+      "grad_norm": 0.2578944265842438,
+      "learning_rate": 1.93381560005865e-06,
+      "loss": 0.3023,
+      "step": 989
+    },
+    {
+      "epoch": 2.219730941704036,
+      "grad_norm": 0.23659422993659973,
+      "learning_rate": 1.923520646486695e-06,
+      "loss": 0.2951,
+      "step": 990
+    },
+    {
+      "epoch": 2.2219730941704037,
+      "grad_norm": 0.24604566395282745,
+      "learning_rate": 1.9132466388872862e-06,
+      "loss": 0.2993,
+      "step": 991
+    },
+    {
+      "epoch": 2.2242152466367715,
+      "grad_norm": 0.2721068859100342,
+      "learning_rate": 1.9029936472102162e-06,
+      "loss": 0.2968,
+      "step": 992
+    },
+    {
+      "epoch": 2.226457399103139,
+      "grad_norm": 0.23521971702575684,
+      "learning_rate": 1.8927617412621825e-06,
+      "loss": 0.2924,
+      "step": 993
+    },
+    {
+      "epoch": 2.2286995515695067,
+      "grad_norm": 0.29433685541152954,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.2988,
+      "step": 994
+    },
+    {
+      "epoch": 2.2309417040358746,
+      "grad_norm": 0.24003687500953674,
+      "learning_rate": 1.8723614650617721e-06,
+      "loss": 0.2949,
+      "step": 995
+    },
+    {
+      "epoch": 2.233183856502242,
+      "grad_norm": 0.23856155574321747,
+      "learning_rate": 1.8621932337031056e-06,
+      "loss": 0.2935,
+      "step": 996
+    },
+    {
+      "epoch": 2.2354260089686098,
+      "grad_norm": 0.2722351551055908,
+      "learning_rate": 1.8520463658599457e-06,
+      "loss": 0.3081,
+      "step": 997
+    },
+    {
+      "epoch": 2.2376681614349776,
+      "grad_norm": 0.2292395830154419,
+      "learning_rate": 1.8419209306164653e-06,
+      "loss": 0.2978,
+      "step": 998
+    },
+    {
+      "epoch": 2.2399103139013454,
+      "grad_norm": 0.24535441398620605,
+      "learning_rate": 1.831816996910908e-06,
+      "loss": 0.2993,
+      "step": 999
+    },
+    {
+      "epoch": 2.242152466367713,
+      "grad_norm": 0.2564886510372162,
+      "learning_rate": 1.821734633535127e-06,
+      "loss": 0.3137,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2443946188340806,
+      "grad_norm": 0.2586636245250702,
+      "learning_rate": 1.8116739091341174e-06,
+      "loss": 0.2795,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2466367713004485,
+      "grad_norm": 0.2690829634666443,
+      "learning_rate": 1.8016348922055448e-06,
+      "loss": 0.3023,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2488789237668163,
+      "grad_norm": 0.3920111358165741,
+      "learning_rate": 1.791617651099281e-06,
+      "loss": 0.3013,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2511210762331837,
+      "grad_norm": 0.2917453348636627,
+      "learning_rate": 1.7816222540169437e-06,
+      "loss": 0.3002,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2533632286995515,
+      "grad_norm": 0.2729063928127289,
+      "learning_rate": 1.7716487690114209e-06,
+      "loss": 0.3058,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2556053811659194,
+      "grad_norm": 0.26944971084594727,
+      "learning_rate": 1.7616972639864166e-06,
+      "loss": 0.2967,
+      "step": 1006
+    },
+    {
+      "epoch": 2.257847533632287,
+      "grad_norm": 0.25574350357055664,
+      "learning_rate": 1.7517678066959864e-06,
+      "loss": 0.3144,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2600896860986546,
+      "grad_norm": 0.26157188415527344,
+      "learning_rate": 1.7418604647440735e-06,
+      "loss": 0.301,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2623318385650224,
+      "grad_norm": 0.2627256214618683,
+      "learning_rate": 1.7319753055840555e-06,
+      "loss": 0.2974,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2645739910313902,
+      "grad_norm": 0.2455906867980957,
+      "learning_rate": 1.7221123965182712e-06,
+      "loss": 0.2987,
+      "step": 1010
+    },
+    {
+      "epoch": 2.266816143497758,
+      "grad_norm": 0.258920282125473,
+      "learning_rate": 1.712271804697581e-06,
+      "loss": 0.3055,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2690582959641254,
+      "grad_norm": 0.24115526676177979,
+      "learning_rate": 1.702453597120895e-06,
+      "loss": 0.2946,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2713004484304933,
+      "grad_norm": 0.2788311243057251,
+      "learning_rate": 1.6926578406347217e-06,
+      "loss": 0.297,
+      "step": 1013
+    },
+    {
+      "epoch": 2.273542600896861,
+      "grad_norm": 0.27021676301956177,
+      "learning_rate": 1.6828846019327128e-06,
+      "loss": 0.2956,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2757847533632285,
+      "grad_norm": 0.29145047068595886,
+      "learning_rate": 1.6731339475552143e-06,
+      "loss": 0.2916,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2780269058295963,
+      "grad_norm": 0.27871477603912354,
+      "learning_rate": 1.6634059438888034e-06,
+      "loss": 0.3024,
+      "step": 1016
+    },
+    {
+      "epoch": 2.280269058295964,
+      "grad_norm": 0.2968088388442993,
+      "learning_rate": 1.6537006571658432e-06,
+      "loss": 0.3026,
+      "step": 1017
+    },
+    {
+      "epoch": 2.282511210762332,
+      "grad_norm": 0.2674208879470825,
+      "learning_rate": 1.6440181534640277e-06,
+      "loss": 0.3185,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2847533632286994,
+      "grad_norm": 0.2779238224029541,
+      "learning_rate": 1.634358498705943e-06,
+      "loss": 0.3123,
+      "step": 1019
+    },
+    {
+      "epoch": 2.286995515695067,
+      "grad_norm": 0.264187216758728,
+      "learning_rate": 1.6247217586585974e-06,
+      "loss": 0.2949,
+      "step": 1020
+    },
+    {
+      "epoch": 2.289237668161435,
+      "grad_norm": 0.2609671354293823,
+      "learning_rate": 1.6151079989329964e-06,
+      "loss": 0.304,
+      "step": 1021
+    },
+    {
+      "epoch": 2.291479820627803,
+      "grad_norm": 0.2503441274166107,
+      "learning_rate": 1.6055172849836826e-06,
+      "loss": 0.3141,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2937219730941703,
+      "grad_norm": 0.270553320646286,
+      "learning_rate": 1.5959496821082905e-06,
+      "loss": 0.2993,
+      "step": 1023
+    },
+    {
+      "epoch": 2.295964125560538,
+      "grad_norm": 0.24788302183151245,
+      "learning_rate": 1.5864052554471066e-06,
+      "loss": 0.3108,
+      "step": 1024
+    },
+    {
+      "epoch": 2.298206278026906,
+      "grad_norm": 0.25432872772216797,
+      "learning_rate": 1.5768840699826266e-06,
+      "loss": 0.2959,
+      "step": 1025
+    },
+    {
+      "epoch": 2.3004484304932733,
+      "grad_norm": 0.25345277786254883,
+      "learning_rate": 1.567386190539107e-06,
+      "loss": 0.316,
+      "step": 1026
+    },
+    {
+      "epoch": 2.302690582959641,
+      "grad_norm": 0.25282248854637146,
+      "learning_rate": 1.5579116817821272e-06,
+      "loss": 0.2997,
+      "step": 1027
+    },
+    {
+      "epoch": 2.304932735426009,
+      "grad_norm": 0.27014443278312683,
+      "learning_rate": 1.548460608218148e-06,
+      "loss": 0.3028,
+      "step": 1028
+    },
+    {
+      "epoch": 2.307174887892377,
+      "grad_norm": 0.23407581448554993,
+      "learning_rate": 1.5390330341940813e-06,
+      "loss": 0.3064,
+      "step": 1029
+    },
+    {
+      "epoch": 2.3094170403587446,
+      "grad_norm": 0.26750046014785767,
+      "learning_rate": 1.5296290238968303e-06,
+      "loss": 0.3016,
+      "step": 1030
+    },
+    {
+      "epoch": 2.311659192825112,
+      "grad_norm": 0.23493976891040802,
+      "learning_rate": 1.5202486413528805e-06,
+      "loss": 0.302,
+      "step": 1031
+    },
+    {
+      "epoch": 2.31390134529148,
+      "grad_norm": 0.25511831045150757,
+      "learning_rate": 1.5108919504278408e-06,
+      "loss": 0.3008,
+      "step": 1032
+    },
+    {
+      "epoch": 2.3161434977578477,
+      "grad_norm": 0.2481481283903122,
+      "learning_rate": 1.5015590148260205e-06,
+      "loss": 0.2939,
+      "step": 1033
+    },
+    {
+      "epoch": 2.318385650224215,
+      "grad_norm": 0.2439967542886734,
+      "learning_rate": 1.4922498980899907e-06,
+      "loss": 0.2967,
+      "step": 1034
+    },
+    {
+      "epoch": 2.320627802690583,
+      "grad_norm": 0.262937068939209,
+      "learning_rate": 1.4829646636001583e-06,
+      "loss": 0.2963,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3228699551569507,
+      "grad_norm": 0.2474174052476883,
+      "learning_rate": 1.4737033745743245e-06,
+      "loss": 0.3073,
+      "step": 1036
+    },
+    {
+      "epoch": 2.3251121076233185,
+      "grad_norm": 0.30798351764678955,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.3167,
+      "step": 1037
+    },
+    {
+      "epoch": 2.327354260089686,
+      "grad_norm": 0.2542681097984314,
+      "learning_rate": 1.4552528849702852e-06,
+      "loss": 0.297,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3295964125560538,
+      "grad_norm": 0.2480105310678482,
+      "learning_rate": 1.446063810010815e-06,
+      "loss": 0.2998,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3318385650224216,
+      "grad_norm": 0.25097596645355225,
+      "learning_rate": 1.4368989317519615e-06,
+      "loss": 0.2907,
+      "step": 1040
+    },
+    {
+      "epoch": 2.3340807174887894,
+      "grad_norm": 0.6831316947937012,
+      "learning_rate": 1.42775831259209e-06,
+      "loss": 0.2961,
+      "step": 1041
+    },
+    {
+      "epoch": 2.336322869955157,
+      "grad_norm": 0.2686893939971924,
+      "learning_rate": 1.4186420147644053e-06,
+      "loss": 0.3046,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3385650224215246,
+      "grad_norm": 0.2736101448535919,
+      "learning_rate": 1.4095501003365163e-06,
+      "loss": 0.2887,
+      "step": 1043
+    },
+    {
+      "epoch": 2.3408071748878925,
+      "grad_norm": 0.24822431802749634,
+      "learning_rate": 1.4004826312100218e-06,
+      "loss": 0.3087,
+      "step": 1044
+    },
+    {
+      "epoch": 2.34304932735426,
+      "grad_norm": 0.25574350357055664,
+      "learning_rate": 1.3914396691200842e-06,
+      "loss": 0.3111,
+      "step": 1045
+    },
+    {
+      "epoch": 2.3452914798206277,
+      "grad_norm": 0.2476636916399002,
+      "learning_rate": 1.3824212756350196e-06,
+      "loss": 0.2918,
+      "step": 1046
+    },
+    {
+      "epoch": 2.3475336322869955,
+      "grad_norm": 0.25612545013427734,
+      "learning_rate": 1.3734275121558582e-06,
+      "loss": 0.3009,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3497757847533634,
+      "grad_norm": 0.2628982663154602,
+      "learning_rate": 1.36445843991595e-06,
+      "loss": 0.3054,
+      "step": 1048
+    },
+    {
+      "epoch": 2.352017937219731,
+      "grad_norm": 0.2909297049045563,
+      "learning_rate": 1.3555141199805305e-06,
+      "loss": 0.3043,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3542600896860986,
+      "grad_norm": 0.26528793573379517,
+      "learning_rate": 1.3465946132463125e-06,
+      "loss": 0.3109,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3565022421524664,
+      "grad_norm": 0.27219653129577637,
+      "learning_rate": 1.337699980441069e-06,
+      "loss": 0.2916,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3587443946188342,
+      "grad_norm": 0.2665805220603943,
+      "learning_rate": 1.3288302821232247e-06,
+      "loss": 0.3023,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3609865470852016,
+      "grad_norm": 0.26604628562927246,
+      "learning_rate": 1.3199855786814353e-06,
+      "loss": 0.3097,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3632286995515694,
+      "grad_norm": 0.2503181993961334,
+      "learning_rate": 1.3111659303341824e-06,
+      "loss": 0.2926,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3654708520179373,
+      "grad_norm": 0.27894577383995056,
+      "learning_rate": 1.30237139712936e-06,
+      "loss": 0.307,
+      "step": 1055
+    },
+    {
+      "epoch": 2.367713004484305,
+      "grad_norm": 0.2794185280799866,
+      "learning_rate": 1.2936020389438753e-06,
+      "loss": 0.3028,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3699551569506725,
+      "grad_norm": 0.25606441497802734,
+      "learning_rate": 1.2848579154832203e-06,
+      "loss": 0.2948,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3721973094170403,
+      "grad_norm": 0.27696335315704346,
+      "learning_rate": 1.2761390862810907e-06,
+      "loss": 0.3073,
+      "step": 1058
+    },
+    {
+      "epoch": 2.374439461883408,
+      "grad_norm": 0.32369160652160645,
+      "learning_rate": 1.2674456106989618e-06,
+      "loss": 0.2959,
+      "step": 1059
+    },
+    {
+      "epoch": 2.376681614349776,
+      "grad_norm": 0.28073742985725403,
+      "learning_rate": 1.258777547925692e-06,
+      "loss": 0.2945,
+      "step": 1060
+    },
+    {
+      "epoch": 2.3789237668161434,
+      "grad_norm": 0.2556833326816559,
+      "learning_rate": 1.2501349569771165e-06,
+      "loss": 0.2986,
+      "step": 1061
+    },
+    {
+      "epoch": 2.381165919282511,
+      "grad_norm": 0.280185729265213,
+      "learning_rate": 1.2415178966956531e-06,
+      "loss": 0.2978,
+      "step": 1062
+    },
+    {
+      "epoch": 2.383408071748879,
+      "grad_norm": 0.2573022246360779,
+      "learning_rate": 1.2329264257498907e-06,
+      "loss": 0.3016,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3856502242152464,
+      "grad_norm": 0.2627410292625427,
+      "learning_rate": 1.2243606026341976e-06,
+      "loss": 0.2914,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3878923766816142,
+      "grad_norm": 0.27249768376350403,
+      "learning_rate": 1.2158204856683176e-06,
+      "loss": 0.3065,
+      "step": 1065
+    },
+    {
+      "epoch": 2.390134529147982,
+      "grad_norm": 0.2667676508426666,
+      "learning_rate": 1.2073061329969843e-06,
+      "loss": 0.3168,
+      "step": 1066
+    },
+    {
+      "epoch": 2.39237668161435,
+      "grad_norm": 0.32781699299812317,
+      "learning_rate": 1.1988176025895059e-06,
+      "loss": 0.3007,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3946188340807173,
+      "grad_norm": 0.2705301344394684,
+      "learning_rate": 1.190354952239391e-06,
+      "loss": 0.2996,
+      "step": 1068
+    },
+    {
+      "epoch": 2.396860986547085,
+      "grad_norm": 0.2654258906841278,
+      "learning_rate": 1.1819182395639406e-06,
+      "loss": 0.301,
+      "step": 1069
+    },
+    {
+      "epoch": 2.399103139013453,
+      "grad_norm": 0.2752063572406769,
+      "learning_rate": 1.1735075220038634e-06,
+      "loss": 0.2904,
+      "step": 1070
+    },
+    {
+      "epoch": 2.401345291479821,
+      "grad_norm": 0.26312655210494995,
+      "learning_rate": 1.1651228568228813e-06,
+      "loss": 0.2999,
+      "step": 1071
+    },
+    {
+      "epoch": 2.403587443946188,
+      "grad_norm": 0.2838899493217468,
+      "learning_rate": 1.1567643011073393e-06,
+      "loss": 0.3091,
+      "step": 1072
+    },
+    {
+      "epoch": 2.405829596412556,
+      "grad_norm": 0.30166107416152954,
+      "learning_rate": 1.1484319117658232e-06,
+      "loss": 0.3037,
+      "step": 1073
+    },
+    {
+      "epoch": 2.408071748878924,
+      "grad_norm": 0.25249266624450684,
+      "learning_rate": 1.1401257455287612e-06,
+      "loss": 0.3017,
+      "step": 1074
+    },
+    {
+      "epoch": 2.4103139013452917,
+      "grad_norm": 0.26185014843940735,
+      "learning_rate": 1.131845858948047e-06,
+      "loss": 0.3073,
+      "step": 1075
+    },
+    {
+      "epoch": 2.412556053811659,
+      "grad_norm": 0.28532183170318604,
+      "learning_rate": 1.1235923083966488e-06,
+      "loss": 0.3001,
+      "step": 1076
+    },
+    {
+      "epoch": 2.414798206278027,
+      "grad_norm": 0.2872476279735565,
+      "learning_rate": 1.1153651500682306e-06,
+      "loss": 0.2999,
+      "step": 1077
+    },
+    {
+      "epoch": 2.4170403587443947,
+      "grad_norm": 0.23972049355506897,
+      "learning_rate": 1.107164439976764e-06,
+      "loss": 0.3041,
+      "step": 1078
+    },
+    {
+      "epoch": 2.4192825112107625,
+      "grad_norm": 0.2524144649505615,
+      "learning_rate": 1.0989902339561554e-06,
+      "loss": 0.3091,
+      "step": 1079
+    },
+    {
+      "epoch": 2.42152466367713,
+      "grad_norm": 0.2719744145870209,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.2883,
+      "step": 1080
+    },
+    {
+      "epoch": 2.4237668161434978,
+      "grad_norm": 0.27270424365997314,
+      "learning_rate": 1.0827215565604775e-06,
+      "loss": 0.2964,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4260089686098656,
+      "grad_norm": 0.25880736112594604,
+      "learning_rate": 1.0746271959494453e-06,
+      "loss": 0.302,
+      "step": 1082
+    },
+    {
+      "epoch": 2.428251121076233,
+      "grad_norm": 0.2571720778942108,
+      "learning_rate": 1.0665595609365892e-06,
+      "loss": 0.3013,
+      "step": 1083
+    },
+    {
+      "epoch": 2.430493273542601,
+      "grad_norm": 0.25926458835601807,
+      "learning_rate": 1.0585187064497753e-06,
+      "loss": 0.3145,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4327354260089686,
+      "grad_norm": 0.25466495752334595,
+      "learning_rate": 1.0505046872345454e-06,
+      "loss": 0.2899,
+      "step": 1085
+    },
+    {
+      "epoch": 2.4349775784753365,
+      "grad_norm": 0.3124285042285919,
+      "learning_rate": 1.04251755785373e-06,
+      "loss": 0.3029,
+      "step": 1086
+    },
+    {
+      "epoch": 2.437219730941704,
+      "grad_norm": 0.2768646478652954,
+      "learning_rate": 1.0345573726870845e-06,
+      "loss": 0.3082,
+      "step": 1087
+    },
+    {
+      "epoch": 2.4394618834080717,
+      "grad_norm": 0.27555832266807556,
+      "learning_rate": 1.0266241859309134e-06,
+      "loss": 0.3024,
+      "step": 1088
+    },
+    {
+      "epoch": 2.4417040358744395,
+      "grad_norm": 0.27056509256362915,
+      "learning_rate": 1.018718051597713e-06,
+      "loss": 0.299,
+      "step": 1089
+    },
+    {
+      "epoch": 2.4439461883408073,
+      "grad_norm": 0.24575532972812653,
+      "learning_rate": 1.0108390235157828e-06,
+      "loss": 0.2827,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4461883408071747,
+      "grad_norm": 0.2585075795650482,
+      "learning_rate": 1.0029871553288828e-06,
+      "loss": 0.2945,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4484304932735426,
+      "grad_norm": 0.2977042496204376,
+      "learning_rate": 9.951625004958498e-07,
+      "loss": 0.3053,
+      "step": 1092
+    },
+    {
+      "epoch": 2.4506726457399104,
+      "grad_norm": 0.2591491639614105,
+      "learning_rate": 9.873651122902472e-07,
+      "loss": 0.2977,
+      "step": 1093
+    },
+    {
+      "epoch": 2.452914798206278,
+      "grad_norm": 0.2563302516937256,
+      "learning_rate": 9.795950437999852e-07,
+      "loss": 0.3057,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4551569506726456,
+      "grad_norm": 0.26039084792137146,
+      "learning_rate": 9.718523479269814e-07,
+      "loss": 0.2946,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4573991031390134,
+      "grad_norm": 0.25933098793029785,
+      "learning_rate": 9.64137077386782e-07,
+      "loss": 0.2981,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4596412556053813,
+      "grad_norm": 0.2648906409740448,
+      "learning_rate": 9.564492847082118e-07,
+      "loss": 0.3073,
+      "step": 1097
+    },
+    {
+      "epoch": 2.461883408071749,
+      "grad_norm": 0.24137040972709656,
+      "learning_rate": 9.487890222330137e-07,
+      "loss": 0.3194,
+      "step": 1098
+    },
+    {
+      "epoch": 2.4641255605381165,
+      "grad_norm": 0.25679558515548706,
+      "learning_rate": 9.411563421154979e-07,
+      "loss": 0.3037,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4663677130044843,
+      "grad_norm": 0.25994250178337097,
+      "learning_rate": 9.335512963221732e-07,
+      "loss": 0.2864,
+      "step": 1100
+    },
+    {
+      "epoch": 2.468609865470852,
+      "grad_norm": 0.2594643235206604,
+      "learning_rate": 9.259739366314125e-07,
+      "loss": 0.3061,
+      "step": 1101
+    },
+    {
+      "epoch": 2.4708520179372195,
+      "grad_norm": 0.24762535095214844,
+      "learning_rate": 9.184243146330829e-07,
+      "loss": 0.3058,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4730941704035874,
+      "grad_norm": 0.2531186640262604,
+      "learning_rate": 9.109024817282042e-07,
+      "loss": 0.3097,
+      "step": 1103
+    },
+    {
+      "epoch": 2.475336322869955,
+      "grad_norm": 0.25676417350769043,
+      "learning_rate": 9.03408489128596e-07,
+      "loss": 0.2963,
+      "step": 1104
+    },
+    {
+      "epoch": 2.477578475336323,
+      "grad_norm": 0.4618217945098877,
+      "learning_rate": 8.959423878565288e-07,
+      "loss": 0.314,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4798206278026904,
+      "grad_norm": 0.25637221336364746,
+      "learning_rate": 8.885042287443785e-07,
+      "loss": 0.3062,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4820627802690582,
+      "grad_norm": 0.2761717438697815,
+      "learning_rate": 8.810940624342784e-07,
+      "loss": 0.2967,
+      "step": 1107
+    },
+    {
+      "epoch": 2.484304932735426,
+      "grad_norm": 0.5241203308105469,
+      "learning_rate": 8.737119393777737e-07,
+      "loss": 0.2959,
+      "step": 1108
+    },
+    {
+      "epoch": 2.486547085201794,
+      "grad_norm": 0.2604654133319855,
+      "learning_rate": 8.663579098354808e-07,
+      "loss": 0.2976,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4887892376681613,
+      "grad_norm": 0.2650544047355652,
+      "learning_rate": 8.590320238767425e-07,
+      "loss": 0.296,
+      "step": 1110
+    },
+    {
+      "epoch": 2.491031390134529,
+      "grad_norm": 0.2814645767211914,
+      "learning_rate": 8.517343313792881e-07,
+      "loss": 0.2976,
+      "step": 1111
+    },
+    {
+      "epoch": 2.493273542600897,
+      "grad_norm": 0.2577002942562103,
+      "learning_rate": 8.444648820288959e-07,
+      "loss": 0.3116,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4955156950672643,
+      "grad_norm": 0.27652761340141296,
+      "learning_rate": 8.372237253190513e-07,
+      "loss": 0.2904,
+      "step": 1113
+    },
+    {
+      "epoch": 2.497757847533632,
+      "grad_norm": 0.3046053647994995,
+      "learning_rate": 8.30010910550611e-07,
+      "loss": 0.2946,
+      "step": 1114
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.26762911677360535,
+      "learning_rate": 8.228264868314673e-07,
+      "loss": 0.2933,
+      "step": 1115
+    },
+    {
+      "epoch": 2.502242152466368,
+      "grad_norm": 0.3415077328681946,
+      "learning_rate": 8.156705030762196e-07,
+      "loss": 0.3155,
+      "step": 1116
+    },
+    {
+      "epoch": 2.5044843049327357,
+      "grad_norm": 0.2586733400821686,
+      "learning_rate": 8.08543008005826e-07,
+      "loss": 0.2905,
+      "step": 1117
+    },
+    {
+      "epoch": 2.506726457399103,
+      "grad_norm": 0.255113810300827,
+      "learning_rate": 8.014440501472909e-07,
+      "loss": 0.2897,
+      "step": 1118
+    },
+    {
+      "epoch": 2.508968609865471,
+      "grad_norm": 0.26700153946876526,
+      "learning_rate": 7.943736778333211e-07,
+      "loss": 0.2936,
+      "step": 1119
+    },
+    {
+      "epoch": 2.5112107623318387,
+      "grad_norm": 0.24239665269851685,
+      "learning_rate": 7.873319392020013e-07,
+      "loss": 0.3057,
+      "step": 1120
+    },
+    {
+      "epoch": 2.513452914798206,
+      "grad_norm": 0.2529595196247101,
+      "learning_rate": 7.803188821964652e-07,
+      "loss": 0.3,
+      "step": 1121
+    },
+    {
+      "epoch": 2.515695067264574,
+      "grad_norm": 0.2428136169910431,
+      "learning_rate": 7.733345545645726e-07,
+      "loss": 0.3062,
+      "step": 1122
+    },
+    {
+      "epoch": 2.5179372197309418,
+      "grad_norm": 0.26277613639831543,
+      "learning_rate": 7.663790038585794e-07,
+      "loss": 0.2978,
+      "step": 1123
+    },
+    {
+      "epoch": 2.520179372197309,
+      "grad_norm": 0.6022162437438965,
+      "learning_rate": 7.594522774348162e-07,
+      "loss": 0.2944,
+      "step": 1124
+    },
+    {
+      "epoch": 2.522421524663677,
+      "grad_norm": 0.25606080889701843,
+      "learning_rate": 7.525544224533643e-07,
+      "loss": 0.2938,
+      "step": 1125
+    },
+    {
+      "epoch": 2.524663677130045,
+      "grad_norm": 0.2809850573539734,
+      "learning_rate": 7.456854858777418e-07,
+      "loss": 0.2937,
+      "step": 1126
+    },
+    {
+      "epoch": 2.5269058295964126,
+      "grad_norm": 0.3383976221084595,
+      "learning_rate": 7.388455144745693e-07,
+      "loss": 0.3022,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5291479820627805,
+      "grad_norm": 0.24517032504081726,
+      "learning_rate": 7.320345548132679e-07,
+      "loss": 0.3012,
+      "step": 1128
+    },
+    {
+      "epoch": 2.531390134529148,
+      "grad_norm": 0.5830869674682617,
+      "learning_rate": 7.252526532657306e-07,
+      "loss": 0.2981,
+      "step": 1129
+    },
+    {
+      "epoch": 2.5336322869955157,
+      "grad_norm": 0.2513468861579895,
+      "learning_rate": 7.184998560060114e-07,
+      "loss": 0.3041,
+      "step": 1130
+    },
+    {
+      "epoch": 2.5358744394618835,
+      "grad_norm": 0.28253576159477234,
+      "learning_rate": 7.117762090100083e-07,
+      "loss": 0.307,
+      "step": 1131
+    },
+    {
+      "epoch": 2.538116591928251,
+      "grad_norm": 0.2715183198451996,
+      "learning_rate": 7.050817580551567e-07,
+      "loss": 0.3172,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5403587443946187,
+      "grad_norm": 0.2764574885368347,
+      "learning_rate": 6.984165487201083e-07,
+      "loss": 0.307,
+      "step": 1133
+    },
+    {
+      "epoch": 2.5426008968609866,
+      "grad_norm": 0.3001376688480377,
+      "learning_rate": 6.917806263844268e-07,
+      "loss": 0.303,
+      "step": 1134
+    },
+    {
+      "epoch": 2.5448430493273544,
+      "grad_norm": 0.29435887932777405,
+      "learning_rate": 6.851740362282788e-07,
+      "loss": 0.3056,
+      "step": 1135
+    },
+    {
+      "epoch": 2.547085201793722,
+      "grad_norm": 0.2583148777484894,
+      "learning_rate": 6.785968232321233e-07,
+      "loss": 0.2954,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5493273542600896,
+      "grad_norm": 0.25455039739608765,
+      "learning_rate": 6.720490321764073e-07,
+      "loss": 0.3054,
+      "step": 1137
+    },
+    {
+      "epoch": 2.5515695067264574,
+      "grad_norm": 0.2625746428966522,
+      "learning_rate": 6.655307076412637e-07,
+      "loss": 0.2849,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5538116591928253,
+      "grad_norm": 0.2622738778591156,
+      "learning_rate": 6.590418940062021e-07,
+      "loss": 0.299,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5560538116591927,
+      "grad_norm": 0.2581905424594879,
+      "learning_rate": 6.525826354498088e-07,
+      "loss": 0.3144,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5582959641255605,
+      "grad_norm": 0.2643068730831146,
+      "learning_rate": 6.461529759494489e-07,
+      "loss": 0.3078,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5605381165919283,
+      "grad_norm": 0.28449636697769165,
+      "learning_rate": 6.397529592809615e-07,
+      "loss": 0.3083,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5627802690582957,
+      "grad_norm": 0.2818273603916168,
+      "learning_rate": 6.333826290183709e-07,
+      "loss": 0.3077,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5650224215246635,
+      "grad_norm": 0.25915291905403137,
+      "learning_rate": 6.270420285335748e-07,
+      "loss": 0.3009,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5672645739910314,
+      "grad_norm": 0.2597827613353729,
+      "learning_rate": 6.20731200996067e-07,
+      "loss": 0.291,
+      "step": 1145
+    },
+    {
+      "epoch": 2.569506726457399,
+      "grad_norm": 0.2516624629497528,
+      "learning_rate": 6.14450189372628e-07,
+      "loss": 0.3046,
+      "step": 1146
+    },
+    {
+      "epoch": 2.571748878923767,
+      "grad_norm": 0.25022467970848083,
+      "learning_rate": 6.081990364270418e-07,
+      "loss": 0.2962,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5739910313901344,
+      "grad_norm": 0.2646273672580719,
+      "learning_rate": 6.019777847198006e-07,
+      "loss": 0.3053,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5762331838565022,
+      "grad_norm": 0.2732788324356079,
+      "learning_rate": 5.957864766078186e-07,
+      "loss": 0.289,
+      "step": 1149
+    },
+    {
+      "epoch": 2.57847533632287,
+      "grad_norm": 0.2976854145526886,
+      "learning_rate": 5.896251542441395e-07,
+      "loss": 0.3114,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5807174887892375,
+      "grad_norm": 0.25121501088142395,
+      "learning_rate": 5.834938595776518e-07,
+      "loss": 0.3099,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5829596412556053,
+      "grad_norm": 0.245786651968956,
+      "learning_rate": 5.773926343528008e-07,
+      "loss": 0.2893,
+      "step": 1152
+    },
+    {
+      "epoch": 2.585201793721973,
+      "grad_norm": 0.25858086347579956,
+      "learning_rate": 5.71321520109312e-07,
+      "loss": 0.3059,
+      "step": 1153
+    },
+    {
+      "epoch": 2.587443946188341,
+      "grad_norm": 0.24982774257659912,
+      "learning_rate": 5.652805581818943e-07,
+      "loss": 0.2912,
+      "step": 1154
+    },
+    {
+      "epoch": 2.589686098654709,
+      "grad_norm": 0.2730509340763092,
+      "learning_rate": 5.592697896999749e-07,
+      "loss": 0.3014,
+      "step": 1155
+    },
+    {
+      "epoch": 2.591928251121076,
+      "grad_norm": 0.24873457849025726,
+      "learning_rate": 5.532892555874059e-07,
+      "loss": 0.2915,
+      "step": 1156
+    },
+    {
+      "epoch": 2.594170403587444,
+      "grad_norm": 0.3189297616481781,
+      "learning_rate": 5.473389965621934e-07,
+      "loss": 0.3022,
+      "step": 1157
+    },
+    {
+      "epoch": 2.596412556053812,
+      "grad_norm": 0.26500681042671204,
+      "learning_rate": 5.414190531362162e-07,
+      "loss": 0.2874,
+      "step": 1158
+    },
+    {
+      "epoch": 2.598654708520179,
+      "grad_norm": 0.2654920220375061,
+      "learning_rate": 5.355294656149545e-07,
+      "loss": 0.3048,
+      "step": 1159
+    },
+    {
+      "epoch": 2.600896860986547,
+      "grad_norm": 0.27304619550704956,
+      "learning_rate": 5.296702740972104e-07,
+      "loss": 0.2997,
+      "step": 1160
+    },
+    {
+      "epoch": 2.603139013452915,
+      "grad_norm": 0.25512734055519104,
+      "learning_rate": 5.23841518474838e-07,
+      "loss": 0.3047,
+      "step": 1161
+    },
+    {
+      "epoch": 2.6053811659192823,
+      "grad_norm": 0.28056371212005615,
+      "learning_rate": 5.180432384324691e-07,
+      "loss": 0.2957,
+      "step": 1162
+    },
+    {
+      "epoch": 2.60762331838565,
+      "grad_norm": 0.26484590768814087,
+      "learning_rate": 5.122754734472496e-07,
+      "loss": 0.3161,
+      "step": 1163
+    },
+    {
+      "epoch": 2.609865470852018,
+      "grad_norm": 0.254194974899292,
+      "learning_rate": 5.065382627885585e-07,
+      "loss": 0.3074,
+      "step": 1164
+    },
+    {
+      "epoch": 2.6121076233183858,
+      "grad_norm": 0.2750493884086609,
+      "learning_rate": 5.008316455177558e-07,
+      "loss": 0.3007,
+      "step": 1165
+    },
+    {
+      "epoch": 2.6143497757847536,
+      "grad_norm": 0.3747583329677582,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.2943,
+      "step": 1166
+    },
+    {
+      "epoch": 2.616591928251121,
+      "grad_norm": 0.2645847797393799,
+      "learning_rate": 4.895103463435108e-07,
+      "loss": 0.3159,
+      "step": 1167
+    },
+    {
+      "epoch": 2.618834080717489,
+      "grad_norm": 0.25880464911460876,
+      "learning_rate": 4.838957415202605e-07,
+      "loss": 0.2958,
+      "step": 1168
+    },
+    {
+      "epoch": 2.6210762331838566,
+      "grad_norm": 0.24956227838993073,
+      "learning_rate": 4.783118842447587e-07,
+      "loss": 0.291,
+      "step": 1169
+    },
+    {
+      "epoch": 2.623318385650224,
+      "grad_norm": 0.2469998449087143,
+      "learning_rate": 4.727588125342669e-07,
+      "loss": 0.3085,
+      "step": 1170
+    },
+    {
+      "epoch": 2.625560538116592,
+      "grad_norm": 0.27497708797454834,
+      "learning_rate": 4.67236564196446e-07,
+      "loss": 0.3124,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6278026905829597,
+      "grad_norm": 0.23264223337173462,
+      "learning_rate": 4.617451768290976e-07,
+      "loss": 0.3073,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6300448430493275,
+      "grad_norm": 0.2628634572029114,
+      "learning_rate": 4.5628468781991007e-07,
+      "loss": 0.2902,
+      "step": 1173
+    },
+    {
+      "epoch": 2.6322869955156953,
+      "grad_norm": 0.2661301791667938,
+      "learning_rate": 4.508551343462014e-07,
+      "loss": 0.3089,
+      "step": 1174
+    },
+    {
+      "epoch": 2.6345291479820627,
+      "grad_norm": 0.2762359082698822,
+      "learning_rate": 4.454565533746674e-07,
+      "loss": 0.2933,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6367713004484306,
+      "grad_norm": 0.26297131180763245,
+      "learning_rate": 4.400889816611315e-07,
+      "loss": 0.299,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6390134529147984,
+      "grad_norm": 0.3095591068267822,
+      "learning_rate": 4.347524557502919e-07,
+      "loss": 0.3008,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6412556053811658,
+      "grad_norm": 0.2516915798187256,
+      "learning_rate": 4.29447011975474e-07,
+      "loss": 0.3081,
+      "step": 1178
+    },
+    {
+      "epoch": 2.6434977578475336,
+      "grad_norm": 0.24786975979804993,
+      "learning_rate": 4.2417268645838117e-07,
+      "loss": 0.3005,
+      "step": 1179
+    },
+    {
+      "epoch": 2.6457399103139014,
+      "grad_norm": 0.3402668535709381,
+      "learning_rate": 4.1892951510885516e-07,
+      "loss": 0.3048,
+      "step": 1180
+    },
+    {
+      "epoch": 2.647982062780269,
+      "grad_norm": 0.25502660870552063,
+      "learning_rate": 4.1371753362462076e-07,
+      "loss": 0.3036,
+      "step": 1181
+    },
+    {
+      "epoch": 2.6502242152466366,
+      "grad_norm": 0.25100746750831604,
+      "learning_rate": 4.0853677749105426e-07,
+      "loss": 0.3033,
+      "step": 1182
+    },
+    {
+      "epoch": 2.6524663677130045,
+      "grad_norm": 0.24611695110797882,
+      "learning_rate": 4.0338728198093424e-07,
+      "loss": 0.3076,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6547085201793723,
+      "grad_norm": 0.2921901345252991,
+      "learning_rate": 3.9826908215420344e-07,
+      "loss": 0.3206,
+      "step": 1184
+    },
+    {
+      "epoch": 2.65695067264574,
+      "grad_norm": 0.25988176465034485,
+      "learning_rate": 3.931822128577301e-07,
+      "loss": 0.2994,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6591928251121075,
+      "grad_norm": 0.2670322358608246,
+      "learning_rate": 3.8812670872507454e-07,
+      "loss": 0.3013,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6614349775784754,
+      "grad_norm": 0.7288517951965332,
+      "learning_rate": 3.831026041762431e-07,
+      "loss": 0.3103,
+      "step": 1187
+    },
+    {
+      "epoch": 2.663677130044843,
+      "grad_norm": 0.2742021977901459,
+      "learning_rate": 3.781099334174676e-07,
+      "loss": 0.2993,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6659192825112106,
+      "grad_norm": 0.2673848271369934,
+      "learning_rate": 3.7314873044095977e-07,
+      "loss": 0.304,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6681614349775784,
+      "grad_norm": 0.26937294006347656,
+      "learning_rate": 3.6821902902469066e-07,
+      "loss": 0.311,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6704035874439462,
+      "grad_norm": 0.27371418476104736,
+      "learning_rate": 3.633208627321483e-07,
+      "loss": 0.3019,
+      "step": 1191
+    },
+    {
+      "epoch": 2.672645739910314,
+      "grad_norm": 0.2817760705947876,
+      "learning_rate": 3.5845426491212366e-07,
+      "loss": 0.3099,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6748878923766815,
+      "grad_norm": 0.2583814561367035,
+      "learning_rate": 3.5361926869847184e-07,
+      "loss": 0.3034,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6771300448430493,
+      "grad_norm": 0.2837837040424347,
+      "learning_rate": 3.4881590700989175e-07,
+      "loss": 0.2896,
+      "step": 1194
+    },
+    {
+      "epoch": 2.679372197309417,
+      "grad_norm": 0.2697250247001648,
+      "learning_rate": 3.440442125497001e-07,
+      "loss": 0.3025,
+      "step": 1195
+    },
+    {
+      "epoch": 2.681614349775785,
+      "grad_norm": 0.26696670055389404,
+      "learning_rate": 3.3930421780561386e-07,
+      "loss": 0.2963,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6838565022421523,
+      "grad_norm": 0.29351720213890076,
+      "learning_rate": 3.345959550495187e-07,
+      "loss": 0.2951,
+      "step": 1197
+    },
+    {
+      "epoch": 2.68609865470852,
+      "grad_norm": 0.2652941346168518,
+      "learning_rate": 3.299194563372604e-07,
+      "loss": 0.298,
+      "step": 1198
+    },
+    {
+      "epoch": 2.688340807174888,
+      "grad_norm": 0.26794031262397766,
+      "learning_rate": 3.2527475350841944e-07,
+      "loss": 0.2962,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6905829596412554,
+      "grad_norm": 0.6158033609390259,
+      "learning_rate": 3.2066187818609705e-07,
+      "loss": 0.3006,
+      "step": 1200
+    },
+    {
+      "epoch": 2.692825112107623,
+      "grad_norm": 0.31049904227256775,
+      "learning_rate": 3.1608086177669915e-07,
+      "loss": 0.306,
+      "step": 1201
+    },
+    {
+      "epoch": 2.695067264573991,
+      "grad_norm": 0.2741469144821167,
+      "learning_rate": 3.1153173546972395e-07,
+      "loss": 0.3006,
+      "step": 1202
+    },
+    {
+      "epoch": 2.697309417040359,
+      "grad_norm": 0.27378717064857483,
+      "learning_rate": 3.07014530237546e-07,
+      "loss": 0.2911,
+      "step": 1203
+    },
+    {
+      "epoch": 2.6995515695067267,
+      "grad_norm": 0.29777902364730835,
+      "learning_rate": 3.025292768352095e-07,
+      "loss": 0.3108,
+      "step": 1204
+    },
+    {
+      "epoch": 2.701793721973094,
+      "grad_norm": 0.25688156485557556,
+      "learning_rate": 2.980760058002163e-07,
+      "loss": 0.3004,
+      "step": 1205
+    },
+    {
+      "epoch": 2.704035874439462,
+      "grad_norm": 0.2554680109024048,
+      "learning_rate": 2.9365474745231935e-07,
+      "loss": 0.2963,
+      "step": 1206
+    },
+    {
+      "epoch": 2.7062780269058297,
+      "grad_norm": 0.24676163494586945,
+      "learning_rate": 2.892655318933152e-07,
+      "loss": 0.296,
+      "step": 1207
+    },
+    {
+      "epoch": 2.708520179372197,
+      "grad_norm": 0.2643212080001831,
+      "learning_rate": 2.8490838900683826e-07,
+      "loss": 0.3054,
+      "step": 1208
+    },
+    {
+      "epoch": 2.710762331838565,
+      "grad_norm": 0.3614406883716583,
+      "learning_rate": 2.8058334845816214e-07,
+      "loss": 0.307,
+      "step": 1209
+    },
+    {
+      "epoch": 2.713004484304933,
+      "grad_norm": 0.29051950573921204,
+      "learning_rate": 2.7629043969399193e-07,
+      "loss": 0.3121,
+      "step": 1210
+    },
+    {
+      "epoch": 2.7152466367713,
+      "grad_norm": 0.2684550881385803,
+      "learning_rate": 2.7202969194226436e-07,
+      "loss": 0.2947,
+      "step": 1211
+    },
+    {
+      "epoch": 2.717488789237668,
+      "grad_norm": 0.32691189646720886,
+      "learning_rate": 2.67801134211953e-07,
+      "loss": 0.2982,
+      "step": 1212
+    },
+    {
+      "epoch": 2.719730941704036,
+      "grad_norm": 0.2630385160446167,
+      "learning_rate": 2.636047952928689e-07,
+      "loss": 0.3068,
+      "step": 1213
+    },
+    {
+      "epoch": 2.7219730941704037,
+      "grad_norm": 0.4160473942756653,
+      "learning_rate": 2.594407037554586e-07,
+      "loss": 0.2979,
+      "step": 1214
+    },
+    {
+      "epoch": 2.7242152466367715,
+      "grad_norm": 0.27399975061416626,
+      "learning_rate": 2.55308887950621e-07,
+      "loss": 0.2915,
+      "step": 1215
+    },
+    {
+      "epoch": 2.726457399103139,
+      "grad_norm": 0.2523746192455292,
+      "learning_rate": 2.5120937600950524e-07,
+      "loss": 0.304,
+      "step": 1216
+    },
+    {
+      "epoch": 2.7286995515695067,
+      "grad_norm": 0.27583104372024536,
+      "learning_rate": 2.471421958433218e-07,
+      "loss": 0.2963,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7309417040358746,
+      "grad_norm": 0.2688456177711487,
+      "learning_rate": 2.431073751431529e-07,
+      "loss": 0.3093,
+      "step": 1218
+    },
+    {
+      "epoch": 2.733183856502242,
+      "grad_norm": 0.24724715948104858,
+      "learning_rate": 2.3910494137976526e-07,
+      "loss": 0.3027,
+      "step": 1219
+    },
+    {
+      "epoch": 2.7354260089686098,
+      "grad_norm": 0.26289764046669006,
+      "learning_rate": 2.3513492180342102e-07,
+      "loss": 0.2996,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7376681614349776,
+      "grad_norm": 0.28143543004989624,
+      "learning_rate": 2.3119734344369095e-07,
+      "loss": 0.3005,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7399103139013454,
+      "grad_norm": 0.2666257917881012,
+      "learning_rate": 2.2729223310927473e-07,
+      "loss": 0.3035,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7421524663677133,
+      "grad_norm": 0.28415727615356445,
+      "learning_rate": 2.2341961738781558e-07,
+      "loss": 0.3057,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7443946188340806,
+      "grad_norm": 0.2540895342826843,
+      "learning_rate": 2.195795226457176e-07,
+      "loss": 0.2961,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7466367713004485,
+      "grad_norm": 0.2929799556732178,
+      "learning_rate": 2.1577197502797133e-07,
+      "loss": 0.2976,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7488789237668163,
+      "grad_norm": 0.2539980709552765,
+      "learning_rate": 2.1199700045797077e-07,
+      "loss": 0.3071,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7511210762331837,
+      "grad_norm": 0.2813381552696228,
+      "learning_rate": 2.0825462463733903e-07,
+      "loss": 0.3044,
+      "step": 1227
+    },
+    {
+      "epoch": 2.7533632286995515,
+      "grad_norm": 0.27382275462150574,
+      "learning_rate": 2.0454487304575277e-07,
+      "loss": 0.3025,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7556053811659194,
+      "grad_norm": 0.3286834955215454,
+      "learning_rate": 2.0086777094077137e-07,
+      "loss": 0.3015,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7578475336322867,
+      "grad_norm": 0.28388431668281555,
+      "learning_rate": 1.9722334335766092e-07,
+      "loss": 0.2993,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7600896860986546,
+      "grad_norm": 0.36208170652389526,
+      "learning_rate": 1.9361161510922432e-07,
+      "loss": 0.2956,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7623318385650224,
+      "grad_norm": 0.2674930989742279,
+      "learning_rate": 1.900326107856354e-07,
+      "loss": 0.2983,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7645739910313902,
+      "grad_norm": 0.4405798614025116,
+      "learning_rate": 1.8648635475427112e-07,
+      "loss": 0.2953,
+      "step": 1233
+    },
+    {
+      "epoch": 2.766816143497758,
+      "grad_norm": 0.2496076226234436,
+      "learning_rate": 1.829728711595391e-07,
+      "loss": 0.3015,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7690582959641254,
+      "grad_norm": 0.33233916759490967,
+      "learning_rate": 1.7949218392272427e-07,
+      "loss": 0.2915,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7713004484304933,
+      "grad_norm": 0.27011215686798096,
+      "learning_rate": 1.7604431674181633e-07,
+      "loss": 0.3111,
+      "step": 1236
+    },
+    {
+      "epoch": 2.773542600896861,
+      "grad_norm": 0.25498443841934204,
+      "learning_rate": 1.7262929309135312e-07,
+      "loss": 0.3147,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7757847533632285,
+      "grad_norm": 0.27067622542381287,
+      "learning_rate": 1.6924713622225975e-07,
+      "loss": 0.3122,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7780269058295963,
+      "grad_norm": 0.29973870515823364,
+      "learning_rate": 1.6589786916169083e-07,
+      "loss": 0.301,
+      "step": 1239
+    },
+    {
+      "epoch": 2.780269058295964,
+      "grad_norm": 0.25834912061691284,
+      "learning_rate": 1.6258151471287397e-07,
+      "loss": 0.296,
+      "step": 1240
+    },
+    {
+      "epoch": 2.782511210762332,
+      "grad_norm": 0.29060620069503784,
+      "learning_rate": 1.592980954549511e-07,
+      "loss": 0.3,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7847533632287,
+      "grad_norm": 0.265201598405838,
+      "learning_rate": 1.5604763374283073e-07,
+      "loss": 0.2961,
+      "step": 1242
+    },
+    {
+      "epoch": 2.786995515695067,
+      "grad_norm": 0.281155526638031,
+      "learning_rate": 1.5283015170702964e-07,
+      "loss": 0.288,
+      "step": 1243
+    },
+    {
+      "epoch": 2.789237668161435,
+      "grad_norm": 0.26797351241111755,
+      "learning_rate": 1.4964567125352714e-07,
+      "loss": 0.2974,
+      "step": 1244
+    },
+    {
+      "epoch": 2.791479820627803,
+      "grad_norm": 0.265396386384964,
+      "learning_rate": 1.4649421406361164e-07,
+      "loss": 0.3019,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7937219730941703,
+      "grad_norm": 0.26489442586898804,
+      "learning_rate": 1.4337580159373864e-07,
+      "loss": 0.3077,
+      "step": 1246
+    },
+    {
+      "epoch": 2.795964125560538,
+      "grad_norm": 0.28167134523391724,
+      "learning_rate": 1.4029045507537696e-07,
+      "loss": 0.2941,
+      "step": 1247
+    },
+    {
+      "epoch": 2.798206278026906,
+      "grad_norm": 0.2779931128025055,
+      "learning_rate": 1.3723819551487217e-07,
+      "loss": 0.2995,
+      "step": 1248
+    },
+    {
+      "epoch": 2.8004484304932733,
+      "grad_norm": 0.24855895340442657,
+      "learning_rate": 1.3421904369329664e-07,
+      "loss": 0.3079,
+      "step": 1249
+    },
+    {
+      "epoch": 2.802690582959641,
+      "grad_norm": 0.3202498257160187,
+      "learning_rate": 1.3123302016631477e-07,
+      "loss": 0.3039,
+      "step": 1250
+    },
+    {
+      "epoch": 2.804932735426009,
+      "grad_norm": 0.25492188334465027,
+      "learning_rate": 1.2828014526403587e-07,
+      "loss": 0.3085,
+      "step": 1251
+    },
+    {
+      "epoch": 2.807174887892377,
+      "grad_norm": 0.2611996531486511,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.3027,
+      "step": 1252
+    },
+    {
+      "epoch": 2.8094170403587446,
+      "grad_norm": 0.24753612279891968,
+      "learning_rate": 1.224739215254478e-07,
+      "loss": 0.3083,
+      "step": 1253
+    },
+    {
+      "epoch": 2.811659192825112,
+      "grad_norm": 0.252591073513031,
+      "learning_rate": 1.196206122203647e-07,
+      "loss": 0.2887,
+      "step": 1254
+    },
+    {
+      "epoch": 2.81390134529148,
+      "grad_norm": 0.2581121027469635,
+      "learning_rate": 1.1680053060216912e-07,
+      "loss": 0.3039,
+      "step": 1255
+    },
+    {
+      "epoch": 2.8161434977578477,
+      "grad_norm": 0.2642206847667694,
+      "learning_rate": 1.1401369587117072e-07,
+      "loss": 0.3044,
+      "step": 1256
+    },
+    {
+      "epoch": 2.818385650224215,
+      "grad_norm": 0.36497628688812256,
+      "learning_rate": 1.1126012700131805e-07,
+      "loss": 0.2978,
+      "step": 1257
+    },
+    {
+      "epoch": 2.820627802690583,
+      "grad_norm": 0.25660938024520874,
+      "learning_rate": 1.0853984274007246e-07,
+      "loss": 0.3154,
+      "step": 1258
+    },
+    {
+      "epoch": 2.8228699551569507,
+      "grad_norm": 0.32555511593818665,
+      "learning_rate": 1.0585286160827935e-07,
+      "loss": 0.3003,
+      "step": 1259
+    },
+    {
+      "epoch": 2.8251121076233185,
+      "grad_norm": 0.5322585701942444,
+      "learning_rate": 1.031992019000444e-07,
+      "loss": 0.2933,
+      "step": 1260
+    },
+    {
+      "epoch": 2.827354260089686,
+      "grad_norm": 0.2479945868253708,
+      "learning_rate": 1.0057888168260311e-07,
+      "loss": 0.3131,
+      "step": 1261
+    },
+    {
+      "epoch": 2.8295964125560538,
+      "grad_norm": 0.2676950693130493,
+      "learning_rate": 9.799191879620474e-08,
+      "loss": 0.294,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8318385650224216,
+      "grad_norm": 0.24628274142742157,
+      "learning_rate": 9.543833085398635e-08,
+      "loss": 0.3054,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8340807174887894,
+      "grad_norm": 0.25654077529907227,
+      "learning_rate": 9.291813524185512e-08,
+      "loss": 0.3102,
+      "step": 1264
+    },
+    {
+      "epoch": 2.836322869955157,
+      "grad_norm": 0.25525566935539246,
+      "learning_rate": 9.043134911836782e-08,
+      "loss": 0.3091,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8385650224215246,
+      "grad_norm": 0.2652760446071625,
+      "learning_rate": 8.797798941461655e-08,
+      "loss": 0.2998,
+      "step": 1266
+    },
+    {
+      "epoch": 2.8408071748878925,
+      "grad_norm": 0.4145089387893677,
+      "learning_rate": 8.555807283411211e-08,
+      "loss": 0.3134,
+      "step": 1267
+    },
+    {
+      "epoch": 2.84304932735426,
+      "grad_norm": 0.26492810249328613,
+      "learning_rate": 8.317161585266964e-08,
+      "loss": 0.2898,
+      "step": 1268
+    },
+    {
+      "epoch": 2.8452914798206277,
+      "grad_norm": 0.2700561285018921,
+      "learning_rate": 8.08186347182971e-08,
+      "loss": 0.3091,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8475336322869955,
+      "grad_norm": 0.2543805241584778,
+      "learning_rate": 7.84991454510864e-08,
+      "loss": 0.3053,
+      "step": 1270
+    },
+    {
+      "epoch": 2.8497757847533634,
+      "grad_norm": 0.2684503495693207,
+      "learning_rate": 7.621316384309962e-08,
+      "loss": 0.301,
+      "step": 1271
+    },
+    {
+      "epoch": 2.852017937219731,
+      "grad_norm": 0.2750353217124939,
+      "learning_rate": 7.396070545826805e-08,
+      "loss": 0.3108,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8542600896860986,
+      "grad_norm": 0.2564678192138672,
+      "learning_rate": 7.174178563228051e-08,
+      "loss": 0.2999,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8565022421524664,
+      "grad_norm": 0.3156406879425049,
+      "learning_rate": 6.955641947248127e-08,
+      "loss": 0.295,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8587443946188342,
+      "grad_norm": 0.26844534277915955,
+      "learning_rate": 6.74046218577673e-08,
+      "loss": 0.2984,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8609865470852016,
+      "grad_norm": 0.25132638216018677,
+      "learning_rate": 6.52864074384868e-08,
+      "loss": 0.3059,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8632286995515694,
+      "grad_norm": 0.2832241654396057,
+      "learning_rate": 6.320179063634024e-08,
+      "loss": 0.3013,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8654708520179373,
+      "grad_norm": 0.2861579656600952,
+      "learning_rate": 6.115078564427946e-08,
+      "loss": 0.3138,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8677130044843047,
+      "grad_norm": 0.2510896325111389,
+      "learning_rate": 5.91334064264143e-08,
+      "loss": 0.2989,
+      "step": 1279
+    },
+    {
+      "epoch": 2.8699551569506725,
+      "grad_norm": 0.26631325483322144,
+      "learning_rate": 5.714966671791555e-08,
+      "loss": 0.2956,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8721973094170403,
+      "grad_norm": 0.2636234760284424,
+      "learning_rate": 5.519958002492221e-08,
+      "loss": 0.3095,
+      "step": 1281
+    },
+    {
+      "epoch": 2.874439461883408,
+      "grad_norm": 0.2975630462169647,
+      "learning_rate": 5.3283159624448745e-08,
+      "loss": 0.2943,
+      "step": 1282
+    },
+    {
+      "epoch": 2.876681614349776,
+      "grad_norm": 0.279818594455719,
+      "learning_rate": 5.1400418564296894e-08,
+      "loss": 0.3023,
+      "step": 1283
+    },
+    {
+      "epoch": 2.8789237668161434,
+      "grad_norm": 0.2595173418521881,
+      "learning_rate": 4.955136966296292e-08,
+      "loss": 0.3015,
+      "step": 1284
+    },
+    {
+      "epoch": 2.881165919282511,
+      "grad_norm": 0.29233136773109436,
+      "learning_rate": 4.773602550955381e-08,
+      "loss": 0.2889,
+      "step": 1285
+    },
+    {
+      "epoch": 2.883408071748879,
+      "grad_norm": 0.25460243225097656,
+      "learning_rate": 4.5954398463700647e-08,
+      "loss": 0.3079,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8856502242152464,
+      "grad_norm": 0.2648269534111023,
+      "learning_rate": 4.420650065547427e-08,
+      "loss": 0.3149,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8878923766816142,
+      "grad_norm": 0.25933533906936646,
+      "learning_rate": 4.249234398530144e-08,
+      "loss": 0.297,
+      "step": 1288
+    },
+    {
+      "epoch": 2.890134529147982,
+      "grad_norm": 0.2393570989370346,
+      "learning_rate": 4.081194012388601e-08,
+      "loss": 0.3102,
+      "step": 1289
+    },
+    {
+      "epoch": 2.89237668161435,
+      "grad_norm": 0.27711760997772217,
+      "learning_rate": 3.916530051212841e-08,
+      "loss": 0.2971,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8946188340807177,
+      "grad_norm": 0.2654544711112976,
+      "learning_rate": 3.7552436361046885e-08,
+      "loss": 0.2992,
+      "step": 1291
+    },
+    {
+      "epoch": 2.896860986547085,
+      "grad_norm": 0.25165459513664246,
+      "learning_rate": 3.5973358651703036e-08,
+      "loss": 0.3154,
+      "step": 1292
+    },
+    {
+      "epoch": 2.899103139013453,
+      "grad_norm": 0.2462768852710724,
+      "learning_rate": 3.442807813512583e-08,
+      "loss": 0.3012,
+      "step": 1293
+    },
+    {
+      "epoch": 2.901345291479821,
+      "grad_norm": 0.2619148790836334,
+      "learning_rate": 3.2916605332238284e-08,
+      "loss": 0.3048,
+      "step": 1294
+    },
+    {
+      "epoch": 2.903587443946188,
+      "grad_norm": 0.24694286286830902,
+      "learning_rate": 3.143895053378698e-08,
+      "loss": 0.2978,
+      "step": 1295
+    },
+    {
+      "epoch": 2.905829596412556,
+      "grad_norm": 0.2698237895965576,
+      "learning_rate": 2.9995123800270476e-08,
+      "loss": 0.2986,
+      "step": 1296
+    },
+    {
+      "epoch": 2.908071748878924,
+      "grad_norm": 0.3392399549484253,
+      "learning_rate": 2.858513496187154e-08,
+      "loss": 0.3029,
+      "step": 1297
+    },
+    {
+      "epoch": 2.910313901345291,
+      "grad_norm": 0.3584289252758026,
+      "learning_rate": 2.7208993618390578e-08,
+      "loss": 0.3091,
+      "step": 1298
+    },
+    {
+      "epoch": 2.912556053811659,
+      "grad_norm": 0.2510569393634796,
+      "learning_rate": 2.5866709139180102e-08,
+      "loss": 0.3058,
+      "step": 1299
+    },
+    {
+      "epoch": 2.914798206278027,
+      "grad_norm": 0.25194698572158813,
+      "learning_rate": 2.4558290663080354e-08,
+      "loss": 0.2971,
+      "step": 1300
+    },
+    {
+      "epoch": 2.9170403587443947,
+      "grad_norm": 0.22809167206287384,
+      "learning_rate": 2.3283747098357678e-08,
+      "loss": 0.3178,
+      "step": 1301
+    },
+    {
+      "epoch": 2.9192825112107625,
+      "grad_norm": 0.2600100636482239,
+      "learning_rate": 2.2043087122644023e-08,
+      "loss": 0.3126,
+      "step": 1302
+    },
+    {
+      "epoch": 2.92152466367713,
+      "grad_norm": 0.26127195358276367,
+      "learning_rate": 2.083631918287643e-08,
+      "loss": 0.3007,
+      "step": 1303
+    },
+    {
+      "epoch": 2.9237668161434978,
+      "grad_norm": 0.2572695314884186,
+      "learning_rate": 1.966345149524207e-08,
+      "loss": 0.3028,
+      "step": 1304
+    },
+    {
+      "epoch": 2.9260089686098656,
+      "grad_norm": 0.264804482460022,
+      "learning_rate": 1.852449204511997e-08,
+      "loss": 0.2961,
+      "step": 1305
+    },
+    {
+      "epoch": 2.928251121076233,
+      "grad_norm": 0.23976270854473114,
+      "learning_rate": 1.741944858702771e-08,
+      "loss": 0.3027,
+      "step": 1306
+    },
+    {
+      "epoch": 2.930493273542601,
+      "grad_norm": 0.29768192768096924,
+      "learning_rate": 1.63483286445687e-08,
+      "loss": 0.2868,
+      "step": 1307
+    },
+    {
+      "epoch": 2.9327354260089686,
+      "grad_norm": 0.2572835385799408,
+      "learning_rate": 1.5311139510380545e-08,
+      "loss": 0.3057,
+      "step": 1308
+    },
+    {
+      "epoch": 2.9349775784753365,
+      "grad_norm": 0.6042338609695435,
+      "learning_rate": 1.430788824608509e-08,
+      "loss": 0.2911,
+      "step": 1309
+    },
+    {
+      "epoch": 2.9372197309417043,
+      "grad_norm": 0.25112152099609375,
+      "learning_rate": 1.333858168224178e-08,
+      "loss": 0.3041,
+      "step": 1310
+    },
+    {
+      "epoch": 2.9394618834080717,
+      "grad_norm": 0.23699337244033813,
+      "learning_rate": 1.2403226418298831e-08,
+      "loss": 0.2932,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9417040358744395,
+      "grad_norm": 0.32967260479927063,
+      "learning_rate": 1.1501828822551575e-08,
+      "loss": 0.3096,
+      "step": 1312
+    },
+    {
+      "epoch": 2.9439461883408073,
+      "grad_norm": 0.25067201256752014,
+      "learning_rate": 1.063439503209529e-08,
+      "loss": 0.2893,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9461883408071747,
+      "grad_norm": 0.25441816449165344,
+      "learning_rate": 9.800930952786336e-09,
+      "loss": 0.3091,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9484304932735426,
+      "grad_norm": 0.26315128803253174,
+      "learning_rate": 9.001442259200522e-09,
+      "loss": 0.3108,
+      "step": 1315
+    },
+    {
+      "epoch": 2.9506726457399104,
+      "grad_norm": 0.26232296228408813,
+      "learning_rate": 8.235934394594802e-09,
+      "loss": 0.2997,
+      "step": 1316
+    },
+    {
+      "epoch": 2.952914798206278,
+      "grad_norm": 0.4478548765182495,
+      "learning_rate": 7.5044125708712e-09,
+      "loss": 0.299,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9551569506726456,
+      "grad_norm": 0.2623569369316101,
+      "learning_rate": 6.806881768539053e-09,
+      "loss": 0.312,
+      "step": 1318
+    },
+    {
+      "epoch": 2.9573991031390134,
+      "grad_norm": 0.25379660725593567,
+      "learning_rate": 6.14334673668282e-09,
+      "loss": 0.2966,
+      "step": 1319
+    },
+    {
+      "epoch": 2.9596412556053813,
+      "grad_norm": 0.28556954860687256,
+      "learning_rate": 5.51381199293044e-09,
+      "loss": 0.3078,
+      "step": 1320
+    },
+    {
+      "epoch": 2.961883408071749,
+      "grad_norm": 0.2840799391269684,
+      "learning_rate": 4.918281823420024e-09,
+      "loss": 0.3019,
+      "step": 1321
+    },
+    {
+      "epoch": 2.9641255605381165,
+      "grad_norm": 0.28264370560646057,
+      "learning_rate": 4.356760282773209e-09,
+      "loss": 0.3037,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9663677130044843,
+      "grad_norm": 0.28895458579063416,
+      "learning_rate": 3.8292511940657415e-09,
+      "loss": 0.2978,
+      "step": 1323
+    },
+    {
+      "epoch": 2.968609865470852,
+      "grad_norm": 0.2634439170360565,
+      "learning_rate": 3.3357581488030476e-09,
+      "loss": 0.3027,
+      "step": 1324
+    },
+    {
+      "epoch": 2.9708520179372195,
+      "grad_norm": 0.2698596715927124,
+      "learning_rate": 2.8762845068941448e-09,
+      "loss": 0.3069,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9730941704035874,
+      "grad_norm": 0.2459186166524887,
+      "learning_rate": 2.4508333966305473e-09,
+      "loss": 0.2986,
+      "step": 1326
+    },
+    {
+      "epoch": 2.975336322869955,
+      "grad_norm": 0.26233017444610596,
+      "learning_rate": 2.059407714662398e-09,
+      "loss": 0.2981,
+      "step": 1327
+    },
+    {
+      "epoch": 2.977578475336323,
+      "grad_norm": 0.255839467048645,
+      "learning_rate": 1.7020101259818123e-09,
+      "loss": 0.3174,
+      "step": 1328
+    },
+    {
+      "epoch": 2.979820627802691,
+      "grad_norm": 0.28446513414382935,
+      "learning_rate": 1.3786430639023406e-09,
+      "loss": 0.2935,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9820627802690582,
+      "grad_norm": 0.2634483575820923,
+      "learning_rate": 1.089308730043981e-09,
+      "loss": 0.3061,
+      "step": 1330
+    },
+    {
+      "epoch": 2.984304932735426,
+      "grad_norm": 0.28108322620391846,
+      "learning_rate": 8.340090943176338e-10,
+      "loss": 0.2997,
+      "step": 1331
+    },
+    {
+      "epoch": 2.986547085201794,
+      "grad_norm": 0.25852110981941223,
+      "learning_rate": 6.127458949106713e-10,
+      "loss": 0.2976,
+      "step": 1332
+    },
+    {
+      "epoch": 2.9887892376681613,
+      "grad_norm": 0.2785802185535431,
+      "learning_rate": 4.255206382769439e-10,
+      "loss": 0.3023,
+      "step": 1333
+    },
+    {
+      "epoch": 2.991031390134529,
+      "grad_norm": 0.28913140296936035,
+      "learning_rate": 2.723345991245685e-10,
+      "loss": 0.3072,
+      "step": 1334
+    },
+    {
+      "epoch": 2.993273542600897,
+      "grad_norm": 0.27687427401542664,
+      "learning_rate": 1.5318882040926685e-10,
+      "loss": 0.2961,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9955156950672643,
+      "grad_norm": 0.2595735788345337,
+      "learning_rate": 6.808411332548393e-11,
+      "loss": 0.2973,
+      "step": 1336
+    },
+    {
+      "epoch": 2.997757847533632,
+      "grad_norm": 0.2668149173259735,
+      "learning_rate": 1.7021057303057143e-11,
+      "loss": 0.3099,
+      "step": 1337
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.28971925377845764,
+      "learning_rate": 0.0,
+      "loss": 0.2815,
+      "step": 1338
+    },
+    {
+      "epoch": 3.0,
+      "step": 1338,
+      "total_flos": 1.7722899549913088e+16,
+      "train_loss": 0.031145622332890827,
+      "train_runtime": 14165.5102,
+      "train_samples_per_second": 24.134,
+      "train_steps_per_second": 0.094
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1338,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7722899549913088e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}