diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36513 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 5210,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009601536245799327,
+      "grad_norm": 4.608575884356107,
+      "learning_rate": 0.0,
+      "loss": 0.9304,
+      "step": 1
+    },
+    {
+      "epoch": 0.0019203072491598655,
+      "grad_norm": 5.267697768316677,
+      "learning_rate": 1.9193857965451055e-08,
+      "loss": 0.8959,
+      "step": 2
+    },
+    {
+      "epoch": 0.0028804608737397984,
+      "grad_norm": 3.674166203239673,
+      "learning_rate": 3.838771593090211e-08,
+      "loss": 0.9946,
+      "step": 3
+    },
+    {
+      "epoch": 0.003840614498319731,
+      "grad_norm": 4.212680693162968,
+      "learning_rate": 5.758157389635317e-08,
+      "loss": 1.0958,
+      "step": 4
+    },
+    {
+      "epoch": 0.004800768122899664,
+      "grad_norm": 4.966685183363455,
+      "learning_rate": 7.677543186180422e-08,
+      "loss": 1.1641,
+      "step": 5
+    },
+    {
+      "epoch": 0.005760921747479597,
+      "grad_norm": 3.778226764973662,
+      "learning_rate": 9.596928982725528e-08,
+      "loss": 0.9914,
+      "step": 6
+    },
+    {
+      "epoch": 0.00672107537205953,
+      "grad_norm": 4.083675255841295,
+      "learning_rate": 1.1516314779270635e-07,
+      "loss": 0.9815,
+      "step": 7
+    },
+    {
+      "epoch": 0.007681228996639462,
+      "grad_norm": 4.457244690216626,
+      "learning_rate": 1.343570057581574e-07,
+      "loss": 1.1257,
+      "step": 8
+    },
+    {
+      "epoch": 0.008641382621219395,
+      "grad_norm": 4.297331061465192,
+      "learning_rate": 1.5355086372360844e-07,
+      "loss": 1.1114,
+      "step": 9
+    },
+    {
+      "epoch": 0.009601536245799328,
+      "grad_norm": 3.475381879941915,
+      "learning_rate": 1.7274472168905953e-07,
+      "loss": 0.9588,
+      "step": 10
+    },
+    {
+      "epoch": 0.01056168987037926,
+      "grad_norm": 4.019411039718911,
+      "learning_rate": 1.9193857965451055e-07,
+      "loss": 1.082,
+      "step": 11
+    },
+    {
+      "epoch": 0.011521843494959194,
+      "grad_norm": 4.298656013398512,
+      "learning_rate": 2.1113243761996164e-07,
+      "loss": 0.9894,
+      "step": 12
+    },
+    {
+      "epoch": 0.012481997119539127,
+      "grad_norm": 4.298841455109918,
+      "learning_rate": 2.303262955854127e-07,
+      "loss": 1.0283,
+      "step": 13
+    },
+    {
+      "epoch": 0.01344215074411906,
+      "grad_norm": 3.8125619523001975,
+      "learning_rate": 2.4952015355086375e-07,
+      "loss": 0.9738,
+      "step": 14
+    },
+    {
+      "epoch": 0.014402304368698993,
+      "grad_norm": 5.948867639867879,
+      "learning_rate": 2.687140115163148e-07,
+      "loss": 1.1343,
+      "step": 15
+    },
+    {
+      "epoch": 0.015362457993278924,
+      "grad_norm": 4.516786504025045,
+      "learning_rate": 2.8790786948176586e-07,
+      "loss": 1.1204,
+      "step": 16
+    },
+    {
+      "epoch": 0.01632261161785886,
+      "grad_norm": 3.6831244206580913,
+      "learning_rate": 3.071017274472169e-07,
+      "loss": 0.9958,
+      "step": 17
+    },
+    {
+      "epoch": 0.01728276524243879,
+      "grad_norm": 3.9360004704440303,
+      "learning_rate": 3.262955854126679e-07,
+      "loss": 1.0483,
+      "step": 18
+    },
+    {
+      "epoch": 0.018242918867018725,
+      "grad_norm": 3.901791783080527,
+      "learning_rate": 3.4548944337811905e-07,
+      "loss": 1.0408,
+      "step": 19
+    },
+    {
+      "epoch": 0.019203072491598656,
+      "grad_norm": 4.3095324049321935,
+      "learning_rate": 3.646833013435701e-07,
+      "loss": 0.965,
+      "step": 20
+    },
+    {
+      "epoch": 0.020163226116178587,
+      "grad_norm": 4.148314319671607,
+      "learning_rate": 3.838771593090211e-07,
+      "loss": 1.1593,
+      "step": 21
+    },
+    {
+      "epoch": 0.02112337974075852,
+      "grad_norm": 3.535029565945762,
+      "learning_rate": 4.0307101727447224e-07,
+      "loss": 0.9418,
+      "step": 22
+    },
+    {
+      "epoch": 0.022083533365338453,
+      "grad_norm": 3.243808885930317,
+      "learning_rate": 4.2226487523992327e-07,
+      "loss": 0.9671,
+      "step": 23
+    },
+    {
+      "epoch": 0.023043686989918388,
+      "grad_norm": 3.1901459239600882,
+      "learning_rate": 4.414587332053743e-07,
+      "loss": 0.9802,
+      "step": 24
+    },
+    {
+      "epoch": 0.02400384061449832,
+      "grad_norm": 3.2819870328486016,
+      "learning_rate": 4.606525911708254e-07,
+      "loss": 1.0308,
+      "step": 25
+    },
+    {
+      "epoch": 0.024963994239078253,
+      "grad_norm": 3.5977491618398236,
+      "learning_rate": 4.798464491362765e-07,
+      "loss": 0.987,
+      "step": 26
+    },
+    {
+      "epoch": 0.025924147863658185,
+      "grad_norm": 3.3874834219166745,
+      "learning_rate": 4.990403071017275e-07,
+      "loss": 0.941,
+      "step": 27
+    },
+    {
+      "epoch": 0.02688430148823812,
+      "grad_norm": 3.240612452815504,
+      "learning_rate": 5.182341650671785e-07,
+      "loss": 1.072,
+      "step": 28
+    },
+    {
+      "epoch": 0.02784445511281805,
+      "grad_norm": 4.48751661697212,
+      "learning_rate": 5.374280230326296e-07,
+      "loss": 1.1463,
+      "step": 29
+    },
+    {
+      "epoch": 0.028804608737397985,
+      "grad_norm": 3.3988821883857896,
+      "learning_rate": 5.566218809980807e-07,
+      "loss": 1.0261,
+      "step": 30
+    },
+    {
+      "epoch": 0.029764762361977917,
+      "grad_norm": 3.2127956160200215,
+      "learning_rate": 5.758157389635317e-07,
+      "loss": 0.8477,
+      "step": 31
+    },
+    {
+      "epoch": 0.030724915986557848,
+      "grad_norm": 3.3352057499764154,
+      "learning_rate": 5.950095969289827e-07,
+      "loss": 0.9437,
+      "step": 32
+    },
+    {
+      "epoch": 0.03168506961113778,
+      "grad_norm": 3.060516597467567,
+      "learning_rate": 6.142034548944338e-07,
+      "loss": 1.0675,
+      "step": 33
+    },
+    {
+      "epoch": 0.03264522323571772,
+      "grad_norm": 2.308625467912329,
+      "learning_rate": 6.333973128598848e-07,
+      "loss": 0.9703,
+      "step": 34
+    },
+    {
+      "epoch": 0.033605376860297645,
+      "grad_norm": 2.187053407523522,
+      "learning_rate": 6.525911708253358e-07,
+      "loss": 1.006,
+      "step": 35
+    },
+    {
+      "epoch": 0.03456553048487758,
+      "grad_norm": 2.8405778838587903,
+      "learning_rate": 6.717850287907871e-07,
+      "loss": 0.8908,
+      "step": 36
+    },
+    {
+      "epoch": 0.035525684109457514,
+      "grad_norm": 2.566314021468516,
+      "learning_rate": 6.909788867562381e-07,
+      "loss": 0.9259,
+      "step": 37
+    },
+    {
+      "epoch": 0.03648583773403745,
+      "grad_norm": 3.0130066930691988,
+      "learning_rate": 7.101727447216891e-07,
+      "loss": 0.9417,
+      "step": 38
+    },
+    {
+      "epoch": 0.03744599135861738,
+      "grad_norm": 2.21081996209867,
+      "learning_rate": 7.293666026871402e-07,
+      "loss": 0.9622,
+      "step": 39
+    },
+    {
+      "epoch": 0.03840614498319731,
+      "grad_norm": 1.8475861673862382,
+      "learning_rate": 7.485604606525912e-07,
+      "loss": 0.7887,
+      "step": 40
+    },
+    {
+      "epoch": 0.039366298607777246,
+      "grad_norm": 2.381750801892437,
+      "learning_rate": 7.677543186180422e-07,
+      "loss": 0.8993,
+      "step": 41
+    },
+    {
+      "epoch": 0.040326452232357174,
+      "grad_norm": 2.295035679157881,
+      "learning_rate": 7.869481765834934e-07,
+      "loss": 1.0287,
+      "step": 42
+    },
+    {
+      "epoch": 0.04128660585693711,
+      "grad_norm": 2.3462274641458474,
+      "learning_rate": 8.061420345489445e-07,
+      "loss": 0.9755,
+      "step": 43
+    },
+    {
+      "epoch": 0.04224675948151704,
+      "grad_norm": 1.6515479761903826,
+      "learning_rate": 8.253358925143955e-07,
+      "loss": 0.8174,
+      "step": 44
+    },
+    {
+      "epoch": 0.04320691310609698,
+      "grad_norm": 1.7674375018904431,
+      "learning_rate": 8.445297504798465e-07,
+      "loss": 0.9022,
+      "step": 45
+    },
+    {
+      "epoch": 0.044167066730676906,
+      "grad_norm": 1.9577421293495636,
+      "learning_rate": 8.637236084452976e-07,
+      "loss": 1.0111,
+      "step": 46
+    },
+    {
+      "epoch": 0.04512722035525684,
+      "grad_norm": 1.7020469055537244,
+      "learning_rate": 8.829174664107486e-07,
+      "loss": 0.819,
+      "step": 47
+    },
+    {
+      "epoch": 0.046087373979836775,
+      "grad_norm": 2.01572674952168,
+      "learning_rate": 9.021113243761997e-07,
+      "loss": 0.9874,
+      "step": 48
+    },
+    {
+      "epoch": 0.04704752760441671,
+      "grad_norm": 2.3600284799271622,
+      "learning_rate": 9.213051823416508e-07,
+      "loss": 1.0881,
+      "step": 49
+    },
+    {
+      "epoch": 0.04800768122899664,
+      "grad_norm": 2.1334878781214934,
+      "learning_rate": 9.404990403071018e-07,
+      "loss": 1.0728,
+      "step": 50
+    },
+    {
+      "epoch": 0.04896783485357657,
+      "grad_norm": 2.081635795467802,
+      "learning_rate": 9.59692898272553e-07,
+      "loss": 1.0563,
+      "step": 51
+    },
+    {
+      "epoch": 0.04992798847815651,
+      "grad_norm": 1.872979323941484,
+      "learning_rate": 9.788867562380039e-07,
+      "loss": 0.7869,
+      "step": 52
+    },
+    {
+      "epoch": 0.050888142102736435,
+      "grad_norm": 2.253948697980142,
+      "learning_rate": 9.98080614203455e-07,
+      "loss": 1.0892,
+      "step": 53
+    },
+    {
+      "epoch": 0.05184829572731637,
+      "grad_norm": 1.7260286729835796,
+      "learning_rate": 1.0172744721689061e-06,
+      "loss": 0.922,
+      "step": 54
+    },
+    {
+      "epoch": 0.052808449351896304,
+      "grad_norm": 1.7645361159227275,
+      "learning_rate": 1.036468330134357e-06,
+      "loss": 0.963,
+      "step": 55
+    },
+    {
+      "epoch": 0.05376860297647624,
+      "grad_norm": 1.8586055165595463,
+      "learning_rate": 1.0556621880998082e-06,
+      "loss": 0.999,
+      "step": 56
+    },
+    {
+      "epoch": 0.05472875660105617,
+      "grad_norm": 1.843097634523443,
+      "learning_rate": 1.074856046065259e-06,
+      "loss": 1.009,
+      "step": 57
+    },
+    {
+      "epoch": 0.0556889102256361,
+      "grad_norm": 1.7610200133615168,
+      "learning_rate": 1.0940499040307102e-06,
+      "loss": 0.9579,
+      "step": 58
+    },
+    {
+      "epoch": 0.056649063850216036,
+      "grad_norm": 1.7138745961014983,
+      "learning_rate": 1.1132437619961614e-06,
+      "loss": 0.9182,
+      "step": 59
+    },
+    {
+      "epoch": 0.05760921747479597,
+      "grad_norm": 2.134408662198101,
+      "learning_rate": 1.1324376199616125e-06,
+      "loss": 0.8898,
+      "step": 60
+    },
+    {
+      "epoch": 0.0585693710993759,
+      "grad_norm": 1.2519878171090029,
+      "learning_rate": 1.1516314779270634e-06,
+      "loss": 0.8702,
+      "step": 61
+    },
+    {
+      "epoch": 0.05952952472395583,
+      "grad_norm": 1.1554566479823363,
+      "learning_rate": 1.1708253358925146e-06,
+      "loss": 0.7871,
+      "step": 62
+    },
+    {
+      "epoch": 0.06048967834853577,
+      "grad_norm": 1.6615875454598308,
+      "learning_rate": 1.1900191938579655e-06,
+      "loss": 0.9018,
+      "step": 63
+    },
+    {
+      "epoch": 0.061449831973115696,
+      "grad_norm": 1.7505284195298252,
+      "learning_rate": 1.2092130518234166e-06,
+      "loss": 1.0219,
+      "step": 64
+    },
+    {
+      "epoch": 0.06240998559769563,
+      "grad_norm": 1.5611445163160693,
+      "learning_rate": 1.2284069097888675e-06,
+      "loss": 0.9937,
+      "step": 65
+    },
+    {
+      "epoch": 0.06337013922227556,
+      "grad_norm": 1.0975898539730802,
+      "learning_rate": 1.2476007677543187e-06,
+      "loss": 0.7717,
+      "step": 66
+    },
+    {
+      "epoch": 0.0643302928468555,
+      "grad_norm": 1.2643294950307615,
+      "learning_rate": 1.2667946257197696e-06,
+      "loss": 0.9343,
+      "step": 67
+    },
+    {
+      "epoch": 0.06529044647143543,
+      "grad_norm": 1.2926314844705193,
+      "learning_rate": 1.2859884836852207e-06,
+      "loss": 0.9389,
+      "step": 68
+    },
+    {
+      "epoch": 0.06625060009601537,
+      "grad_norm": 1.5008851206841063,
+      "learning_rate": 1.3051823416506717e-06,
+      "loss": 0.9279,
+      "step": 69
+    },
+    {
+      "epoch": 0.06721075372059529,
+      "grad_norm": 1.4376243421140087,
+      "learning_rate": 1.324376199616123e-06,
+      "loss": 1.0277,
+      "step": 70
+    },
+    {
+      "epoch": 0.06817090734517522,
+      "grad_norm": 1.6332295422438075,
+      "learning_rate": 1.3435700575815741e-06,
+      "loss": 0.9154,
+      "step": 71
+    },
+    {
+      "epoch": 0.06913106096975516,
+      "grad_norm": 1.1937192619764858,
+      "learning_rate": 1.362763915547025e-06,
+      "loss": 0.865,
+      "step": 72
+    },
+    {
+      "epoch": 0.0700912145943351,
+      "grad_norm": 1.6160555706501898,
+      "learning_rate": 1.3819577735124762e-06,
+      "loss": 0.8731,
+      "step": 73
+    },
+    {
+      "epoch": 0.07105136821891503,
+      "grad_norm": 1.2828558000342634,
+      "learning_rate": 1.4011516314779271e-06,
+      "loss": 0.8508,
+      "step": 74
+    },
+    {
+      "epoch": 0.07201152184349496,
+      "grad_norm": 1.1433690942334571,
+      "learning_rate": 1.4203454894433783e-06,
+      "loss": 0.7113,
+      "step": 75
+    },
+    {
+      "epoch": 0.0729716754680749,
+      "grad_norm": 1.3498602336942411,
+      "learning_rate": 1.4395393474088292e-06,
+      "loss": 0.8771,
+      "step": 76
+    },
+    {
+      "epoch": 0.07393182909265482,
+      "grad_norm": 1.1090404504798739,
+      "learning_rate": 1.4587332053742803e-06,
+      "loss": 0.8389,
+      "step": 77
+    },
+    {
+      "epoch": 0.07489198271723475,
+      "grad_norm": 1.3371889492769014,
+      "learning_rate": 1.4779270633397312e-06,
+      "loss": 0.9674,
+      "step": 78
+    },
+    {
+      "epoch": 0.07585213634181469,
+      "grad_norm": 1.004920469127281,
+      "learning_rate": 1.4971209213051824e-06,
+      "loss": 0.8359,
+      "step": 79
+    },
+    {
+      "epoch": 0.07681228996639462,
+      "grad_norm": 1.1309480698032008,
+      "learning_rate": 1.5163147792706335e-06,
+      "loss": 0.871,
+      "step": 80
+    },
+    {
+      "epoch": 0.07777244359097456,
+      "grad_norm": 1.0926935444754795,
+      "learning_rate": 1.5355086372360844e-06,
+      "loss": 0.831,
+      "step": 81
+    },
+    {
+      "epoch": 0.07873259721555449,
+      "grad_norm": 1.3243569741514198,
+      "learning_rate": 1.5547024952015358e-06,
+      "loss": 0.929,
+      "step": 82
+    },
+    {
+      "epoch": 0.07969275084013443,
+      "grad_norm": 1.1459157332069554,
+      "learning_rate": 1.5738963531669867e-06,
+      "loss": 0.8642,
+      "step": 83
+    },
+    {
+      "epoch": 0.08065290446471435,
+      "grad_norm": 1.6430972393724526,
+      "learning_rate": 1.5930902111324378e-06,
+      "loss": 0.8614,
+      "step": 84
+    },
+    {
+      "epoch": 0.08161305808929428,
+      "grad_norm": 0.9635181271227865,
+      "learning_rate": 1.612284069097889e-06,
+      "loss": 0.8327,
+      "step": 85
+    },
+    {
+      "epoch": 0.08257321171387422,
+      "grad_norm": 1.0793717741613553,
+      "learning_rate": 1.63147792706334e-06,
+      "loss": 0.8913,
+      "step": 86
+    },
+    {
+      "epoch": 0.08353336533845415,
+      "grad_norm": 1.0314798311669489,
+      "learning_rate": 1.650671785028791e-06,
+      "loss": 0.7473,
+      "step": 87
+    },
+    {
+      "epoch": 0.08449351896303409,
+      "grad_norm": 1.135066098383753,
+      "learning_rate": 1.669865642994242e-06,
+      "loss": 0.8715,
+      "step": 88
+    },
+    {
+      "epoch": 0.08545367258761402,
+      "grad_norm": 1.1920535091940676,
+      "learning_rate": 1.689059500959693e-06,
+      "loss": 0.9432,
+      "step": 89
+    },
+    {
+      "epoch": 0.08641382621219396,
+      "grad_norm": 1.5782636211463792,
+      "learning_rate": 1.708253358925144e-06,
+      "loss": 1.0285,
+      "step": 90
+    },
+    {
+      "epoch": 0.08737397983677389,
+      "grad_norm": 1.3099591913322186,
+      "learning_rate": 1.7274472168905951e-06,
+      "loss": 0.9003,
+      "step": 91
+    },
+    {
+      "epoch": 0.08833413346135381,
+      "grad_norm": 1.1075263821827226,
+      "learning_rate": 1.746641074856046e-06,
+      "loss": 0.9024,
+      "step": 92
+    },
+    {
+      "epoch": 0.08929428708593375,
+      "grad_norm": 1.0596908140580912,
+      "learning_rate": 1.7658349328214972e-06,
+      "loss": 0.8997,
+      "step": 93
+    },
+    {
+      "epoch": 0.09025444071051368,
+      "grad_norm": 1.6824209174093476,
+      "learning_rate": 1.7850287907869481e-06,
+      "loss": 0.943,
+      "step": 94
+    },
+    {
+      "epoch": 0.09121459433509362,
+      "grad_norm": 1.4178621073994107,
+      "learning_rate": 1.8042226487523995e-06,
+      "loss": 0.9556,
+      "step": 95
+    },
+    {
+      "epoch": 0.09217474795967355,
+      "grad_norm": 1.9323428454897225,
+      "learning_rate": 1.8234165067178506e-06,
+      "loss": 0.9296,
+      "step": 96
+    },
+    {
+      "epoch": 0.09313490158425348,
+      "grad_norm": 1.113187686000337,
+      "learning_rate": 1.8426103646833015e-06,
+      "loss": 0.8704,
+      "step": 97
+    },
+    {
+      "epoch": 0.09409505520883342,
+      "grad_norm": 1.2127175436399928,
+      "learning_rate": 1.8618042226487527e-06,
+      "loss": 0.9178,
+      "step": 98
+    },
+    {
+      "epoch": 0.09505520883341334,
+      "grad_norm": 1.4205983901354309,
+      "learning_rate": 1.8809980806142036e-06,
+      "loss": 0.7902,
+      "step": 99
+    },
+    {
+      "epoch": 0.09601536245799328,
+      "grad_norm": 1.2336160656741093,
+      "learning_rate": 1.9001919385796547e-06,
+      "loss": 0.9751,
+      "step": 100
+    },
+    {
+      "epoch": 0.09697551608257321,
+      "grad_norm": 1.32218061984364,
+      "learning_rate": 1.919385796545106e-06,
+      "loss": 0.9422,
+      "step": 101
+    },
+    {
+      "epoch": 0.09793566970715314,
+      "grad_norm": 1.4349730066378146,
+      "learning_rate": 1.9385796545105568e-06,
+      "loss": 0.9295,
+      "step": 102
+    },
+    {
+      "epoch": 0.09889582333173308,
+      "grad_norm": 0.9816231563605403,
+      "learning_rate": 1.9577735124760077e-06,
+      "loss": 0.7518,
+      "step": 103
+    },
+    {
+      "epoch": 0.09985597695631301,
+      "grad_norm": 1.1071092246727834,
+      "learning_rate": 1.9769673704414586e-06,
+      "loss": 0.8828,
+      "step": 104
+    },
+    {
+      "epoch": 0.10081613058089295,
+      "grad_norm": 1.0429061975034752,
+      "learning_rate": 1.99616122840691e-06,
+      "loss": 0.7532,
+      "step": 105
+    },
+    {
+      "epoch": 0.10177628420547287,
+      "grad_norm": 1.3352134212660092,
+      "learning_rate": 2.015355086372361e-06,
+      "loss": 0.9469,
+      "step": 106
+    },
+    {
+      "epoch": 0.1027364378300528,
+      "grad_norm": 0.9859105142366051,
+      "learning_rate": 2.0345489443378122e-06,
+      "loss": 0.7679,
+      "step": 107
+    },
+    {
+      "epoch": 0.10369659145463274,
+      "grad_norm": 0.92010241449416,
+      "learning_rate": 2.053742802303263e-06,
+      "loss": 0.7839,
+      "step": 108
+    },
+    {
+      "epoch": 0.10465674507921267,
+      "grad_norm": 0.959546321803882,
+      "learning_rate": 2.072936660268714e-06,
+      "loss": 0.8704,
+      "step": 109
+    },
+    {
+      "epoch": 0.10561689870379261,
+      "grad_norm": 1.0121164230279138,
+      "learning_rate": 2.0921305182341654e-06,
+      "loss": 0.7748,
+      "step": 110
+    },
+    {
+      "epoch": 0.10657705232837254,
+      "grad_norm": 0.8983272418888552,
+      "learning_rate": 2.1113243761996164e-06,
+      "loss": 0.8037,
+      "step": 111
+    },
+    {
+      "epoch": 0.10753720595295248,
+      "grad_norm": 1.2574271355202917,
+      "learning_rate": 2.1305182341650673e-06,
+      "loss": 0.9004,
+      "step": 112
+    },
+    {
+      "epoch": 0.1084973595775324,
+      "grad_norm": 1.093188744338786,
+      "learning_rate": 2.149712092130518e-06,
+      "loss": 0.7215,
+      "step": 113
+    },
+    {
+      "epoch": 0.10945751320211233,
+      "grad_norm": 1.0677246649240915,
+      "learning_rate": 2.1689059500959696e-06,
+      "loss": 0.7169,
+      "step": 114
+    },
+    {
+      "epoch": 0.11041766682669227,
+      "grad_norm": 1.2362607538290646,
+      "learning_rate": 2.1880998080614205e-06,
+      "loss": 0.8536,
+      "step": 115
+    },
+    {
+      "epoch": 0.1113778204512722,
+      "grad_norm": 1.7753777515012845,
+      "learning_rate": 2.2072936660268714e-06,
+      "loss": 0.9374,
+      "step": 116
+    },
+    {
+      "epoch": 0.11233797407585214,
+      "grad_norm": 0.8545104864205748,
+      "learning_rate": 2.2264875239923228e-06,
+      "loss": 0.7226,
+      "step": 117
+    },
+    {
+      "epoch": 0.11329812770043207,
+      "grad_norm": 0.8855990202815591,
+      "learning_rate": 2.2456813819577737e-06,
+      "loss": 0.8334,
+      "step": 118
+    },
+    {
+      "epoch": 0.114258281325012,
+      "grad_norm": 1.1549401776774397,
+      "learning_rate": 2.264875239923225e-06,
+      "loss": 0.8969,
+      "step": 119
+    },
+    {
+      "epoch": 0.11521843494959194,
+      "grad_norm": 0.9811642711182177,
+      "learning_rate": 2.284069097888676e-06,
+      "loss": 0.9129,
+      "step": 120
+    },
+    {
+      "epoch": 0.11617858857417186,
+      "grad_norm": 1.533573497571935,
+      "learning_rate": 2.303262955854127e-06,
+      "loss": 0.9064,
+      "step": 121
+    },
+    {
+      "epoch": 0.1171387421987518,
+      "grad_norm": 0.9402456230334737,
+      "learning_rate": 2.322456813819578e-06,
+      "loss": 0.7372,
+      "step": 122
+    },
+    {
+      "epoch": 0.11809889582333173,
+      "grad_norm": 1.1086864304645334,
+      "learning_rate": 2.341650671785029e-06,
+      "loss": 0.875,
+      "step": 123
+    },
+    {
+      "epoch": 0.11905904944791167,
+      "grad_norm": 1.2002234675956465,
+      "learning_rate": 2.36084452975048e-06,
+      "loss": 0.7128,
+      "step": 124
+    },
+    {
+      "epoch": 0.1200192030724916,
+      "grad_norm": 0.8158104333736143,
+      "learning_rate": 2.380038387715931e-06,
+      "loss": 0.6694,
+      "step": 125
+    },
+    {
+      "epoch": 0.12097935669707154,
+      "grad_norm": 1.0877193848670241,
+      "learning_rate": 2.3992322456813823e-06,
+      "loss": 0.7247,
+      "step": 126
+    },
+    {
+      "epoch": 0.12193951032165147,
+      "grad_norm": 1.719738648966655,
+      "learning_rate": 2.4184261036468333e-06,
+      "loss": 0.7878,
+      "step": 127
+    },
+    {
+      "epoch": 0.12289966394623139,
+      "grad_norm": 1.0309225906761206,
+      "learning_rate": 2.437619961612284e-06,
+      "loss": 0.9741,
+      "step": 128
+    },
+    {
+      "epoch": 0.12385981757081133,
+      "grad_norm": 1.2089797111899467,
+      "learning_rate": 2.456813819577735e-06,
+      "loss": 0.9194,
+      "step": 129
+    },
+    {
+      "epoch": 0.12481997119539126,
+      "grad_norm": 0.9567385863393127,
+      "learning_rate": 2.4760076775431864e-06,
+      "loss": 0.7369,
+      "step": 130
+    },
+    {
+      "epoch": 0.1257801248199712,
+      "grad_norm": 1.142430713832724,
+      "learning_rate": 2.4952015355086374e-06,
+      "loss": 0.8294,
+      "step": 131
+    },
+    {
+      "epoch": 0.12674027844455113,
+      "grad_norm": 1.3090923807049384,
+      "learning_rate": 2.5143953934740883e-06,
+      "loss": 0.8456,
+      "step": 132
+    },
+    {
+      "epoch": 0.12770043206913106,
+      "grad_norm": 1.2503457366705912,
+      "learning_rate": 2.5335892514395392e-06,
+      "loss": 0.8744,
+      "step": 133
+    },
+    {
+      "epoch": 0.128660585693711,
+      "grad_norm": 1.0711467403091606,
+      "learning_rate": 2.5527831094049906e-06,
+      "loss": 0.8338,
+      "step": 134
+    },
+    {
+      "epoch": 0.12962073931829093,
+      "grad_norm": 1.6084156917236838,
+      "learning_rate": 2.5719769673704415e-06,
+      "loss": 0.8592,
+      "step": 135
+    },
+    {
+      "epoch": 0.13058089294287087,
+      "grad_norm": 1.3308971115078625,
+      "learning_rate": 2.5911708253358924e-06,
+      "loss": 0.7858,
+      "step": 136
+    },
+    {
+      "epoch": 0.1315410465674508,
+      "grad_norm": 1.0828775067399747,
+      "learning_rate": 2.6103646833013433e-06,
+      "loss": 0.9026,
+      "step": 137
+    },
+    {
+      "epoch": 0.13250120019203074,
+      "grad_norm": 1.2085159913760772,
+      "learning_rate": 2.629558541266795e-06,
+      "loss": 0.8446,
+      "step": 138
+    },
+    {
+      "epoch": 0.13346135381661065,
+      "grad_norm": 1.3699534639223174,
+      "learning_rate": 2.648752399232246e-06,
+      "loss": 0.8861,
+      "step": 139
+    },
+    {
+      "epoch": 0.13442150744119058,
+      "grad_norm": 1.1738872763207,
+      "learning_rate": 2.6679462571976974e-06,
+      "loss": 0.7103,
+      "step": 140
+    },
+    {
+      "epoch": 0.13538166106577051,
+      "grad_norm": 1.0481635182090046,
+      "learning_rate": 2.6871401151631483e-06,
+      "loss": 0.9215,
+      "step": 141
+    },
+    {
+      "epoch": 0.13634181469035045,
+      "grad_norm": 0.8045116032807893,
+      "learning_rate": 2.7063339731285992e-06,
+      "loss": 0.7805,
+      "step": 142
+    },
+    {
+      "epoch": 0.13730196831493038,
+      "grad_norm": 0.9818798002309104,
+      "learning_rate": 2.72552783109405e-06,
+      "loss": 0.7844,
+      "step": 143
+    },
+    {
+      "epoch": 0.13826212193951032,
+      "grad_norm": 1.1680069006683926,
+      "learning_rate": 2.7447216890595015e-06,
+      "loss": 0.7744,
+      "step": 144
+    },
+    {
+      "epoch": 0.13922227556409025,
+      "grad_norm": 1.2164677300673201,
+      "learning_rate": 2.7639155470249524e-06,
+      "loss": 0.825,
+      "step": 145
+    },
+    {
+      "epoch": 0.1401824291886702,
+      "grad_norm": 1.2465951929177923,
+      "learning_rate": 2.7831094049904033e-06,
+      "loss": 0.6901,
+      "step": 146
+    },
+    {
+      "epoch": 0.14114258281325012,
+      "grad_norm": 1.100866715811699,
+      "learning_rate": 2.8023032629558543e-06,
+      "loss": 0.8722,
+      "step": 147
+    },
+    {
+      "epoch": 0.14210273643783006,
+      "grad_norm": 1.6066181507941228,
+      "learning_rate": 2.8214971209213056e-06,
+      "loss": 0.8334,
+      "step": 148
+    },
+    {
+      "epoch": 0.14306289006241,
+      "grad_norm": 1.3751884963973906,
+      "learning_rate": 2.8406909788867565e-06,
+      "loss": 0.7968,
+      "step": 149
+    },
+    {
+      "epoch": 0.14402304368698993,
+      "grad_norm": 1.1715432702710724,
+      "learning_rate": 2.8598848368522074e-06,
+      "loss": 0.8436,
+      "step": 150
+    },
+    {
+      "epoch": 0.14498319731156986,
+      "grad_norm": 1.0127103261881165,
+      "learning_rate": 2.8790786948176584e-06,
+      "loss": 0.7558,
+      "step": 151
+    },
+    {
+      "epoch": 0.1459433509361498,
+      "grad_norm": 0.9868103566572486,
+      "learning_rate": 2.8982725527831097e-06,
+      "loss": 0.7393,
+      "step": 152
+    },
+    {
+      "epoch": 0.1469035045607297,
+      "grad_norm": 1.2398536101083277,
+      "learning_rate": 2.9174664107485606e-06,
+      "loss": 0.873,
+      "step": 153
+    },
+    {
+      "epoch": 0.14786365818530964,
+      "grad_norm": 1.1999787944229654,
+      "learning_rate": 2.9366602687140116e-06,
+      "loss": 0.9016,
+      "step": 154
+    },
+    {
+      "epoch": 0.14882381180988957,
+      "grad_norm": 1.1324449796486242,
+      "learning_rate": 2.9558541266794625e-06,
+      "loss": 0.7667,
+      "step": 155
+    },
+    {
+      "epoch": 0.1497839654344695,
+      "grad_norm": 1.3120704929861124,
+      "learning_rate": 2.975047984644914e-06,
+      "loss": 0.8744,
+      "step": 156
+    },
+    {
+      "epoch": 0.15074411905904944,
+      "grad_norm": 0.9062163071328126,
+      "learning_rate": 2.9942418426103648e-06,
+      "loss": 0.7798,
+      "step": 157
+    },
+    {
+      "epoch": 0.15170427268362938,
+      "grad_norm": 1.4076582801284412,
+      "learning_rate": 3.0134357005758157e-06,
+      "loss": 0.8565,
+      "step": 158
+    },
+    {
+      "epoch": 0.1526644263082093,
+      "grad_norm": 1.4673082584603003,
+      "learning_rate": 3.032629558541267e-06,
+      "loss": 0.8813,
+      "step": 159
+    },
+    {
+      "epoch": 0.15362457993278925,
+      "grad_norm": 1.374671366145908,
+      "learning_rate": 3.051823416506718e-06,
+      "loss": 0.8445,
+      "step": 160
+    },
+    {
+      "epoch": 0.15458473355736918,
+      "grad_norm": 1.2868007929715295,
+      "learning_rate": 3.071017274472169e-06,
+      "loss": 0.8934,
+      "step": 161
+    },
+    {
+      "epoch": 0.15554488718194912,
+      "grad_norm": 1.4464722202223055,
+      "learning_rate": 3.09021113243762e-06,
+      "loss": 0.7582,
+      "step": 162
+    },
+    {
+      "epoch": 0.15650504080652905,
+      "grad_norm": 1.105469935141118,
+      "learning_rate": 3.1094049904030716e-06,
+      "loss": 0.7808,
+      "step": 163
+    },
+    {
+      "epoch": 0.15746519443110898,
+      "grad_norm": 1.0911900597803557,
+      "learning_rate": 3.1285988483685225e-06,
+      "loss": 0.796,
+      "step": 164
+    },
+    {
+      "epoch": 0.15842534805568892,
+      "grad_norm": 1.3309231366723955,
+      "learning_rate": 3.1477927063339734e-06,
+      "loss": 0.8544,
+      "step": 165
+    },
+    {
+      "epoch": 0.15938550168026885,
+      "grad_norm": 1.2548221881193011,
+      "learning_rate": 3.1669865642994248e-06,
+      "loss": 0.8361,
+      "step": 166
+    },
+    {
+      "epoch": 0.1603456553048488,
+      "grad_norm": 1.7659278560067024,
+      "learning_rate": 3.1861804222648757e-06,
+      "loss": 0.8906,
+      "step": 167
+    },
+    {
+      "epoch": 0.1613058089294287,
+      "grad_norm": 1.1655867571269192,
+      "learning_rate": 3.2053742802303266e-06,
+      "loss": 0.894,
+      "step": 168
+    },
+    {
+      "epoch": 0.16226596255400863,
+      "grad_norm": 1.2433543508842448,
+      "learning_rate": 3.224568138195778e-06,
+      "loss": 0.8004,
+      "step": 169
+    },
+    {
+      "epoch": 0.16322611617858857,
+      "grad_norm": 0.9872268083143309,
+      "learning_rate": 3.243761996161229e-06,
+      "loss": 0.822,
+      "step": 170
+    },
+    {
+      "epoch": 0.1641862698031685,
+      "grad_norm": 1.3753973529861458,
+      "learning_rate": 3.26295585412668e-06,
+      "loss": 0.9343,
+      "step": 171
+    },
+    {
+      "epoch": 0.16514642342774843,
+      "grad_norm": 1.0586321759599753,
+      "learning_rate": 3.2821497120921307e-06,
+      "loss": 0.844,
+      "step": 172
+    },
+    {
+      "epoch": 0.16610657705232837,
+      "grad_norm": 0.9715401565614465,
+      "learning_rate": 3.301343570057582e-06,
+      "loss": 0.7735,
+      "step": 173
+    },
+    {
+      "epoch": 0.1670667306769083,
+      "grad_norm": 1.2748668842297577,
+      "learning_rate": 3.320537428023033e-06,
+      "loss": 0.7567,
+      "step": 174
+    },
+    {
+      "epoch": 0.16802688430148824,
+      "grad_norm": 0.9412191459151686,
+      "learning_rate": 3.339731285988484e-06,
+      "loss": 0.6285,
+      "step": 175
+    },
+    {
+      "epoch": 0.16898703792606817,
+      "grad_norm": 1.1052323344585788,
+      "learning_rate": 3.358925143953935e-06,
+      "loss": 0.9069,
+      "step": 176
+    },
+    {
+      "epoch": 0.1699471915506481,
+      "grad_norm": 1.08861582464332,
+      "learning_rate": 3.378119001919386e-06,
+      "loss": 0.7559,
+      "step": 177
+    },
+    {
+      "epoch": 0.17090734517522804,
+      "grad_norm": 0.9319805506601659,
+      "learning_rate": 3.397312859884837e-06,
+      "loss": 0.7105,
+      "step": 178
+    },
+    {
+      "epoch": 0.17186749879980798,
+      "grad_norm": 0.9771730792010799,
+      "learning_rate": 3.416506717850288e-06,
+      "loss": 0.8552,
+      "step": 179
+    },
+    {
+      "epoch": 0.1728276524243879,
+      "grad_norm": 0.8470050375728845,
+      "learning_rate": 3.435700575815739e-06,
+      "loss": 0.7615,
+      "step": 180
+    },
+    {
+      "epoch": 0.17378780604896785,
+      "grad_norm": 1.0959154608915553,
+      "learning_rate": 3.4548944337811903e-06,
+      "loss": 0.7262,
+      "step": 181
+    },
+    {
+      "epoch": 0.17474795967354778,
+      "grad_norm": 1.4625375688138138,
+      "learning_rate": 3.4740882917466412e-06,
+      "loss": 0.8587,
+      "step": 182
+    },
+    {
+      "epoch": 0.1757081132981277,
+      "grad_norm": 1.0000628986768312,
+      "learning_rate": 3.493282149712092e-06,
+      "loss": 0.7761,
+      "step": 183
+    },
+    {
+      "epoch": 0.17666826692270762,
+      "grad_norm": 1.0005373628571768,
+      "learning_rate": 3.512476007677543e-06,
+      "loss": 0.6796,
+      "step": 184
+    },
+    {
+      "epoch": 0.17762842054728756,
+      "grad_norm": 0.9968604487993922,
+      "learning_rate": 3.5316698656429944e-06,
+      "loss": 0.8201,
+      "step": 185
+    },
+    {
+      "epoch": 0.1785885741718675,
+      "grad_norm": 1.4993901938369656,
+      "learning_rate": 3.5508637236084453e-06,
+      "loss": 0.9407,
+      "step": 186
+    },
+    {
+      "epoch": 0.17954872779644743,
+      "grad_norm": 0.8704017460553644,
+      "learning_rate": 3.5700575815738963e-06,
+      "loss": 0.6974,
+      "step": 187
+    },
+    {
+      "epoch": 0.18050888142102736,
+      "grad_norm": 0.9761668674958587,
+      "learning_rate": 3.589251439539348e-06,
+      "loss": 0.8016,
+      "step": 188
+    },
+    {
+      "epoch": 0.1814690350456073,
+      "grad_norm": 1.0515756767474949,
+      "learning_rate": 3.608445297504799e-06,
+      "loss": 0.7232,
+      "step": 189
+    },
+    {
+      "epoch": 0.18242918867018723,
+      "grad_norm": 1.555490418358551,
+      "learning_rate": 3.62763915547025e-06,
+      "loss": 0.6891,
+      "step": 190
+    },
+    {
+      "epoch": 0.18338934229476717,
+      "grad_norm": 0.9343252294840784,
+      "learning_rate": 3.6468330134357012e-06,
+      "loss": 0.7632,
+      "step": 191
+    },
+    {
+      "epoch": 0.1843494959193471,
+      "grad_norm": 1.0717238045522897,
+      "learning_rate": 3.666026871401152e-06,
+      "loss": 0.7967,
+      "step": 192
+    },
+    {
+      "epoch": 0.18530964954392704,
+      "grad_norm": 0.9023425167215238,
+      "learning_rate": 3.685220729366603e-06,
+      "loss": 0.6851,
+      "step": 193
+    },
+    {
+      "epoch": 0.18626980316850697,
+      "grad_norm": 1.0354539206924438,
+      "learning_rate": 3.704414587332054e-06,
+      "loss": 0.6541,
+      "step": 194
+    },
+    {
+      "epoch": 0.1872299567930869,
+      "grad_norm": 0.9154002359123815,
+      "learning_rate": 3.7236084452975053e-06,
+      "loss": 0.6118,
+      "step": 195
+    },
+    {
+      "epoch": 0.18819011041766684,
+      "grad_norm": 1.4834012311015687,
+      "learning_rate": 3.7428023032629563e-06,
+      "loss": 0.7373,
+      "step": 196
+    },
+    {
+      "epoch": 0.18915026404224675,
+      "grad_norm": 0.9395875480988187,
+      "learning_rate": 3.761996161228407e-06,
+      "loss": 0.7418,
+      "step": 197
+    },
+    {
+      "epoch": 0.19011041766682668,
+      "grad_norm": 0.8228848151380725,
+      "learning_rate": 3.781190019193858e-06,
+      "loss": 0.639,
+      "step": 198
+    },
+    {
+      "epoch": 0.19107057129140662,
+      "grad_norm": 1.0222902437157457,
+      "learning_rate": 3.8003838771593095e-06,
+      "loss": 0.7843,
+      "step": 199
+    },
+    {
+      "epoch": 0.19203072491598655,
+      "grad_norm": 1.1971998828421855,
+      "learning_rate": 3.81957773512476e-06,
+      "loss": 0.8042,
+      "step": 200
+    },
+    {
+      "epoch": 0.19299087854056649,
+      "grad_norm": 1.3436354605563,
+      "learning_rate": 3.838771593090212e-06,
+      "loss": 0.6823,
+      "step": 201
+    },
+    {
+      "epoch": 0.19395103216514642,
+      "grad_norm": 1.0432691374129013,
+      "learning_rate": 3.857965451055662e-06,
+      "loss": 0.8763,
+      "step": 202
+    },
+    {
+      "epoch": 0.19491118578972635,
+      "grad_norm": 1.4598651306676347,
+      "learning_rate": 3.8771593090211136e-06,
+      "loss": 0.7037,
+      "step": 203
+    },
+    {
+      "epoch": 0.1958713394143063,
+      "grad_norm": 1.2815645347506386,
+      "learning_rate": 3.896353166986565e-06,
+      "loss": 0.7283,
+      "step": 204
+    },
+    {
+      "epoch": 0.19683149303888622,
+      "grad_norm": 1.0833856307385523,
+      "learning_rate": 3.915547024952015e-06,
+      "loss": 0.8133,
+      "step": 205
+    },
+    {
+      "epoch": 0.19779164666346616,
+      "grad_norm": 1.3786574048365472,
+      "learning_rate": 3.934740882917467e-06,
+      "loss": 0.9112,
+      "step": 206
+    },
+    {
+      "epoch": 0.1987518002880461,
+      "grad_norm": 1.2875346340348794,
+      "learning_rate": 3.953934740882917e-06,
+      "loss": 0.6257,
+      "step": 207
+    },
+    {
+      "epoch": 0.19971195391262603,
+      "grad_norm": 1.0242596133728246,
+      "learning_rate": 3.973128598848369e-06,
+      "loss": 0.8232,
+      "step": 208
+    },
+    {
+      "epoch": 0.20067210753720596,
+      "grad_norm": 0.9870530206057417,
+      "learning_rate": 3.99232245681382e-06,
+      "loss": 0.8147,
+      "step": 209
+    },
+    {
+      "epoch": 0.2016322611617859,
+      "grad_norm": 1.4004219647275036,
+      "learning_rate": 4.0115163147792705e-06,
+      "loss": 0.7407,
+      "step": 210
+    },
+    {
+      "epoch": 0.20259241478636583,
+      "grad_norm": 1.9647525624689024,
+      "learning_rate": 4.030710172744722e-06,
+      "loss": 0.9065,
+      "step": 211
+    },
+    {
+      "epoch": 0.20355256841094574,
+      "grad_norm": 1.1758284793178084,
+      "learning_rate": 4.049904030710173e-06,
+      "loss": 0.7354,
+      "step": 212
+    },
+    {
+      "epoch": 0.20451272203552567,
+      "grad_norm": 0.9399628887884863,
+      "learning_rate": 4.0690978886756245e-06,
+      "loss": 0.7094,
+      "step": 213
+    },
+    {
+      "epoch": 0.2054728756601056,
+      "grad_norm": 1.229532902404846,
+      "learning_rate": 4.088291746641076e-06,
+      "loss": 0.7294,
+      "step": 214
+    },
+    {
+      "epoch": 0.20643302928468554,
+      "grad_norm": 1.285481349569688,
+      "learning_rate": 4.107485604606526e-06,
+      "loss": 0.7874,
+      "step": 215
+    },
+    {
+      "epoch": 0.20739318290926548,
+      "grad_norm": 1.7048926937913806,
+      "learning_rate": 4.126679462571978e-06,
+      "loss": 0.9107,
+      "step": 216
+    },
+    {
+      "epoch": 0.2083533365338454,
+      "grad_norm": 1.2789559622430384,
+      "learning_rate": 4.145873320537428e-06,
+      "loss": 0.7613,
+      "step": 217
+    },
+    {
+      "epoch": 0.20931349015842535,
+      "grad_norm": 1.099423101490443,
+      "learning_rate": 4.1650671785028795e-06,
+      "loss": 0.8418,
+      "step": 218
+    },
+    {
+      "epoch": 0.21027364378300528,
+      "grad_norm": 1.0902575563764023,
+      "learning_rate": 4.184261036468331e-06,
+      "loss": 0.6956,
+      "step": 219
+    },
+    {
+      "epoch": 0.21123379740758522,
+      "grad_norm": 1.272728305370189,
+      "learning_rate": 4.203454894433781e-06,
+      "loss": 0.9046,
+      "step": 220
+    },
+    {
+      "epoch": 0.21219395103216515,
+      "grad_norm": 0.9179036321827532,
+      "learning_rate": 4.222648752399233e-06,
+      "loss": 0.8334,
+      "step": 221
+    },
+    {
+      "epoch": 0.21315410465674509,
+      "grad_norm": 1.1608689862055543,
+      "learning_rate": 4.241842610364684e-06,
+      "loss": 0.7988,
+      "step": 222
+    },
+    {
+      "epoch": 0.21411425828132502,
+      "grad_norm": 0.9424134482088689,
+      "learning_rate": 4.2610364683301346e-06,
+      "loss": 0.7275,
+      "step": 223
+    },
+    {
+      "epoch": 0.21507441190590496,
+      "grad_norm": 1.0056322925801187,
+      "learning_rate": 4.280230326295586e-06,
+      "loss": 0.8358,
+      "step": 224
+    },
+    {
+      "epoch": 0.2160345655304849,
+      "grad_norm": 1.358420329516284,
+      "learning_rate": 4.299424184261036e-06,
+      "loss": 0.789,
+      "step": 225
+    },
+    {
+      "epoch": 0.2169947191550648,
+      "grad_norm": 1.0604793977460594,
+      "learning_rate": 4.318618042226488e-06,
+      "loss": 0.7716,
+      "step": 226
+    },
+    {
+      "epoch": 0.21795487277964473,
+      "grad_norm": 1.3579766865243843,
+      "learning_rate": 4.337811900191939e-06,
+      "loss": 0.7773,
+      "step": 227
+    },
+    {
+      "epoch": 0.21891502640422467,
+      "grad_norm": 1.1969285333538193,
+      "learning_rate": 4.35700575815739e-06,
+      "loss": 0.7988,
+      "step": 228
+    },
+    {
+      "epoch": 0.2198751800288046,
+      "grad_norm": 1.0977349545103885,
+      "learning_rate": 4.376199616122841e-06,
+      "loss": 0.8328,
+      "step": 229
+    },
+    {
+      "epoch": 0.22083533365338454,
+      "grad_norm": 1.1029551255090795,
+      "learning_rate": 4.395393474088292e-06,
+      "loss": 0.686,
+      "step": 230
+    },
+    {
+      "epoch": 0.22179548727796447,
+      "grad_norm": 1.4552381896353486,
+      "learning_rate": 4.414587332053743e-06,
+      "loss": 0.8195,
+      "step": 231
+    },
+    {
+      "epoch": 0.2227556409025444,
+      "grad_norm": 0.894667659418859,
+      "learning_rate": 4.433781190019194e-06,
+      "loss": 0.7275,
+      "step": 232
+    },
+    {
+      "epoch": 0.22371579452712434,
+      "grad_norm": 0.8202068548985902,
+      "learning_rate": 4.4529750479846455e-06,
+      "loss": 0.5861,
+      "step": 233
+    },
+    {
+      "epoch": 0.22467594815170427,
+      "grad_norm": 1.029707146111314,
+      "learning_rate": 4.472168905950096e-06,
+      "loss": 0.6985,
+      "step": 234
+    },
+    {
+      "epoch": 0.2256361017762842,
+      "grad_norm": 1.0990681024322204,
+      "learning_rate": 4.491362763915547e-06,
+      "loss": 0.68,
+      "step": 235
+    },
+    {
+      "epoch": 0.22659625540086414,
+      "grad_norm": 1.4012799413122012,
+      "learning_rate": 4.510556621880998e-06,
+      "loss": 0.8869,
+      "step": 236
+    },
+    {
+      "epoch": 0.22755640902544408,
+      "grad_norm": 1.1973821399483648,
+      "learning_rate": 4.52975047984645e-06,
+      "loss": 0.7971,
+      "step": 237
+    },
+    {
+      "epoch": 0.228516562650024,
+      "grad_norm": 1.4686234982760031,
+      "learning_rate": 4.5489443378119005e-06,
+      "loss": 0.7288,
+      "step": 238
+    },
+    {
+      "epoch": 0.22947671627460395,
+      "grad_norm": 1.032604350539258,
+      "learning_rate": 4.568138195777352e-06,
+      "loss": 0.7507,
+      "step": 239
+    },
+    {
+      "epoch": 0.23043686989918388,
+      "grad_norm": 1.0822190269808825,
+      "learning_rate": 4.587332053742803e-06,
+      "loss": 0.8673,
+      "step": 240
+    },
+    {
+      "epoch": 0.2313970235237638,
+      "grad_norm": 1.0856558538512318,
+      "learning_rate": 4.606525911708254e-06,
+      "loss": 0.7822,
+      "step": 241
+    },
+    {
+      "epoch": 0.23235717714834372,
+      "grad_norm": 1.253257076933008,
+      "learning_rate": 4.625719769673705e-06,
+      "loss": 0.8852,
+      "step": 242
+    },
+    {
+      "epoch": 0.23331733077292366,
+      "grad_norm": 1.0533352620131817,
+      "learning_rate": 4.644913627639156e-06,
+      "loss": 0.8494,
+      "step": 243
+    },
+    {
+      "epoch": 0.2342774843975036,
+      "grad_norm": 1.2104402492231694,
+      "learning_rate": 4.664107485604607e-06,
+      "loss": 0.8848,
+      "step": 244
+    },
+    {
+      "epoch": 0.23523763802208353,
+      "grad_norm": 1.0309254759547493,
+      "learning_rate": 4.683301343570058e-06,
+      "loss": 0.7899,
+      "step": 245
+    },
+    {
+      "epoch": 0.23619779164666346,
+      "grad_norm": 0.9228274298271464,
+      "learning_rate": 4.702495201535509e-06,
+      "loss": 0.7684,
+      "step": 246
+    },
+    {
+      "epoch": 0.2371579452712434,
+      "grad_norm": 1.082494978756675,
+      "learning_rate": 4.72168905950096e-06,
+      "loss": 0.8015,
+      "step": 247
+    },
+    {
+      "epoch": 0.23811809889582333,
+      "grad_norm": 1.1404265500842226,
+      "learning_rate": 4.7408829174664115e-06,
+      "loss": 0.8995,
+      "step": 248
+    },
+    {
+      "epoch": 0.23907825252040327,
+      "grad_norm": 1.1383637841992147,
+      "learning_rate": 4.760076775431862e-06,
+      "loss": 0.7914,
+      "step": 249
+    },
+    {
+      "epoch": 0.2400384061449832,
+      "grad_norm": 1.2275079047335584,
+      "learning_rate": 4.779270633397313e-06,
+      "loss": 0.8338,
+      "step": 250
+    },
+    {
+      "epoch": 0.24099855976956314,
+      "grad_norm": 1.1011103569744343,
+      "learning_rate": 4.798464491362765e-06,
+      "loss": 0.7431,
+      "step": 251
+    },
+    {
+      "epoch": 0.24195871339414307,
+      "grad_norm": 1.0587489444236706,
+      "learning_rate": 4.817658349328215e-06,
+      "loss": 0.7682,
+      "step": 252
+    },
+    {
+      "epoch": 0.242918867018723,
+      "grad_norm": 1.26897029416747,
+      "learning_rate": 4.8368522072936665e-06,
+      "loss": 0.7562,
+      "step": 253
+    },
+    {
+      "epoch": 0.24387902064330294,
+      "grad_norm": 0.8705948431481193,
+      "learning_rate": 4.856046065259117e-06,
+      "loss": 0.7452,
+      "step": 254
+    },
+    {
+      "epoch": 0.24483917426788285,
+      "grad_norm": 1.092814530275756,
+      "learning_rate": 4.875239923224568e-06,
+      "loss": 0.7531,
+      "step": 255
+    },
+    {
+      "epoch": 0.24579932789246278,
+      "grad_norm": 1.297369303442484,
+      "learning_rate": 4.89443378119002e-06,
+      "loss": 0.7553,
+      "step": 256
+    },
+    {
+      "epoch": 0.24675948151704272,
+      "grad_norm": 1.0620363460937197,
+      "learning_rate": 4.91362763915547e-06,
+      "loss": 0.778,
+      "step": 257
+    },
+    {
+      "epoch": 0.24771963514162265,
+      "grad_norm": 1.007077263281022,
+      "learning_rate": 4.9328214971209215e-06,
+      "loss": 0.8044,
+      "step": 258
+    },
+    {
+      "epoch": 0.2486797887662026,
+      "grad_norm": 1.2092919265067725,
+      "learning_rate": 4.952015355086373e-06,
+      "loss": 0.7579,
+      "step": 259
+    },
+    {
+      "epoch": 0.24963994239078252,
+      "grad_norm": 1.0123125970450146,
+      "learning_rate": 4.971209213051823e-06,
+      "loss": 0.7101,
+      "step": 260
+    },
+    {
+      "epoch": 0.2506000960153625,
+      "grad_norm": 0.8793511419440708,
+      "learning_rate": 4.990403071017275e-06,
+      "loss": 0.6262,
+      "step": 261
+    },
+    {
+      "epoch": 0.2515602496399424,
+      "grad_norm": 1.078809487237499,
+      "learning_rate": 5.009596928982726e-06,
+      "loss": 0.6783,
+      "step": 262
+    },
+    {
+      "epoch": 0.2525204032645223,
+      "grad_norm": 0.941420750891124,
+      "learning_rate": 5.028790786948177e-06,
+      "loss": 0.6403,
+      "step": 263
+    },
+    {
+      "epoch": 0.25348055688910226,
+      "grad_norm": 1.3575222071854605,
+      "learning_rate": 5.047984644913628e-06,
+      "loss": 0.6863,
+      "step": 264
+    },
+    {
+      "epoch": 0.25444071051368217,
+      "grad_norm": 0.9902478107580348,
+      "learning_rate": 5.0671785028790784e-06,
+      "loss": 0.6048,
+      "step": 265
+    },
+    {
+      "epoch": 0.25540086413826213,
+      "grad_norm": 0.9383297421101107,
+      "learning_rate": 5.086372360844531e-06,
+      "loss": 0.7624,
+      "step": 266
+    },
+    {
+      "epoch": 0.25636101776284204,
+      "grad_norm": 0.9745420306969944,
+      "learning_rate": 5.105566218809981e-06,
+      "loss": 0.726,
+      "step": 267
+    },
+    {
+      "epoch": 0.257321171387422,
+      "grad_norm": 0.8823186085065814,
+      "learning_rate": 5.1247600767754325e-06,
+      "loss": 0.6667,
+      "step": 268
+    },
+    {
+      "epoch": 0.2582813250120019,
+      "grad_norm": 1.3077109124843154,
+      "learning_rate": 5.143953934740883e-06,
+      "loss": 0.7825,
+      "step": 269
+    },
+    {
+      "epoch": 0.25924147863658187,
+      "grad_norm": 0.8701718980494662,
+      "learning_rate": 5.163147792706334e-06,
+      "loss": 0.7284,
+      "step": 270
+    },
+    {
+      "epoch": 0.2602016322611618,
+      "grad_norm": 1.0947520139625904,
+      "learning_rate": 5.182341650671785e-06,
+      "loss": 0.7205,
+      "step": 271
+    },
+    {
+      "epoch": 0.26116178588574174,
+      "grad_norm": 0.9035421633997119,
+      "learning_rate": 5.201535508637236e-06,
+      "loss": 0.707,
+      "step": 272
+    },
+    {
+      "epoch": 0.26212193951032164,
+      "grad_norm": 1.2393989616692116,
+      "learning_rate": 5.220729366602687e-06,
+      "loss": 0.8318,
+      "step": 273
+    },
+    {
+      "epoch": 0.2630820931349016,
+      "grad_norm": 1.0820647264342638,
+      "learning_rate": 5.239923224568139e-06,
+      "loss": 0.7086,
+      "step": 274
+    },
+    {
+      "epoch": 0.2640422467594815,
+      "grad_norm": 0.9102334415716998,
+      "learning_rate": 5.25911708253359e-06,
+      "loss": 0.686,
+      "step": 275
+    },
+    {
+      "epoch": 0.2650024003840615,
+      "grad_norm": 0.9891112292775399,
+      "learning_rate": 5.278310940499041e-06,
+      "loss": 0.6594,
+      "step": 276
+    },
+    {
+      "epoch": 0.2659625540086414,
+      "grad_norm": 1.6096713844116006,
+      "learning_rate": 5.297504798464492e-06,
+      "loss": 0.7862,
+      "step": 277
+    },
+    {
+      "epoch": 0.2669227076332213,
+      "grad_norm": 1.1225227418118746,
+      "learning_rate": 5.3166986564299425e-06,
+      "loss": 0.7328,
+      "step": 278
+    },
+    {
+      "epoch": 0.26788286125780125,
+      "grad_norm": 1.0231886402817059,
+      "learning_rate": 5.335892514395395e-06,
+      "loss": 0.6849,
+      "step": 279
+    },
+    {
+      "epoch": 0.26884301488238116,
+      "grad_norm": 1.2155240110282657,
+      "learning_rate": 5.355086372360845e-06,
+      "loss": 0.7907,
+      "step": 280
+    },
+    {
+      "epoch": 0.2698031685069611,
+      "grad_norm": 1.292857617189349,
+      "learning_rate": 5.374280230326297e-06,
+      "loss": 0.726,
+      "step": 281
+    },
+    {
+      "epoch": 0.27076332213154103,
+      "grad_norm": 1.272186258694057,
+      "learning_rate": 5.393474088291747e-06,
+      "loss": 0.6674,
+      "step": 282
+    },
+    {
+      "epoch": 0.271723475756121,
+      "grad_norm": 1.3078425645435416,
+      "learning_rate": 5.4126679462571984e-06,
+      "loss": 0.7705,
+      "step": 283
+    },
+    {
+      "epoch": 0.2726836293807009,
+      "grad_norm": 1.0078428876763987,
+      "learning_rate": 5.431861804222649e-06,
+      "loss": 0.7072,
+      "step": 284
+    },
+    {
+      "epoch": 0.27364378300528086,
+      "grad_norm": 1.2618657788714938,
+      "learning_rate": 5.4510556621881e-06,
+      "loss": 0.8385,
+      "step": 285
+    },
+    {
+      "epoch": 0.27460393662986077,
+      "grad_norm": 1.1965521394945369,
+      "learning_rate": 5.470249520153551e-06,
+      "loss": 0.7379,
+      "step": 286
+    },
+    {
+      "epoch": 0.27556409025444073,
+      "grad_norm": 1.0040579235952913,
+      "learning_rate": 5.489443378119003e-06,
+      "loss": 0.6312,
+      "step": 287
+    },
+    {
+      "epoch": 0.27652424387902064,
+      "grad_norm": 1.1226365966223,
+      "learning_rate": 5.5086372360844535e-06,
+      "loss": 0.7333,
+      "step": 288
+    },
+    {
+      "epoch": 0.2774843975036006,
+      "grad_norm": 1.0426235237494172,
+      "learning_rate": 5.527831094049905e-06,
+      "loss": 0.8043,
+      "step": 289
+    },
+    {
+      "epoch": 0.2784445511281805,
+      "grad_norm": 1.1282354419614937,
+      "learning_rate": 5.547024952015355e-06,
+      "loss": 0.6533,
+      "step": 290
+    },
+    {
+      "epoch": 0.27940470475276047,
+      "grad_norm": 1.0876477193458896,
+      "learning_rate": 5.566218809980807e-06,
+      "loss": 0.8533,
+      "step": 291
+    },
+    {
+      "epoch": 0.2803648583773404,
+      "grad_norm": 1.2267395738659088,
+      "learning_rate": 5.585412667946257e-06,
+      "loss": 0.7215,
+      "step": 292
+    },
+    {
+      "epoch": 0.2813250120019203,
+      "grad_norm": 1.3886867771152398,
+      "learning_rate": 5.6046065259117085e-06,
+      "loss": 0.7892,
+      "step": 293
+    },
+    {
+      "epoch": 0.28228516562650025,
+      "grad_norm": 1.2328491859061401,
+      "learning_rate": 5.623800383877159e-06,
+      "loss": 0.8432,
+      "step": 294
+    },
+    {
+      "epoch": 0.28324531925108015,
+      "grad_norm": 1.518901224201875,
+      "learning_rate": 5.642994241842611e-06,
+      "loss": 0.8301,
+      "step": 295
+    },
+    {
+      "epoch": 0.2842054728756601,
+      "grad_norm": 1.456338581338527,
+      "learning_rate": 5.662188099808062e-06,
+      "loss": 0.8818,
+      "step": 296
+    },
+    {
+      "epoch": 0.28516562650024,
+      "grad_norm": 1.3027020402970058,
+      "learning_rate": 5.681381957773513e-06,
+      "loss": 0.8254,
+      "step": 297
+    },
+    {
+      "epoch": 0.28612578012482,
+      "grad_norm": 0.9080934518090018,
+      "learning_rate": 5.7005758157389635e-06,
+      "loss": 0.6895,
+      "step": 298
+    },
+    {
+      "epoch": 0.2870859337493999,
+      "grad_norm": 1.2130812480748785,
+      "learning_rate": 5.719769673704415e-06,
+      "loss": 0.7949,
+      "step": 299
+    },
+    {
+      "epoch": 0.28804608737397985,
+      "grad_norm": 1.1160052811250392,
+      "learning_rate": 5.738963531669866e-06,
+      "loss": 0.8164,
+      "step": 300
+    },
+    {
+      "epoch": 0.28900624099855976,
+      "grad_norm": 1.1210266919814806,
+      "learning_rate": 5.758157389635317e-06,
+      "loss": 0.649,
+      "step": 301
+    },
+    {
+      "epoch": 0.2899663946231397,
+      "grad_norm": 1.0879864278375668,
+      "learning_rate": 5.777351247600769e-06,
+      "loss": 0.8905,
+      "step": 302
+    },
+    {
+      "epoch": 0.29092654824771963,
+      "grad_norm": 1.0213016500599061,
+      "learning_rate": 5.7965451055662194e-06,
+      "loss": 0.7535,
+      "step": 303
+    },
+    {
+      "epoch": 0.2918867018722996,
+      "grad_norm": 0.865511806914633,
+      "learning_rate": 5.815738963531671e-06,
+      "loss": 0.7009,
+      "step": 304
+    },
+    {
+      "epoch": 0.2928468554968795,
+      "grad_norm": 1.0762956326446182,
+      "learning_rate": 5.834932821497121e-06,
+      "loss": 0.8394,
+      "step": 305
+    },
+    {
+      "epoch": 0.2938070091214594,
+      "grad_norm": 1.1011468772663233,
+      "learning_rate": 5.854126679462573e-06,
+      "loss": 0.7994,
+      "step": 306
+    },
+    {
+      "epoch": 0.29476716274603937,
+      "grad_norm": 0.9358929440958041,
+      "learning_rate": 5.873320537428023e-06,
+      "loss": 0.6915,
+      "step": 307
+    },
+    {
+      "epoch": 0.2957273163706193,
+      "grad_norm": 1.1361983938783253,
+      "learning_rate": 5.892514395393475e-06,
+      "loss": 0.7465,
+      "step": 308
+    },
+    {
+      "epoch": 0.29668746999519924,
+      "grad_norm": 0.9706557310474103,
+      "learning_rate": 5.911708253358925e-06,
+      "loss": 0.7263,
+      "step": 309
+    },
+    {
+      "epoch": 0.29764762361977914,
+      "grad_norm": 0.9957406689560145,
+      "learning_rate": 5.930902111324377e-06,
+      "loss": 0.6731,
+      "step": 310
+    },
+    {
+      "epoch": 0.2986077772443591,
+      "grad_norm": 0.8297692761648142,
+      "learning_rate": 5.950095969289828e-06,
+      "loss": 0.6627,
+      "step": 311
+    },
+    {
+      "epoch": 0.299567930868939,
+      "grad_norm": 1.0823838214362094,
+      "learning_rate": 5.969289827255279e-06,
+      "loss": 0.7685,
+      "step": 312
+    },
+    {
+      "epoch": 0.300528084493519,
+      "grad_norm": 1.075826853937253,
+      "learning_rate": 5.9884836852207295e-06,
+      "loss": 0.7684,
+      "step": 313
+    },
+    {
+      "epoch": 0.3014882381180989,
+      "grad_norm": 1.3059389423523045,
+      "learning_rate": 6.007677543186181e-06,
+      "loss": 0.8944,
+      "step": 314
+    },
+    {
+      "epoch": 0.30244839174267885,
+      "grad_norm": 1.083104462062558,
+      "learning_rate": 6.026871401151631e-06,
+      "loss": 0.7052,
+      "step": 315
+    },
+    {
+      "epoch": 0.30340854536725875,
+      "grad_norm": 0.9209470027209099,
+      "learning_rate": 6.0460652591170836e-06,
+      "loss": 0.6925,
+      "step": 316
+    },
+    {
+      "epoch": 0.3043686989918387,
+      "grad_norm": 1.5506192977662627,
+      "learning_rate": 6.065259117082534e-06,
+      "loss": 0.8079,
+      "step": 317
+    },
+    {
+      "epoch": 0.3053288526164186,
+      "grad_norm": 1.1809708521141837,
+      "learning_rate": 6.084452975047985e-06,
+      "loss": 0.8492,
+      "step": 318
+    },
+    {
+      "epoch": 0.3062890062409986,
+      "grad_norm": 0.888021044899069,
+      "learning_rate": 6.103646833013436e-06,
+      "loss": 0.6798,
+      "step": 319
+    },
+    {
+      "epoch": 0.3072491598655785,
+      "grad_norm": 1.059234718818979,
+      "learning_rate": 6.122840690978887e-06,
+      "loss": 0.7214,
+      "step": 320
+    },
+    {
+      "epoch": 0.3082093134901584,
+      "grad_norm": 1.2972122688066008,
+      "learning_rate": 6.142034548944338e-06,
+      "loss": 0.8314,
+      "step": 321
+    },
+    {
+      "epoch": 0.30916946711473836,
+      "grad_norm": 1.026669162294941,
+      "learning_rate": 6.161228406909789e-06,
+      "loss": 0.6907,
+      "step": 322
+    },
+    {
+      "epoch": 0.31012962073931827,
+      "grad_norm": 1.0317826337977176,
+      "learning_rate": 6.18042226487524e-06,
+      "loss": 0.6827,
+      "step": 323
+    },
+    {
+      "epoch": 0.31108977436389823,
+      "grad_norm": 1.1500376036788722,
+      "learning_rate": 6.199616122840692e-06,
+      "loss": 0.8111,
+      "step": 324
+    },
+    {
+      "epoch": 0.31204992798847814,
+      "grad_norm": 0.9380007217093529,
+      "learning_rate": 6.218809980806143e-06,
+      "loss": 0.5933,
+      "step": 325
+    },
+    {
+      "epoch": 0.3130100816130581,
+      "grad_norm": 1.5376009293846107,
+      "learning_rate": 6.238003838771594e-06,
+      "loss": 0.8007,
+      "step": 326
+    },
+    {
+      "epoch": 0.313970235237638,
+      "grad_norm": 1.4599272669739034,
+      "learning_rate": 6.257197696737045e-06,
+      "loss": 0.7693,
+      "step": 327
+    },
+    {
+      "epoch": 0.31493038886221797,
+      "grad_norm": 1.5064515194175732,
+      "learning_rate": 6.2763915547024955e-06,
+      "loss": 0.9682,
+      "step": 328
+    },
+    {
+      "epoch": 0.3158905424867979,
+      "grad_norm": 0.9818886506201384,
+      "learning_rate": 6.295585412667947e-06,
+      "loss": 0.6798,
+      "step": 329
+    },
+    {
+      "epoch": 0.31685069611137784,
+      "grad_norm": 0.9712226276725909,
+      "learning_rate": 6.314779270633397e-06,
+      "loss": 0.743,
+      "step": 330
+    },
+    {
+      "epoch": 0.31781084973595775,
+      "grad_norm": 1.401672652996209,
+      "learning_rate": 6.3339731285988495e-06,
+      "loss": 0.7874,
+      "step": 331
+    },
+    {
+      "epoch": 0.3187710033605377,
+      "grad_norm": 1.0783950280605588,
+      "learning_rate": 6.3531669865643e-06,
+      "loss": 0.8842,
+      "step": 332
+    },
+    {
+      "epoch": 0.3197311569851176,
+      "grad_norm": 1.066363631638222,
+      "learning_rate": 6.372360844529751e-06,
+      "loss": 0.7388,
+      "step": 333
+    },
+    {
+      "epoch": 0.3206913106096976,
+      "grad_norm": 1.2209253548266827,
+      "learning_rate": 6.391554702495202e-06,
+      "loss": 0.7856,
+      "step": 334
+    },
+    {
+      "epoch": 0.3216514642342775,
+      "grad_norm": 1.2441561987844856,
+      "learning_rate": 6.410748560460653e-06,
+      "loss": 0.746,
+      "step": 335
+    },
+    {
+      "epoch": 0.3226116178588574,
+      "grad_norm": 1.2247304882054608,
+      "learning_rate": 6.429942418426104e-06,
+      "loss": 0.6776,
+      "step": 336
+    },
+    {
+      "epoch": 0.32357177148343735,
+      "grad_norm": 1.0856409119088077,
+      "learning_rate": 6.449136276391556e-06,
+      "loss": 0.721,
+      "step": 337
+    },
+    {
+      "epoch": 0.32453192510801726,
+      "grad_norm": 1.0796733699008674,
+      "learning_rate": 6.4683301343570056e-06,
+      "loss": 0.6634,
+      "step": 338
+    },
+    {
+      "epoch": 0.3254920787325972,
+      "grad_norm": 0.953864425810896,
+      "learning_rate": 6.487523992322458e-06,
+      "loss": 0.658,
+      "step": 339
+    },
+    {
+      "epoch": 0.32645223235717713,
+      "grad_norm": 1.071183113729068,
+      "learning_rate": 6.506717850287908e-06,
+      "loss": 0.8166,
+      "step": 340
+    },
+    {
+      "epoch": 0.3274123859817571,
+      "grad_norm": 0.877389486714033,
+      "learning_rate": 6.52591170825336e-06,
+      "loss": 0.5952,
+      "step": 341
+    },
+    {
+      "epoch": 0.328372539606337,
+      "grad_norm": 1.0020588864869746,
+      "learning_rate": 6.54510556621881e-06,
+      "loss": 0.6962,
+      "step": 342
+    },
+    {
+      "epoch": 0.32933269323091696,
+      "grad_norm": 1.3953195365139277,
+      "learning_rate": 6.5642994241842614e-06,
+      "loss": 0.9079,
+      "step": 343
+    },
+    {
+      "epoch": 0.33029284685549687,
+      "grad_norm": 1.1863169477349746,
+      "learning_rate": 6.583493282149712e-06,
+      "loss": 0.7766,
+      "step": 344
+    },
+    {
+      "epoch": 0.33125300048007683,
+      "grad_norm": 1.199300644800143,
+      "learning_rate": 6.602687140115164e-06,
+      "loss": 0.8119,
+      "step": 345
+    },
+    {
+      "epoch": 0.33221315410465674,
+      "grad_norm": 1.0315407252952944,
+      "learning_rate": 6.621880998080615e-06,
+      "loss": 0.6961,
+      "step": 346
+    },
+    {
+      "epoch": 0.3331733077292367,
+      "grad_norm": 0.8332104483111213,
+      "learning_rate": 6.641074856046066e-06,
+      "loss": 0.6525,
+      "step": 347
+    },
+    {
+      "epoch": 0.3341334613538166,
+      "grad_norm": 1.0502384293543738,
+      "learning_rate": 6.6602687140115165e-06,
+      "loss": 0.7915,
+      "step": 348
+    },
+    {
+      "epoch": 0.33509361497839657,
+      "grad_norm": 1.75292168897863,
+      "learning_rate": 6.679462571976968e-06,
+      "loss": 0.9489,
+      "step": 349
+    },
+    {
+      "epoch": 0.3360537686029765,
+      "grad_norm": 1.0394537445812293,
+      "learning_rate": 6.698656429942419e-06,
+      "loss": 0.7422,
+      "step": 350
+    },
+    {
+      "epoch": 0.3370139222275564,
+      "grad_norm": 1.3457781986708817,
+      "learning_rate": 6.71785028790787e-06,
+      "loss": 0.7588,
+      "step": 351
+    },
+    {
+      "epoch": 0.33797407585213635,
+      "grad_norm": 1.1385632722570405,
+      "learning_rate": 6.737044145873322e-06,
+      "loss": 0.7602,
+      "step": 352
+    },
+    {
+      "epoch": 0.33893422947671625,
+      "grad_norm": 1.6067517847274677,
+      "learning_rate": 6.756238003838772e-06,
+      "loss": 0.8005,
+      "step": 353
+    },
+    {
+      "epoch": 0.3398943831012962,
+      "grad_norm": 1.233975400494269,
+      "learning_rate": 6.775431861804224e-06,
+      "loss": 0.7682,
+      "step": 354
+    },
+    {
+      "epoch": 0.3408545367258761,
+      "grad_norm": 1.4526697439532525,
+      "learning_rate": 6.794625719769674e-06,
+      "loss": 0.75,
+      "step": 355
+    },
+    {
+      "epoch": 0.3418146903504561,
+      "grad_norm": 0.9323361856360292,
+      "learning_rate": 6.8138195777351256e-06,
+      "loss": 0.6922,
+      "step": 356
+    },
+    {
+      "epoch": 0.342774843975036,
+      "grad_norm": 1.0609558520672935,
+      "learning_rate": 6.833013435700576e-06,
+      "loss": 0.8037,
+      "step": 357
+    },
+    {
+      "epoch": 0.34373499759961595,
+      "grad_norm": 0.9283068391125257,
+      "learning_rate": 6.852207293666027e-06,
+      "loss": 0.8207,
+      "step": 358
+    },
+    {
+      "epoch": 0.34469515122419586,
+      "grad_norm": 1.2741384759296661,
+      "learning_rate": 6.871401151631478e-06,
+      "loss": 0.7734,
+      "step": 359
+    },
+    {
+      "epoch": 0.3456553048487758,
+      "grad_norm": 1.2226624613325408,
+      "learning_rate": 6.89059500959693e-06,
+      "loss": 0.7153,
+      "step": 360
+    },
+    {
+      "epoch": 0.34661545847335573,
+      "grad_norm": 0.9392211836640727,
+      "learning_rate": 6.909788867562381e-06,
+      "loss": 0.6772,
+      "step": 361
+    },
+    {
+      "epoch": 0.3475756120979357,
+      "grad_norm": 1.4176149408001522,
+      "learning_rate": 6.928982725527832e-06,
+      "loss": 0.7478,
+      "step": 362
+    },
+    {
+      "epoch": 0.3485357657225156,
+      "grad_norm": 1.055027957209198,
+      "learning_rate": 6.9481765834932824e-06,
+      "loss": 0.7025,
+      "step": 363
+    },
+    {
+      "epoch": 0.34949591934709556,
+      "grad_norm": 0.9237996976063603,
+      "learning_rate": 6.967370441458734e-06,
+      "loss": 0.7641,
+      "step": 364
+    },
+    {
+      "epoch": 0.35045607297167547,
+      "grad_norm": 1.0023494135672188,
+      "learning_rate": 6.986564299424184e-06,
+      "loss": 0.773,
+      "step": 365
+    },
+    {
+      "epoch": 0.3514162265962554,
+      "grad_norm": 1.1170315829753363,
+      "learning_rate": 7.005758157389636e-06,
+      "loss": 0.7737,
+      "step": 366
+    },
+    {
+      "epoch": 0.35237638022083534,
+      "grad_norm": 1.6571294664693028,
+      "learning_rate": 7.024952015355086e-06,
+      "loss": 0.7678,
+      "step": 367
+    },
+    {
+      "epoch": 0.35333653384541525,
+      "grad_norm": 1.1463208677096601,
+      "learning_rate": 7.044145873320538e-06,
+      "loss": 0.7038,
+      "step": 368
+    },
+    {
+      "epoch": 0.3542966874699952,
+      "grad_norm": 1.2041496401683789,
+      "learning_rate": 7.063339731285989e-06,
+      "loss": 0.8187,
+      "step": 369
+    },
+    {
+      "epoch": 0.3552568410945751,
+      "grad_norm": 1.3123729805107278,
+      "learning_rate": 7.08253358925144e-06,
+      "loss": 0.7559,
+      "step": 370
+    },
+    {
+      "epoch": 0.3562169947191551,
+      "grad_norm": 1.2531666737811658,
+      "learning_rate": 7.101727447216891e-06,
+      "loss": 0.8967,
+      "step": 371
+    },
+    {
+      "epoch": 0.357177148343735,
+      "grad_norm": 1.6044073246344652,
+      "learning_rate": 7.120921305182342e-06,
+      "loss": 0.8522,
+      "step": 372
+    },
+    {
+      "epoch": 0.35813730196831495,
+      "grad_norm": 1.46386764159694,
+      "learning_rate": 7.1401151631477925e-06,
+      "loss": 0.8204,
+      "step": 373
+    },
+    {
+      "epoch": 0.35909745559289485,
+      "grad_norm": 1.124678220450824,
+      "learning_rate": 7.159309021113245e-06,
+      "loss": 0.7096,
+      "step": 374
+    },
+    {
+      "epoch": 0.3600576092174748,
+      "grad_norm": 1.0907290673814356,
+      "learning_rate": 7.178502879078696e-06,
+      "loss": 0.6543,
+      "step": 375
+    },
+    {
+      "epoch": 0.3610177628420547,
+      "grad_norm": 1.2365841743837318,
+      "learning_rate": 7.1976967370441466e-06,
+      "loss": 0.7116,
+      "step": 376
+    },
+    {
+      "epoch": 0.3619779164666347,
+      "grad_norm": 1.1952106366739865,
+      "learning_rate": 7.216890595009598e-06,
+      "loss": 0.6948,
+      "step": 377
+    },
+    {
+      "epoch": 0.3629380700912146,
+      "grad_norm": 1.156231992977645,
+      "learning_rate": 7.236084452975048e-06,
+      "loss": 0.7459,
+      "step": 378
+    },
+    {
+      "epoch": 0.3638982237157945,
+      "grad_norm": 1.0787123026665089,
+      "learning_rate": 7.2552783109405e-06,
+      "loss": 0.7374,
+      "step": 379
+    },
+    {
+      "epoch": 0.36485837734037446,
+      "grad_norm": 1.161408465298755,
+      "learning_rate": 7.27447216890595e-06,
+      "loss": 0.7205,
+      "step": 380
+    },
+    {
+      "epoch": 0.36581853096495437,
+      "grad_norm": 1.155974667611427,
+      "learning_rate": 7.2936660268714024e-06,
+      "loss": 0.8232,
+      "step": 381
+    },
+    {
+      "epoch": 0.36677868458953433,
+      "grad_norm": 1.2962147768859738,
+      "learning_rate": 7.312859884836853e-06,
+      "loss": 0.6994,
+      "step": 382
+    },
+    {
+      "epoch": 0.36773883821411424,
+      "grad_norm": 1.0432828620281644,
+      "learning_rate": 7.332053742802304e-06,
+      "loss": 0.7713,
+      "step": 383
+    },
+    {
+      "epoch": 0.3686989918386942,
+      "grad_norm": 1.6400306383833076,
+      "learning_rate": 7.351247600767755e-06,
+      "loss": 0.7455,
+      "step": 384
+    },
+    {
+      "epoch": 0.3696591454632741,
+      "grad_norm": 1.8687163253976649,
+      "learning_rate": 7.370441458733206e-06,
+      "loss": 0.8921,
+      "step": 385
+    },
+    {
+      "epoch": 0.37061929908785407,
+      "grad_norm": 1.15552226222218,
+      "learning_rate": 7.389635316698657e-06,
+      "loss": 0.8523,
+      "step": 386
+    },
+    {
+      "epoch": 0.371579452712434,
+      "grad_norm": 1.2197160624499277,
+      "learning_rate": 7.408829174664108e-06,
+      "loss": 0.8379,
+      "step": 387
+    },
+    {
+      "epoch": 0.37253960633701394,
+      "grad_norm": 0.9473852646791292,
+      "learning_rate": 7.4280230326295585e-06,
+      "loss": 0.6417,
+      "step": 388
+    },
+    {
+      "epoch": 0.37349975996159385,
+      "grad_norm": 1.0627377340766202,
+      "learning_rate": 7.447216890595011e-06,
+      "loss": 0.7168,
+      "step": 389
+    },
+    {
+      "epoch": 0.3744599135861738,
+      "grad_norm": 1.9613488950786884,
+      "learning_rate": 7.466410748560461e-06,
+      "loss": 0.832,
+      "step": 390
+    },
+    {
+      "epoch": 0.3754200672107537,
+      "grad_norm": 0.9768112116462052,
+      "learning_rate": 7.4856046065259125e-06,
+      "loss": 0.7722,
+      "step": 391
+    },
+    {
+      "epoch": 0.3763802208353337,
+      "grad_norm": 1.1088139902732483,
+      "learning_rate": 7.504798464491363e-06,
+      "loss": 0.7741,
+      "step": 392
+    },
+    {
+      "epoch": 0.3773403744599136,
+      "grad_norm": 1.1277644988362703,
+      "learning_rate": 7.523992322456814e-06,
+      "loss": 0.7548,
+      "step": 393
+    },
+    {
+      "epoch": 0.3783005280844935,
+      "grad_norm": 1.3490775794268282,
+      "learning_rate": 7.543186180422265e-06,
+      "loss": 0.7274,
+      "step": 394
+    },
+    {
+      "epoch": 0.37926068170907346,
+      "grad_norm": 1.2752102918651995,
+      "learning_rate": 7.562380038387716e-06,
+      "loss": 0.7634,
+      "step": 395
+    },
+    {
+      "epoch": 0.38022083533365336,
+      "grad_norm": 1.3248095204390724,
+      "learning_rate": 7.581573896353167e-06,
+      "loss": 0.7598,
+      "step": 396
+    },
+    {
+      "epoch": 0.3811809889582333,
+      "grad_norm": 1.4388469294151665,
+      "learning_rate": 7.600767754318619e-06,
+      "loss": 0.8855,
+      "step": 397
+    },
+    {
+      "epoch": 0.38214114258281323,
+      "grad_norm": 1.0323243137009235,
+      "learning_rate": 7.61996161228407e-06,
+      "loss": 0.6274,
+      "step": 398
+    },
+    {
+      "epoch": 0.3831012962073932,
+      "grad_norm": 1.4550500195070324,
+      "learning_rate": 7.63915547024952e-06,
+      "loss": 0.7161,
+      "step": 399
+    },
+    {
+      "epoch": 0.3840614498319731,
+      "grad_norm": 1.0927734198673136,
+      "learning_rate": 7.658349328214972e-06,
+      "loss": 0.6599,
+      "step": 400
+    },
+    {
+      "epoch": 0.38502160345655306,
+      "grad_norm": 1.0386436633896041,
+      "learning_rate": 7.677543186180423e-06,
+      "loss": 0.6468,
+      "step": 401
+    },
+    {
+      "epoch": 0.38598175708113297,
+      "grad_norm": 1.2385527094382942,
+      "learning_rate": 7.696737044145875e-06,
+      "loss": 0.6902,
+      "step": 402
+    },
+    {
+      "epoch": 0.38694191070571293,
+      "grad_norm": 1.128496898995166,
+      "learning_rate": 7.715930902111324e-06,
+      "loss": 0.8124,
+      "step": 403
+    },
+    {
+      "epoch": 0.38790206433029284,
+      "grad_norm": 0.8982281997977789,
+      "learning_rate": 7.735124760076776e-06,
+      "loss": 0.6971,
+      "step": 404
+    },
+    {
+      "epoch": 0.3888622179548728,
+      "grad_norm": 1.3087047484167935,
+      "learning_rate": 7.754318618042227e-06,
+      "loss": 0.7527,
+      "step": 405
+    },
+    {
+      "epoch": 0.3898223715794527,
+      "grad_norm": 1.0148715452035035,
+      "learning_rate": 7.773512476007678e-06,
+      "loss": 0.6074,
+      "step": 406
+    },
+    {
+      "epoch": 0.39078252520403267,
+      "grad_norm": 1.2181305573923982,
+      "learning_rate": 7.79270633397313e-06,
+      "loss": 0.7455,
+      "step": 407
+    },
+    {
+      "epoch": 0.3917426788286126,
+      "grad_norm": 1.257923950398552,
+      "learning_rate": 7.811900191938581e-06,
+      "loss": 0.7506,
+      "step": 408
+    },
+    {
+      "epoch": 0.3927028324531925,
+      "grad_norm": 0.9385542220228643,
+      "learning_rate": 7.83109404990403e-06,
+      "loss": 0.7203,
+      "step": 409
+    },
+    {
+      "epoch": 0.39366298607777245,
+      "grad_norm": 0.9489668803662427,
+      "learning_rate": 7.850287907869482e-06,
+      "loss": 0.6797,
+      "step": 410
+    },
+    {
+      "epoch": 0.39462313970235235,
+      "grad_norm": 1.138399536824701,
+      "learning_rate": 7.869481765834934e-06,
+      "loss": 0.721,
+      "step": 411
+    },
+    {
+      "epoch": 0.3955832933269323,
+      "grad_norm": 0.8761756089098698,
+      "learning_rate": 7.888675623800385e-06,
+      "loss": 0.7422,
+      "step": 412
+    },
+    {
+      "epoch": 0.3965434469515122,
+      "grad_norm": 1.136293201008876,
+      "learning_rate": 7.907869481765835e-06,
+      "loss": 0.7577,
+      "step": 413
+    },
+    {
+      "epoch": 0.3975036005760922,
+      "grad_norm": 0.9088338251669745,
+      "learning_rate": 7.927063339731288e-06,
+      "loss": 0.6826,
+      "step": 414
+    },
+    {
+      "epoch": 0.3984637542006721,
+      "grad_norm": 1.4226305549508294,
+      "learning_rate": 7.946257197696737e-06,
+      "loss": 0.7768,
+      "step": 415
+    },
+    {
+      "epoch": 0.39942390782525206,
+      "grad_norm": 1.6031042617246214,
+      "learning_rate": 7.965451055662189e-06,
+      "loss": 0.9425,
+      "step": 416
+    },
+    {
+      "epoch": 0.40038406144983196,
+      "grad_norm": 1.1182897641253295,
+      "learning_rate": 7.98464491362764e-06,
+      "loss": 0.744,
+      "step": 417
+    },
+    {
+      "epoch": 0.4013442150744119,
+      "grad_norm": 1.1624749434686374,
+      "learning_rate": 8.003838771593091e-06,
+      "loss": 0.8097,
+      "step": 418
+    },
+    {
+      "epoch": 0.40230436869899183,
+      "grad_norm": 1.0640000938896768,
+      "learning_rate": 8.023032629558541e-06,
+      "loss": 0.7217,
+      "step": 419
+    },
+    {
+      "epoch": 0.4032645223235718,
+      "grad_norm": 1.029486325121184,
+      "learning_rate": 8.042226487523992e-06,
+      "loss": 0.734,
+      "step": 420
+    },
+    {
+      "epoch": 0.4042246759481517,
+      "grad_norm": 1.4704604374496304,
+      "learning_rate": 8.061420345489444e-06,
+      "loss": 0.8084,
+      "step": 421
+    },
+    {
+      "epoch": 0.40518482957273166,
+      "grad_norm": 0.93094639130799,
+      "learning_rate": 8.080614203454895e-06,
+      "loss": 0.6201,
+      "step": 422
+    },
+    {
+      "epoch": 0.40614498319731157,
+      "grad_norm": 1.4281187881326085,
+      "learning_rate": 8.099808061420346e-06,
+      "loss": 0.8938,
+      "step": 423
+    },
+    {
+      "epoch": 0.4071051368218915,
+      "grad_norm": 1.147134853007333,
+      "learning_rate": 8.119001919385798e-06,
+      "loss": 0.7301,
+      "step": 424
+    },
+    {
+      "epoch": 0.40806529044647144,
+      "grad_norm": 1.0973175759148972,
+      "learning_rate": 8.138195777351249e-06,
+      "loss": 0.8135,
+      "step": 425
+    },
+    {
+      "epoch": 0.40902544407105135,
+      "grad_norm": 1.4733627401673446,
+      "learning_rate": 8.157389635316699e-06,
+      "loss": 0.9395,
+      "step": 426
+    },
+    {
+      "epoch": 0.4099855976956313,
+      "grad_norm": 0.9007463265785779,
+      "learning_rate": 8.176583493282152e-06,
+      "loss": 0.718,
+      "step": 427
+    },
+    {
+      "epoch": 0.4109457513202112,
+      "grad_norm": 0.9380859442585787,
+      "learning_rate": 8.195777351247601e-06,
+      "loss": 0.7025,
+      "step": 428
+    },
+    {
+      "epoch": 0.4119059049447912,
+      "grad_norm": 1.1911460339975852,
+      "learning_rate": 8.214971209213053e-06,
+      "loss": 0.8301,
+      "step": 429
+    },
+    {
+      "epoch": 0.4128660585693711,
+      "grad_norm": 1.12943551602526,
+      "learning_rate": 8.234165067178504e-06,
+      "loss": 0.8313,
+      "step": 430
+    },
+    {
+      "epoch": 0.41382621219395105,
+      "grad_norm": 1.185153963997797,
+      "learning_rate": 8.253358925143955e-06,
+      "loss": 0.7322,
+      "step": 431
+    },
+    {
+      "epoch": 0.41478636581853096,
+      "grad_norm": 0.9500772402083794,
+      "learning_rate": 8.272552783109405e-06,
+      "loss": 0.7539,
+      "step": 432
+    },
+    {
+      "epoch": 0.4157465194431109,
+      "grad_norm": 0.9685064250353801,
+      "learning_rate": 8.291746641074856e-06,
+      "loss": 0.7116,
+      "step": 433
+    },
+    {
+      "epoch": 0.4167066730676908,
+      "grad_norm": 1.439697940069654,
+      "learning_rate": 8.310940499040308e-06,
+      "loss": 0.8389,
+      "step": 434
+    },
+    {
+      "epoch": 0.4176668266922708,
+      "grad_norm": 1.650060931842286,
+      "learning_rate": 8.330134357005759e-06,
+      "loss": 0.8147,
+      "step": 435
+    },
+    {
+      "epoch": 0.4186269803168507,
+      "grad_norm": 1.5097188148620866,
+      "learning_rate": 8.34932821497121e-06,
+      "loss": 0.8671,
+      "step": 436
+    },
+    {
+      "epoch": 0.4195871339414306,
+      "grad_norm": 1.0491157820672372,
+      "learning_rate": 8.368522072936662e-06,
+      "loss": 0.7125,
+      "step": 437
+    },
+    {
+      "epoch": 0.42054728756601056,
+      "grad_norm": 1.2715868974375322,
+      "learning_rate": 8.387715930902111e-06,
+      "loss": 0.7865,
+      "step": 438
+    },
+    {
+      "epoch": 0.42150744119059047,
+      "grad_norm": 1.091876813620556,
+      "learning_rate": 8.406909788867563e-06,
+      "loss": 0.7577,
+      "step": 439
+    },
+    {
+      "epoch": 0.42246759481517043,
+      "grad_norm": 1.2597352342543193,
+      "learning_rate": 8.426103646833014e-06,
+      "loss": 0.823,
+      "step": 440
+    },
+    {
+      "epoch": 0.42342774843975034,
+      "grad_norm": 1.2434732628786342,
+      "learning_rate": 8.445297504798465e-06,
+      "loss": 0.7093,
+      "step": 441
+    },
+    {
+      "epoch": 0.4243879020643303,
+      "grad_norm": 1.1444923505918632,
+      "learning_rate": 8.464491362763915e-06,
+      "loss": 0.6886,
+      "step": 442
+    },
+    {
+      "epoch": 0.4253480556889102,
+      "grad_norm": 1.0409174548837277,
+      "learning_rate": 8.483685220729368e-06,
+      "loss": 0.7463,
+      "step": 443
+    },
+    {
+      "epoch": 0.42630820931349017,
+      "grad_norm": 1.2007743151913164,
+      "learning_rate": 8.502879078694818e-06,
+      "loss": 0.7989,
+      "step": 444
+    },
+    {
+      "epoch": 0.4272683629380701,
+      "grad_norm": 1.2718837493354382,
+      "learning_rate": 8.522072936660269e-06,
+      "loss": 0.7426,
+      "step": 445
+    },
+    {
+      "epoch": 0.42822851656265004,
+      "grad_norm": 1.2568901583053014,
+      "learning_rate": 8.54126679462572e-06,
+      "loss": 0.8255,
+      "step": 446
+    },
+    {
+      "epoch": 0.42918867018722995,
+      "grad_norm": 0.9969486622008054,
+      "learning_rate": 8.560460652591172e-06,
+      "loss": 0.6967,
+      "step": 447
+    },
+    {
+      "epoch": 0.4301488238118099,
+      "grad_norm": 1.1899514801382536,
+      "learning_rate": 8.579654510556623e-06,
+      "loss": 0.6996,
+      "step": 448
+    },
+    {
+      "epoch": 0.4311089774363898,
+      "grad_norm": 1.0152504780424239,
+      "learning_rate": 8.598848368522073e-06,
+      "loss": 0.7519,
+      "step": 449
+    },
+    {
+      "epoch": 0.4320691310609698,
+      "grad_norm": 1.2901043367604972,
+      "learning_rate": 8.618042226487526e-06,
+      "loss": 0.7782,
+      "step": 450
+    },
+    {
+      "epoch": 0.4330292846855497,
+      "grad_norm": 0.9785594955317184,
+      "learning_rate": 8.637236084452976e-06,
+      "loss": 0.6279,
+      "step": 451
+    },
+    {
+      "epoch": 0.4339894383101296,
+      "grad_norm": 0.9454490728083296,
+      "learning_rate": 8.656429942418427e-06,
+      "loss": 0.6849,
+      "step": 452
+    },
+    {
+      "epoch": 0.43494959193470956,
+      "grad_norm": 1.4515223195648956,
+      "learning_rate": 8.675623800383878e-06,
+      "loss": 0.8921,
+      "step": 453
+    },
+    {
+      "epoch": 0.43590974555928946,
+      "grad_norm": 1.1925663852839914,
+      "learning_rate": 8.69481765834933e-06,
+      "loss": 0.7893,
+      "step": 454
+    },
+    {
+      "epoch": 0.4368698991838694,
+      "grad_norm": 0.9817277941974228,
+      "learning_rate": 8.71401151631478e-06,
+      "loss": 0.7494,
+      "step": 455
+    },
+    {
+      "epoch": 0.43783005280844933,
+      "grad_norm": 1.073831518730329,
+      "learning_rate": 8.73320537428023e-06,
+      "loss": 0.6855,
+      "step": 456
+    },
+    {
+      "epoch": 0.4387902064330293,
+      "grad_norm": 1.1640819618236569,
+      "learning_rate": 8.752399232245682e-06,
+      "loss": 0.8602,
+      "step": 457
+    },
+    {
+      "epoch": 0.4397503600576092,
+      "grad_norm": 0.9732047167295669,
+      "learning_rate": 8.771593090211133e-06,
+      "loss": 0.7405,
+      "step": 458
+    },
+    {
+      "epoch": 0.44071051368218916,
+      "grad_norm": 1.4253504559694403,
+      "learning_rate": 8.790786948176585e-06,
+      "loss": 0.8094,
+      "step": 459
+    },
+    {
+      "epoch": 0.44167066730676907,
+      "grad_norm": 1.1120566176209299,
+      "learning_rate": 8.809980806142036e-06,
+      "loss": 0.8219,
+      "step": 460
+    },
+    {
+      "epoch": 0.44263082093134903,
+      "grad_norm": 1.1476137020908006,
+      "learning_rate": 8.829174664107486e-06,
+      "loss": 0.7989,
+      "step": 461
+    },
+    {
+      "epoch": 0.44359097455592894,
+      "grad_norm": 1.0376270729267838,
+      "learning_rate": 8.848368522072937e-06,
+      "loss": 0.7844,
+      "step": 462
+    },
+    {
+      "epoch": 0.4445511281805089,
+      "grad_norm": 1.88847392184355,
+      "learning_rate": 8.867562380038388e-06,
+      "loss": 0.8719,
+      "step": 463
+    },
+    {
+      "epoch": 0.4455112818050888,
+      "grad_norm": 1.1570205261545414,
+      "learning_rate": 8.88675623800384e-06,
+      "loss": 0.7337,
+      "step": 464
+    },
+    {
+      "epoch": 0.4464714354296688,
+      "grad_norm": 1.2953565641587041,
+      "learning_rate": 8.905950095969291e-06,
+      "loss": 0.6254,
+      "step": 465
+    },
+    {
+      "epoch": 0.4474315890542487,
+      "grad_norm": 1.1688420038847154,
+      "learning_rate": 8.925143953934742e-06,
+      "loss": 0.8725,
+      "step": 466
+    },
+    {
+      "epoch": 0.4483917426788286,
+      "grad_norm": 1.0335620781982393,
+      "learning_rate": 8.944337811900192e-06,
+      "loss": 0.7754,
+      "step": 467
+    },
+    {
+      "epoch": 0.44935189630340855,
+      "grad_norm": 0.8619939885192579,
+      "learning_rate": 8.963531669865643e-06,
+      "loss": 0.6979,
+      "step": 468
+    },
+    {
+      "epoch": 0.45031204992798846,
+      "grad_norm": 1.2976131274666975,
+      "learning_rate": 8.982725527831095e-06,
+      "loss": 0.667,
+      "step": 469
+    },
+    {
+      "epoch": 0.4512722035525684,
+      "grad_norm": 1.4021578605304392,
+      "learning_rate": 9.001919385796546e-06,
+      "loss": 0.8609,
+      "step": 470
+    },
+    {
+      "epoch": 0.4522323571771483,
+      "grad_norm": 1.2024029118089836,
+      "learning_rate": 9.021113243761996e-06,
+      "loss": 0.8361,
+      "step": 471
+    },
+    {
+      "epoch": 0.4531925108017283,
+      "grad_norm": 1.0460626952391077,
+      "learning_rate": 9.040307101727449e-06,
+      "loss": 0.7147,
+      "step": 472
+    },
+    {
+      "epoch": 0.4541526644263082,
+      "grad_norm": 1.2908116152032034,
+      "learning_rate": 9.0595009596929e-06,
+      "loss": 0.8197,
+      "step": 473
+    },
+    {
+      "epoch": 0.45511281805088816,
+      "grad_norm": 1.03318946346981,
+      "learning_rate": 9.07869481765835e-06,
+      "loss": 0.768,
+      "step": 474
+    },
+    {
+      "epoch": 0.45607297167546806,
+      "grad_norm": 1.3220638446805137,
+      "learning_rate": 9.097888675623801e-06,
+      "loss": 0.8007,
+      "step": 475
+    },
+    {
+      "epoch": 0.457033125300048,
+      "grad_norm": 1.1554417914733812,
+      "learning_rate": 9.117082533589252e-06,
+      "loss": 0.7636,
+      "step": 476
+    },
+    {
+      "epoch": 0.45799327892462793,
+      "grad_norm": 1.1310593331314387,
+      "learning_rate": 9.136276391554704e-06,
+      "loss": 0.6219,
+      "step": 477
+    },
+    {
+      "epoch": 0.4589534325492079,
+      "grad_norm": 1.0875448265190137,
+      "learning_rate": 9.155470249520153e-06,
+      "loss": 0.6831,
+      "step": 478
+    },
+    {
+      "epoch": 0.4599135861737878,
+      "grad_norm": 1.4031599853280303,
+      "learning_rate": 9.174664107485606e-06,
+      "loss": 0.7808,
+      "step": 479
+    },
+    {
+      "epoch": 0.46087373979836777,
+      "grad_norm": 1.391170940885926,
+      "learning_rate": 9.193857965451056e-06,
+      "loss": 0.7529,
+      "step": 480
+    },
+    {
+      "epoch": 0.46183389342294767,
+      "grad_norm": 1.0087271205949113,
+      "learning_rate": 9.213051823416507e-06,
+      "loss": 0.6919,
+      "step": 481
+    },
+    {
+      "epoch": 0.4627940470475276,
+      "grad_norm": 0.9830620697275546,
+      "learning_rate": 9.232245681381959e-06,
+      "loss": 0.6466,
+      "step": 482
+    },
+    {
+      "epoch": 0.46375420067210754,
+      "grad_norm": 1.1083503739334748,
+      "learning_rate": 9.25143953934741e-06,
+      "loss": 0.6824,
+      "step": 483
+    },
+    {
+      "epoch": 0.46471435429668745,
+      "grad_norm": 1.290594688103715,
+      "learning_rate": 9.27063339731286e-06,
+      "loss": 0.7947,
+      "step": 484
+    },
+    {
+      "epoch": 0.4656745079212674,
+      "grad_norm": 1.0041716569519412,
+      "learning_rate": 9.289827255278311e-06,
+      "loss": 0.6783,
+      "step": 485
+    },
+    {
+      "epoch": 0.4666346615458473,
+      "grad_norm": 1.1598686934846796,
+      "learning_rate": 9.309021113243763e-06,
+      "loss": 0.6613,
+      "step": 486
+    },
+    {
+      "epoch": 0.4675948151704273,
+      "grad_norm": 1.260111175387368,
+      "learning_rate": 9.328214971209214e-06,
+      "loss": 0.7378,
+      "step": 487
+    },
+    {
+      "epoch": 0.4685549687950072,
+      "grad_norm": 0.9393424317838525,
+      "learning_rate": 9.347408829174665e-06,
+      "loss": 0.8005,
+      "step": 488
+    },
+    {
+      "epoch": 0.46951512241958715,
+      "grad_norm": 1.1210853185169674,
+      "learning_rate": 9.366602687140117e-06,
+      "loss": 0.6807,
+      "step": 489
+    },
+    {
+      "epoch": 0.47047527604416706,
+      "grad_norm": 1.4438230017220477,
+      "learning_rate": 9.385796545105566e-06,
+      "loss": 0.7997,
+      "step": 490
+    },
+    {
+      "epoch": 0.471435429668747,
+      "grad_norm": 1.3277569402674447,
+      "learning_rate": 9.404990403071018e-06,
+      "loss": 0.7416,
+      "step": 491
+    },
+    {
+      "epoch": 0.4723955832933269,
+      "grad_norm": 1.2432681770756384,
+      "learning_rate": 9.424184261036469e-06,
+      "loss": 0.7542,
+      "step": 492
+    },
+    {
+      "epoch": 0.4733557369179069,
+      "grad_norm": 0.9683846009176716,
+      "learning_rate": 9.44337811900192e-06,
+      "loss": 0.6364,
+      "step": 493
+    },
+    {
+      "epoch": 0.4743158905424868,
+      "grad_norm": 1.0629872575186827,
+      "learning_rate": 9.46257197696737e-06,
+      "loss": 0.7682,
+      "step": 494
+    },
+    {
+      "epoch": 0.47527604416706676,
+      "grad_norm": 1.0658891130678316,
+      "learning_rate": 9.481765834932823e-06,
+      "loss": 0.7502,
+      "step": 495
+    },
+    {
+      "epoch": 0.47623619779164666,
+      "grad_norm": 1.3363058386178244,
+      "learning_rate": 9.500959692898273e-06,
+      "loss": 0.8112,
+      "step": 496
+    },
+    {
+      "epoch": 0.47719635141622657,
+      "grad_norm": 1.0265023768250465,
+      "learning_rate": 9.520153550863724e-06,
+      "loss": 0.7149,
+      "step": 497
+    },
+    {
+      "epoch": 0.47815650504080653,
+      "grad_norm": 1.2889831472177289,
+      "learning_rate": 9.539347408829175e-06,
+      "loss": 0.7936,
+      "step": 498
+    },
+    {
+      "epoch": 0.47911665866538644,
+      "grad_norm": 1.250136499572121,
+      "learning_rate": 9.558541266794627e-06,
+      "loss": 0.6048,
+      "step": 499
+    },
+    {
+      "epoch": 0.4800768122899664,
+      "grad_norm": 1.3111050762625585,
+      "learning_rate": 9.577735124760078e-06,
+      "loss": 0.728,
+      "step": 500
+    },
+    {
+      "epoch": 0.4810369659145463,
+      "grad_norm": 1.7608381448641053,
+      "learning_rate": 9.59692898272553e-06,
+      "loss": 0.7994,
+      "step": 501
+    },
+    {
+      "epoch": 0.4819971195391263,
+      "grad_norm": 1.2290182303499815,
+      "learning_rate": 9.61612284069098e-06,
+      "loss": 0.8003,
+      "step": 502
+    },
+    {
+      "epoch": 0.4829572731637062,
+      "grad_norm": 1.6135633512832115,
+      "learning_rate": 9.63531669865643e-06,
+      "loss": 0.838,
+      "step": 503
+    },
+    {
+      "epoch": 0.48391742678828614,
+      "grad_norm": 1.164478249652597,
+      "learning_rate": 9.654510556621882e-06,
+      "loss": 0.7259,
+      "step": 504
+    },
+    {
+      "epoch": 0.48487758041286605,
+      "grad_norm": 0.9993561066591795,
+      "learning_rate": 9.673704414587333e-06,
+      "loss": 0.7748,
+      "step": 505
+    },
+    {
+      "epoch": 0.485837734037446,
+      "grad_norm": 0.8358295488869283,
+      "learning_rate": 9.692898272552784e-06,
+      "loss": 0.6693,
+      "step": 506
+    },
+    {
+      "epoch": 0.4867978876620259,
+      "grad_norm": 1.1106122084032795,
+      "learning_rate": 9.712092130518234e-06,
+      "loss": 0.721,
+      "step": 507
+    },
+    {
+      "epoch": 0.4877580412866059,
+      "grad_norm": 1.487188285181112,
+      "learning_rate": 9.731285988483687e-06,
+      "loss": 0.7509,
+      "step": 508
+    },
+    {
+      "epoch": 0.4887181949111858,
+      "grad_norm": 1.1402411926952243,
+      "learning_rate": 9.750479846449137e-06,
+      "loss": 0.6647,
+      "step": 509
+    },
+    {
+      "epoch": 0.4896783485357657,
+      "grad_norm": 1.0367107322127078,
+      "learning_rate": 9.769673704414588e-06,
+      "loss": 0.7887,
+      "step": 510
+    },
+    {
+      "epoch": 0.49063850216034566,
+      "grad_norm": 1.3987993604220361,
+      "learning_rate": 9.78886756238004e-06,
+      "loss": 0.7736,
+      "step": 511
+    },
+    {
+      "epoch": 0.49159865578492556,
+      "grad_norm": 1.259632180608351,
+      "learning_rate": 9.80806142034549e-06,
+      "loss": 0.7329,
+      "step": 512
+    },
+    {
+      "epoch": 0.4925588094095055,
+      "grad_norm": 1.1142094683577137,
+      "learning_rate": 9.82725527831094e-06,
+      "loss": 0.6632,
+      "step": 513
+    },
+    {
+      "epoch": 0.49351896303408543,
+      "grad_norm": 1.0929712059339454,
+      "learning_rate": 9.846449136276392e-06,
+      "loss": 0.7179,
+      "step": 514
+    },
+    {
+      "epoch": 0.4944791166586654,
+      "grad_norm": 1.306135215088198,
+      "learning_rate": 9.865642994241843e-06,
+      "loss": 0.6585,
+      "step": 515
+    },
+    {
+      "epoch": 0.4954392702832453,
+      "grad_norm": 1.1252208319095651,
+      "learning_rate": 9.884836852207294e-06,
+      "loss": 0.8159,
+      "step": 516
+    },
+    {
+      "epoch": 0.49639942390782527,
+      "grad_norm": 1.486805087876587,
+      "learning_rate": 9.904030710172746e-06,
+      "loss": 0.8046,
+      "step": 517
+    },
+    {
+      "epoch": 0.4973595775324052,
+      "grad_norm": 1.1393464745618724,
+      "learning_rate": 9.923224568138197e-06,
+      "loss": 0.7898,
+      "step": 518
+    },
+    {
+      "epoch": 0.49831973115698514,
+      "grad_norm": 1.224267542983065,
+      "learning_rate": 9.942418426103647e-06,
+      "loss": 0.7973,
+      "step": 519
+    },
+    {
+      "epoch": 0.49927988478156504,
+      "grad_norm": 1.1859907199010784,
+      "learning_rate": 9.961612284069098e-06,
+      "loss": 0.8521,
+      "step": 520
+    },
+    {
+      "epoch": 0.500240038406145,
+      "grad_norm": 1.5259121980424533,
+      "learning_rate": 9.98080614203455e-06,
+      "loss": 0.8218,
+      "step": 521
+    },
+    {
+      "epoch": 0.501200192030725,
+      "grad_norm": 1.2072523535698836,
+      "learning_rate": 1e-05,
+      "loss": 0.7709,
+      "step": 522
+    },
+    {
+      "epoch": 0.5021603456553049,
+      "grad_norm": 1.2171256402748378,
+      "learning_rate": 9.999998877776725e-06,
+      "loss": 0.652,
+      "step": 523
+    },
+    {
+      "epoch": 0.5031204992798848,
+      "grad_norm": 1.049515981094851,
+      "learning_rate": 9.999995511107399e-06,
+      "loss": 0.7062,
+      "step": 524
+    },
+    {
+      "epoch": 0.5040806529044647,
+      "grad_norm": 1.0998316431712438,
+      "learning_rate": 9.999989899993536e-06,
+      "loss": 0.7867,
+      "step": 525
+    },
+    {
+      "epoch": 0.5050408065290446,
+      "grad_norm": 1.1011145838187344,
+      "learning_rate": 9.999982044437652e-06,
+      "loss": 0.6948,
+      "step": 526
+    },
+    {
+      "epoch": 0.5060009601536246,
+      "grad_norm": 1.0336696003692087,
+      "learning_rate": 9.999971944443276e-06,
+      "loss": 0.702,
+      "step": 527
+    },
+    {
+      "epoch": 0.5069611137782045,
+      "grad_norm": 1.1680220201586808,
+      "learning_rate": 9.999959600014942e-06,
+      "loss": 0.7674,
+      "step": 528
+    },
+    {
+      "epoch": 0.5079212674027844,
+      "grad_norm": 0.9829065057909755,
+      "learning_rate": 9.999945011158189e-06,
+      "loss": 0.7185,
+      "step": 529
+    },
+    {
+      "epoch": 0.5088814210273643,
+      "grad_norm": 1.0062085732737889,
+      "learning_rate": 9.999928177879567e-06,
+      "loss": 0.7118,
+      "step": 530
+    },
+    {
+      "epoch": 0.5098415746519444,
+      "grad_norm": 1.399248665460084,
+      "learning_rate": 9.999909100186634e-06,
+      "loss": 0.8042,
+      "step": 531
+    },
+    {
+      "epoch": 0.5108017282765243,
+      "grad_norm": 0.9818054618777152,
+      "learning_rate": 9.999887778087948e-06,
+      "loss": 0.7058,
+      "step": 532
+    },
+    {
+      "epoch": 0.5117618819011042,
+      "grad_norm": 1.165173517671421,
+      "learning_rate": 9.999864211593088e-06,
+      "loss": 0.7926,
+      "step": 533
+    },
+    {
+      "epoch": 0.5127220355256841,
+      "grad_norm": 1.1505524949085641,
+      "learning_rate": 9.999838400712627e-06,
+      "loss": 0.8351,
+      "step": 534
+    },
+    {
+      "epoch": 0.5136821891502641,
+      "grad_norm": 1.1208313391537508,
+      "learning_rate": 9.999810345458153e-06,
+      "loss": 0.605,
+      "step": 535
+    },
+    {
+      "epoch": 0.514642342774844,
+      "grad_norm": 1.1968339780644133,
+      "learning_rate": 9.999780045842262e-06,
+      "loss": 0.644,
+      "step": 536
+    },
+    {
+      "epoch": 0.5156024963994239,
+      "grad_norm": 1.0693441645105062,
+      "learning_rate": 9.999747501878551e-06,
+      "loss": 0.6703,
+      "step": 537
+    },
+    {
+      "epoch": 0.5165626500240038,
+      "grad_norm": 1.2133540246571193,
+      "learning_rate": 9.999712713581634e-06,
+      "loss": 0.8123,
+      "step": 538
+    },
+    {
+      "epoch": 0.5175228036485837,
+      "grad_norm": 1.0453334308646136,
+      "learning_rate": 9.99967568096712e-06,
+      "loss": 0.6976,
+      "step": 539
+    },
+    {
+      "epoch": 0.5184829572731637,
+      "grad_norm": 1.3264475966595732,
+      "learning_rate": 9.999636404051638e-06,
+      "loss": 0.7837,
+      "step": 540
+    },
+    {
+      "epoch": 0.5194431108977436,
+      "grad_norm": 1.0861707447698705,
+      "learning_rate": 9.999594882852817e-06,
+      "loss": 0.7799,
+      "step": 541
+    },
+    {
+      "epoch": 0.5204032645223235,
+      "grad_norm": 1.2946964966740842,
+      "learning_rate": 9.999551117389298e-06,
+      "loss": 0.6334,
+      "step": 542
+    },
+    {
+      "epoch": 0.5213634181469035,
+      "grad_norm": 0.929657847384208,
+      "learning_rate": 9.999505107680722e-06,
+      "loss": 0.8083,
+      "step": 543
+    },
+    {
+      "epoch": 0.5223235717714835,
+      "grad_norm": 1.1401440990502336,
+      "learning_rate": 9.999456853747744e-06,
+      "loss": 0.7132,
+      "step": 544
+    },
+    {
+      "epoch": 0.5232837253960634,
+      "grad_norm": 1.0340277545375338,
+      "learning_rate": 9.999406355612029e-06,
+      "loss": 0.7124,
+      "step": 545
+    },
+    {
+      "epoch": 0.5242438790206433,
+      "grad_norm": 1.247007016370199,
+      "learning_rate": 9.99935361329624e-06,
+      "loss": 0.7065,
+      "step": 546
+    },
+    {
+      "epoch": 0.5252040326452232,
+      "grad_norm": 0.9998755107014989,
+      "learning_rate": 9.999298626824053e-06,
+      "loss": 0.7451,
+      "step": 547
+    },
+    {
+      "epoch": 0.5261641862698032,
+      "grad_norm": 1.2332553479270054,
+      "learning_rate": 9.999241396220152e-06,
+      "loss": 0.7216,
+      "step": 548
+    },
+    {
+      "epoch": 0.5271243398943831,
+      "grad_norm": 1.2155641200696106,
+      "learning_rate": 9.999181921510229e-06,
+      "loss": 0.782,
+      "step": 549
+    },
+    {
+      "epoch": 0.528084493518963,
+      "grad_norm": 1.51168636775896,
+      "learning_rate": 9.999120202720978e-06,
+      "loss": 0.9307,
+      "step": 550
+    },
+    {
+      "epoch": 0.5290446471435429,
+      "grad_norm": 1.0359215075770465,
+      "learning_rate": 9.999056239880107e-06,
+      "loss": 0.6104,
+      "step": 551
+    },
+    {
+      "epoch": 0.530004800768123,
+      "grad_norm": 1.0464957447369558,
+      "learning_rate": 9.998990033016327e-06,
+      "loss": 0.6045,
+      "step": 552
+    },
+    {
+      "epoch": 0.5309649543927029,
+      "grad_norm": 1.1114456620613888,
+      "learning_rate": 9.998921582159357e-06,
+      "loss": 0.8023,
+      "step": 553
+    },
+    {
+      "epoch": 0.5319251080172828,
+      "grad_norm": 1.5494709814113228,
+      "learning_rate": 9.998850887339923e-06,
+      "loss": 0.7696,
+      "step": 554
+    },
+    {
+      "epoch": 0.5328852616418627,
+      "grad_norm": 1.4890090358538963,
+      "learning_rate": 9.998777948589762e-06,
+      "loss": 0.7569,
+      "step": 555
+    },
+    {
+      "epoch": 0.5338454152664426,
+      "grad_norm": 0.8865139886047781,
+      "learning_rate": 9.998702765941613e-06,
+      "loss": 0.7251,
+      "step": 556
+    },
+    {
+      "epoch": 0.5348055688910226,
+      "grad_norm": 1.3058829182513438,
+      "learning_rate": 9.998625339429227e-06,
+      "loss": 0.7223,
+      "step": 557
+    },
+    {
+      "epoch": 0.5357657225156025,
+      "grad_norm": 1.276478715271317,
+      "learning_rate": 9.998545669087357e-06,
+      "loss": 0.7673,
+      "step": 558
+    },
+    {
+      "epoch": 0.5367258761401824,
+      "grad_norm": 1.1676751331206674,
+      "learning_rate": 9.998463754951767e-06,
+      "loss": 0.7617,
+      "step": 559
+    },
+    {
+      "epoch": 0.5376860297647623,
+      "grad_norm": 1.2123803522040177,
+      "learning_rate": 9.998379597059228e-06,
+      "loss": 0.7129,
+      "step": 560
+    },
+    {
+      "epoch": 0.5386461833893423,
+      "grad_norm": 1.3814639588806619,
+      "learning_rate": 9.99829319544752e-06,
+      "loss": 0.8845,
+      "step": 561
+    },
+    {
+      "epoch": 0.5396063370139222,
+      "grad_norm": 1.11055749603962,
+      "learning_rate": 9.998204550155424e-06,
+      "loss": 0.7557,
+      "step": 562
+    },
+    {
+      "epoch": 0.5405664906385022,
+      "grad_norm": 1.0674759443075517,
+      "learning_rate": 9.998113661222732e-06,
+      "loss": 0.7595,
+      "step": 563
+    },
+    {
+      "epoch": 0.5415266442630821,
+      "grad_norm": 0.994763687652783,
+      "learning_rate": 9.998020528690247e-06,
+      "loss": 0.814,
+      "step": 564
+    },
+    {
+      "epoch": 0.5424867978876621,
+      "grad_norm": 1.0417008849188154,
+      "learning_rate": 9.99792515259977e-06,
+      "loss": 0.7379,
+      "step": 565
+    },
+    {
+      "epoch": 0.543446951512242,
+      "grad_norm": 1.1478987486458458,
+      "learning_rate": 9.997827532994118e-06,
+      "loss": 0.7426,
+      "step": 566
+    },
+    {
+      "epoch": 0.5444071051368219,
+      "grad_norm": 1.1565190817285003,
+      "learning_rate": 9.99772766991711e-06,
+      "loss": 0.7203,
+      "step": 567
+    },
+    {
+      "epoch": 0.5453672587614018,
+      "grad_norm": 1.1192363371512206,
+      "learning_rate": 9.997625563413574e-06,
+      "loss": 0.7739,
+      "step": 568
+    },
+    {
+      "epoch": 0.5463274123859817,
+      "grad_norm": 0.9567002043036419,
+      "learning_rate": 9.997521213529345e-06,
+      "loss": 0.705,
+      "step": 569
+    },
+    {
+      "epoch": 0.5472875660105617,
+      "grad_norm": 1.0897891589405941,
+      "learning_rate": 9.997414620311262e-06,
+      "loss": 0.6637,
+      "step": 570
+    },
+    {
+      "epoch": 0.5482477196351416,
+      "grad_norm": 0.9563438824899532,
+      "learning_rate": 9.997305783807176e-06,
+      "loss": 0.7579,
+      "step": 571
+    },
+    {
+      "epoch": 0.5492078732597215,
+      "grad_norm": 1.1554010195538116,
+      "learning_rate": 9.997194704065942e-06,
+      "loss": 0.8239,
+      "step": 572
+    },
+    {
+      "epoch": 0.5501680268843014,
+      "grad_norm": 1.2004353682200677,
+      "learning_rate": 9.997081381137424e-06,
+      "loss": 0.8628,
+      "step": 573
+    },
+    {
+      "epoch": 0.5511281805088815,
+      "grad_norm": 1.0220163262467359,
+      "learning_rate": 9.996965815072489e-06,
+      "loss": 0.6323,
+      "step": 574
+    },
+    {
+      "epoch": 0.5520883341334614,
+      "grad_norm": 1.6768057570791945,
+      "learning_rate": 9.996848005923012e-06,
+      "loss": 0.7965,
+      "step": 575
+    },
+    {
+      "epoch": 0.5530484877580413,
+      "grad_norm": 1.1602600016134357,
+      "learning_rate": 9.996727953741881e-06,
+      "loss": 0.6791,
+      "step": 576
+    },
+    {
+      "epoch": 0.5540086413826212,
+      "grad_norm": 0.9931168407930752,
+      "learning_rate": 9.996605658582982e-06,
+      "loss": 0.7462,
+      "step": 577
+    },
+    {
+      "epoch": 0.5549687950072012,
+      "grad_norm": 1.4796154459879218,
+      "learning_rate": 9.996481120501214e-06,
+      "loss": 0.8238,
+      "step": 578
+    },
+    {
+      "epoch": 0.5559289486317811,
+      "grad_norm": 1.381992426279988,
+      "learning_rate": 9.996354339552482e-06,
+      "loss": 0.7427,
+      "step": 579
+    },
+    {
+      "epoch": 0.556889102256361,
+      "grad_norm": 1.000051907716779,
+      "learning_rate": 9.996225315793692e-06,
+      "loss": 0.6611,
+      "step": 580
+    },
+    {
+      "epoch": 0.5578492558809409,
+      "grad_norm": 1.8016378676915354,
+      "learning_rate": 9.996094049282768e-06,
+      "loss": 0.7868,
+      "step": 581
+    },
+    {
+      "epoch": 0.5588094095055209,
+      "grad_norm": 1.1278980971599617,
+      "learning_rate": 9.995960540078627e-06,
+      "loss": 0.7274,
+      "step": 582
+    },
+    {
+      "epoch": 0.5597695631301008,
+      "grad_norm": 1.7484961964443118,
+      "learning_rate": 9.995824788241206e-06,
+      "loss": 0.9602,
+      "step": 583
+    },
+    {
+      "epoch": 0.5607297167546808,
+      "grad_norm": 1.145929824155346,
+      "learning_rate": 9.995686793831438e-06,
+      "loss": 0.7076,
+      "step": 584
+    },
+    {
+      "epoch": 0.5616898703792607,
+      "grad_norm": 1.1004359314897234,
+      "learning_rate": 9.995546556911271e-06,
+      "loss": 0.681,
+      "step": 585
+    },
+    {
+      "epoch": 0.5626500240038406,
+      "grad_norm": 1.0748063299411916,
+      "learning_rate": 9.995404077543654e-06,
+      "loss": 0.7148,
+      "step": 586
+    },
+    {
+      "epoch": 0.5636101776284206,
+      "grad_norm": 0.821838615755815,
+      "learning_rate": 9.995259355792544e-06,
+      "loss": 0.6614,
+      "step": 587
+    },
+    {
+      "epoch": 0.5645703312530005,
+      "grad_norm": 1.2133270777619678,
+      "learning_rate": 9.995112391722905e-06,
+      "loss": 0.7483,
+      "step": 588
+    },
+    {
+      "epoch": 0.5655304848775804,
+      "grad_norm": 1.2328883410395755,
+      "learning_rate": 9.994963185400708e-06,
+      "loss": 0.8195,
+      "step": 589
+    },
+    {
+      "epoch": 0.5664906385021603,
+      "grad_norm": 1.0076087642529332,
+      "learning_rate": 9.994811736892931e-06,
+      "loss": 0.7355,
+      "step": 590
+    },
+    {
+      "epoch": 0.5674507921267403,
+      "grad_norm": 1.2919954902291888,
+      "learning_rate": 9.994658046267556e-06,
+      "loss": 0.833,
+      "step": 591
+    },
+    {
+      "epoch": 0.5684109457513202,
+      "grad_norm": 1.0294781538999418,
+      "learning_rate": 9.994502113593575e-06,
+      "loss": 0.8137,
+      "step": 592
+    },
+    {
+      "epoch": 0.5693710993759001,
+      "grad_norm": 1.2539612701534901,
+      "learning_rate": 9.994343938940983e-06,
+      "loss": 0.6718,
+      "step": 593
+    },
+    {
+      "epoch": 0.57033125300048,
+      "grad_norm": 1.0280669880358662,
+      "learning_rate": 9.994183522380784e-06,
+      "loss": 0.7441,
+      "step": 594
+    },
+    {
+      "epoch": 0.5712914066250601,
+      "grad_norm": 0.9019541853420437,
+      "learning_rate": 9.994020863984986e-06,
+      "loss": 0.6743,
+      "step": 595
+    },
+    {
+      "epoch": 0.57225156024964,
+      "grad_norm": 1.058143387523447,
+      "learning_rate": 9.993855963826605e-06,
+      "loss": 0.6711,
+      "step": 596
+    },
+    {
+      "epoch": 0.5732117138742199,
+      "grad_norm": 1.0307534340163422,
+      "learning_rate": 9.993688821979664e-06,
+      "loss": 0.6419,
+      "step": 597
+    },
+    {
+      "epoch": 0.5741718674987998,
+      "grad_norm": 1.0157479655127266,
+      "learning_rate": 9.993519438519191e-06,
+      "loss": 0.7381,
+      "step": 598
+    },
+    {
+      "epoch": 0.5751320211233797,
+      "grad_norm": 1.5360923531404511,
+      "learning_rate": 9.993347813521219e-06,
+      "loss": 0.8652,
+      "step": 599
+    },
+    {
+      "epoch": 0.5760921747479597,
+      "grad_norm": 0.8794359882129259,
+      "learning_rate": 9.993173947062789e-06,
+      "loss": 0.6855,
+      "step": 600
+    },
+    {
+      "epoch": 0.5770523283725396,
+      "grad_norm": 1.2191476435446476,
+      "learning_rate": 9.992997839221947e-06,
+      "loss": 0.7825,
+      "step": 601
+    },
+    {
+      "epoch": 0.5780124819971195,
+      "grad_norm": 1.0797985506718768,
+      "learning_rate": 9.99281949007775e-06,
+      "loss": 0.8174,
+      "step": 602
+    },
+    {
+      "epoch": 0.5789726356216994,
+      "grad_norm": 1.397720342441156,
+      "learning_rate": 9.992638899710252e-06,
+      "loss": 0.8272,
+      "step": 603
+    },
+    {
+      "epoch": 0.5799327892462794,
+      "grad_norm": 0.9830573117461331,
+      "learning_rate": 9.99245606820052e-06,
+      "loss": 0.6114,
+      "step": 604
+    },
+    {
+      "epoch": 0.5808929428708594,
+      "grad_norm": 1.0176480970254405,
+      "learning_rate": 9.992270995630628e-06,
+      "loss": 0.7211,
+      "step": 605
+    },
+    {
+      "epoch": 0.5818530964954393,
+      "grad_norm": 1.2533849123137955,
+      "learning_rate": 9.992083682083649e-06,
+      "loss": 0.8221,
+      "step": 606
+    },
+    {
+      "epoch": 0.5828132501200192,
+      "grad_norm": 1.004562598726377,
+      "learning_rate": 9.991894127643666e-06,
+      "loss": 0.6359,
+      "step": 607
+    },
+    {
+      "epoch": 0.5837734037445992,
+      "grad_norm": 1.3104279757416193,
+      "learning_rate": 9.991702332395771e-06,
+      "loss": 0.9004,
+      "step": 608
+    },
+    {
+      "epoch": 0.5847335573691791,
+      "grad_norm": 1.1907424868300012,
+      "learning_rate": 9.991508296426057e-06,
+      "loss": 0.639,
+      "step": 609
+    },
+    {
+      "epoch": 0.585693710993759,
+      "grad_norm": 1.3716403450449266,
+      "learning_rate": 9.991312019821626e-06,
+      "loss": 0.8258,
+      "step": 610
+    },
+    {
+      "epoch": 0.5866538646183389,
+      "grad_norm": 0.9895375215410548,
+      "learning_rate": 9.991113502670583e-06,
+      "loss": 0.6116,
+      "step": 611
+    },
+    {
+      "epoch": 0.5876140182429188,
+      "grad_norm": 0.9266557576594376,
+      "learning_rate": 9.99091274506204e-06,
+      "loss": 0.6317,
+      "step": 612
+    },
+    {
+      "epoch": 0.5885741718674988,
+      "grad_norm": 0.9225656836771144,
+      "learning_rate": 9.990709747086118e-06,
+      "loss": 0.6896,
+      "step": 613
+    },
+    {
+      "epoch": 0.5895343254920787,
+      "grad_norm": 0.9279076070689715,
+      "learning_rate": 9.990504508833936e-06,
+      "loss": 0.6593,
+      "step": 614
+    },
+    {
+      "epoch": 0.5904944791166586,
+      "grad_norm": 1.3615841081527988,
+      "learning_rate": 9.990297030397626e-06,
+      "loss": 0.9277,
+      "step": 615
+    },
+    {
+      "epoch": 0.5914546327412386,
+      "grad_norm": 1.0888009181578198,
+      "learning_rate": 9.990087311870322e-06,
+      "loss": 0.6918,
+      "step": 616
+    },
+    {
+      "epoch": 0.5924147863658186,
+      "grad_norm": 1.2019128759256583,
+      "learning_rate": 9.989875353346164e-06,
+      "loss": 0.7467,
+      "step": 617
+    },
+    {
+      "epoch": 0.5933749399903985,
+      "grad_norm": 1.1162187027509973,
+      "learning_rate": 9.9896611549203e-06,
+      "loss": 0.7803,
+      "step": 618
+    },
+    {
+      "epoch": 0.5943350936149784,
+      "grad_norm": 1.0995846179986166,
+      "learning_rate": 9.98944471668888e-06,
+      "loss": 0.6535,
+      "step": 619
+    },
+    {
+      "epoch": 0.5952952472395583,
+      "grad_norm": 1.2874007059819845,
+      "learning_rate": 9.98922603874906e-06,
+      "loss": 0.8015,
+      "step": 620
+    },
+    {
+      "epoch": 0.5962554008641383,
+      "grad_norm": 1.1709057648541077,
+      "learning_rate": 9.989005121199004e-06,
+      "loss": 0.8659,
+      "step": 621
+    },
+    {
+      "epoch": 0.5972155544887182,
+      "grad_norm": 1.1298017335305068,
+      "learning_rate": 9.98878196413788e-06,
+      "loss": 0.8084,
+      "step": 622
+    },
+    {
+      "epoch": 0.5981757081132981,
+      "grad_norm": 1.1273397470795112,
+      "learning_rate": 9.988556567665858e-06,
+      "loss": 0.6616,
+      "step": 623
+    },
+    {
+      "epoch": 0.599135861737878,
+      "grad_norm": 1.1287215761694498,
+      "learning_rate": 9.988328931884116e-06,
+      "loss": 0.8314,
+      "step": 624
+    },
+    {
+      "epoch": 0.600096015362458,
+      "grad_norm": 1.2200338008797798,
+      "learning_rate": 9.98809905689484e-06,
+      "loss": 0.8232,
+      "step": 625
+    },
+    {
+      "epoch": 0.601056168987038,
+      "grad_norm": 1.0124851725821313,
+      "learning_rate": 9.987866942801216e-06,
+      "loss": 0.6169,
+      "step": 626
+    },
+    {
+      "epoch": 0.6020163226116179,
+      "grad_norm": 0.8272113332478752,
+      "learning_rate": 9.987632589707441e-06,
+      "loss": 0.6627,
+      "step": 627
+    },
+    {
+      "epoch": 0.6029764762361978,
+      "grad_norm": 0.9571583436346156,
+      "learning_rate": 9.98739599771871e-06,
+      "loss": 0.7432,
+      "step": 628
+    },
+    {
+      "epoch": 0.6039366298607777,
+      "grad_norm": 1.3672229056284988,
+      "learning_rate": 9.987157166941228e-06,
+      "loss": 0.8827,
+      "step": 629
+    },
+    {
+      "epoch": 0.6048967834853577,
+      "grad_norm": 1.174360049598899,
+      "learning_rate": 9.986916097482204e-06,
+      "loss": 0.7367,
+      "step": 630
+    },
+    {
+      "epoch": 0.6058569371099376,
+      "grad_norm": 0.8517725810532072,
+      "learning_rate": 9.98667278944985e-06,
+      "loss": 0.5979,
+      "step": 631
+    },
+    {
+      "epoch": 0.6068170907345175,
+      "grad_norm": 1.1119512062132852,
+      "learning_rate": 9.986427242953387e-06,
+      "loss": 0.7079,
+      "step": 632
+    },
+    {
+      "epoch": 0.6077772443590974,
+      "grad_norm": 1.6624527590685678,
+      "learning_rate": 9.986179458103036e-06,
+      "loss": 0.8425,
+      "step": 633
+    },
+    {
+      "epoch": 0.6087373979836774,
+      "grad_norm": 0.8868536609149565,
+      "learning_rate": 9.985929435010024e-06,
+      "loss": 0.6338,
+      "step": 634
+    },
+    {
+      "epoch": 0.6096975516082573,
+      "grad_norm": 1.062767851898318,
+      "learning_rate": 9.985677173786587e-06,
+      "loss": 0.6638,
+      "step": 635
+    },
+    {
+      "epoch": 0.6106577052328372,
+      "grad_norm": 1.130431234420582,
+      "learning_rate": 9.98542267454596e-06,
+      "loss": 0.7227,
+      "step": 636
+    },
+    {
+      "epoch": 0.6116178588574172,
+      "grad_norm": 1.0969498581452322,
+      "learning_rate": 9.985165937402386e-06,
+      "loss": 0.7648,
+      "step": 637
+    },
+    {
+      "epoch": 0.6125780124819972,
+      "grad_norm": 1.3792237075028597,
+      "learning_rate": 9.98490696247111e-06,
+      "loss": 0.8187,
+      "step": 638
+    },
+    {
+      "epoch": 0.6135381661065771,
+      "grad_norm": 1.60062571530598,
+      "learning_rate": 9.984645749868386e-06,
+      "loss": 0.7237,
+      "step": 639
+    },
+    {
+      "epoch": 0.614498319731157,
+      "grad_norm": 1.0548834013904989,
+      "learning_rate": 9.984382299711466e-06,
+      "loss": 0.7315,
+      "step": 640
+    },
+    {
+      "epoch": 0.6154584733557369,
+      "grad_norm": 1.6070075540077156,
+      "learning_rate": 9.984116612118612e-06,
+      "loss": 0.793,
+      "step": 641
+    },
+    {
+      "epoch": 0.6164186269803168,
+      "grad_norm": 1.027726907615115,
+      "learning_rate": 9.98384868720909e-06,
+      "loss": 0.7115,
+      "step": 642
+    },
+    {
+      "epoch": 0.6173787806048968,
+      "grad_norm": 0.9533472183683033,
+      "learning_rate": 9.983578525103166e-06,
+      "loss": 0.7205,
+      "step": 643
+    },
+    {
+      "epoch": 0.6183389342294767,
+      "grad_norm": 1.278072786735959,
+      "learning_rate": 9.983306125922111e-06,
+      "loss": 0.7248,
+      "step": 644
+    },
+    {
+      "epoch": 0.6192990878540566,
+      "grad_norm": 1.30698682005401,
+      "learning_rate": 9.983031489788208e-06,
+      "loss": 0.7187,
+      "step": 645
+    },
+    {
+      "epoch": 0.6202592414786365,
+      "grad_norm": 1.1294142037988335,
+      "learning_rate": 9.982754616824733e-06,
+      "loss": 0.7404,
+      "step": 646
+    },
+    {
+      "epoch": 0.6212193951032166,
+      "grad_norm": 1.1339707486746415,
+      "learning_rate": 9.982475507155974e-06,
+      "loss": 0.7069,
+      "step": 647
+    },
+    {
+      "epoch": 0.6221795487277965,
+      "grad_norm": 1.045473010633134,
+      "learning_rate": 9.98219416090722e-06,
+      "loss": 0.6399,
+      "step": 648
+    },
+    {
+      "epoch": 0.6231397023523764,
+      "grad_norm": 1.1491372204255086,
+      "learning_rate": 9.981910578204764e-06,
+      "loss": 0.6836,
+      "step": 649
+    },
+    {
+      "epoch": 0.6240998559769563,
+      "grad_norm": 1.3019640496272942,
+      "learning_rate": 9.981624759175902e-06,
+      "loss": 0.7818,
+      "step": 650
+    },
+    {
+      "epoch": 0.6250600096015363,
+      "grad_norm": 0.9641995848565628,
+      "learning_rate": 9.981336703948936e-06,
+      "loss": 0.7384,
+      "step": 651
+    },
+    {
+      "epoch": 0.6260201632261162,
+      "grad_norm": 1.0424494813372942,
+      "learning_rate": 9.98104641265317e-06,
+      "loss": 0.7581,
+      "step": 652
+    },
+    {
+      "epoch": 0.6269803168506961,
+      "grad_norm": 1.1749410105746338,
+      "learning_rate": 9.980753885418916e-06,
+      "loss": 0.8092,
+      "step": 653
+    },
+    {
+      "epoch": 0.627940470475276,
+      "grad_norm": 1.0608257188772605,
+      "learning_rate": 9.980459122377484e-06,
+      "loss": 0.7052,
+      "step": 654
+    },
+    {
+      "epoch": 0.628900624099856,
+      "grad_norm": 0.9851354208777588,
+      "learning_rate": 9.98016212366119e-06,
+      "loss": 0.7334,
+      "step": 655
+    },
+    {
+      "epoch": 0.6298607777244359,
+      "grad_norm": 1.1309322468621306,
+      "learning_rate": 9.979862889403353e-06,
+      "loss": 0.7993,
+      "step": 656
+    },
+    {
+      "epoch": 0.6308209313490158,
+      "grad_norm": 1.4423042663996517,
+      "learning_rate": 9.979561419738297e-06,
+      "loss": 0.8021,
+      "step": 657
+    },
+    {
+      "epoch": 0.6317810849735958,
+      "grad_norm": 1.1726897326686694,
+      "learning_rate": 9.979257714801348e-06,
+      "loss": 0.7577,
+      "step": 658
+    },
+    {
+      "epoch": 0.6327412385981757,
+      "grad_norm": 0.9727137981998019,
+      "learning_rate": 9.978951774728837e-06,
+      "loss": 0.728,
+      "step": 659
+    },
+    {
+      "epoch": 0.6337013922227557,
+      "grad_norm": 0.9433976828393242,
+      "learning_rate": 9.978643599658095e-06,
+      "loss": 0.705,
+      "step": 660
+    },
+    {
+      "epoch": 0.6346615458473356,
+      "grad_norm": 0.958260167898449,
+      "learning_rate": 9.97833318972746e-06,
+      "loss": 0.7253,
+      "step": 661
+    },
+    {
+      "epoch": 0.6356216994719155,
+      "grad_norm": 1.1728067197573666,
+      "learning_rate": 9.978020545076274e-06,
+      "loss": 0.7876,
+      "step": 662
+    },
+    {
+      "epoch": 0.6365818530964954,
+      "grad_norm": 1.0887980881688628,
+      "learning_rate": 9.977705665844875e-06,
+      "loss": 0.7163,
+      "step": 663
+    },
+    {
+      "epoch": 0.6375420067210754,
+      "grad_norm": 1.1071907542436588,
+      "learning_rate": 9.977388552174611e-06,
+      "loss": 0.7235,
+      "step": 664
+    },
+    {
+      "epoch": 0.6385021603456553,
+      "grad_norm": 1.1929947896189024,
+      "learning_rate": 9.977069204207832e-06,
+      "loss": 0.7432,
+      "step": 665
+    },
+    {
+      "epoch": 0.6394623139702352,
+      "grad_norm": 1.6705352459935405,
+      "learning_rate": 9.97674762208789e-06,
+      "loss": 0.7442,
+      "step": 666
+    },
+    {
+      "epoch": 0.6404224675948151,
+      "grad_norm": 0.9169215619256575,
+      "learning_rate": 9.976423805959138e-06,
+      "loss": 0.74,
+      "step": 667
+    },
+    {
+      "epoch": 0.6413826212193952,
+      "grad_norm": 1.6167894713612192,
+      "learning_rate": 9.976097755966935e-06,
+      "loss": 0.7734,
+      "step": 668
+    },
+    {
+      "epoch": 0.6423427748439751,
+      "grad_norm": 0.8516747743106268,
+      "learning_rate": 9.975769472257642e-06,
+      "loss": 0.6725,
+      "step": 669
+    },
+    {
+      "epoch": 0.643302928468555,
+      "grad_norm": 1.1960384862782092,
+      "learning_rate": 9.97543895497862e-06,
+      "loss": 0.7296,
+      "step": 670
+    },
+    {
+      "epoch": 0.6442630820931349,
+      "grad_norm": 1.151368014455306,
+      "learning_rate": 9.975106204278235e-06,
+      "loss": 0.7847,
+      "step": 671
+    },
+    {
+      "epoch": 0.6452232357177148,
+      "grad_norm": 1.4878206628916213,
+      "learning_rate": 9.974771220305855e-06,
+      "loss": 0.7559,
+      "step": 672
+    },
+    {
+      "epoch": 0.6461833893422948,
+      "grad_norm": 1.2396185589531643,
+      "learning_rate": 9.974434003211853e-06,
+      "loss": 0.7136,
+      "step": 673
+    },
+    {
+      "epoch": 0.6471435429668747,
+      "grad_norm": 1.1392041183252792,
+      "learning_rate": 9.974094553147599e-06,
+      "loss": 0.7088,
+      "step": 674
+    },
+    {
+      "epoch": 0.6481036965914546,
+      "grad_norm": 1.0042368852303203,
+      "learning_rate": 9.973752870265473e-06,
+      "loss": 0.7617,
+      "step": 675
+    },
+    {
+      "epoch": 0.6490638502160345,
+      "grad_norm": 1.0109453482022366,
+      "learning_rate": 9.973408954718848e-06,
+      "loss": 0.6427,
+      "step": 676
+    },
+    {
+      "epoch": 0.6500240038406145,
+      "grad_norm": 1.4767642145792161,
+      "learning_rate": 9.973062806662106e-06,
+      "loss": 0.7787,
+      "step": 677
+    },
+    {
+      "epoch": 0.6509841574651944,
+      "grad_norm": 1.149817371556,
+      "learning_rate": 9.97271442625063e-06,
+      "loss": 0.7017,
+      "step": 678
+    },
+    {
+      "epoch": 0.6519443110897744,
+      "grad_norm": 1.0927365492730468,
+      "learning_rate": 9.972363813640804e-06,
+      "loss": 0.6969,
+      "step": 679
+    },
+    {
+      "epoch": 0.6529044647143543,
+      "grad_norm": 1.1557208626189814,
+      "learning_rate": 9.972010968990012e-06,
+      "loss": 0.777,
+      "step": 680
+    },
+    {
+      "epoch": 0.6538646183389343,
+      "grad_norm": 1.0242126820972108,
+      "learning_rate": 9.971655892456646e-06,
+      "loss": 0.5974,
+      "step": 681
+    },
+    {
+      "epoch": 0.6548247719635142,
+      "grad_norm": 1.4722262917903646,
+      "learning_rate": 9.971298584200093e-06,
+      "loss": 0.7496,
+      "step": 682
+    },
+    {
+      "epoch": 0.6557849255880941,
+      "grad_norm": 1.228142897886962,
+      "learning_rate": 9.970939044380743e-06,
+      "loss": 0.7637,
+      "step": 683
+    },
+    {
+      "epoch": 0.656745079212674,
+      "grad_norm": 1.0246575470997343,
+      "learning_rate": 9.970577273159995e-06,
+      "loss": 0.73,
+      "step": 684
+    },
+    {
+      "epoch": 0.6577052328372539,
+      "grad_norm": 1.1064980469670687,
+      "learning_rate": 9.970213270700241e-06,
+      "loss": 0.7003,
+      "step": 685
+    },
+    {
+      "epoch": 0.6586653864618339,
+      "grad_norm": 0.9485775501229478,
+      "learning_rate": 9.969847037164881e-06,
+      "loss": 0.7478,
+      "step": 686
+    },
+    {
+      "epoch": 0.6596255400864138,
+      "grad_norm": 1.0209641451595586,
+      "learning_rate": 9.969478572718307e-06,
+      "loss": 0.7107,
+      "step": 687
+    },
+    {
+      "epoch": 0.6605856937109937,
+      "grad_norm": 1.2777317659195842,
+      "learning_rate": 9.969107877525925e-06,
+      "loss": 0.6346,
+      "step": 688
+    },
+    {
+      "epoch": 0.6615458473355736,
+      "grad_norm": 1.1304395876489817,
+      "learning_rate": 9.968734951754134e-06,
+      "loss": 0.716,
+      "step": 689
+    },
+    {
+      "epoch": 0.6625060009601537,
+      "grad_norm": 1.0589643707841647,
+      "learning_rate": 9.968359795570333e-06,
+      "loss": 0.7335,
+      "step": 690
+    },
+    {
+      "epoch": 0.6634661545847336,
+      "grad_norm": 1.312163556334331,
+      "learning_rate": 9.96798240914293e-06,
+      "loss": 0.7644,
+      "step": 691
+    },
+    {
+      "epoch": 0.6644263082093135,
+      "grad_norm": 1.4654703740824706,
+      "learning_rate": 9.96760279264133e-06,
+      "loss": 0.7075,
+      "step": 692
+    },
+    {
+      "epoch": 0.6653864618338934,
+      "grad_norm": 1.2308530665209227,
+      "learning_rate": 9.967220946235933e-06,
+      "loss": 0.6366,
+      "step": 693
+    },
+    {
+      "epoch": 0.6663466154584734,
+      "grad_norm": 1.4689991042569908,
+      "learning_rate": 9.966836870098153e-06,
+      "loss": 0.8619,
+      "step": 694
+    },
+    {
+      "epoch": 0.6673067690830533,
+      "grad_norm": 1.1544376204402063,
+      "learning_rate": 9.966450564400396e-06,
+      "loss": 0.5347,
+      "step": 695
+    },
+    {
+      "epoch": 0.6682669227076332,
+      "grad_norm": 1.018684998631377,
+      "learning_rate": 9.966062029316067e-06,
+      "loss": 0.7297,
+      "step": 696
+    },
+    {
+      "epoch": 0.6692270763322131,
+      "grad_norm": 1.2731963424973747,
+      "learning_rate": 9.965671265019576e-06,
+      "loss": 0.7425,
+      "step": 697
+    },
+    {
+      "epoch": 0.6701872299567931,
+      "grad_norm": 1.2094633115207627,
+      "learning_rate": 9.965278271686335e-06,
+      "loss": 0.834,
+      "step": 698
+    },
+    {
+      "epoch": 0.671147383581373,
+      "grad_norm": 1.298156294985173,
+      "learning_rate": 9.964883049492755e-06,
+      "loss": 0.7799,
+      "step": 699
+    },
+    {
+      "epoch": 0.672107537205953,
+      "grad_norm": 0.932403307235914,
+      "learning_rate": 9.964485598616244e-06,
+      "loss": 0.6305,
+      "step": 700
+    },
+    {
+      "epoch": 0.6730676908305329,
+      "grad_norm": 1.348840185082681,
+      "learning_rate": 9.964085919235217e-06,
+      "loss": 0.7148,
+      "step": 701
+    },
+    {
+      "epoch": 0.6740278444551128,
+      "grad_norm": 1.0730348632411175,
+      "learning_rate": 9.963684011529084e-06,
+      "loss": 0.7093,
+      "step": 702
+    },
+    {
+      "epoch": 0.6749879980796928,
+      "grad_norm": 1.7379797117287121,
+      "learning_rate": 9.963279875678256e-06,
+      "loss": 0.7236,
+      "step": 703
+    },
+    {
+      "epoch": 0.6759481517042727,
+      "grad_norm": 1.0101795335218668,
+      "learning_rate": 9.962873511864146e-06,
+      "loss": 0.7541,
+      "step": 704
+    },
+    {
+      "epoch": 0.6769083053288526,
+      "grad_norm": 1.095236103584098,
+      "learning_rate": 9.962464920269167e-06,
+      "loss": 0.6931,
+      "step": 705
+    },
+    {
+      "epoch": 0.6778684589534325,
+      "grad_norm": 0.9888949939493659,
+      "learning_rate": 9.962054101076733e-06,
+      "loss": 0.7213,
+      "step": 706
+    },
+    {
+      "epoch": 0.6788286125780125,
+      "grad_norm": 1.1704736246244622,
+      "learning_rate": 9.961641054471252e-06,
+      "loss": 0.7301,
+      "step": 707
+    },
+    {
+      "epoch": 0.6797887662025924,
+      "grad_norm": 1.1322459449428894,
+      "learning_rate": 9.96122578063814e-06,
+      "loss": 0.7355,
+      "step": 708
+    },
+    {
+      "epoch": 0.6807489198271723,
+      "grad_norm": 1.3295227832959442,
+      "learning_rate": 9.960808279763807e-06,
+      "loss": 0.7722,
+      "step": 709
+    },
+    {
+      "epoch": 0.6817090734517522,
+      "grad_norm": 1.2022742602302552,
+      "learning_rate": 9.960388552035666e-06,
+      "loss": 0.6774,
+      "step": 710
+    },
+    {
+      "epoch": 0.6826692270763323,
+      "grad_norm": 1.088527690096997,
+      "learning_rate": 9.959966597642127e-06,
+      "loss": 0.7082,
+      "step": 711
+    },
+    {
+      "epoch": 0.6836293807009122,
+      "grad_norm": 1.4359245485883494,
+      "learning_rate": 9.9595424167726e-06,
+      "loss": 0.7906,
+      "step": 712
+    },
+    {
+      "epoch": 0.6845895343254921,
+      "grad_norm": 1.0712636033705696,
+      "learning_rate": 9.9591160096175e-06,
+      "loss": 0.6228,
+      "step": 713
+    },
+    {
+      "epoch": 0.685549687950072,
+      "grad_norm": 1.1286122698085486,
+      "learning_rate": 9.95868737636823e-06,
+      "loss": 0.6131,
+      "step": 714
+    },
+    {
+      "epoch": 0.6865098415746519,
+      "grad_norm": 1.754001739111913,
+      "learning_rate": 9.958256517217204e-06,
+      "loss": 0.8415,
+      "step": 715
+    },
+    {
+      "epoch": 0.6874699951992319,
+      "grad_norm": 1.2415521559789686,
+      "learning_rate": 9.95782343235783e-06,
+      "loss": 0.6967,
+      "step": 716
+    },
+    {
+      "epoch": 0.6884301488238118,
+      "grad_norm": 0.9746405707963949,
+      "learning_rate": 9.95738812198451e-06,
+      "loss": 0.6819,
+      "step": 717
+    },
+    {
+      "epoch": 0.6893903024483917,
+      "grad_norm": 1.1467656093166074,
+      "learning_rate": 9.956950586292655e-06,
+      "loss": 0.7994,
+      "step": 718
+    },
+    {
+      "epoch": 0.6903504560729716,
+      "grad_norm": 1.1143672934666407,
+      "learning_rate": 9.956510825478669e-06,
+      "loss": 0.6832,
+      "step": 719
+    },
+    {
+      "epoch": 0.6913106096975516,
+      "grad_norm": 1.3721570448044658,
+      "learning_rate": 9.956068839739955e-06,
+      "loss": 0.8298,
+      "step": 720
+    },
+    {
+      "epoch": 0.6922707633221316,
+      "grad_norm": 1.6265371445701913,
+      "learning_rate": 9.955624629274916e-06,
+      "loss": 0.9974,
+      "step": 721
+    },
+    {
+      "epoch": 0.6932309169467115,
+      "grad_norm": 1.4662804140862407,
+      "learning_rate": 9.955178194282953e-06,
+      "loss": 0.7463,
+      "step": 722
+    },
+    {
+      "epoch": 0.6941910705712914,
+      "grad_norm": 0.8255635339080193,
+      "learning_rate": 9.954729534964469e-06,
+      "loss": 0.5698,
+      "step": 723
+    },
+    {
+      "epoch": 0.6951512241958714,
+      "grad_norm": 1.3585911943208118,
+      "learning_rate": 9.954278651520859e-06,
+      "loss": 0.7732,
+      "step": 724
+    },
+    {
+      "epoch": 0.6961113778204513,
+      "grad_norm": 1.060688981132975,
+      "learning_rate": 9.953825544154519e-06,
+      "loss": 0.7416,
+      "step": 725
+    },
+    {
+      "epoch": 0.6970715314450312,
+      "grad_norm": 1.2736780912673817,
+      "learning_rate": 9.953370213068847e-06,
+      "loss": 0.6838,
+      "step": 726
+    },
+    {
+      "epoch": 0.6980316850696111,
+      "grad_norm": 1.3360461372150172,
+      "learning_rate": 9.952912658468235e-06,
+      "loss": 0.7093,
+      "step": 727
+    },
+    {
+      "epoch": 0.6989918386941911,
+      "grad_norm": 1.0287316815152034,
+      "learning_rate": 9.952452880558073e-06,
+      "loss": 0.8231,
+      "step": 728
+    },
+    {
+      "epoch": 0.699951992318771,
+      "grad_norm": 1.1300889405704038,
+      "learning_rate": 9.951990879544753e-06,
+      "loss": 0.729,
+      "step": 729
+    },
+    {
+      "epoch": 0.7009121459433509,
+      "grad_norm": 0.9505504150767853,
+      "learning_rate": 9.95152665563566e-06,
+      "loss": 0.7065,
+      "step": 730
+    },
+    {
+      "epoch": 0.7018722995679308,
+      "grad_norm": 1.038257585330584,
+      "learning_rate": 9.951060209039182e-06,
+      "loss": 0.7397,
+      "step": 731
+    },
+    {
+      "epoch": 0.7028324531925108,
+      "grad_norm": 0.8443922390769166,
+      "learning_rate": 9.950591539964698e-06,
+      "loss": 0.706,
+      "step": 732
+    },
+    {
+      "epoch": 0.7037926068170908,
+      "grad_norm": 1.2080688483001045,
+      "learning_rate": 9.950120648622592e-06,
+      "loss": 0.7449,
+      "step": 733
+    },
+    {
+      "epoch": 0.7047527604416707,
+      "grad_norm": 0.9072170255903059,
+      "learning_rate": 9.94964753522424e-06,
+      "loss": 0.7127,
+      "step": 734
+    },
+    {
+      "epoch": 0.7057129140662506,
+      "grad_norm": 1.2648229256791117,
+      "learning_rate": 9.949172199982019e-06,
+      "loss": 0.7527,
+      "step": 735
+    },
+    {
+      "epoch": 0.7066730676908305,
+      "grad_norm": 1.3483477493599871,
+      "learning_rate": 9.9486946431093e-06,
+      "loss": 0.955,
+      "step": 736
+    },
+    {
+      "epoch": 0.7076332213154105,
+      "grad_norm": 1.1352257922072784,
+      "learning_rate": 9.948214864820455e-06,
+      "loss": 0.7398,
+      "step": 737
+    },
+    {
+      "epoch": 0.7085933749399904,
+      "grad_norm": 1.3783392370323646,
+      "learning_rate": 9.94773286533085e-06,
+      "loss": 0.8405,
+      "step": 738
+    },
+    {
+      "epoch": 0.7095535285645703,
+      "grad_norm": 1.0389249982585742,
+      "learning_rate": 9.947248644856852e-06,
+      "loss": 0.6672,
+      "step": 739
+    },
+    {
+      "epoch": 0.7105136821891502,
+      "grad_norm": 1.3151653766211417,
+      "learning_rate": 9.946762203615818e-06,
+      "loss": 0.764,
+      "step": 740
+    },
+    {
+      "epoch": 0.7114738358137302,
+      "grad_norm": 0.9917853380663585,
+      "learning_rate": 9.94627354182611e-06,
+      "loss": 0.7454,
+      "step": 741
+    },
+    {
+      "epoch": 0.7124339894383102,
+      "grad_norm": 1.1532058497358957,
+      "learning_rate": 9.945782659707082e-06,
+      "loss": 0.6812,
+      "step": 742
+    },
+    {
+      "epoch": 0.7133941430628901,
+      "grad_norm": 1.1572527961716201,
+      "learning_rate": 9.945289557479086e-06,
+      "loss": 0.7281,
+      "step": 743
+    },
+    {
+      "epoch": 0.71435429668747,
+      "grad_norm": 1.176236714703303,
+      "learning_rate": 9.944794235363468e-06,
+      "loss": 0.6686,
+      "step": 744
+    },
+    {
+      "epoch": 0.7153144503120499,
+      "grad_norm": 1.004038515625951,
+      "learning_rate": 9.944296693582576e-06,
+      "loss": 0.7013,
+      "step": 745
+    },
+    {
+      "epoch": 0.7162746039366299,
+      "grad_norm": 1.43231333048898,
+      "learning_rate": 9.94379693235975e-06,
+      "loss": 0.6745,
+      "step": 746
+    },
+    {
+      "epoch": 0.7172347575612098,
+      "grad_norm": 0.8278900895756431,
+      "learning_rate": 9.943294951919327e-06,
+      "loss": 0.5708,
+      "step": 747
+    },
+    {
+      "epoch": 0.7181949111857897,
+      "grad_norm": 1.4659914938897542,
+      "learning_rate": 9.94279075248664e-06,
+      "loss": 0.8242,
+      "step": 748
+    },
+    {
+      "epoch": 0.7191550648103696,
+      "grad_norm": 1.4706142820300785,
+      "learning_rate": 9.942284334288021e-06,
+      "loss": 0.8457,
+      "step": 749
+    },
+    {
+      "epoch": 0.7201152184349496,
+      "grad_norm": 1.1984466642743257,
+      "learning_rate": 9.941775697550796e-06,
+      "loss": 0.7123,
+      "step": 750
+    },
+    {
+      "epoch": 0.7210753720595295,
+      "grad_norm": 1.1050073348933755,
+      "learning_rate": 9.941264842503282e-06,
+      "loss": 0.8425,
+      "step": 751
+    },
+    {
+      "epoch": 0.7220355256841094,
+      "grad_norm": 0.8687768913712395,
+      "learning_rate": 9.9407517693748e-06,
+      "loss": 0.6265,
+      "step": 752
+    },
+    {
+      "epoch": 0.7229956793086894,
+      "grad_norm": 0.9769568173569657,
+      "learning_rate": 9.940236478395662e-06,
+      "loss": 0.724,
+      "step": 753
+    },
+    {
+      "epoch": 0.7239558329332694,
+      "grad_norm": 1.6207874366444537,
+      "learning_rate": 9.939718969797178e-06,
+      "loss": 0.8284,
+      "step": 754
+    },
+    {
+      "epoch": 0.7249159865578493,
+      "grad_norm": 0.9805328249455393,
+      "learning_rate": 9.939199243811653e-06,
+      "loss": 0.6822,
+      "step": 755
+    },
+    {
+      "epoch": 0.7258761401824292,
+      "grad_norm": 1.246024930996379,
+      "learning_rate": 9.938677300672382e-06,
+      "loss": 0.7775,
+      "step": 756
+    },
+    {
+      "epoch": 0.7268362938070091,
+      "grad_norm": 1.198028879321762,
+      "learning_rate": 9.938153140613662e-06,
+      "loss": 0.7028,
+      "step": 757
+    },
+    {
+      "epoch": 0.727796447431589,
+      "grad_norm": 0.9973811279361954,
+      "learning_rate": 9.937626763870784e-06,
+      "loss": 0.6884,
+      "step": 758
+    },
+    {
+      "epoch": 0.728756601056169,
+      "grad_norm": 1.03169869259323,
+      "learning_rate": 9.937098170680034e-06,
+      "loss": 0.6303,
+      "step": 759
+    },
+    {
+      "epoch": 0.7297167546807489,
+      "grad_norm": 0.9247880187978105,
+      "learning_rate": 9.936567361278687e-06,
+      "loss": 0.6576,
+      "step": 760
+    },
+    {
+      "epoch": 0.7306769083053288,
+      "grad_norm": 0.8417058658300991,
+      "learning_rate": 9.93603433590502e-06,
+      "loss": 0.6361,
+      "step": 761
+    },
+    {
+      "epoch": 0.7316370619299087,
+      "grad_norm": 1.03716134734401,
+      "learning_rate": 9.935499094798305e-06,
+      "loss": 0.7885,
+      "step": 762
+    },
+    {
+      "epoch": 0.7325972155544888,
+      "grad_norm": 1.2074321143830193,
+      "learning_rate": 9.934961638198804e-06,
+      "loss": 0.6944,
+      "step": 763
+    },
+    {
+      "epoch": 0.7335573691790687,
+      "grad_norm": 1.1848604524508257,
+      "learning_rate": 9.934421966347773e-06,
+      "loss": 0.8124,
+      "step": 764
+    },
+    {
+      "epoch": 0.7345175228036486,
+      "grad_norm": 1.0429348427785348,
+      "learning_rate": 9.93388007948747e-06,
+      "loss": 0.7377,
+      "step": 765
+    },
+    {
+      "epoch": 0.7354776764282285,
+      "grad_norm": 0.8888707066106554,
+      "learning_rate": 9.933335977861138e-06,
+      "loss": 0.6764,
+      "step": 766
+    },
+    {
+      "epoch": 0.7364378300528085,
+      "grad_norm": 1.0626278519468342,
+      "learning_rate": 9.93278966171302e-06,
+      "loss": 0.6276,
+      "step": 767
+    },
+    {
+      "epoch": 0.7373979836773884,
+      "grad_norm": 0.8962912358229971,
+      "learning_rate": 9.932241131288352e-06,
+      "loss": 0.6641,
+      "step": 768
+    },
+    {
+      "epoch": 0.7383581373019683,
+      "grad_norm": 0.8941146837643744,
+      "learning_rate": 9.931690386833363e-06,
+      "loss": 0.7262,
+      "step": 769
+    },
+    {
+      "epoch": 0.7393182909265482,
+      "grad_norm": 1.0424749755401177,
+      "learning_rate": 9.931137428595276e-06,
+      "loss": 0.7582,
+      "step": 770
+    },
+    {
+      "epoch": 0.7402784445511282,
+      "grad_norm": 1.5239003387542835,
+      "learning_rate": 9.930582256822307e-06,
+      "loss": 0.7967,
+      "step": 771
+    },
+    {
+      "epoch": 0.7412385981757081,
+      "grad_norm": 0.9789285534367823,
+      "learning_rate": 9.93002487176367e-06,
+      "loss": 0.7711,
+      "step": 772
+    },
+    {
+      "epoch": 0.742198751800288,
+      "grad_norm": 0.9798122544351332,
+      "learning_rate": 9.929465273669567e-06,
+      "loss": 0.7138,
+      "step": 773
+    },
+    {
+      "epoch": 0.743158905424868,
+      "grad_norm": 1.0089252998813754,
+      "learning_rate": 9.928903462791195e-06,
+      "loss": 0.7679,
+      "step": 774
+    },
+    {
+      "epoch": 0.7441190590494479,
+      "grad_norm": 1.129686771537608,
+      "learning_rate": 9.928339439380745e-06,
+      "loss": 0.7867,
+      "step": 775
+    },
+    {
+      "epoch": 0.7450792126740279,
+      "grad_norm": 1.0680489943573477,
+      "learning_rate": 9.927773203691403e-06,
+      "loss": 0.7233,
+      "step": 776
+    },
+    {
+      "epoch": 0.7460393662986078,
+      "grad_norm": 1.1991253915669609,
+      "learning_rate": 9.927204755977344e-06,
+      "loss": 0.8214,
+      "step": 777
+    },
+    {
+      "epoch": 0.7469995199231877,
+      "grad_norm": 1.1810464281071666,
+      "learning_rate": 9.926634096493738e-06,
+      "loss": 0.7266,
+      "step": 778
+    },
+    {
+      "epoch": 0.7479596735477676,
+      "grad_norm": 1.1918237303571164,
+      "learning_rate": 9.92606122549675e-06,
+      "loss": 0.7091,
+      "step": 779
+    },
+    {
+      "epoch": 0.7489198271723476,
+      "grad_norm": 1.4809878657655684,
+      "learning_rate": 9.925486143243534e-06,
+      "loss": 0.8235,
+      "step": 780
+    },
+    {
+      "epoch": 0.7498799807969275,
+      "grad_norm": 1.0623502206306763,
+      "learning_rate": 9.92490884999224e-06,
+      "loss": 0.6431,
+      "step": 781
+    },
+    {
+      "epoch": 0.7508401344215074,
+      "grad_norm": 1.1728978803988328,
+      "learning_rate": 9.924329346002007e-06,
+      "loss": 0.7072,
+      "step": 782
+    },
+    {
+      "epoch": 0.7518002880460873,
+      "grad_norm": 1.2661200193104183,
+      "learning_rate": 9.923747631532968e-06,
+      "loss": 0.7299,
+      "step": 783
+    },
+    {
+      "epoch": 0.7527604416706674,
+      "grad_norm": 1.369779764076495,
+      "learning_rate": 9.92316370684625e-06,
+      "loss": 0.6906,
+      "step": 784
+    },
+    {
+      "epoch": 0.7537205952952473,
+      "grad_norm": 0.9954497341903171,
+      "learning_rate": 9.922577572203969e-06,
+      "loss": 0.6794,
+      "step": 785
+    },
+    {
+      "epoch": 0.7546807489198272,
+      "grad_norm": 1.4017643770239192,
+      "learning_rate": 9.921989227869236e-06,
+      "loss": 0.8379,
+      "step": 786
+    },
+    {
+      "epoch": 0.7556409025444071,
+      "grad_norm": 1.3015683095250705,
+      "learning_rate": 9.921398674106152e-06,
+      "loss": 0.826,
+      "step": 787
+    },
+    {
+      "epoch": 0.756601056168987,
+      "grad_norm": 1.2155557789621108,
+      "learning_rate": 9.92080591117981e-06,
+      "loss": 0.8202,
+      "step": 788
+    },
+    {
+      "epoch": 0.757561209793567,
+      "grad_norm": 1.0881907984132828,
+      "learning_rate": 9.920210939356294e-06,
+      "loss": 0.7415,
+      "step": 789
+    },
+    {
+      "epoch": 0.7585213634181469,
+      "grad_norm": 1.2935427695919515,
+      "learning_rate": 9.919613758902682e-06,
+      "loss": 0.7854,
+      "step": 790
+    },
+    {
+      "epoch": 0.7594815170427268,
+      "grad_norm": 1.2385800002831058,
+      "learning_rate": 9.91901437008704e-06,
+      "loss": 0.7114,
+      "step": 791
+    },
+    {
+      "epoch": 0.7604416706673067,
+      "grad_norm": 1.001176807057794,
+      "learning_rate": 9.918412773178431e-06,
+      "loss": 0.7033,
+      "step": 792
+    },
+    {
+      "epoch": 0.7614018242918867,
+      "grad_norm": 1.1181435549505114,
+      "learning_rate": 9.917808968446903e-06,
+      "loss": 0.7477,
+      "step": 793
+    },
+    {
+      "epoch": 0.7623619779164666,
+      "grad_norm": 1.265201274814283,
+      "learning_rate": 9.917202956163496e-06,
+      "loss": 0.7502,
+      "step": 794
+    },
+    {
+      "epoch": 0.7633221315410466,
+      "grad_norm": 1.0814941759105932,
+      "learning_rate": 9.916594736600244e-06,
+      "loss": 0.7752,
+      "step": 795
+    },
+    {
+      "epoch": 0.7642822851656265,
+      "grad_norm": 1.015411031942487,
+      "learning_rate": 9.915984310030172e-06,
+      "loss": 0.8252,
+      "step": 796
+    },
+    {
+      "epoch": 0.7652424387902065,
+      "grad_norm": 1.0752951651724534,
+      "learning_rate": 9.91537167672729e-06,
+      "loss": 0.6898,
+      "step": 797
+    },
+    {
+      "epoch": 0.7662025924147864,
+      "grad_norm": 0.8978312152634739,
+      "learning_rate": 9.914756836966604e-06,
+      "loss": 0.6568,
+      "step": 798
+    },
+    {
+      "epoch": 0.7671627460393663,
+      "grad_norm": 1.4174184006325772,
+      "learning_rate": 9.914139791024112e-06,
+      "loss": 0.7441,
+      "step": 799
+    },
+    {
+      "epoch": 0.7681228996639462,
+      "grad_norm": 1.0309137281263778,
+      "learning_rate": 9.913520539176797e-06,
+      "loss": 0.8065,
+      "step": 800
+    },
+    {
+      "epoch": 0.7690830532885262,
+      "grad_norm": 1.1056726761800266,
+      "learning_rate": 9.912899081702633e-06,
+      "loss": 0.7709,
+      "step": 801
+    },
+    {
+      "epoch": 0.7700432069131061,
+      "grad_norm": 1.1237710277127422,
+      "learning_rate": 9.912275418880588e-06,
+      "loss": 0.7291,
+      "step": 802
+    },
+    {
+      "epoch": 0.771003360537686,
+      "grad_norm": 0.8081645629455623,
+      "learning_rate": 9.911649550990618e-06,
+      "loss": 0.6145,
+      "step": 803
+    },
+    {
+      "epoch": 0.7719635141622659,
+      "grad_norm": 0.8747329876189975,
+      "learning_rate": 9.911021478313665e-06,
+      "loss": 0.6091,
+      "step": 804
+    },
+    {
+      "epoch": 0.7729236677868458,
+      "grad_norm": 0.8873911870516857,
+      "learning_rate": 9.910391201131668e-06,
+      "loss": 0.6622,
+      "step": 805
+    },
+    {
+      "epoch": 0.7738838214114259,
+      "grad_norm": 0.9666313834432418,
+      "learning_rate": 9.909758719727549e-06,
+      "loss": 0.5906,
+      "step": 806
+    },
+    {
+      "epoch": 0.7748439750360058,
+      "grad_norm": 0.9382478033157138,
+      "learning_rate": 9.909124034385225e-06,
+      "loss": 0.6475,
+      "step": 807
+    },
+    {
+      "epoch": 0.7758041286605857,
+      "grad_norm": 1.4249896826413047,
+      "learning_rate": 9.908487145389595e-06,
+      "loss": 0.9343,
+      "step": 808
+    },
+    {
+      "epoch": 0.7767642822851656,
+      "grad_norm": 1.1015378414410202,
+      "learning_rate": 9.907848053026554e-06,
+      "loss": 0.6708,
+      "step": 809
+    },
+    {
+      "epoch": 0.7777244359097456,
+      "grad_norm": 0.8335101639979812,
+      "learning_rate": 9.907206757582987e-06,
+      "loss": 0.5111,
+      "step": 810
+    },
+    {
+      "epoch": 0.7786845895343255,
+      "grad_norm": 1.0698637825517112,
+      "learning_rate": 9.90656325934676e-06,
+      "loss": 0.7194,
+      "step": 811
+    },
+    {
+      "epoch": 0.7796447431589054,
+      "grad_norm": 1.3467599291721108,
+      "learning_rate": 9.905917558606734e-06,
+      "loss": 0.8593,
+      "step": 812
+    },
+    {
+      "epoch": 0.7806048967834853,
+      "grad_norm": 0.9265083094260931,
+      "learning_rate": 9.905269655652758e-06,
+      "loss": 0.691,
+      "step": 813
+    },
+    {
+      "epoch": 0.7815650504080653,
+      "grad_norm": 0.9569516294394528,
+      "learning_rate": 9.904619550775668e-06,
+      "loss": 0.6904,
+      "step": 814
+    },
+    {
+      "epoch": 0.7825252040326452,
+      "grad_norm": 1.2033247575360706,
+      "learning_rate": 9.903967244267287e-06,
+      "loss": 0.7906,
+      "step": 815
+    },
+    {
+      "epoch": 0.7834853576572252,
+      "grad_norm": 1.2045148381420874,
+      "learning_rate": 9.903312736420433e-06,
+      "loss": 0.7269,
+      "step": 816
+    },
+    {
+      "epoch": 0.7844455112818051,
+      "grad_norm": 0.9815529060942425,
+      "learning_rate": 9.902656027528903e-06,
+      "loss": 0.7521,
+      "step": 817
+    },
+    {
+      "epoch": 0.785405664906385,
+      "grad_norm": 1.0169730117521143,
+      "learning_rate": 9.90199711788749e-06,
+      "loss": 0.6917,
+      "step": 818
+    },
+    {
+      "epoch": 0.786365818530965,
+      "grad_norm": 1.279845677472938,
+      "learning_rate": 9.901336007791971e-06,
+      "loss": 0.8229,
+      "step": 819
+    },
+    {
+      "epoch": 0.7873259721555449,
+      "grad_norm": 1.1504766635141583,
+      "learning_rate": 9.900672697539108e-06,
+      "loss": 0.7362,
+      "step": 820
+    },
+    {
+      "epoch": 0.7882861257801248,
+      "grad_norm": 1.0363323444139323,
+      "learning_rate": 9.900007187426659e-06,
+      "loss": 0.6574,
+      "step": 821
+    },
+    {
+      "epoch": 0.7892462794047047,
+      "grad_norm": 1.0099042824557933,
+      "learning_rate": 9.89933947775336e-06,
+      "loss": 0.6754,
+      "step": 822
+    },
+    {
+      "epoch": 0.7902064330292847,
+      "grad_norm": 0.9730742596836239,
+      "learning_rate": 9.898669568818942e-06,
+      "loss": 0.6871,
+      "step": 823
+    },
+    {
+      "epoch": 0.7911665866538646,
+      "grad_norm": 0.7856584341577615,
+      "learning_rate": 9.897997460924118e-06,
+      "loss": 0.5745,
+      "step": 824
+    },
+    {
+      "epoch": 0.7921267402784445,
+      "grad_norm": 1.2340542727190076,
+      "learning_rate": 9.89732315437059e-06,
+      "loss": 0.7881,
+      "step": 825
+    },
+    {
+      "epoch": 0.7930868939030244,
+      "grad_norm": 1.0505728583822727,
+      "learning_rate": 9.896646649461047e-06,
+      "loss": 0.7125,
+      "step": 826
+    },
+    {
+      "epoch": 0.7940470475276045,
+      "grad_norm": 0.9100796308195849,
+      "learning_rate": 9.895967946499165e-06,
+      "loss": 0.6839,
+      "step": 827
+    },
+    {
+      "epoch": 0.7950072011521844,
+      "grad_norm": 0.9586535381174625,
+      "learning_rate": 9.895287045789608e-06,
+      "loss": 0.7303,
+      "step": 828
+    },
+    {
+      "epoch": 0.7959673547767643,
+      "grad_norm": 1.3749908323684445,
+      "learning_rate": 9.894603947638023e-06,
+      "loss": 0.7061,
+      "step": 829
+    },
+    {
+      "epoch": 0.7969275084013442,
+      "grad_norm": 1.1785565342144533,
+      "learning_rate": 9.893918652351048e-06,
+      "loss": 0.7537,
+      "step": 830
+    },
+    {
+      "epoch": 0.7978876620259241,
+      "grad_norm": 1.0812103342300967,
+      "learning_rate": 9.893231160236303e-06,
+      "loss": 0.8364,
+      "step": 831
+    },
+    {
+      "epoch": 0.7988478156505041,
+      "grad_norm": 0.9867997910803941,
+      "learning_rate": 9.892541471602395e-06,
+      "loss": 0.6845,
+      "step": 832
+    },
+    {
+      "epoch": 0.799807969275084,
+      "grad_norm": 0.9732673403408759,
+      "learning_rate": 9.891849586758918e-06,
+      "loss": 0.7633,
+      "step": 833
+    },
+    {
+      "epoch": 0.8007681228996639,
+      "grad_norm": 0.9872640710831527,
+      "learning_rate": 9.891155506016452e-06,
+      "loss": 0.7817,
+      "step": 834
+    },
+    {
+      "epoch": 0.8017282765242438,
+      "grad_norm": 1.055306679841214,
+      "learning_rate": 9.890459229686564e-06,
+      "loss": 0.6603,
+      "step": 835
+    },
+    {
+      "epoch": 0.8026884301488239,
+      "grad_norm": 0.9568869069835517,
+      "learning_rate": 9.889760758081803e-06,
+      "loss": 0.6943,
+      "step": 836
+    },
+    {
+      "epoch": 0.8036485837734038,
+      "grad_norm": 1.0072131187344215,
+      "learning_rate": 9.889060091515707e-06,
+      "loss": 0.6164,
+      "step": 837
+    },
+    {
+      "epoch": 0.8046087373979837,
+      "grad_norm": 0.9978428928688824,
+      "learning_rate": 9.8883572303028e-06,
+      "loss": 0.7224,
+      "step": 838
+    },
+    {
+      "epoch": 0.8055688910225636,
+      "grad_norm": 1.2332002879202255,
+      "learning_rate": 9.887652174758581e-06,
+      "loss": 0.6889,
+      "step": 839
+    },
+    {
+      "epoch": 0.8065290446471436,
+      "grad_norm": 1.0961807544454492,
+      "learning_rate": 9.88694492519955e-06,
+      "loss": 0.6717,
+      "step": 840
+    },
+    {
+      "epoch": 0.8074891982717235,
+      "grad_norm": 1.1307172390388518,
+      "learning_rate": 9.88623548194318e-06,
+      "loss": 0.8255,
+      "step": 841
+    },
+    {
+      "epoch": 0.8084493518963034,
+      "grad_norm": 0.9297760721920874,
+      "learning_rate": 9.88552384530793e-06,
+      "loss": 0.666,
+      "step": 842
+    },
+    {
+      "epoch": 0.8094095055208833,
+      "grad_norm": 1.1095655791690846,
+      "learning_rate": 9.884810015613255e-06,
+      "loss": 0.7067,
+      "step": 843
+    },
+    {
+      "epoch": 0.8103696591454633,
+      "grad_norm": 1.1551923464569889,
+      "learning_rate": 9.884093993179575e-06,
+      "loss": 0.7585,
+      "step": 844
+    },
+    {
+      "epoch": 0.8113298127700432,
+      "grad_norm": 0.9387136764430609,
+      "learning_rate": 9.883375778328312e-06,
+      "loss": 0.7272,
+      "step": 845
+    },
+    {
+      "epoch": 0.8122899663946231,
+      "grad_norm": 1.0708716761811787,
+      "learning_rate": 9.882655371381861e-06,
+      "loss": 0.7268,
+      "step": 846
+    },
+    {
+      "epoch": 0.813250120019203,
+      "grad_norm": 0.948634004314031,
+      "learning_rate": 9.881932772663609e-06,
+      "loss": 0.7238,
+      "step": 847
+    },
+    {
+      "epoch": 0.814210273643783,
+      "grad_norm": 1.185852375482035,
+      "learning_rate": 9.881207982497917e-06,
+      "loss": 0.7256,
+      "step": 848
+    },
+    {
+      "epoch": 0.815170427268363,
+      "grad_norm": 1.0054289883736454,
+      "learning_rate": 9.880481001210142e-06,
+      "loss": 0.6653,
+      "step": 849
+    },
+    {
+      "epoch": 0.8161305808929429,
+      "grad_norm": 0.9993636485340736,
+      "learning_rate": 9.879751829126612e-06,
+      "loss": 0.6556,
+      "step": 850
+    },
+    {
+      "epoch": 0.8170907345175228,
+      "grad_norm": 0.9645142898485913,
+      "learning_rate": 9.87902046657465e-06,
+      "loss": 0.6754,
+      "step": 851
+    },
+    {
+      "epoch": 0.8180508881421027,
+      "grad_norm": 0.9891246316672525,
+      "learning_rate": 9.878286913882553e-06,
+      "loss": 0.6649,
+      "step": 852
+    },
+    {
+      "epoch": 0.8190110417666827,
+      "grad_norm": 1.3063650699400922,
+      "learning_rate": 9.877551171379606e-06,
+      "loss": 0.8185,
+      "step": 853
+    },
+    {
+      "epoch": 0.8199711953912626,
+      "grad_norm": 1.1918128804072938,
+      "learning_rate": 9.876813239396079e-06,
+      "loss": 0.7668,
+      "step": 854
+    },
+    {
+      "epoch": 0.8209313490158425,
+      "grad_norm": 0.8694013244785461,
+      "learning_rate": 9.876073118263216e-06,
+      "loss": 0.6434,
+      "step": 855
+    },
+    {
+      "epoch": 0.8218915026404224,
+      "grad_norm": 1.092408433099053,
+      "learning_rate": 9.875330808313254e-06,
+      "loss": 0.7972,
+      "step": 856
+    },
+    {
+      "epoch": 0.8228516562650025,
+      "grad_norm": 1.412526926967191,
+      "learning_rate": 9.874586309879406e-06,
+      "loss": 0.656,
+      "step": 857
+    },
+    {
+      "epoch": 0.8238118098895824,
+      "grad_norm": 1.1217090893050323,
+      "learning_rate": 9.87383962329587e-06,
+      "loss": 0.8015,
+      "step": 858
+    },
+    {
+      "epoch": 0.8247719635141623,
+      "grad_norm": 1.1657462651565507,
+      "learning_rate": 9.873090748897826e-06,
+      "loss": 0.7275,
+      "step": 859
+    },
+    {
+      "epoch": 0.8257321171387422,
+      "grad_norm": 1.1787793782521532,
+      "learning_rate": 9.872339687021434e-06,
+      "loss": 0.7413,
+      "step": 860
+    },
+    {
+      "epoch": 0.8266922707633221,
+      "grad_norm": 1.340616231597411,
+      "learning_rate": 9.871586438003841e-06,
+      "loss": 0.7786,
+      "step": 861
+    },
+    {
+      "epoch": 0.8276524243879021,
+      "grad_norm": 1.179602094426698,
+      "learning_rate": 9.870831002183168e-06,
+      "loss": 0.7823,
+      "step": 862
+    },
+    {
+      "epoch": 0.828612578012482,
+      "grad_norm": 1.1223593173529196,
+      "learning_rate": 9.870073379898526e-06,
+      "loss": 0.7104,
+      "step": 863
+    },
+    {
+      "epoch": 0.8295727316370619,
+      "grad_norm": 1.400705044676731,
+      "learning_rate": 9.869313571490002e-06,
+      "loss": 0.6805,
+      "step": 864
+    },
+    {
+      "epoch": 0.8305328852616418,
+      "grad_norm": 1.3950024166572066,
+      "learning_rate": 9.868551577298665e-06,
+      "loss": 0.7823,
+      "step": 865
+    },
+    {
+      "epoch": 0.8314930388862218,
+      "grad_norm": 1.256792972347036,
+      "learning_rate": 9.867787397666568e-06,
+      "loss": 0.6954,
+      "step": 866
+    },
+    {
+      "epoch": 0.8324531925108017,
+      "grad_norm": 0.8067378155239507,
+      "learning_rate": 9.86702103293674e-06,
+      "loss": 0.6488,
+      "step": 867
+    },
+    {
+      "epoch": 0.8334133461353816,
+      "grad_norm": 0.9597200041625673,
+      "learning_rate": 9.866252483453199e-06,
+      "loss": 0.7559,
+      "step": 868
+    },
+    {
+      "epoch": 0.8343734997599616,
+      "grad_norm": 1.3440961886731113,
+      "learning_rate": 9.865481749560933e-06,
+      "loss": 0.8602,
+      "step": 869
+    },
+    {
+      "epoch": 0.8353336533845416,
+      "grad_norm": 0.8939860482760397,
+      "learning_rate": 9.864708831605918e-06,
+      "loss": 0.6142,
+      "step": 870
+    },
+    {
+      "epoch": 0.8362938070091215,
+      "grad_norm": 1.3127632405498024,
+      "learning_rate": 9.863933729935112e-06,
+      "loss": 0.779,
+      "step": 871
+    },
+    {
+      "epoch": 0.8372539606337014,
+      "grad_norm": 1.4737052784454916,
+      "learning_rate": 9.863156444896446e-06,
+      "loss": 0.7681,
+      "step": 872
+    },
+    {
+      "epoch": 0.8382141142582813,
+      "grad_norm": 1.2860252668236365,
+      "learning_rate": 9.862376976838835e-06,
+      "loss": 0.6934,
+      "step": 873
+    },
+    {
+      "epoch": 0.8391742678828612,
+      "grad_norm": 1.1395463857271173,
+      "learning_rate": 9.861595326112176e-06,
+      "loss": 0.7219,
+      "step": 874
+    },
+    {
+      "epoch": 0.8401344215074412,
+      "grad_norm": 1.009070130534368,
+      "learning_rate": 9.860811493067344e-06,
+      "loss": 0.7276,
+      "step": 875
+    },
+    {
+      "epoch": 0.8410945751320211,
+      "grad_norm": 1.0647868874105912,
+      "learning_rate": 9.86002547805619e-06,
+      "loss": 0.6948,
+      "step": 876
+    },
+    {
+      "epoch": 0.842054728756601,
+      "grad_norm": 0.8266180898239189,
+      "learning_rate": 9.85923728143155e-06,
+      "loss": 0.7082,
+      "step": 877
+    },
+    {
+      "epoch": 0.8430148823811809,
+      "grad_norm": 0.9831565698075821,
+      "learning_rate": 9.858446903547239e-06,
+      "loss": 0.7596,
+      "step": 878
+    },
+    {
+      "epoch": 0.843975036005761,
+      "grad_norm": 1.4438296502868566,
+      "learning_rate": 9.857654344758044e-06,
+      "loss": 0.8311,
+      "step": 879
+    },
+    {
+      "epoch": 0.8449351896303409,
+      "grad_norm": 1.563787214147985,
+      "learning_rate": 9.85685960541974e-06,
+      "loss": 0.8639,
+      "step": 880
+    },
+    {
+      "epoch": 0.8458953432549208,
+      "grad_norm": 1.0946544146378487,
+      "learning_rate": 9.856062685889076e-06,
+      "loss": 0.7273,
+      "step": 881
+    },
+    {
+      "epoch": 0.8468554968795007,
+      "grad_norm": 1.1508085335490117,
+      "learning_rate": 9.855263586523782e-06,
+      "loss": 0.763,
+      "step": 882
+    },
+    {
+      "epoch": 0.8478156505040807,
+      "grad_norm": 0.7477159491211194,
+      "learning_rate": 9.854462307682561e-06,
+      "loss": 0.627,
+      "step": 883
+    },
+    {
+      "epoch": 0.8487758041286606,
+      "grad_norm": 1.3831521429558586,
+      "learning_rate": 9.853658849725105e-06,
+      "loss": 0.8919,
+      "step": 884
+    },
+    {
+      "epoch": 0.8497359577532405,
+      "grad_norm": 0.9523410225588861,
+      "learning_rate": 9.852853213012072e-06,
+      "loss": 0.6652,
+      "step": 885
+    },
+    {
+      "epoch": 0.8506961113778204,
+      "grad_norm": 0.8440621024279631,
+      "learning_rate": 9.852045397905106e-06,
+      "loss": 0.6793,
+      "step": 886
+    },
+    {
+      "epoch": 0.8516562650024004,
+      "grad_norm": 0.9628712617820975,
+      "learning_rate": 9.851235404766827e-06,
+      "loss": 0.7682,
+      "step": 887
+    },
+    {
+      "epoch": 0.8526164186269803,
+      "grad_norm": 1.10112303741811,
+      "learning_rate": 9.850423233960831e-06,
+      "loss": 0.764,
+      "step": 888
+    },
+    {
+      "epoch": 0.8535765722515603,
+      "grad_norm": 0.8290764739831752,
+      "learning_rate": 9.849608885851693e-06,
+      "loss": 0.6713,
+      "step": 889
+    },
+    {
+      "epoch": 0.8545367258761402,
+      "grad_norm": 1.187481809310174,
+      "learning_rate": 9.848792360804966e-06,
+      "loss": 0.7793,
+      "step": 890
+    },
+    {
+      "epoch": 0.8554968795007201,
+      "grad_norm": 1.2014014665986943,
+      "learning_rate": 9.84797365918718e-06,
+      "loss": 0.7957,
+      "step": 891
+    },
+    {
+      "epoch": 0.8564570331253001,
+      "grad_norm": 1.5352983734519552,
+      "learning_rate": 9.847152781365839e-06,
+      "loss": 0.8209,
+      "step": 892
+    },
+    {
+      "epoch": 0.85741718674988,
+      "grad_norm": 1.0656341787458772,
+      "learning_rate": 9.846329727709429e-06,
+      "loss": 0.7326,
+      "step": 893
+    },
+    {
+      "epoch": 0.8583773403744599,
+      "grad_norm": 0.89403431681965,
+      "learning_rate": 9.845504498587406e-06,
+      "loss": 0.709,
+      "step": 894
+    },
+    {
+      "epoch": 0.8593374939990398,
+      "grad_norm": 1.015401224155412,
+      "learning_rate": 9.844677094370212e-06,
+      "loss": 0.6282,
+      "step": 895
+    },
+    {
+      "epoch": 0.8602976476236198,
+      "grad_norm": 0.938644230385865,
+      "learning_rate": 9.843847515429256e-06,
+      "loss": 0.7081,
+      "step": 896
+    },
+    {
+      "epoch": 0.8612578012481997,
+      "grad_norm": 1.1323586524295224,
+      "learning_rate": 9.843015762136927e-06,
+      "loss": 0.7606,
+      "step": 897
+    },
+    {
+      "epoch": 0.8622179548727796,
+      "grad_norm": 1.1903531592349523,
+      "learning_rate": 9.842181834866593e-06,
+      "loss": 0.8757,
+      "step": 898
+    },
+    {
+      "epoch": 0.8631781084973595,
+      "grad_norm": 1.1879468386238274,
+      "learning_rate": 9.841345733992594e-06,
+      "loss": 0.6575,
+      "step": 899
+    },
+    {
+      "epoch": 0.8641382621219396,
+      "grad_norm": 1.087325624822077,
+      "learning_rate": 9.840507459890244e-06,
+      "loss": 0.6507,
+      "step": 900
+    },
+    {
+      "epoch": 0.8650984157465195,
+      "grad_norm": 0.9657237185910814,
+      "learning_rate": 9.839667012935838e-06,
+      "loss": 0.7462,
+      "step": 901
+    },
+    {
+      "epoch": 0.8660585693710994,
+      "grad_norm": 1.0414975352895643,
+      "learning_rate": 9.838824393506644e-06,
+      "loss": 0.7267,
+      "step": 902
+    },
+    {
+      "epoch": 0.8670187229956793,
+      "grad_norm": 0.9438593248708991,
+      "learning_rate": 9.837979601980902e-06,
+      "loss": 0.6806,
+      "step": 903
+    },
+    {
+      "epoch": 0.8679788766202592,
+      "grad_norm": 1.2800688399586553,
+      "learning_rate": 9.837132638737834e-06,
+      "loss": 0.7785,
+      "step": 904
+    },
+    {
+      "epoch": 0.8689390302448392,
+      "grad_norm": 1.0178473453974097,
+      "learning_rate": 9.83628350415763e-06,
+      "loss": 0.6255,
+      "step": 905
+    },
+    {
+      "epoch": 0.8698991838694191,
+      "grad_norm": 1.2880937767052882,
+      "learning_rate": 9.835432198621457e-06,
+      "loss": 0.6661,
+      "step": 906
+    },
+    {
+      "epoch": 0.870859337493999,
+      "grad_norm": 0.9608735784060488,
+      "learning_rate": 9.83457872251146e-06,
+      "loss": 0.6812,
+      "step": 907
+    },
+    {
+      "epoch": 0.8718194911185789,
+      "grad_norm": 1.0969764334445848,
+      "learning_rate": 9.833723076210751e-06,
+      "loss": 0.7732,
+      "step": 908
+    },
+    {
+      "epoch": 0.8727796447431589,
+      "grad_norm": 0.7722051052853002,
+      "learning_rate": 9.832865260103423e-06,
+      "loss": 0.5588,
+      "step": 909
+    },
+    {
+      "epoch": 0.8737397983677389,
+      "grad_norm": 1.5185930389007303,
+      "learning_rate": 9.832005274574541e-06,
+      "loss": 0.6865,
+      "step": 910
+    },
+    {
+      "epoch": 0.8746999519923188,
+      "grad_norm": 1.0466221007331493,
+      "learning_rate": 9.831143120010143e-06,
+      "loss": 0.7048,
+      "step": 911
+    },
+    {
+      "epoch": 0.8756601056168987,
+      "grad_norm": 1.2581603579966176,
+      "learning_rate": 9.830278796797239e-06,
+      "loss": 0.6606,
+      "step": 912
+    },
+    {
+      "epoch": 0.8766202592414787,
+      "grad_norm": 1.0031861202706405,
+      "learning_rate": 9.829412305323818e-06,
+      "loss": 0.7588,
+      "step": 913
+    },
+    {
+      "epoch": 0.8775804128660586,
+      "grad_norm": 1.5109231034827575,
+      "learning_rate": 9.828543645978835e-06,
+      "loss": 0.7686,
+      "step": 914
+    },
+    {
+      "epoch": 0.8785405664906385,
+      "grad_norm": 0.9541242793362736,
+      "learning_rate": 9.827672819152223e-06,
+      "loss": 0.7057,
+      "step": 915
+    },
+    {
+      "epoch": 0.8795007201152184,
+      "grad_norm": 0.930848836653686,
+      "learning_rate": 9.82679982523489e-06,
+      "loss": 0.8373,
+      "step": 916
+    },
+    {
+      "epoch": 0.8804608737397984,
+      "grad_norm": 0.8351213660877761,
+      "learning_rate": 9.82592466461871e-06,
+      "loss": 0.6362,
+      "step": 917
+    },
+    {
+      "epoch": 0.8814210273643783,
+      "grad_norm": 1.6665042808958377,
+      "learning_rate": 9.825047337696532e-06,
+      "loss": 0.8145,
+      "step": 918
+    },
+    {
+      "epoch": 0.8823811809889582,
+      "grad_norm": 1.3889228139305345,
+      "learning_rate": 9.824167844862184e-06,
+      "loss": 0.7571,
+      "step": 919
+    },
+    {
+      "epoch": 0.8833413346135381,
+      "grad_norm": 1.0552246147514572,
+      "learning_rate": 9.823286186510456e-06,
+      "loss": 0.6295,
+      "step": 920
+    },
+    {
+      "epoch": 0.884301488238118,
+      "grad_norm": 1.0503134578542208,
+      "learning_rate": 9.822402363037118e-06,
+      "loss": 0.7932,
+      "step": 921
+    },
+    {
+      "epoch": 0.8852616418626981,
+      "grad_norm": 1.1169129482469298,
+      "learning_rate": 9.82151637483891e-06,
+      "loss": 0.8022,
+      "step": 922
+    },
+    {
+      "epoch": 0.886221795487278,
+      "grad_norm": 1.0773362308593635,
+      "learning_rate": 9.820628222313535e-06,
+      "loss": 0.7533,
+      "step": 923
+    },
+    {
+      "epoch": 0.8871819491118579,
+      "grad_norm": 0.9450074031059397,
+      "learning_rate": 9.819737905859685e-06,
+      "loss": 0.6949,
+      "step": 924
+    },
+    {
+      "epoch": 0.8881421027364378,
+      "grad_norm": 0.9031119730881773,
+      "learning_rate": 9.818845425877008e-06,
+      "loss": 0.7283,
+      "step": 925
+    },
+    {
+      "epoch": 0.8891022563610178,
+      "grad_norm": 0.865666045566144,
+      "learning_rate": 9.81795078276613e-06,
+      "loss": 0.5821,
+      "step": 926
+    },
+    {
+      "epoch": 0.8900624099855977,
+      "grad_norm": 1.0100469478753122,
+      "learning_rate": 9.817053976928646e-06,
+      "loss": 0.7533,
+      "step": 927
+    },
+    {
+      "epoch": 0.8910225636101776,
+      "grad_norm": 1.0527936122369437,
+      "learning_rate": 9.816155008767123e-06,
+      "loss": 0.6932,
+      "step": 928
+    },
+    {
+      "epoch": 0.8919827172347575,
+      "grad_norm": 0.8333309807679432,
+      "learning_rate": 9.8152538786851e-06,
+      "loss": 0.6628,
+      "step": 929
+    },
+    {
+      "epoch": 0.8929428708593375,
+      "grad_norm": 1.366332958709172,
+      "learning_rate": 9.814350587087083e-06,
+      "loss": 0.7792,
+      "step": 930
+    },
+    {
+      "epoch": 0.8939030244839175,
+      "grad_norm": 0.9127210005336927,
+      "learning_rate": 9.813445134378548e-06,
+      "loss": 0.5963,
+      "step": 931
+    },
+    {
+      "epoch": 0.8948631781084974,
+      "grad_norm": 0.9688156315477393,
+      "learning_rate": 9.812537520965946e-06,
+      "loss": 0.722,
+      "step": 932
+    },
+    {
+      "epoch": 0.8958233317330773,
+      "grad_norm": 0.8734856874686121,
+      "learning_rate": 9.811627747256694e-06,
+      "loss": 0.6533,
+      "step": 933
+    },
+    {
+      "epoch": 0.8967834853576572,
+      "grad_norm": 1.1996538417789966,
+      "learning_rate": 9.81071581365918e-06,
+      "loss": 0.8021,
+      "step": 934
+    },
+    {
+      "epoch": 0.8977436389822372,
+      "grad_norm": 0.7782880837157595,
+      "learning_rate": 9.809801720582761e-06,
+      "loss": 0.6337,
+      "step": 935
+    },
+    {
+      "epoch": 0.8987037926068171,
+      "grad_norm": 1.241375510903179,
+      "learning_rate": 9.808885468437764e-06,
+      "loss": 0.8058,
+      "step": 936
+    },
+    {
+      "epoch": 0.899663946231397,
+      "grad_norm": 0.9984549357600859,
+      "learning_rate": 9.807967057635484e-06,
+      "loss": 0.6715,
+      "step": 937
+    },
+    {
+      "epoch": 0.9006240998559769,
+      "grad_norm": 1.612263573576596,
+      "learning_rate": 9.807046488588186e-06,
+      "loss": 0.7684,
+      "step": 938
+    },
+    {
+      "epoch": 0.9015842534805569,
+      "grad_norm": 0.9560953567318755,
+      "learning_rate": 9.806123761709103e-06,
+      "loss": 0.7007,
+      "step": 939
+    },
+    {
+      "epoch": 0.9025444071051368,
+      "grad_norm": 1.1696828477556798,
+      "learning_rate": 9.80519887741244e-06,
+      "loss": 0.7824,
+      "step": 940
+    },
+    {
+      "epoch": 0.9035045607297167,
+      "grad_norm": 1.2451480990025774,
+      "learning_rate": 9.804271836113364e-06,
+      "loss": 0.816,
+      "step": 941
+    },
+    {
+      "epoch": 0.9044647143542967,
+      "grad_norm": 0.9536640280225801,
+      "learning_rate": 9.803342638228014e-06,
+      "loss": 0.6372,
+      "step": 942
+    },
+    {
+      "epoch": 0.9054248679788767,
+      "grad_norm": 0.7391240801814135,
+      "learning_rate": 9.802411284173502e-06,
+      "loss": 0.612,
+      "step": 943
+    },
+    {
+      "epoch": 0.9063850216034566,
+      "grad_norm": 1.089277700491067,
+      "learning_rate": 9.801477774367898e-06,
+      "loss": 0.7988,
+      "step": 944
+    },
+    {
+      "epoch": 0.9073451752280365,
+      "grad_norm": 0.8675997079996093,
+      "learning_rate": 9.800542109230247e-06,
+      "loss": 0.6562,
+      "step": 945
+    },
+    {
+      "epoch": 0.9083053288526164,
+      "grad_norm": 1.1584731501395555,
+      "learning_rate": 9.799604289180556e-06,
+      "loss": 0.8249,
+      "step": 946
+    },
+    {
+      "epoch": 0.9092654824771963,
+      "grad_norm": 0.8931861272908558,
+      "learning_rate": 9.798664314639805e-06,
+      "loss": 0.6226,
+      "step": 947
+    },
+    {
+      "epoch": 0.9102256361017763,
+      "grad_norm": 0.7958676768875341,
+      "learning_rate": 9.79772218602994e-06,
+      "loss": 0.5626,
+      "step": 948
+    },
+    {
+      "epoch": 0.9111857897263562,
+      "grad_norm": 1.1274446320695066,
+      "learning_rate": 9.796777903773868e-06,
+      "loss": 0.7325,
+      "step": 949
+    },
+    {
+      "epoch": 0.9121459433509361,
+      "grad_norm": 1.6436918793587711,
+      "learning_rate": 9.795831468295472e-06,
+      "loss": 0.7576,
+      "step": 950
+    },
+    {
+      "epoch": 0.913106096975516,
+      "grad_norm": 0.9519199622538829,
+      "learning_rate": 9.794882880019594e-06,
+      "loss": 0.6777,
+      "step": 951
+    },
+    {
+      "epoch": 0.914066250600096,
+      "grad_norm": 1.3511811414293768,
+      "learning_rate": 9.793932139372045e-06,
+      "loss": 0.6944,
+      "step": 952
+    },
+    {
+      "epoch": 0.915026404224676,
+      "grad_norm": 1.6853823596747193,
+      "learning_rate": 9.792979246779602e-06,
+      "loss": 0.8167,
+      "step": 953
+    },
+    {
+      "epoch": 0.9159865578492559,
+      "grad_norm": 0.9679278304656862,
+      "learning_rate": 9.79202420267001e-06,
+      "loss": 0.6716,
+      "step": 954
+    },
+    {
+      "epoch": 0.9169467114738358,
+      "grad_norm": 0.8875846722638174,
+      "learning_rate": 9.791067007471977e-06,
+      "loss": 0.6776,
+      "step": 955
+    },
+    {
+      "epoch": 0.9179068650984158,
+      "grad_norm": 1.1772177263021604,
+      "learning_rate": 9.790107661615179e-06,
+      "loss": 0.7093,
+      "step": 956
+    },
+    {
+      "epoch": 0.9188670187229957,
+      "grad_norm": 1.1549620029922816,
+      "learning_rate": 9.789146165530255e-06,
+      "loss": 0.6411,
+      "step": 957
+    },
+    {
+      "epoch": 0.9198271723475756,
+      "grad_norm": 1.1434851514172506,
+      "learning_rate": 9.78818251964881e-06,
+      "loss": 0.7028,
+      "step": 958
+    },
+    {
+      "epoch": 0.9207873259721555,
+      "grad_norm": 0.8836920295059776,
+      "learning_rate": 9.787216724403414e-06,
+      "loss": 0.5954,
+      "step": 959
+    },
+    {
+      "epoch": 0.9217474795967355,
+      "grad_norm": 1.0899293586144565,
+      "learning_rate": 9.786248780227604e-06,
+      "loss": 0.7355,
+      "step": 960
+    },
+    {
+      "epoch": 0.9227076332213154,
+      "grad_norm": 1.190062067333841,
+      "learning_rate": 9.785278687555879e-06,
+      "loss": 0.6962,
+      "step": 961
+    },
+    {
+      "epoch": 0.9236677868458953,
+      "grad_norm": 1.0814165951280394,
+      "learning_rate": 9.784306446823702e-06,
+      "loss": 0.769,
+      "step": 962
+    },
+    {
+      "epoch": 0.9246279404704753,
+      "grad_norm": 1.1042095834041685,
+      "learning_rate": 9.783332058467502e-06,
+      "loss": 0.6473,
+      "step": 963
+    },
+    {
+      "epoch": 0.9255880940950552,
+      "grad_norm": 1.0550430585383388,
+      "learning_rate": 9.782355522924674e-06,
+      "loss": 0.7857,
+      "step": 964
+    },
+    {
+      "epoch": 0.9265482477196352,
+      "grad_norm": 1.0585925318826632,
+      "learning_rate": 9.78137684063357e-06,
+      "loss": 0.7238,
+      "step": 965
+    },
+    {
+      "epoch": 0.9275084013442151,
+      "grad_norm": 1.3454075899995044,
+      "learning_rate": 9.780396012033514e-06,
+      "loss": 0.7132,
+      "step": 966
+    },
+    {
+      "epoch": 0.928468554968795,
+      "grad_norm": 1.574673131762867,
+      "learning_rate": 9.779413037564787e-06,
+      "loss": 0.769,
+      "step": 967
+    },
+    {
+      "epoch": 0.9294287085933749,
+      "grad_norm": 0.8652122236817356,
+      "learning_rate": 9.778427917668636e-06,
+      "loss": 0.6542,
+      "step": 968
+    },
+    {
+      "epoch": 0.9303888622179549,
+      "grad_norm": 0.8921018355811567,
+      "learning_rate": 9.777440652787273e-06,
+      "loss": 0.6678,
+      "step": 969
+    },
+    {
+      "epoch": 0.9313490158425348,
+      "grad_norm": 1.1528185965641355,
+      "learning_rate": 9.776451243363866e-06,
+      "loss": 0.787,
+      "step": 970
+    },
+    {
+      "epoch": 0.9323091694671147,
+      "grad_norm": 1.1132653970572794,
+      "learning_rate": 9.775459689842558e-06,
+      "loss": 0.7405,
+      "step": 971
+    },
+    {
+      "epoch": 0.9332693230916946,
+      "grad_norm": 1.2393363136160531,
+      "learning_rate": 9.774465992668438e-06,
+      "loss": 0.7021,
+      "step": 972
+    },
+    {
+      "epoch": 0.9342294767162747,
+      "grad_norm": 0.9355066306206749,
+      "learning_rate": 9.773470152287573e-06,
+      "loss": 0.7447,
+      "step": 973
+    },
+    {
+      "epoch": 0.9351896303408546,
+      "grad_norm": 0.9415751218932975,
+      "learning_rate": 9.772472169146977e-06,
+      "loss": 0.6681,
+      "step": 974
+    },
+    {
+      "epoch": 0.9361497839654345,
+      "grad_norm": 1.0843407626427144,
+      "learning_rate": 9.771472043694645e-06,
+      "loss": 0.7294,
+      "step": 975
+    },
+    {
+      "epoch": 0.9371099375900144,
+      "grad_norm": 1.0410943714404997,
+      "learning_rate": 9.770469776379514e-06,
+      "loss": 0.7667,
+      "step": 976
+    },
+    {
+      "epoch": 0.9380700912145943,
+      "grad_norm": 0.8336942772176519,
+      "learning_rate": 9.769465367651494e-06,
+      "loss": 0.6097,
+      "step": 977
+    },
+    {
+      "epoch": 0.9390302448391743,
+      "grad_norm": 1.2244204622837596,
+      "learning_rate": 9.768458817961455e-06,
+      "loss": 0.6858,
+      "step": 978
+    },
+    {
+      "epoch": 0.9399903984637542,
+      "grad_norm": 0.8684284584573115,
+      "learning_rate": 9.767450127761223e-06,
+      "loss": 0.6282,
+      "step": 979
+    },
+    {
+      "epoch": 0.9409505520883341,
+      "grad_norm": 0.917142452064741,
+      "learning_rate": 9.766439297503589e-06,
+      "loss": 0.6797,
+      "step": 980
+    },
+    {
+      "epoch": 0.941910705712914,
+      "grad_norm": 1.0197470571452965,
+      "learning_rate": 9.765426327642306e-06,
+      "loss": 0.7103,
+      "step": 981
+    },
+    {
+      "epoch": 0.942870859337494,
+      "grad_norm": 1.3988824865834113,
+      "learning_rate": 9.764411218632085e-06,
+      "loss": 0.7972,
+      "step": 982
+    },
+    {
+      "epoch": 0.943831012962074,
+      "grad_norm": 1.1531592311801235,
+      "learning_rate": 9.763393970928593e-06,
+      "loss": 0.6764,
+      "step": 983
+    },
+    {
+      "epoch": 0.9447911665866539,
+      "grad_norm": 1.225196826877668,
+      "learning_rate": 9.76237458498847e-06,
+      "loss": 0.6423,
+      "step": 984
+    },
+    {
+      "epoch": 0.9457513202112338,
+      "grad_norm": 0.9255053798709998,
+      "learning_rate": 9.7613530612693e-06,
+      "loss": 0.6171,
+      "step": 985
+    },
+    {
+      "epoch": 0.9467114738358138,
+      "grad_norm": 1.1231221742640263,
+      "learning_rate": 9.760329400229639e-06,
+      "loss": 0.7753,
+      "step": 986
+    },
+    {
+      "epoch": 0.9476716274603937,
+      "grad_norm": 0.9184948985050901,
+      "learning_rate": 9.759303602328993e-06,
+      "loss": 0.7141,
+      "step": 987
+    },
+    {
+      "epoch": 0.9486317810849736,
+      "grad_norm": 0.9349566038732365,
+      "learning_rate": 9.758275668027834e-06,
+      "loss": 0.7295,
+      "step": 988
+    },
+    {
+      "epoch": 0.9495919347095535,
+      "grad_norm": 0.9080066148908356,
+      "learning_rate": 9.757245597787592e-06,
+      "loss": 0.7392,
+      "step": 989
+    },
+    {
+      "epoch": 0.9505520883341335,
+      "grad_norm": 1.269783362818564,
+      "learning_rate": 9.756213392070654e-06,
+      "loss": 0.7644,
+      "step": 990
+    },
+    {
+      "epoch": 0.9515122419587134,
+      "grad_norm": 0.8583944516713555,
+      "learning_rate": 9.755179051340365e-06,
+      "loss": 0.6627,
+      "step": 991
+    },
+    {
+      "epoch": 0.9524723955832933,
+      "grad_norm": 0.9155158800596387,
+      "learning_rate": 9.75414257606103e-06,
+      "loss": 0.6376,
+      "step": 992
+    },
+    {
+      "epoch": 0.9534325492078732,
+      "grad_norm": 1.208151950924738,
+      "learning_rate": 9.75310396669791e-06,
+      "loss": 0.7503,
+      "step": 993
+    },
+    {
+      "epoch": 0.9543927028324531,
+      "grad_norm": 1.033989677497992,
+      "learning_rate": 9.752063223717227e-06,
+      "loss": 0.7704,
+      "step": 994
+    },
+    {
+      "epoch": 0.9553528564570332,
+      "grad_norm": 1.210710566475916,
+      "learning_rate": 9.75102034758616e-06,
+      "loss": 0.5415,
+      "step": 995
+    },
+    {
+      "epoch": 0.9563130100816131,
+      "grad_norm": 1.7618543655335772,
+      "learning_rate": 9.749975338772846e-06,
+      "loss": 0.8394,
+      "step": 996
+    },
+    {
+      "epoch": 0.957273163706193,
+      "grad_norm": 0.9979885638324807,
+      "learning_rate": 9.748928197746376e-06,
+      "loss": 0.5887,
+      "step": 997
+    },
+    {
+      "epoch": 0.9582333173307729,
+      "grad_norm": 1.4603713511847616,
+      "learning_rate": 9.7478789249768e-06,
+      "loss": 0.7873,
+      "step": 998
+    },
+    {
+      "epoch": 0.9591934709553529,
+      "grad_norm": 0.8390123183988712,
+      "learning_rate": 9.746827520935129e-06,
+      "loss": 0.6434,
+      "step": 999
+    },
+    {
+      "epoch": 0.9601536245799328,
+      "grad_norm": 1.0913093884651996,
+      "learning_rate": 9.745773986093325e-06,
+      "loss": 0.6331,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9611137782045127,
+      "grad_norm": 1.7215798127009005,
+      "learning_rate": 9.744718320924307e-06,
+      "loss": 0.8343,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9620739318290926,
+      "grad_norm": 1.0625049746786475,
+      "learning_rate": 9.743660525901952e-06,
+      "loss": 0.6828,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9630340854536726,
+      "grad_norm": 1.0133495556448957,
+      "learning_rate": 9.742600601501095e-06,
+      "loss": 0.6493,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9639942390782525,
+      "grad_norm": 0.9451001954187663,
+      "learning_rate": 9.741538548197522e-06,
+      "loss": 0.7463,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9649543927028325,
+      "grad_norm": 1.3277949374129938,
+      "learning_rate": 9.74047436646798e-06,
+      "loss": 0.7634,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9659145463274124,
+      "grad_norm": 1.3966822263033054,
+      "learning_rate": 9.739408056790168e-06,
+      "loss": 0.7821,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9668746999519923,
+      "grad_norm": 1.5078577294435678,
+      "learning_rate": 9.73833961964274e-06,
+      "loss": 0.8854,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9678348535765723,
+      "grad_norm": 1.4087987293027775,
+      "learning_rate": 9.737269055505308e-06,
+      "loss": 0.7581,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9687950072011522,
+      "grad_norm": 1.0960917211859005,
+      "learning_rate": 9.736196364858435e-06,
+      "loss": 0.6522,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9697551608257321,
+      "grad_norm": 1.3430973961906536,
+      "learning_rate": 9.735121548183642e-06,
+      "loss": 0.797,
+      "step": 1010
+    },
+    {
+      "epoch": 0.970715314450312,
+      "grad_norm": 0.9293505165863168,
+      "learning_rate": 9.7340446059634e-06,
+      "loss": 0.5644,
+      "step": 1011
+    },
+    {
+      "epoch": 0.971675468074892,
+      "grad_norm": 1.030867247805389,
+      "learning_rate": 9.732965538681139e-06,
+      "loss": 0.7227,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9726356216994719,
+      "grad_norm": 0.9622507208004649,
+      "learning_rate": 9.73188434682124e-06,
+      "loss": 0.6653,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9735957753240518,
+      "grad_norm": 1.2255800005686928,
+      "learning_rate": 9.73080103086904e-06,
+      "loss": 0.8339,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9745559289486317,
+      "grad_norm": 1.0563989893236743,
+      "learning_rate": 9.729715591310824e-06,
+      "loss": 0.645,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9755160825732118,
+      "grad_norm": 0.9398333861628091,
+      "learning_rate": 9.728628028633839e-06,
+      "loss": 0.7346,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9764762361977917,
+      "grad_norm": 1.0391255896214462,
+      "learning_rate": 9.727538343326277e-06,
+      "loss": 0.8132,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9774363898223716,
+      "grad_norm": 1.469070299585559,
+      "learning_rate": 9.72644653587729e-06,
+      "loss": 0.9106,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9783965434469515,
+      "grad_norm": 0.9981697225057418,
+      "learning_rate": 9.725352606776973e-06,
+      "loss": 0.7238,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9793566970715314,
+      "grad_norm": 0.7857074918295711,
+      "learning_rate": 9.724256556516383e-06,
+      "loss": 0.6378,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9803168506961114,
+      "grad_norm": 1.3651912217413142,
+      "learning_rate": 9.723158385587525e-06,
+      "loss": 0.7986,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9812770043206913,
+      "grad_norm": 1.255367872502659,
+      "learning_rate": 9.722058094483355e-06,
+      "loss": 0.7143,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9822371579452712,
+      "grad_norm": 0.9567562529683361,
+      "learning_rate": 9.72095568369778e-06,
+      "loss": 0.6067,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9831973115698511,
+      "grad_norm": 1.135525880037657,
+      "learning_rate": 9.719851153725668e-06,
+      "loss": 0.7057,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9841574651944311,
+      "grad_norm": 1.0612664802665412,
+      "learning_rate": 9.718744505062822e-06,
+      "loss": 0.6853,
+      "step": 1025
+    },
+    {
+      "epoch": 0.985117618819011,
+      "grad_norm": 1.2485057106455786,
+      "learning_rate": 9.717635738206007e-06,
+      "loss": 0.6469,
+      "step": 1026
+    },
+    {
+      "epoch": 0.986077772443591,
+      "grad_norm": 1.0333640641052955,
+      "learning_rate": 9.716524853652942e-06,
+      "loss": 0.7068,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9870379260681709,
+      "grad_norm": 1.1254910643496672,
+      "learning_rate": 9.715411851902284e-06,
+      "loss": 0.8391,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9879980796927509,
+      "grad_norm": 1.1589838064442606,
+      "learning_rate": 9.71429673345365e-06,
+      "loss": 0.7025,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9889582333173308,
+      "grad_norm": 1.3250429014521803,
+      "learning_rate": 9.713179498807607e-06,
+      "loss": 0.7709,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9899183869419107,
+      "grad_norm": 1.4416705150751818,
+      "learning_rate": 9.712060148465669e-06,
+      "loss": 0.6959,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9908785405664906,
+      "grad_norm": 1.1092477155984324,
+      "learning_rate": 9.710938682930298e-06,
+      "loss": 0.6921,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9918386941910706,
+      "grad_norm": 1.142382995678008,
+      "learning_rate": 9.709815102704909e-06,
+      "loss": 0.6984,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9927988478156505,
+      "grad_norm": 1.2806339587976485,
+      "learning_rate": 9.708689408293866e-06,
+      "loss": 0.8489,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9937590014402304,
+      "grad_norm": 1.5699760322221297,
+      "learning_rate": 9.707561600202481e-06,
+      "loss": 0.7757,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9947191550648103,
+      "grad_norm": 0.9806876764853485,
+      "learning_rate": 9.706431678937015e-06,
+      "loss": 0.6846,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9956793086893903,
+      "grad_norm": 1.0528504760370094,
+      "learning_rate": 9.705299645004679e-06,
+      "loss": 0.6947,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9966394623139703,
+      "grad_norm": 0.9003852775733797,
+      "learning_rate": 9.704165498913624e-06,
+      "loss": 0.6969,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9975996159385502,
+      "grad_norm": 1.0114923779312535,
+      "learning_rate": 9.703029241172965e-06,
+      "loss": 0.6475,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9985597695631301,
+      "grad_norm": 1.3534167661452765,
+      "learning_rate": 9.701890872292753e-06,
+      "loss": 0.7962,
+      "step": 1040
+    },
+    {
+      "epoch": 0.99951992318771,
+      "grad_norm": 1.2702910771854647,
+      "learning_rate": 9.700750392783986e-06,
+      "loss": 0.6732,
+      "step": 1041
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2702910771854647,
+      "learning_rate": 9.699607803158618e-06,
+      "loss": 0.532,
+      "step": 1042
+    },
+    {
+      "epoch": 1.00096015362458,
+      "grad_norm": 1.5110487943252837,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.5891,
+      "step": 1043
+    },
+    {
+      "epoch": 1.0019203072491598,
+      "grad_norm": 0.7969303706215065,
+      "learning_rate": 9.697316295610604e-06,
+      "loss": 0.5777,
+      "step": 1044
+    },
+    {
+      "epoch": 1.0028804608737398,
+      "grad_norm": 1.5293806184918348,
+      "learning_rate": 9.696167378716591e-06,
+      "loss": 0.7407,
+      "step": 1045
+    },
+    {
+      "epoch": 1.0038406144983196,
+      "grad_norm": 1.1236794712179317,
+      "learning_rate": 9.695016353763241e-06,
+      "loss": 0.632,
+      "step": 1046
+    },
+    {
+      "epoch": 1.0048007681228996,
+      "grad_norm": 1.1510769348363972,
+      "learning_rate": 9.693863221267238e-06,
+      "loss": 0.601,
+      "step": 1047
+    },
+    {
+      "epoch": 1.0057609217474797,
+      "grad_norm": 1.0897212700058942,
+      "learning_rate": 9.692707981746208e-06,
+      "loss": 0.6952,
+      "step": 1048
+    },
+    {
+      "epoch": 1.0067210753720595,
+      "grad_norm": 1.0611847600479067,
+      "learning_rate": 9.691550635718729e-06,
+      "loss": 0.6969,
+      "step": 1049
+    },
+    {
+      "epoch": 1.0076812289966395,
+      "grad_norm": 1.207102458712394,
+      "learning_rate": 9.690391183704318e-06,
+      "loss": 0.6293,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0086413826212195,
+      "grad_norm": 0.9751133947510949,
+      "learning_rate": 9.689229626223441e-06,
+      "loss": 0.7174,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0096015362457993,
+      "grad_norm": 1.000000656715612,
+      "learning_rate": 9.688065963797512e-06,
+      "loss": 0.5724,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0105616898703793,
+      "grad_norm": 1.4651037042965558,
+      "learning_rate": 9.686900196948886e-06,
+      "loss": 0.6057,
+      "step": 1053
+    },
+    {
+      "epoch": 1.011521843494959,
+      "grad_norm": 1.0289676237018328,
+      "learning_rate": 9.685732326200859e-06,
+      "loss": 0.6361,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0124819971195391,
+      "grad_norm": 1.3618297408799271,
+      "learning_rate": 9.684562352077679e-06,
+      "loss": 0.7034,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0134421507441191,
+      "grad_norm": 1.0337606076511678,
+      "learning_rate": 9.683390275104533e-06,
+      "loss": 0.6149,
+      "step": 1056
+    },
+    {
+      "epoch": 1.014402304368699,
+      "grad_norm": 1.3210540444281758,
+      "learning_rate": 9.682216095807559e-06,
+      "loss": 0.7235,
+      "step": 1057
+    },
+    {
+      "epoch": 1.015362457993279,
+      "grad_norm": 1.2917468489810855,
+      "learning_rate": 9.681039814713827e-06,
+      "loss": 0.635,
+      "step": 1058
+    },
+    {
+      "epoch": 1.0163226116178588,
+      "grad_norm": 1.3111956888965486,
+      "learning_rate": 9.67986143235136e-06,
+      "loss": 0.7257,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0172827652424388,
+      "grad_norm": 1.149137628716981,
+      "learning_rate": 9.67868094924912e-06,
+      "loss": 0.6367,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0182429188670188,
+      "grad_norm": 1.0565001299585075,
+      "learning_rate": 9.677498365937017e-06,
+      "loss": 0.5431,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0192030724915986,
+      "grad_norm": 1.000820884997687,
+      "learning_rate": 9.676313682945896e-06,
+      "loss": 0.6542,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0201632261161786,
+      "grad_norm": 1.3601671794282044,
+      "learning_rate": 9.675126900807548e-06,
+      "loss": 0.6957,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0211233797407586,
+      "grad_norm": 1.084951527919392,
+      "learning_rate": 9.673938020054709e-06,
+      "loss": 0.5765,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0220835333653384,
+      "grad_norm": 1.166456668075392,
+      "learning_rate": 9.672747041221055e-06,
+      "loss": 0.6789,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0230436869899184,
+      "grad_norm": 0.887458603305072,
+      "learning_rate": 9.671553964841206e-06,
+      "loss": 0.5923,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0240038406144982,
+      "grad_norm": 0.6994294691373585,
+      "learning_rate": 9.670358791450713e-06,
+      "loss": 0.6856,
+      "step": 1067
+    },
+    {
+      "epoch": 1.0249639942390782,
+      "grad_norm": 1.2050996724454637,
+      "learning_rate": 9.669161521586086e-06,
+      "loss": 0.7988,
+      "step": 1068
+    },
+    {
+      "epoch": 1.0259241478636583,
+      "grad_norm": 1.3281032244939457,
+      "learning_rate": 9.66796215578476e-06,
+      "loss": 0.7569,
+      "step": 1069
+    },
+    {
+      "epoch": 1.026884301488238,
+      "grad_norm": 1.3492642159780655,
+      "learning_rate": 9.666760694585122e-06,
+      "loss": 0.6256,
+      "step": 1070
+    },
+    {
+      "epoch": 1.027844455112818,
+      "grad_norm": 1.3745429981764834,
+      "learning_rate": 9.665557138526492e-06,
+      "loss": 0.7721,
+      "step": 1071
+    },
+    {
+      "epoch": 1.0288046087373979,
+      "grad_norm": 0.9822068570711285,
+      "learning_rate": 9.664351488149136e-06,
+      "loss": 0.7684,
+      "step": 1072
+    },
+    {
+      "epoch": 1.029764762361978,
+      "grad_norm": 1.2721803352495713,
+      "learning_rate": 9.663143743994254e-06,
+      "loss": 0.6653,
+      "step": 1073
+    },
+    {
+      "epoch": 1.030724915986558,
+      "grad_norm": 1.041552222503309,
+      "learning_rate": 9.661933906603994e-06,
+      "loss": 0.7034,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0316850696111377,
+      "grad_norm": 1.1837751802635352,
+      "learning_rate": 9.660721976521433e-06,
+      "loss": 0.617,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0326452232357177,
+      "grad_norm": 1.4451863025436755,
+      "learning_rate": 9.659507954290599e-06,
+      "loss": 0.5942,
+      "step": 1076
+    },
+    {
+      "epoch": 1.0336053768602977,
+      "grad_norm": 0.8662904432223953,
+      "learning_rate": 9.658291840456454e-06,
+      "loss": 0.4988,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0345655304848775,
+      "grad_norm": 0.9639625129611855,
+      "learning_rate": 9.657073635564893e-06,
+      "loss": 0.5872,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0355256841094576,
+      "grad_norm": 0.98202749070384,
+      "learning_rate": 9.65585334016276e-06,
+      "loss": 0.6139,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0364858377340374,
+      "grad_norm": 1.0033748612039537,
+      "learning_rate": 9.65463095479783e-06,
+      "loss": 0.526,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0374459913586174,
+      "grad_norm": 1.0286083707592373,
+      "learning_rate": 9.653406480018819e-06,
+      "loss": 0.6296,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0384061449831974,
+      "grad_norm": 1.4186018864785659,
+      "learning_rate": 9.652179916375382e-06,
+      "loss": 0.6323,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0393662986077772,
+      "grad_norm": 1.048987295491667,
+      "learning_rate": 9.65095126441811e-06,
+      "loss": 0.5984,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0403264522323572,
+      "grad_norm": 1.1790098582257498,
+      "learning_rate": 9.649720524698532e-06,
+      "loss": 0.6916,
+      "step": 1084
+    },
+    {
+      "epoch": 1.041286605856937,
+      "grad_norm": 0.9556042829892353,
+      "learning_rate": 9.648487697769112e-06,
+      "loss": 0.643,
+      "step": 1085
+    },
+    {
+      "epoch": 1.042246759481517,
+      "grad_norm": 0.8133611348102179,
+      "learning_rate": 9.647252784183254e-06,
+      "loss": 0.6535,
+      "step": 1086
+    },
+    {
+      "epoch": 1.043206913106097,
+      "grad_norm": 0.9897981422575274,
+      "learning_rate": 9.646015784495297e-06,
+      "loss": 0.6047,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0441670667306768,
+      "grad_norm": 1.2267752645305483,
+      "learning_rate": 9.64477669926052e-06,
+      "loss": 0.6075,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0451272203552568,
+      "grad_norm": 1.3238292218908252,
+      "learning_rate": 9.643535529035131e-06,
+      "loss": 0.7051,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0460873739798369,
+      "grad_norm": 1.5632993144952576,
+      "learning_rate": 9.64229227437628e-06,
+      "loss": 0.6622,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0470475276044167,
+      "grad_norm": 1.2863923805489401,
+      "learning_rate": 9.641046935842048e-06,
+      "loss": 0.5829,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0480076812289967,
+      "grad_norm": 0.9758855494658143,
+      "learning_rate": 9.63979951399146e-06,
+      "loss": 0.671,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0489678348535765,
+      "grad_norm": 1.22400635676477,
+      "learning_rate": 9.638550009384467e-06,
+      "loss": 0.7383,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0499279884781565,
+      "grad_norm": 0.8371531266770208,
+      "learning_rate": 9.637298422581955e-06,
+      "loss": 0.5952,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0508881421027365,
+      "grad_norm": 0.7397192551218993,
+      "learning_rate": 9.636044754145753e-06,
+      "loss": 0.6145,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0518482957273163,
+      "grad_norm": 1.2804590644356055,
+      "learning_rate": 9.634789004638618e-06,
+      "loss": 0.6465,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0528084493518963,
+      "grad_norm": 0.7206744162417552,
+      "learning_rate": 9.63353117462424e-06,
+      "loss": 0.6666,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0537686029764763,
+      "grad_norm": 1.1434195814155435,
+      "learning_rate": 9.632271264667249e-06,
+      "loss": 0.587,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0547287566010561,
+      "grad_norm": 1.0290623304729012,
+      "learning_rate": 9.631009275333204e-06,
+      "loss": 0.6467,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0556889102256362,
+      "grad_norm": 1.1360460171994695,
+      "learning_rate": 9.629745207188596e-06,
+      "loss": 0.8371,
+      "step": 1100
+    },
+    {
+      "epoch": 1.056649063850216,
+      "grad_norm": 1.1453988893096245,
+      "learning_rate": 9.628479060800855e-06,
+      "loss": 0.6032,
+      "step": 1101
+    },
+    {
+      "epoch": 1.057609217474796,
+      "grad_norm": 0.7128857923854224,
+      "learning_rate": 9.627210836738338e-06,
+      "loss": 0.6731,
+      "step": 1102
+    },
+    {
+      "epoch": 1.058569371099376,
+      "grad_norm": 1.466527768079591,
+      "learning_rate": 9.62594053557034e-06,
+      "loss": 0.7217,
+      "step": 1103
+    },
+    {
+      "epoch": 1.0595295247239558,
+      "grad_norm": 1.416984961465442,
+      "learning_rate": 9.624668157867084e-06,
+      "loss": 0.6269,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0604896783485358,
+      "grad_norm": 1.3810371757039088,
+      "learning_rate": 9.623393704199727e-06,
+      "loss": 0.5632,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0614498319731156,
+      "grad_norm": 0.9934662594978854,
+      "learning_rate": 9.622117175140357e-06,
+      "loss": 0.7161,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0624099855976956,
+      "grad_norm": 1.2094727838376098,
+      "learning_rate": 9.620838571261995e-06,
+      "loss": 0.7962,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0633701392222756,
+      "grad_norm": 1.2298114570943566,
+      "learning_rate": 9.619557893138593e-06,
+      "loss": 0.6097,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0643302928468554,
+      "grad_norm": 1.6030139988398944,
+      "learning_rate": 9.618275141345032e-06,
+      "loss": 0.6341,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0652904464714354,
+      "grad_norm": 1.4528068307728832,
+      "learning_rate": 9.616990316457127e-06,
+      "loss": 0.6742,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0662506000960155,
+      "grad_norm": 1.4810340725441336,
+      "learning_rate": 9.615703419051622e-06,
+      "loss": 0.6646,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0672107537205953,
+      "grad_norm": 1.2063301052611641,
+      "learning_rate": 9.614414449706191e-06,
+      "loss": 0.5886,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0681709073451753,
+      "grad_norm": 0.9906898992808743,
+      "learning_rate": 9.61312340899944e-06,
+      "loss": 0.6759,
+      "step": 1113
+    },
+    {
+      "epoch": 1.069131060969755,
+      "grad_norm": 0.8954179234598504,
+      "learning_rate": 9.6118302975109e-06,
+      "loss": 0.6811,
+      "step": 1114
+    },
+    {
+      "epoch": 1.070091214594335,
+      "grad_norm": 1.0562686078833756,
+      "learning_rate": 9.61053511582104e-06,
+      "loss": 0.6808,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0710513682189151,
+      "grad_norm": 1.151123613333362,
+      "learning_rate": 9.609237864511249e-06,
+      "loss": 0.7211,
+      "step": 1116
+    },
+    {
+      "epoch": 1.072011521843495,
+      "grad_norm": 1.0924995827667974,
+      "learning_rate": 9.607938544163851e-06,
+      "loss": 0.751,
+      "step": 1117
+    },
+    {
+      "epoch": 1.072971675468075,
+      "grad_norm": 1.3204474803168247,
+      "learning_rate": 9.606637155362095e-06,
+      "loss": 0.6516,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0739318290926547,
+      "grad_norm": 1.2692506872886318,
+      "learning_rate": 9.605333698690164e-06,
+      "loss": 0.6404,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0748919827172347,
+      "grad_norm": 1.1176340089741394,
+      "learning_rate": 9.604028174733164e-06,
+      "loss": 0.6189,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0758521363418148,
+      "grad_norm": 1.001771172149186,
+      "learning_rate": 9.60272058407713e-06,
+      "loss": 0.6097,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0768122899663946,
+      "grad_norm": 0.9915205510414012,
+      "learning_rate": 9.601410927309026e-06,
+      "loss": 0.5683,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0777724435909746,
+      "grad_norm": 1.2719858441510843,
+      "learning_rate": 9.600099205016744e-06,
+      "loss": 0.6787,
+      "step": 1123
+    },
+    {
+      "epoch": 1.0787325972155546,
+      "grad_norm": 1.3448172903889624,
+      "learning_rate": 9.5987854177891e-06,
+      "loss": 0.6033,
+      "step": 1124
+    },
+    {
+      "epoch": 1.0796927508401344,
+      "grad_norm": 1.3101282348556873,
+      "learning_rate": 9.597469566215841e-06,
+      "loss": 0.6452,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0806529044647144,
+      "grad_norm": 0.9126392619552218,
+      "learning_rate": 9.596151650887637e-06,
+      "loss": 0.744,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0816130580892942,
+      "grad_norm": 1.096678883600679,
+      "learning_rate": 9.594831672396088e-06,
+      "loss": 0.5663,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0825732117138742,
+      "grad_norm": 1.2189680502188556,
+      "learning_rate": 9.593509631333717e-06,
+      "loss": 0.547,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0835333653384542,
+      "grad_norm": 1.0132906792885186,
+      "learning_rate": 9.592185528293975e-06,
+      "loss": 0.5943,
+      "step": 1129
+    },
+    {
+      "epoch": 1.084493518963034,
+      "grad_norm": 1.4576556473526279,
+      "learning_rate": 9.590859363871237e-06,
+      "loss": 0.6347,
+      "step": 1130
+    },
+    {
+      "epoch": 1.085453672587614,
+      "grad_norm": 1.401688560160359,
+      "learning_rate": 9.589531138660803e-06,
+      "loss": 0.6151,
+      "step": 1131
+    },
+    {
+      "epoch": 1.086413826212194,
+      "grad_norm": 0.9310716344825233,
+      "learning_rate": 9.588200853258901e-06,
+      "loss": 0.6459,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0873739798367739,
+      "grad_norm": 1.4211724125100058,
+      "learning_rate": 9.586868508262681e-06,
+      "loss": 0.6429,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0883341334613539,
+      "grad_norm": 1.3769271746505671,
+      "learning_rate": 9.58553410427022e-06,
+      "loss": 0.597,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0892942870859337,
+      "grad_norm": 0.8611070201311517,
+      "learning_rate": 9.584197641880513e-06,
+      "loss": 0.6302,
+      "step": 1135
+    },
+    {
+      "epoch": 1.0902544407105137,
+      "grad_norm": 0.8510601621483593,
+      "learning_rate": 9.582859121693488e-06,
+      "loss": 0.6772,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0912145943350937,
+      "grad_norm": 0.8837900327645241,
+      "learning_rate": 9.581518544309993e-06,
+      "loss": 0.6534,
+      "step": 1137
+    },
+    {
+      "epoch": 1.0921747479596735,
+      "grad_norm": 1.294948526970718,
+      "learning_rate": 9.580175910331797e-06,
+      "loss": 0.6335,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0931349015842535,
+      "grad_norm": 0.8466550562239562,
+      "learning_rate": 9.578831220361593e-06,
+      "loss": 0.7251,
+      "step": 1139
+    },
+    {
+      "epoch": 1.0940950552088333,
+      "grad_norm": 1.1271084988915,
+      "learning_rate": 9.577484475003e-06,
+      "loss": 0.6197,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0950552088334133,
+      "grad_norm": 1.0209448673116264,
+      "learning_rate": 9.576135674860557e-06,
+      "loss": 0.5503,
+      "step": 1141
+    },
+    {
+      "epoch": 1.0960153624579934,
+      "grad_norm": 1.0666726258807853,
+      "learning_rate": 9.574784820539725e-06,
+      "loss": 0.7134,
+      "step": 1142
+    },
+    {
+      "epoch": 1.0969755160825732,
+      "grad_norm": 1.3085202162459826,
+      "learning_rate": 9.573431912646888e-06,
+      "loss": 0.5638,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0979356697071532,
+      "grad_norm": 1.1263614260396178,
+      "learning_rate": 9.572076951789353e-06,
+      "loss": 0.6075,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0988958233317332,
+      "grad_norm": 1.1508304501793614,
+      "learning_rate": 9.570719938575349e-06,
+      "loss": 0.7034,
+      "step": 1145
+    },
+    {
+      "epoch": 1.099855976956313,
+      "grad_norm": 1.2457271466884658,
+      "learning_rate": 9.56936087361402e-06,
+      "loss": 0.664,
+      "step": 1146
+    },
+    {
+      "epoch": 1.100816130580893,
+      "grad_norm": 1.0148341390353592,
+      "learning_rate": 9.56799975751544e-06,
+      "loss": 0.5465,
+      "step": 1147
+    },
+    {
+      "epoch": 1.1017762842054728,
+      "grad_norm": 0.972231722233613,
+      "learning_rate": 9.566636590890598e-06,
+      "loss": 0.7838,
+      "step": 1148
+    },
+    {
+      "epoch": 1.1027364378300528,
+      "grad_norm": 1.118725517314477,
+      "learning_rate": 9.565271374351405e-06,
+      "loss": 0.6724,
+      "step": 1149
+    },
+    {
+      "epoch": 1.1036965914546328,
+      "grad_norm": 1.1280201861194867,
+      "learning_rate": 9.563904108510693e-06,
+      "loss": 0.6967,
+      "step": 1150
+    },
+    {
+      "epoch": 1.1046567450792126,
+      "grad_norm": 0.9010003252714744,
+      "learning_rate": 9.562534793982212e-06,
+      "loss": 0.6369,
+      "step": 1151
+    },
+    {
+      "epoch": 1.1056168987037926,
+      "grad_norm": 1.0858740201525623,
+      "learning_rate": 9.56116343138063e-06,
+      "loss": 0.6405,
+      "step": 1152
+    },
+    {
+      "epoch": 1.1065770523283724,
+      "grad_norm": 1.4044943630930733,
+      "learning_rate": 9.559790021321544e-06,
+      "loss": 0.7943,
+      "step": 1153
+    },
+    {
+      "epoch": 1.1075372059529525,
+      "grad_norm": 1.4113371109108261,
+      "learning_rate": 9.558414564421456e-06,
+      "loss": 0.6222,
+      "step": 1154
+    },
+    {
+      "epoch": 1.1084973595775325,
+      "grad_norm": 1.0207132568930994,
+      "learning_rate": 9.557037061297799e-06,
+      "loss": 0.594,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1094575132021123,
+      "grad_norm": 1.0170877698543306,
+      "learning_rate": 9.555657512568916e-06,
+      "loss": 0.62,
+      "step": 1156
+    },
+    {
+      "epoch": 1.1104176668266923,
+      "grad_norm": 1.1261083897157445,
+      "learning_rate": 9.554275918854073e-06,
+      "loss": 0.6561,
+      "step": 1157
+    },
+    {
+      "epoch": 1.1113778204512723,
+      "grad_norm": 1.105312343576896,
+      "learning_rate": 9.55289228077345e-06,
+      "loss": 0.6724,
+      "step": 1158
+    },
+    {
+      "epoch": 1.112337974075852,
+      "grad_norm": 1.1372175467826011,
+      "learning_rate": 9.551506598948152e-06,
+      "loss": 0.6643,
+      "step": 1159
+    },
+    {
+      "epoch": 1.1132981277004321,
+      "grad_norm": 1.2311915822855724,
+      "learning_rate": 9.550118874000196e-06,
+      "loss": 0.5947,
+      "step": 1160
+    },
+    {
+      "epoch": 1.114258281325012,
+      "grad_norm": 0.8672811136561209,
+      "learning_rate": 9.548729106552515e-06,
+      "loss": 0.6416,
+      "step": 1161
+    },
+    {
+      "epoch": 1.115218434949592,
+      "grad_norm": 1.0203158135277313,
+      "learning_rate": 9.54733729722896e-06,
+      "loss": 0.6898,
+      "step": 1162
+    },
+    {
+      "epoch": 1.116178588574172,
+      "grad_norm": 1.4296343327952086,
+      "learning_rate": 9.5459434466543e-06,
+      "loss": 0.5914,
+      "step": 1163
+    },
+    {
+      "epoch": 1.1171387421987518,
+      "grad_norm": 1.0449614731472696,
+      "learning_rate": 9.544547555454222e-06,
+      "loss": 0.6339,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1180988958233318,
+      "grad_norm": 1.1184734718295897,
+      "learning_rate": 9.543149624255326e-06,
+      "loss": 0.6805,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1190590494479116,
+      "grad_norm": 1.5050757753662851,
+      "learning_rate": 9.541749653685124e-06,
+      "loss": 0.6398,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1200192030724916,
+      "grad_norm": 1.0627861251187731,
+      "learning_rate": 9.540347644372054e-06,
+      "loss": 0.6408,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1209793566970716,
+      "grad_norm": 1.1010833442250243,
+      "learning_rate": 9.538943596945459e-06,
+      "loss": 0.7185,
+      "step": 1168
+    },
+    {
+      "epoch": 1.1219395103216514,
+      "grad_norm": 1.2756547962854146,
+      "learning_rate": 9.5375375120356e-06,
+      "loss": 0.5925,
+      "step": 1169
+    },
+    {
+      "epoch": 1.1228996639462314,
+      "grad_norm": 0.8752539321638535,
+      "learning_rate": 9.536129390273659e-06,
+      "loss": 0.6565,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1238598175708114,
+      "grad_norm": 1.0638578949842945,
+      "learning_rate": 9.53471923229172e-06,
+      "loss": 0.5382,
+      "step": 1171
+    },
+    {
+      "epoch": 1.1248199711953912,
+      "grad_norm": 1.2180757484115217,
+      "learning_rate": 9.533307038722792e-06,
+      "loss": 0.7057,
+      "step": 1172
+    },
+    {
+      "epoch": 1.1257801248199713,
+      "grad_norm": 0.8609033731653876,
+      "learning_rate": 9.531892810200793e-06,
+      "loss": 0.6092,
+      "step": 1173
+    },
+    {
+      "epoch": 1.126740278444551,
+      "grad_norm": 0.8704390821138364,
+      "learning_rate": 9.530476547360555e-06,
+      "loss": 0.576,
+      "step": 1174
+    },
+    {
+      "epoch": 1.127700432069131,
+      "grad_norm": 1.1661236177930903,
+      "learning_rate": 9.529058250837823e-06,
+      "loss": 0.6458,
+      "step": 1175
+    },
+    {
+      "epoch": 1.128660585693711,
+      "grad_norm": 0.9868914515322151,
+      "learning_rate": 9.527637921269256e-06,
+      "loss": 0.616,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1296207393182909,
+      "grad_norm": 1.1547561709579266,
+      "learning_rate": 9.526215559292423e-06,
+      "loss": 0.6516,
+      "step": 1177
+    },
+    {
+      "epoch": 1.130580892942871,
+      "grad_norm": 1.2761874113530798,
+      "learning_rate": 9.524791165545808e-06,
+      "loss": 0.6422,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1315410465674507,
+      "grad_norm": 0.7061153931590968,
+      "learning_rate": 9.523364740668805e-06,
+      "loss": 0.6609,
+      "step": 1179
+    },
+    {
+      "epoch": 1.1325012001920307,
+      "grad_norm": 0.9106969571888363,
+      "learning_rate": 9.521936285301723e-06,
+      "loss": 0.6723,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1334613538166107,
+      "grad_norm": 1.2465677960183414,
+      "learning_rate": 9.52050580008578e-06,
+      "loss": 0.6674,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1344215074411905,
+      "grad_norm": 1.1124485532076975,
+      "learning_rate": 9.519073285663103e-06,
+      "loss": 0.6539,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1353816610657705,
+      "grad_norm": 1.0718994744889718,
+      "learning_rate": 9.517638742676736e-06,
+      "loss": 0.6634,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1363418146903506,
+      "grad_norm": 1.123650272367286,
+      "learning_rate": 9.516202171770627e-06,
+      "loss": 0.7377,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1373019683149304,
+      "grad_norm": 1.2325714599866064,
+      "learning_rate": 9.51476357358964e-06,
+      "loss": 0.763,
+      "step": 1185
+    },
+    {
+      "epoch": 1.1382621219395104,
+      "grad_norm": 0.7080124445418774,
+      "learning_rate": 9.513322948779544e-06,
+      "loss": 0.6175,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1392222755640902,
+      "grad_norm": 1.1380266027166621,
+      "learning_rate": 9.511880297987022e-06,
+      "loss": 0.6257,
+      "step": 1187
+    },
+    {
+      "epoch": 1.1401824291886702,
+      "grad_norm": 1.0476252524937764,
+      "learning_rate": 9.510435621859662e-06,
+      "loss": 0.6239,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1411425828132502,
+      "grad_norm": 0.9012962777039913,
+      "learning_rate": 9.508988921045967e-06,
+      "loss": 0.6829,
+      "step": 1189
+    },
+    {
+      "epoch": 1.14210273643783,
+      "grad_norm": 1.401581170885142,
+      "learning_rate": 9.507540196195342e-06,
+      "loss": 0.5707,
+      "step": 1190
+    },
+    {
+      "epoch": 1.14306289006241,
+      "grad_norm": 0.8515026664947627,
+      "learning_rate": 9.506089447958108e-06,
+      "loss": 0.6342,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1440230436869898,
+      "grad_norm": 1.4240129750591695,
+      "learning_rate": 9.504636676985486e-06,
+      "loss": 0.7272,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1449831973115698,
+      "grad_norm": 1.0806415361440416,
+      "learning_rate": 9.503181883929613e-06,
+      "loss": 0.6605,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1459433509361499,
+      "grad_norm": 1.0537884629020036,
+      "learning_rate": 9.501725069443528e-06,
+      "loss": 0.5705,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1469035045607296,
+      "grad_norm": 0.9072811020520438,
+      "learning_rate": 9.50026623418118e-06,
+      "loss": 0.6128,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1478636581853097,
+      "grad_norm": 1.2138946145038918,
+      "learning_rate": 9.498805378797427e-06,
+      "loss": 0.6345,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1488238118098897,
+      "grad_norm": 1.2551533599064106,
+      "learning_rate": 9.497342503948027e-06,
+      "loss": 0.6823,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1497839654344695,
+      "grad_norm": 1.3521081824818293,
+      "learning_rate": 9.49587761028965e-06,
+      "loss": 0.5514,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1507441190590495,
+      "grad_norm": 1.0771335823289896,
+      "learning_rate": 9.494410698479873e-06,
+      "loss": 0.7498,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1517042726836293,
+      "grad_norm": 0.9826695036864329,
+      "learning_rate": 9.492941769177176e-06,
+      "loss": 0.6652,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1526644263082093,
+      "grad_norm": 1.0275626147254378,
+      "learning_rate": 9.491470823040944e-06,
+      "loss": 0.6806,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1536245799327893,
+      "grad_norm": 0.9593951529609802,
+      "learning_rate": 9.489997860731475e-06,
+      "loss": 0.6689,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1545847335573691,
+      "grad_norm": 1.02339546311265,
+      "learning_rate": 9.488522882909959e-06,
+      "loss": 0.598,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1555448871819491,
+      "grad_norm": 1.2636604557566715,
+      "learning_rate": 9.4870458902385e-06,
+      "loss": 0.6347,
+      "step": 1204
+    },
+    {
+      "epoch": 1.156505040806529,
+      "grad_norm": 1.40261618222338,
+      "learning_rate": 9.485566883380105e-06,
+      "loss": 0.7028,
+      "step": 1205
+    },
+    {
+      "epoch": 1.157465194431109,
+      "grad_norm": 1.0663856930850502,
+      "learning_rate": 9.484085862998685e-06,
+      "loss": 0.5947,
+      "step": 1206
+    },
+    {
+      "epoch": 1.158425348055689,
+      "grad_norm": 1.4723561780039622,
+      "learning_rate": 9.482602829759052e-06,
+      "loss": 0.6393,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1593855016802688,
+      "grad_norm": 1.1637330649963575,
+      "learning_rate": 9.481117784326927e-06,
+      "loss": 0.5944,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1603456553048488,
+      "grad_norm": 1.032629931610833,
+      "learning_rate": 9.479630727368927e-06,
+      "loss": 0.5966,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1613058089294288,
+      "grad_norm": 0.8392640016288045,
+      "learning_rate": 9.47814165955258e-06,
+      "loss": 0.6257,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1622659625540086,
+      "grad_norm": 1.2493910091463432,
+      "learning_rate": 9.47665058154631e-06,
+      "loss": 0.681,
+      "step": 1211
+    },
+    {
+      "epoch": 1.1632261161785886,
+      "grad_norm": 0.98863249817308,
+      "learning_rate": 9.475157494019445e-06,
+      "loss": 0.4885,
+      "step": 1212
+    },
+    {
+      "epoch": 1.1641862698031684,
+      "grad_norm": 1.2539998735969364,
+      "learning_rate": 9.473662397642219e-06,
+      "loss": 0.7908,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1651464234277484,
+      "grad_norm": 0.9570118437597045,
+      "learning_rate": 9.472165293085765e-06,
+      "loss": 0.5632,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1661065770523285,
+      "grad_norm": 1.3070034260563663,
+      "learning_rate": 9.470666181022114e-06,
+      "loss": 0.8193,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1670667306769082,
+      "grad_norm": 0.9907941564598948,
+      "learning_rate": 9.469165062124203e-06,
+      "loss": 0.6063,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1680268843014883,
+      "grad_norm": 1.2502899623231871,
+      "learning_rate": 9.467661937065869e-06,
+      "loss": 0.6428,
+      "step": 1217
+    },
+    {
+      "epoch": 1.168987037926068,
+      "grad_norm": 0.9611806461190867,
+      "learning_rate": 9.466156806521847e-06,
+      "loss": 0.6312,
+      "step": 1218
+    },
+    {
+      "epoch": 1.169947191550648,
+      "grad_norm": 0.9348007304355214,
+      "learning_rate": 9.464649671167777e-06,
+      "loss": 0.6423,
+      "step": 1219
+    },
+    {
+      "epoch": 1.170907345175228,
+      "grad_norm": 1.0777819064389393,
+      "learning_rate": 9.463140531680193e-06,
+      "loss": 0.6771,
+      "step": 1220
+    },
+    {
+      "epoch": 1.171867498799808,
+      "grad_norm": 1.180486328750565,
+      "learning_rate": 9.461629388736534e-06,
+      "loss": 0.687,
+      "step": 1221
+    },
+    {
+      "epoch": 1.172827652424388,
+      "grad_norm": 0.9790522632115352,
+      "learning_rate": 9.460116243015131e-06,
+      "loss": 0.6485,
+      "step": 1222
+    },
+    {
+      "epoch": 1.173787806048968,
+      "grad_norm": 1.4658160863700345,
+      "learning_rate": 9.458601095195225e-06,
+      "loss": 0.7031,
+      "step": 1223
+    },
+    {
+      "epoch": 1.1747479596735477,
+      "grad_norm": 0.9591722122695792,
+      "learning_rate": 9.45708394595695e-06,
+      "loss": 0.5725,
+      "step": 1224
+    },
+    {
+      "epoch": 1.1757081132981277,
+      "grad_norm": 0.9493091678959932,
+      "learning_rate": 9.455564795981331e-06,
+      "loss": 0.5762,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1766682669227075,
+      "grad_norm": 0.9288315445755897,
+      "learning_rate": 9.454043645950303e-06,
+      "loss": 0.7323,
+      "step": 1226
+    },
+    {
+      "epoch": 1.1776284205472876,
+      "grad_norm": 1.3989016722710288,
+      "learning_rate": 9.452520496546693e-06,
+      "loss": 0.742,
+      "step": 1227
+    },
+    {
+      "epoch": 1.1785885741718676,
+      "grad_norm": 1.7939405129425552,
+      "learning_rate": 9.450995348454228e-06,
+      "loss": 0.7528,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1795487277964474,
+      "grad_norm": 1.4912406316983478,
+      "learning_rate": 9.44946820235753e-06,
+      "loss": 0.7062,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1805088814210274,
+      "grad_norm": 1.4089671456132038,
+      "learning_rate": 9.447939058942119e-06,
+      "loss": 0.7029,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1814690350456072,
+      "grad_norm": 1.130975861916411,
+      "learning_rate": 9.446407918894408e-06,
+      "loss": 0.7265,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1824291886701872,
+      "grad_norm": 1.4660041943418791,
+      "learning_rate": 9.444874782901713e-06,
+      "loss": 0.6933,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1833893422947672,
+      "grad_norm": 1.1556454893411412,
+      "learning_rate": 9.443339651652241e-06,
+      "loss": 0.5965,
+      "step": 1233
+    },
+    {
+      "epoch": 1.184349495919347,
+      "grad_norm": 0.9125541225468079,
+      "learning_rate": 9.441802525835095e-06,
+      "loss": 0.663,
+      "step": 1234
+    },
+    {
+      "epoch": 1.185309649543927,
+      "grad_norm": 0.9050441635898483,
+      "learning_rate": 9.440263406140276e-06,
+      "loss": 0.6862,
+      "step": 1235
+    },
+    {
+      "epoch": 1.186269803168507,
+      "grad_norm": 1.0466940814511771,
+      "learning_rate": 9.438722293258678e-06,
+      "loss": 0.5805,
+      "step": 1236
+    },
+    {
+      "epoch": 1.1872299567930868,
+      "grad_norm": 1.0008127872163366,
+      "learning_rate": 9.43717918788209e-06,
+      "loss": 0.6455,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1881901104176669,
+      "grad_norm": 1.113676512163136,
+      "learning_rate": 9.435634090703195e-06,
+      "loss": 0.6278,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1891502640422467,
+      "grad_norm": 1.2375996678281171,
+      "learning_rate": 9.434087002415571e-06,
+      "loss": 0.6556,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1901104176668267,
+      "grad_norm": 0.8065055434730838,
+      "learning_rate": 9.432537923713689e-06,
+      "loss": 0.5956,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1910705712914067,
+      "grad_norm": 1.1982682927770063,
+      "learning_rate": 9.430986855292916e-06,
+      "loss": 0.6755,
+      "step": 1241
+    },
+    {
+      "epoch": 1.1920307249159865,
+      "grad_norm": 0.9063500599021055,
+      "learning_rate": 9.429433797849508e-06,
+      "loss": 0.6875,
+      "step": 1242
+    },
+    {
+      "epoch": 1.1929908785405665,
+      "grad_norm": 1.1804643784727347,
+      "learning_rate": 9.427878752080613e-06,
+      "loss": 0.7355,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1939510321651463,
+      "grad_norm": 1.0823916368736504,
+      "learning_rate": 9.426321718684282e-06,
+      "loss": 0.6471,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1949111857897263,
+      "grad_norm": 1.3597785702813805,
+      "learning_rate": 9.424762698359442e-06,
+      "loss": 0.6655,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1958713394143063,
+      "grad_norm": 0.8598556636072228,
+      "learning_rate": 9.423201691805927e-06,
+      "loss": 0.5505,
+      "step": 1246
+    },
+    {
+      "epoch": 1.1968314930388861,
+      "grad_norm": 0.995912244969561,
+      "learning_rate": 9.421638699724453e-06,
+      "loss": 0.4913,
+      "step": 1247
+    },
+    {
+      "epoch": 1.1977916466634662,
+      "grad_norm": 0.8356833278020595,
+      "learning_rate": 9.420073722816634e-06,
+      "loss": 0.7199,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1987518002880462,
+      "grad_norm": 1.5643960156153773,
+      "learning_rate": 9.418506761784967e-06,
+      "loss": 0.7136,
+      "step": 1249
+    },
+    {
+      "epoch": 1.199711953912626,
+      "grad_norm": 0.9077061515511775,
+      "learning_rate": 9.416937817332846e-06,
+      "loss": 0.6886,
+      "step": 1250
+    },
+    {
+      "epoch": 1.200672107537206,
+      "grad_norm": 0.9405685661726997,
+      "learning_rate": 9.415366890164553e-06,
+      "loss": 0.5351,
+      "step": 1251
+    },
+    {
+      "epoch": 1.201632261161786,
+      "grad_norm": 1.1406637017982817,
+      "learning_rate": 9.41379398098526e-06,
+      "loss": 0.74,
+      "step": 1252
+    },
+    {
+      "epoch": 1.2025924147863658,
+      "grad_norm": 1.0823585567677922,
+      "learning_rate": 9.41221909050103e-06,
+      "loss": 0.6435,
+      "step": 1253
+    },
+    {
+      "epoch": 1.2035525684109458,
+      "grad_norm": 0.935137312913638,
+      "learning_rate": 9.410642219418816e-06,
+      "loss": 0.645,
+      "step": 1254
+    },
+    {
+      "epoch": 1.2045127220355256,
+      "grad_norm": 1.1245325934238484,
+      "learning_rate": 9.409063368446454e-06,
+      "loss": 0.6627,
+      "step": 1255
+    },
+    {
+      "epoch": 1.2054728756601056,
+      "grad_norm": 0.8878362745856924,
+      "learning_rate": 9.407482538292679e-06,
+      "loss": 0.78,
+      "step": 1256
+    },
+    {
+      "epoch": 1.2064330292846854,
+      "grad_norm": 1.2618439111980329,
+      "learning_rate": 9.405899729667105e-06,
+      "loss": 0.7082,
+      "step": 1257
+    },
+    {
+      "epoch": 1.2073931829092654,
+      "grad_norm": 1.0146361786520939,
+      "learning_rate": 9.404314943280238e-06,
+      "loss": 0.6167,
+      "step": 1258
+    },
+    {
+      "epoch": 1.2083533365338455,
+      "grad_norm": 1.089613875702965,
+      "learning_rate": 9.402728179843472e-06,
+      "loss": 0.732,
+      "step": 1259
+    },
+    {
+      "epoch": 1.2093134901584253,
+      "grad_norm": 0.8991377642980493,
+      "learning_rate": 9.40113944006909e-06,
+      "loss": 0.5676,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2102736437830053,
+      "grad_norm": 1.0111428725977558,
+      "learning_rate": 9.39954872467026e-06,
+      "loss": 0.6712,
+      "step": 1261
+    },
+    {
+      "epoch": 1.2112337974075853,
+      "grad_norm": 0.9909065293109987,
+      "learning_rate": 9.397956034361033e-06,
+      "loss": 0.6683,
+      "step": 1262
+    },
+    {
+      "epoch": 1.212193951032165,
+      "grad_norm": 1.1665268239010138,
+      "learning_rate": 9.396361369856356e-06,
+      "loss": 0.6287,
+      "step": 1263
+    },
+    {
+      "epoch": 1.2131541046567451,
+      "grad_norm": 1.054290739854059,
+      "learning_rate": 9.394764731872054e-06,
+      "loss": 0.6621,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2141142582813251,
+      "grad_norm": 1.0898386457870217,
+      "learning_rate": 9.393166121124841e-06,
+      "loss": 0.5746,
+      "step": 1265
+    },
+    {
+      "epoch": 1.215074411905905,
+      "grad_norm": 0.9566676778756585,
+      "learning_rate": 9.391565538332317e-06,
+      "loss": 0.614,
+      "step": 1266
+    },
+    {
+      "epoch": 1.216034565530485,
+      "grad_norm": 0.8400299196696328,
+      "learning_rate": 9.389962984212966e-06,
+      "loss": 0.6769,
+      "step": 1267
+    },
+    {
+      "epoch": 1.2169947191550647,
+      "grad_norm": 1.2005798867695259,
+      "learning_rate": 9.388358459486156e-06,
+      "loss": 0.7829,
+      "step": 1268
+    },
+    {
+      "epoch": 1.2179548727796448,
+      "grad_norm": 1.2021657494576357,
+      "learning_rate": 9.386751964872144e-06,
+      "loss": 0.5903,
+      "step": 1269
+    },
+    {
+      "epoch": 1.2189150264042246,
+      "grad_norm": 1.1025463800635396,
+      "learning_rate": 9.385143501092066e-06,
+      "loss": 0.667,
+      "step": 1270
+    },
+    {
+      "epoch": 1.2198751800288046,
+      "grad_norm": 1.2542456165397613,
+      "learning_rate": 9.383533068867947e-06,
+      "loss": 0.5988,
+      "step": 1271
+    },
+    {
+      "epoch": 1.2208353336533846,
+      "grad_norm": 1.0585888442495783,
+      "learning_rate": 9.381920668922688e-06,
+      "loss": 0.707,
+      "step": 1272
+    },
+    {
+      "epoch": 1.2217954872779644,
+      "grad_norm": 0.9638408484966245,
+      "learning_rate": 9.380306301980082e-06,
+      "loss": 0.6781,
+      "step": 1273
+    },
+    {
+      "epoch": 1.2227556409025444,
+      "grad_norm": 1.0296720167271556,
+      "learning_rate": 9.378689968764801e-06,
+      "loss": 0.5824,
+      "step": 1274
+    },
+    {
+      "epoch": 1.2237157945271244,
+      "grad_norm": 0.8171310165074823,
+      "learning_rate": 9.377071670002398e-06,
+      "loss": 0.6445,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2246759481517042,
+      "grad_norm": 1.1465404921282132,
+      "learning_rate": 9.37545140641931e-06,
+      "loss": 0.7128,
+      "step": 1276
+    },
+    {
+      "epoch": 1.2256361017762842,
+      "grad_norm": 1.2379480031020065,
+      "learning_rate": 9.373829178742857e-06,
+      "loss": 0.6726,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2265962554008643,
+      "grad_norm": 0.9153325601276985,
+      "learning_rate": 9.37220498770124e-06,
+      "loss": 0.5812,
+      "step": 1278
+    },
+    {
+      "epoch": 1.227556409025444,
+      "grad_norm": 1.0918854791705137,
+      "learning_rate": 9.370578834023539e-06,
+      "loss": 0.5427,
+      "step": 1279
+    },
+    {
+      "epoch": 1.228516562650024,
+      "grad_norm": 1.107125474420226,
+      "learning_rate": 9.368950718439718e-06,
+      "loss": 0.7187,
+      "step": 1280
+    },
+    {
+      "epoch": 1.2294767162746039,
+      "grad_norm": 0.8644454521860808,
+      "learning_rate": 9.367320641680621e-06,
+      "loss": 0.6019,
+      "step": 1281
+    },
+    {
+      "epoch": 1.2304368698991839,
+      "grad_norm": 1.0501231561275055,
+      "learning_rate": 9.365688604477974e-06,
+      "loss": 0.6184,
+      "step": 1282
+    },
+    {
+      "epoch": 1.2313970235237637,
+      "grad_norm": 1.0060824341294845,
+      "learning_rate": 9.364054607564376e-06,
+      "loss": 0.6439,
+      "step": 1283
+    },
+    {
+      "epoch": 1.2323571771483437,
+      "grad_norm": 0.7443161665525969,
+      "learning_rate": 9.362418651673317e-06,
+      "loss": 0.5219,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2333173307729237,
+      "grad_norm": 0.9703118618007911,
+      "learning_rate": 9.360780737539155e-06,
+      "loss": 0.6886,
+      "step": 1285
+    },
+    {
+      "epoch": 1.2342774843975035,
+      "grad_norm": 1.1677659082945249,
+      "learning_rate": 9.359140865897136e-06,
+      "loss": 0.6069,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2352376380220835,
+      "grad_norm": 1.0592077878875672,
+      "learning_rate": 9.357499037483377e-06,
+      "loss": 0.544,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2361977916466635,
+      "grad_norm": 0.8785899027528711,
+      "learning_rate": 9.35585525303488e-06,
+      "loss": 0.6578,
+      "step": 1288
+    },
+    {
+      "epoch": 1.2371579452712433,
+      "grad_norm": 1.4517995476557288,
+      "learning_rate": 9.35420951328952e-06,
+      "loss": 0.7605,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2381180988958234,
+      "grad_norm": 1.2961008752247485,
+      "learning_rate": 9.352561818986056e-06,
+      "loss": 0.6525,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2390782525204034,
+      "grad_norm": 1.4183167203059521,
+      "learning_rate": 9.350912170864116e-06,
+      "loss": 0.6921,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2400384061449832,
+      "grad_norm": 1.6474252803532177,
+      "learning_rate": 9.349260569664212e-06,
+      "loss": 0.6892,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2409985597695632,
+      "grad_norm": 0.7534610842432165,
+      "learning_rate": 9.347607016127728e-06,
+      "loss": 0.6579,
+      "step": 1293
+    },
+    {
+      "epoch": 1.241958713394143,
+      "grad_norm": 1.3560586177855898,
+      "learning_rate": 9.34595151099693e-06,
+      "loss": 0.5546,
+      "step": 1294
+    },
+    {
+      "epoch": 1.242918867018723,
+      "grad_norm": 1.2039054130735463,
+      "learning_rate": 9.344294055014952e-06,
+      "loss": 0.6263,
+      "step": 1295
+    },
+    {
+      "epoch": 1.243879020643303,
+      "grad_norm": 1.2559966060931804,
+      "learning_rate": 9.342634648925813e-06,
+      "loss": 0.7259,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2448391742678828,
+      "grad_norm": 1.126033627440921,
+      "learning_rate": 9.340973293474397e-06,
+      "loss": 0.6782,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2457993278924628,
+      "grad_norm": 1.2534282014031837,
+      "learning_rate": 9.339309989406474e-06,
+      "loss": 0.6615,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2467594815170426,
+      "grad_norm": 1.0497702541081584,
+      "learning_rate": 9.337644737468682e-06,
+      "loss": 0.5481,
+      "step": 1299
+    },
+    {
+      "epoch": 1.2477196351416227,
+      "grad_norm": 0.6864879499147802,
+      "learning_rate": 9.335977538408533e-06,
+      "loss": 0.6377,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2486797887662027,
+      "grad_norm": 1.004307893115873,
+      "learning_rate": 9.334308392974417e-06,
+      "loss": 0.5459,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2496399423907825,
+      "grad_norm": 1.0347046897868124,
+      "learning_rate": 9.332637301915594e-06,
+      "loss": 0.6139,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2506000960153625,
+      "grad_norm": 1.0273603906823152,
+      "learning_rate": 9.3309642659822e-06,
+      "loss": 0.5295,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2515602496399425,
+      "grad_norm": 0.9256038563076083,
+      "learning_rate": 9.329289285925242e-06,
+      "loss": 0.6627,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2525204032645223,
+      "grad_norm": 1.0665551164613651,
+      "learning_rate": 9.327612362496601e-06,
+      "loss": 0.5403,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2534805568891023,
+      "grad_norm": 1.0496992791473883,
+      "learning_rate": 9.325933496449031e-06,
+      "loss": 0.6226,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2544407105136821,
+      "grad_norm": 1.0038462694094532,
+      "learning_rate": 9.324252688536156e-06,
+      "loss": 0.6301,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2554008641382621,
+      "grad_norm": 0.990212670969131,
+      "learning_rate": 9.322569939512472e-06,
+      "loss": 0.6669,
+      "step": 1308
+    },
+    {
+      "epoch": 1.256361017762842,
+      "grad_norm": 1.0100661698775886,
+      "learning_rate": 9.320885250133347e-06,
+      "loss": 0.6576,
+      "step": 1309
+    },
+    {
+      "epoch": 1.257321171387422,
+      "grad_norm": 0.9595990366799756,
+      "learning_rate": 9.319198621155022e-06,
+      "loss": 0.6108,
+      "step": 1310
+    },
+    {
+      "epoch": 1.258281325012002,
+      "grad_norm": 0.9932039061329729,
+      "learning_rate": 9.317510053334604e-06,
+      "loss": 0.7106,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2592414786365818,
+      "grad_norm": 0.827379518177105,
+      "learning_rate": 9.315819547430077e-06,
+      "loss": 0.7086,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2602016322611618,
+      "grad_norm": 0.9826599022429947,
+      "learning_rate": 9.314127104200287e-06,
+      "loss": 0.6515,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2611617858857418,
+      "grad_norm": 0.88254688167717,
+      "learning_rate": 9.312432724404957e-06,
+      "loss": 0.5874,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2621219395103216,
+      "grad_norm": 0.9292760810081079,
+      "learning_rate": 9.310736408804674e-06,
+      "loss": 0.621,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2630820931349016,
+      "grad_norm": 1.1642100883411828,
+      "learning_rate": 9.309038158160896e-06,
+      "loss": 0.486,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2640422467594816,
+      "grad_norm": 0.9674174348083485,
+      "learning_rate": 9.30733797323595e-06,
+      "loss": 0.6359,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2650024003840614,
+      "grad_norm": 1.12103017950127,
+      "learning_rate": 9.305635854793031e-06,
+      "loss": 0.6539,
+      "step": 1318
+    },
+    {
+      "epoch": 1.2659625540086414,
+      "grad_norm": 0.9958390684899106,
+      "learning_rate": 9.3039318035962e-06,
+      "loss": 0.6644,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2669227076332212,
+      "grad_norm": 1.7022252513712437,
+      "learning_rate": 9.30222582041039e-06,
+      "loss": 0.5404,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2678828612578013,
+      "grad_norm": 1.134706650262642,
+      "learning_rate": 9.300517906001399e-06,
+      "loss": 0.7979,
+      "step": 1321
+    },
+    {
+      "epoch": 1.268843014882381,
+      "grad_norm": 1.2286866546646331,
+      "learning_rate": 9.298808061135887e-06,
+      "loss": 0.5412,
+      "step": 1322
+    },
+    {
+      "epoch": 1.269803168506961,
+      "grad_norm": 0.964610599153338,
+      "learning_rate": 9.29709628658139e-06,
+      "loss": 0.5561,
+      "step": 1323
+    },
+    {
+      "epoch": 1.270763322131541,
+      "grad_norm": 1.0421977021558635,
+      "learning_rate": 9.295382583106303e-06,
+      "loss": 0.596,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2717234757561209,
+      "grad_norm": 1.1132921252971018,
+      "learning_rate": 9.29366695147989e-06,
+      "loss": 0.5767,
+      "step": 1325
+    },
+    {
+      "epoch": 1.272683629380701,
+      "grad_norm": 1.3239230868033052,
+      "learning_rate": 9.291949392472276e-06,
+      "loss": 0.6072,
+      "step": 1326
+    },
+    {
+      "epoch": 1.273643783005281,
+      "grad_norm": 1.576167318967835,
+      "learning_rate": 9.290229906854462e-06,
+      "loss": 0.7313,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2746039366298607,
+      "grad_norm": 1.3610779506237412,
+      "learning_rate": 9.288508495398301e-06,
+      "loss": 0.5892,
+      "step": 1328
+    },
+    {
+      "epoch": 1.2755640902544407,
+      "grad_norm": 1.0519195317352252,
+      "learning_rate": 9.28678515887652e-06,
+      "loss": 0.6665,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2765242438790207,
+      "grad_norm": 0.8333354204993307,
+      "learning_rate": 9.285059898062701e-06,
+      "loss": 0.5669,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2774843975036005,
+      "grad_norm": 1.1052555150098762,
+      "learning_rate": 9.2833327137313e-06,
+      "loss": 0.7173,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2784445511281806,
+      "grad_norm": 1.2239098417947563,
+      "learning_rate": 9.281603606657632e-06,
+      "loss": 0.5458,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2794047047527606,
+      "grad_norm": 0.9706516508907063,
+      "learning_rate": 9.27987257761787e-06,
+      "loss": 0.6695,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2803648583773404,
+      "grad_norm": 1.043908515727729,
+      "learning_rate": 9.27813962738906e-06,
+      "loss": 0.6605,
+      "step": 1334
+    },
+    {
+      "epoch": 1.2813250120019202,
+      "grad_norm": 0.9817932598491614,
+      "learning_rate": 9.2764047567491e-06,
+      "loss": 0.5666,
+      "step": 1335
+    },
+    {
+      "epoch": 1.2822851656265002,
+      "grad_norm": 1.3581727939152737,
+      "learning_rate": 9.274667966476757e-06,
+      "loss": 0.5455,
+      "step": 1336
+    },
+    {
+      "epoch": 1.2832453192510802,
+      "grad_norm": 0.9419152885129211,
+      "learning_rate": 9.272929257351659e-06,
+      "loss": 0.5651,
+      "step": 1337
+    },
+    {
+      "epoch": 1.28420547287566,
+      "grad_norm": 1.3104273037342988,
+      "learning_rate": 9.27118863015429e-06,
+      "loss": 0.6209,
+      "step": 1338
+    },
+    {
+      "epoch": 1.28516562650024,
+      "grad_norm": 1.0614631606828682,
+      "learning_rate": 9.269446085666002e-06,
+      "loss": 0.6588,
+      "step": 1339
+    },
+    {
+      "epoch": 1.28612578012482,
+      "grad_norm": 1.2821567442640949,
+      "learning_rate": 9.267701624669006e-06,
+      "loss": 0.6519,
+      "step": 1340
+    },
+    {
+      "epoch": 1.2870859337493998,
+      "grad_norm": 1.111204239556066,
+      "learning_rate": 9.265955247946369e-06,
+      "loss": 0.6144,
+      "step": 1341
+    },
+    {
+      "epoch": 1.2880460873739799,
+      "grad_norm": 1.095390947221734,
+      "learning_rate": 9.26420695628202e-06,
+      "loss": 0.7597,
+      "step": 1342
+    },
+    {
+      "epoch": 1.2890062409985599,
+      "grad_norm": 1.2377076600006347,
+      "learning_rate": 9.262456750460754e-06,
+      "loss": 0.6322,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2899663946231397,
+      "grad_norm": 0.9512879954462602,
+      "learning_rate": 9.260704631268211e-06,
+      "loss": 0.6935,
+      "step": 1344
+    },
+    {
+      "epoch": 1.2909265482477197,
+      "grad_norm": 1.3026950771394807,
+      "learning_rate": 9.258950599490906e-06,
+      "loss": 0.6922,
+      "step": 1345
+    },
+    {
+      "epoch": 1.2918867018722997,
+      "grad_norm": 1.4224545115172562,
+      "learning_rate": 9.257194655916202e-06,
+      "loss": 0.6217,
+      "step": 1346
+    },
+    {
+      "epoch": 1.2928468554968795,
+      "grad_norm": 0.9423072453719439,
+      "learning_rate": 9.255436801332324e-06,
+      "loss": 0.6512,
+      "step": 1347
+    },
+    {
+      "epoch": 1.2938070091214593,
+      "grad_norm": 1.2651014767302182,
+      "learning_rate": 9.253677036528356e-06,
+      "loss": 0.7078,
+      "step": 1348
+    },
+    {
+      "epoch": 1.2947671627460393,
+      "grad_norm": 1.2056034429549998,
+      "learning_rate": 9.251915362294233e-06,
+      "loss": 0.6946,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2957273163706193,
+      "grad_norm": 1.1626763350371463,
+      "learning_rate": 9.250151779420756e-06,
+      "loss": 0.6759,
+      "step": 1350
+    },
+    {
+      "epoch": 1.2966874699951991,
+      "grad_norm": 1.0954103915987892,
+      "learning_rate": 9.248386288699575e-06,
+      "loss": 0.712,
+      "step": 1351
+    },
+    {
+      "epoch": 1.2976476236197791,
+      "grad_norm": 1.308612743368172,
+      "learning_rate": 9.246618890923203e-06,
+      "loss": 0.6832,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2986077772443592,
+      "grad_norm": 1.039110620513189,
+      "learning_rate": 9.244849586885005e-06,
+      "loss": 0.6366,
+      "step": 1353
+    },
+    {
+      "epoch": 1.299567930868939,
+      "grad_norm": 0.9978311166737291,
+      "learning_rate": 9.243078377379202e-06,
+      "loss": 0.6312,
+      "step": 1354
+    },
+    {
+      "epoch": 1.300528084493519,
+      "grad_norm": 0.9165498261707806,
+      "learning_rate": 9.24130526320087e-06,
+      "loss": 0.6052,
+      "step": 1355
+    },
+    {
+      "epoch": 1.301488238118099,
+      "grad_norm": 0.735063514985535,
+      "learning_rate": 9.239530245145945e-06,
+      "loss": 0.6612,
+      "step": 1356
+    },
+    {
+      "epoch": 1.3024483917426788,
+      "grad_norm": 1.0692782405428205,
+      "learning_rate": 9.237753324011208e-06,
+      "loss": 0.6754,
+      "step": 1357
+    },
+    {
+      "epoch": 1.3034085453672588,
+      "grad_norm": 1.5840209243711638,
+      "learning_rate": 9.235974500594305e-06,
+      "loss": 0.6874,
+      "step": 1358
+    },
+    {
+      "epoch": 1.3043686989918388,
+      "grad_norm": 1.3359743683261467,
+      "learning_rate": 9.234193775693728e-06,
+      "loss": 0.7557,
+      "step": 1359
+    },
+    {
+      "epoch": 1.3053288526164186,
+      "grad_norm": 1.3974632602774053,
+      "learning_rate": 9.232411150108826e-06,
+      "loss": 0.6611,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3062890062409986,
+      "grad_norm": 1.0801888072428878,
+      "learning_rate": 9.2306266246398e-06,
+      "loss": 0.6115,
+      "step": 1361
+    },
+    {
+      "epoch": 1.3072491598655784,
+      "grad_norm": 1.1295783496536176,
+      "learning_rate": 9.228840200087708e-06,
+      "loss": 0.6626,
+      "step": 1362
+    },
+    {
+      "epoch": 1.3082093134901585,
+      "grad_norm": 1.261770127027789,
+      "learning_rate": 9.227051877254452e-06,
+      "loss": 0.6509,
+      "step": 1363
+    },
+    {
+      "epoch": 1.3091694671147383,
+      "grad_norm": 1.219139014858422,
+      "learning_rate": 9.225261656942792e-06,
+      "loss": 0.6737,
+      "step": 1364
+    },
+    {
+      "epoch": 1.3101296207393183,
+      "grad_norm": 1.3184165602906317,
+      "learning_rate": 9.22346953995634e-06,
+      "loss": 0.6225,
+      "step": 1365
+    },
+    {
+      "epoch": 1.3110897743638983,
+      "grad_norm": 1.002937899496408,
+      "learning_rate": 9.22167552709956e-06,
+      "loss": 0.6944,
+      "step": 1366
+    },
+    {
+      "epoch": 1.312049927988478,
+      "grad_norm": 1.4336954896809986,
+      "learning_rate": 9.219879619177761e-06,
+      "loss": 0.7621,
+      "step": 1367
+    },
+    {
+      "epoch": 1.313010081613058,
+      "grad_norm": 0.8527614630941144,
+      "learning_rate": 9.21808181699711e-06,
+      "loss": 0.5507,
+      "step": 1368
+    },
+    {
+      "epoch": 1.3139702352376381,
+      "grad_norm": 0.8727290588806392,
+      "learning_rate": 9.21628212136462e-06,
+      "loss": 0.632,
+      "step": 1369
+    },
+    {
+      "epoch": 1.314930388862218,
+      "grad_norm": 1.4499825499908416,
+      "learning_rate": 9.214480533088154e-06,
+      "loss": 0.7411,
+      "step": 1370
+    },
+    {
+      "epoch": 1.315890542486798,
+      "grad_norm": 1.3665233860507873,
+      "learning_rate": 9.212677052976428e-06,
+      "loss": 0.7025,
+      "step": 1371
+    },
+    {
+      "epoch": 1.316850696111378,
+      "grad_norm": 1.3370211780233987,
+      "learning_rate": 9.210871681839006e-06,
+      "loss": 0.6297,
+      "step": 1372
+    },
+    {
+      "epoch": 1.3178108497359577,
+      "grad_norm": 0.8810290539270131,
+      "learning_rate": 9.209064420486296e-06,
+      "loss": 0.5714,
+      "step": 1373
+    },
+    {
+      "epoch": 1.3187710033605378,
+      "grad_norm": 0.8972394920514724,
+      "learning_rate": 9.207255269729559e-06,
+      "loss": 0.5948,
+      "step": 1374
+    },
+    {
+      "epoch": 1.3197311569851176,
+      "grad_norm": 1.4565885202363469,
+      "learning_rate": 9.205444230380904e-06,
+      "loss": 0.7471,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3206913106096976,
+      "grad_norm": 1.3602857157125754,
+      "learning_rate": 9.203631303253292e-06,
+      "loss": 0.6143,
+      "step": 1376
+    },
+    {
+      "epoch": 1.3216514642342774,
+      "grad_norm": 1.2348121608011493,
+      "learning_rate": 9.201816489160518e-06,
+      "loss": 0.7407,
+      "step": 1377
+    },
+    {
+      "epoch": 1.3226116178588574,
+      "grad_norm": 1.1677390321780012,
+      "learning_rate": 9.199999788917238e-06,
+      "loss": 0.7074,
+      "step": 1378
+    },
+    {
+      "epoch": 1.3235717714834374,
+      "grad_norm": 1.6220671969702665,
+      "learning_rate": 9.198181203338949e-06,
+      "loss": 0.672,
+      "step": 1379
+    },
+    {
+      "epoch": 1.3245319251080172,
+      "grad_norm": 1.55816953971552,
+      "learning_rate": 9.196360733241992e-06,
+      "loss": 0.5813,
+      "step": 1380
+    },
+    {
+      "epoch": 1.3254920787325972,
+      "grad_norm": 0.6031797398946008,
+      "learning_rate": 9.194538379443561e-06,
+      "loss": 0.5942,
+      "step": 1381
+    },
+    {
+      "epoch": 1.3264522323571772,
+      "grad_norm": 1.0209739299888865,
+      "learning_rate": 9.192714142761687e-06,
+      "loss": 0.5466,
+      "step": 1382
+    },
+    {
+      "epoch": 1.327412385981757,
+      "grad_norm": 1.0685918514092403,
+      "learning_rate": 9.190888024015252e-06,
+      "loss": 0.7386,
+      "step": 1383
+    },
+    {
+      "epoch": 1.328372539606337,
+      "grad_norm": 2.4654106901615123,
+      "learning_rate": 9.189060024023981e-06,
+      "loss": 0.6167,
+      "step": 1384
+    },
+    {
+      "epoch": 1.329332693230917,
+      "grad_norm": 0.9254052338620629,
+      "learning_rate": 9.187230143608445e-06,
+      "loss": 0.7293,
+      "step": 1385
+    },
+    {
+      "epoch": 1.3302928468554969,
+      "grad_norm": 0.8997768367015149,
+      "learning_rate": 9.185398383590056e-06,
+      "loss": 0.6735,
+      "step": 1386
+    },
+    {
+      "epoch": 1.3312530004800769,
+      "grad_norm": 0.8765280760171718,
+      "learning_rate": 9.18356474479107e-06,
+      "loss": 0.7112,
+      "step": 1387
+    },
+    {
+      "epoch": 1.3322131541046567,
+      "grad_norm": 1.0199429321891622,
+      "learning_rate": 9.181729228034593e-06,
+      "loss": 0.6859,
+      "step": 1388
+    },
+    {
+      "epoch": 1.3331733077292367,
+      "grad_norm": 1.156182337546054,
+      "learning_rate": 9.179891834144565e-06,
+      "loss": 0.5132,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3341334613538165,
+      "grad_norm": 1.1127871788424046,
+      "learning_rate": 9.178052563945773e-06,
+      "loss": 0.6018,
+      "step": 1390
+    },
+    {
+      "epoch": 1.3350936149783965,
+      "grad_norm": 1.0718387568033312,
+      "learning_rate": 9.176211418263845e-06,
+      "loss": 0.6811,
+      "step": 1391
+    },
+    {
+      "epoch": 1.3360537686029765,
+      "grad_norm": 1.6370091625480498,
+      "learning_rate": 9.174368397925254e-06,
+      "loss": 0.5976,
+      "step": 1392
+    },
+    {
+      "epoch": 1.3370139222275563,
+      "grad_norm": 1.3393311738083604,
+      "learning_rate": 9.17252350375731e-06,
+      "loss": 0.5573,
+      "step": 1393
+    },
+    {
+      "epoch": 1.3379740758521363,
+      "grad_norm": 0.9440285509468719,
+      "learning_rate": 9.170676736588167e-06,
+      "loss": 0.6261,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3389342294767164,
+      "grad_norm": 0.8518148709205811,
+      "learning_rate": 9.168828097246819e-06,
+      "loss": 0.5317,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3398943831012962,
+      "grad_norm": 1.001207334694378,
+      "learning_rate": 9.1669775865631e-06,
+      "loss": 0.647,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3408545367258762,
+      "grad_norm": 0.8855183613979828,
+      "learning_rate": 9.165125205367685e-06,
+      "loss": 0.5526,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3418146903504562,
+      "grad_norm": 0.8646748349945449,
+      "learning_rate": 9.16327095449209e-06,
+      "loss": 0.6218,
+      "step": 1398
+    },
+    {
+      "epoch": 1.342774843975036,
+      "grad_norm": 0.9207173830398178,
+      "learning_rate": 9.161414834768662e-06,
+      "loss": 0.679,
+      "step": 1399
+    },
+    {
+      "epoch": 1.343734997599616,
+      "grad_norm": 1.0246811568814804,
+      "learning_rate": 9.159556847030602e-06,
+      "loss": 0.6315,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3446951512241958,
+      "grad_norm": 1.0399275691034016,
+      "learning_rate": 9.157696992111935e-06,
+      "loss": 0.5397,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3456553048487758,
+      "grad_norm": 1.0881315389498556,
+      "learning_rate": 9.15583527084753e-06,
+      "loss": 0.7734,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3466154584733556,
+      "grad_norm": 1.1227693560178058,
+      "learning_rate": 9.153971684073098e-06,
+      "loss": 0.6873,
+      "step": 1403
+    },
+    {
+      "epoch": 1.3475756120979356,
+      "grad_norm": 1.1593023174123618,
+      "learning_rate": 9.15210623262518e-06,
+      "loss": 0.7518,
+      "step": 1404
+    },
+    {
+      "epoch": 1.3485357657225157,
+      "grad_norm": 1.3285923577202956,
+      "learning_rate": 9.150238917341156e-06,
+      "loss": 0.6111,
+      "step": 1405
+    },
+    {
+      "epoch": 1.3494959193470955,
+      "grad_norm": 1.3982348325270315,
+      "learning_rate": 9.148369739059247e-06,
+      "loss": 0.6599,
+      "step": 1406
+    },
+    {
+      "epoch": 1.3504560729716755,
+      "grad_norm": 1.2849899368598579,
+      "learning_rate": 9.146498698618507e-06,
+      "loss": 0.6613,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3514162265962555,
+      "grad_norm": 1.335510666991283,
+      "learning_rate": 9.14462579685882e-06,
+      "loss": 0.5636,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3523763802208353,
+      "grad_norm": 1.1259159335740894,
+      "learning_rate": 9.142751034620923e-06,
+      "loss": 0.539,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3533365338454153,
+      "grad_norm": 1.423562344991987,
+      "learning_rate": 9.140874412746365e-06,
+      "loss": 0.6782,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3542966874699953,
+      "grad_norm": 0.9815155519301481,
+      "learning_rate": 9.138995932077549e-06,
+      "loss": 0.5981,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3552568410945751,
+      "grad_norm": 1.5764550806390785,
+      "learning_rate": 9.137115593457702e-06,
+      "loss": 0.7785,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3562169947191551,
+      "grad_norm": 1.3578567713502319,
+      "learning_rate": 9.135233397730889e-06,
+      "loss": 0.6807,
+      "step": 1413
+    },
+    {
+      "epoch": 1.357177148343735,
+      "grad_norm": 1.4123462218723584,
+      "learning_rate": 9.133349345742007e-06,
+      "loss": 0.8146,
+      "step": 1414
+    },
+    {
+      "epoch": 1.358137301968315,
+      "grad_norm": 1.3656080836677802,
+      "learning_rate": 9.131463438336785e-06,
+      "loss": 0.6289,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3590974555928947,
+      "grad_norm": 0.984971772562819,
+      "learning_rate": 9.12957567636179e-06,
+      "loss": 0.6799,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3600576092174748,
+      "grad_norm": 1.0615120744089255,
+      "learning_rate": 9.127686060664414e-06,
+      "loss": 0.645,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3610177628420548,
+      "grad_norm": 0.7357874833677214,
+      "learning_rate": 9.125794592092891e-06,
+      "loss": 0.6548,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3619779164666346,
+      "grad_norm": 0.8780975777003137,
+      "learning_rate": 9.123901271496276e-06,
+      "loss": 0.5611,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3629380700912146,
+      "grad_norm": 0.8112244447711271,
+      "learning_rate": 9.122006099724463e-06,
+      "loss": 0.5917,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3638982237157946,
+      "grad_norm": 1.221144802300544,
+      "learning_rate": 9.120109077628175e-06,
+      "loss": 0.6999,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3648583773403744,
+      "grad_norm": 1.1403038947834057,
+      "learning_rate": 9.118210206058961e-06,
+      "loss": 0.6721,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3658185309649544,
+      "grad_norm": 0.9730116515747882,
+      "learning_rate": 9.116309485869207e-06,
+      "loss": 0.6044,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3667786845895344,
+      "grad_norm": 0.9493687999626979,
+      "learning_rate": 9.114406917912129e-06,
+      "loss": 0.6536,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3677388382141142,
+      "grad_norm": 1.1547885016150892,
+      "learning_rate": 9.112502503041763e-06,
+      "loss": 0.5292,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3686989918386943,
+      "grad_norm": 1.1910868300541848,
+      "learning_rate": 9.110596242112985e-06,
+      "loss": 0.6713,
+      "step": 1426
+    },
+    {
+      "epoch": 1.369659145463274,
+      "grad_norm": 0.9915731479060182,
+      "learning_rate": 9.108688135981494e-06,
+      "loss": 0.5338,
+      "step": 1427
+    },
+    {
+      "epoch": 1.370619299087854,
+      "grad_norm": 1.1633320943761714,
+      "learning_rate": 9.106778185503816e-06,
+      "loss": 0.6207,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3715794527124339,
+      "grad_norm": 1.1229228094327428,
+      "learning_rate": 9.10486639153731e-06,
+      "loss": 0.6682,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3725396063370139,
+      "grad_norm": 1.1359696559124128,
+      "learning_rate": 9.102952754940162e-06,
+      "loss": 0.6433,
+      "step": 1430
+    },
+    {
+      "epoch": 1.373499759961594,
+      "grad_norm": 1.1452071263196368,
+      "learning_rate": 9.101037276571378e-06,
+      "loss": 0.7461,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3744599135861737,
+      "grad_norm": 1.0217787726374161,
+      "learning_rate": 9.099119957290798e-06,
+      "loss": 0.6545,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3754200672107537,
+      "grad_norm": 1.1024321646787556,
+      "learning_rate": 9.097200797959089e-06,
+      "loss": 0.6808,
+      "step": 1433
+    },
+    {
+      "epoch": 1.3763802208353337,
+      "grad_norm": 1.2041112611388614,
+      "learning_rate": 9.095279799437737e-06,
+      "loss": 0.6022,
+      "step": 1434
+    },
+    {
+      "epoch": 1.3773403744599135,
+      "grad_norm": 1.245185695040958,
+      "learning_rate": 9.09335696258906e-06,
+      "loss": 0.6416,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3783005280844935,
+      "grad_norm": 1.5245819538936007,
+      "learning_rate": 9.091432288276196e-06,
+      "loss": 0.6677,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3792606817090736,
+      "grad_norm": 1.1315279970015388,
+      "learning_rate": 9.089505777363114e-06,
+      "loss": 0.5676,
+      "step": 1437
+    },
+    {
+      "epoch": 1.3802208353336534,
+      "grad_norm": 1.0880311134579044,
+      "learning_rate": 9.087577430714603e-06,
+      "loss": 0.5323,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3811809889582334,
+      "grad_norm": 0.7997114854628721,
+      "learning_rate": 9.085647249196275e-06,
+      "loss": 0.6705,
+      "step": 1439
+    },
+    {
+      "epoch": 1.3821411425828132,
+      "grad_norm": 1.1168457432733025,
+      "learning_rate": 9.083715233674572e-06,
+      "loss": 0.5536,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3831012962073932,
+      "grad_norm": 1.1030453093536399,
+      "learning_rate": 9.081781385016754e-06,
+      "loss": 0.6965,
+      "step": 1441
+    },
+    {
+      "epoch": 1.384061449831973,
+      "grad_norm": 1.112953806581517,
+      "learning_rate": 9.079845704090903e-06,
+      "loss": 0.6036,
+      "step": 1442
+    },
+    {
+      "epoch": 1.385021603456553,
+      "grad_norm": 1.089864089498288,
+      "learning_rate": 9.077908191765925e-06,
+      "loss": 0.676,
+      "step": 1443
+    },
+    {
+      "epoch": 1.385981757081133,
+      "grad_norm": 0.7941147593055814,
+      "learning_rate": 9.075968848911553e-06,
+      "loss": 0.6765,
+      "step": 1444
+    },
+    {
+      "epoch": 1.3869419107057128,
+      "grad_norm": 1.1538336794588986,
+      "learning_rate": 9.074027676398333e-06,
+      "loss": 0.5535,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3879020643302928,
+      "grad_norm": 0.899873624460216,
+      "learning_rate": 9.07208467509764e-06,
+      "loss": 0.5086,
+      "step": 1446
+    },
+    {
+      "epoch": 1.3888622179548729,
+      "grad_norm": 1.2441712252685813,
+      "learning_rate": 9.07013984588166e-06,
+      "loss": 0.8172,
+      "step": 1447
+    },
+    {
+      "epoch": 1.3898223715794527,
+      "grad_norm": 1.1567695863868372,
+      "learning_rate": 9.068193189623412e-06,
+      "loss": 0.6214,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3907825252040327,
+      "grad_norm": 1.1708998754473148,
+      "learning_rate": 9.06624470719673e-06,
+      "loss": 0.5282,
+      "step": 1449
+    },
+    {
+      "epoch": 1.3917426788286127,
+      "grad_norm": 0.9026111959152648,
+      "learning_rate": 9.064294399476265e-06,
+      "loss": 0.5809,
+      "step": 1450
+    },
+    {
+      "epoch": 1.3927028324531925,
+      "grad_norm": 1.2040360577373892,
+      "learning_rate": 9.062342267337487e-06,
+      "loss": 0.5999,
+      "step": 1451
+    },
+    {
+      "epoch": 1.3936629860777725,
+      "grad_norm": 1.1096935773411545,
+      "learning_rate": 9.060388311656689e-06,
+      "loss": 0.6106,
+      "step": 1452
+    },
+    {
+      "epoch": 1.3946231397023523,
+      "grad_norm": 1.0629999388500475,
+      "learning_rate": 9.058432533310982e-06,
+      "loss": 0.6841,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3955832933269323,
+      "grad_norm": 1.2674360481064615,
+      "learning_rate": 9.056474933178293e-06,
+      "loss": 0.7552,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3965434469515121,
+      "grad_norm": 1.3704949314290888,
+      "learning_rate": 9.054515512137367e-06,
+      "loss": 0.5866,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3975036005760921,
+      "grad_norm": 0.9070807201679557,
+      "learning_rate": 9.052554271067771e-06,
+      "loss": 0.5786,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3984637542006721,
+      "grad_norm": 1.0367508024077463,
+      "learning_rate": 9.05059121084988e-06,
+      "loss": 0.5443,
+      "step": 1457
+    },
+    {
+      "epoch": 1.399423907825252,
+      "grad_norm": 0.8228263048244029,
+      "learning_rate": 9.048626332364891e-06,
+      "loss": 0.665,
+      "step": 1458
+    },
+    {
+      "epoch": 1.400384061449832,
+      "grad_norm": 1.0817115873933376,
+      "learning_rate": 9.046659636494821e-06,
+      "loss": 0.5638,
+      "step": 1459
+    },
+    {
+      "epoch": 1.401344215074412,
+      "grad_norm": 1.1415393174650825,
+      "learning_rate": 9.044691124122496e-06,
+      "loss": 0.7456,
+      "step": 1460
+    },
+    {
+      "epoch": 1.4023043686989918,
+      "grad_norm": 1.042080496976787,
+      "learning_rate": 9.042720796131561e-06,
+      "loss": 0.6686,
+      "step": 1461
+    },
+    {
+      "epoch": 1.4032645223235718,
+      "grad_norm": 1.1222340889332822,
+      "learning_rate": 9.040748653406475e-06,
+      "loss": 0.752,
+      "step": 1462
+    },
+    {
+      "epoch": 1.4042246759481518,
+      "grad_norm": 1.3848173780245217,
+      "learning_rate": 9.038774696832511e-06,
+      "loss": 0.807,
+      "step": 1463
+    },
+    {
+      "epoch": 1.4051848295727316,
+      "grad_norm": 1.1288431218857278,
+      "learning_rate": 9.036798927295757e-06,
+      "loss": 0.7102,
+      "step": 1464
+    },
+    {
+      "epoch": 1.4061449831973116,
+      "grad_norm": 1.01517480041356,
+      "learning_rate": 9.034821345683117e-06,
+      "loss": 0.5792,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4071051368218914,
+      "grad_norm": 1.0674838336963253,
+      "learning_rate": 9.032841952882306e-06,
+      "loss": 0.6886,
+      "step": 1466
+    },
+    {
+      "epoch": 1.4080652904464714,
+      "grad_norm": 1.2617241953266387,
+      "learning_rate": 9.030860749781848e-06,
+      "loss": 0.6348,
+      "step": 1467
+    },
+    {
+      "epoch": 1.4090254440710512,
+      "grad_norm": 1.034571044333529,
+      "learning_rate": 9.028877737271089e-06,
+      "loss": 0.5829,
+      "step": 1468
+    },
+    {
+      "epoch": 1.4099855976956313,
+      "grad_norm": 1.0033035253032856,
+      "learning_rate": 9.02689291624018e-06,
+      "loss": 0.7316,
+      "step": 1469
+    },
+    {
+      "epoch": 1.4109457513202113,
+      "grad_norm": 1.4016124783738984,
+      "learning_rate": 9.024906287580085e-06,
+      "loss": 0.6421,
+      "step": 1470
+    },
+    {
+      "epoch": 1.411905904944791,
+      "grad_norm": 0.9863253720673301,
+      "learning_rate": 9.022917852182582e-06,
+      "loss": 0.5401,
+      "step": 1471
+    },
+    {
+      "epoch": 1.412866058569371,
+      "grad_norm": 1.754467437274913,
+      "learning_rate": 9.020927610940256e-06,
+      "loss": 0.7443,
+      "step": 1472
+    },
+    {
+      "epoch": 1.413826212193951,
+      "grad_norm": 1.2353265052336058,
+      "learning_rate": 9.018935564746509e-06,
+      "loss": 0.6328,
+      "step": 1473
+    },
+    {
+      "epoch": 1.414786365818531,
+      "grad_norm": 1.0963681138104033,
+      "learning_rate": 9.016941714495546e-06,
+      "loss": 0.6851,
+      "step": 1474
+    },
+    {
+      "epoch": 1.415746519443111,
+      "grad_norm": 1.3878482660616152,
+      "learning_rate": 9.014946061082385e-06,
+      "loss": 0.747,
+      "step": 1475
+    },
+    {
+      "epoch": 1.416706673067691,
+      "grad_norm": 1.0494981175572002,
+      "learning_rate": 9.012948605402856e-06,
+      "loss": 0.6898,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4176668266922707,
+      "grad_norm": 1.2387679013385462,
+      "learning_rate": 9.010949348353595e-06,
+      "loss": 0.6006,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4186269803168507,
+      "grad_norm": 1.2177631267117774,
+      "learning_rate": 9.008948290832046e-06,
+      "loss": 0.8044,
+      "step": 1478
+    },
+    {
+      "epoch": 1.4195871339414305,
+      "grad_norm": 1.1236670782980807,
+      "learning_rate": 9.006945433736461e-06,
+      "loss": 0.4909,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4205472875660106,
+      "grad_norm": 1.1811146156755659,
+      "learning_rate": 9.004940777965903e-06,
+      "loss": 0.6744,
+      "step": 1480
+    },
+    {
+      "epoch": 1.4215074411905904,
+      "grad_norm": 1.2952915847830806,
+      "learning_rate": 9.002934324420242e-06,
+      "loss": 0.8284,
+      "step": 1481
+    },
+    {
+      "epoch": 1.4224675948151704,
+      "grad_norm": 1.3186626780304642,
+      "learning_rate": 9.000926074000149e-06,
+      "loss": 0.6409,
+      "step": 1482
+    },
+    {
+      "epoch": 1.4234277484397504,
+      "grad_norm": 1.1083243673963934,
+      "learning_rate": 8.998916027607111e-06,
+      "loss": 0.6272,
+      "step": 1483
+    },
+    {
+      "epoch": 1.4243879020643302,
+      "grad_norm": 1.1486576286510621,
+      "learning_rate": 8.996904186143414e-06,
+      "loss": 0.6471,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4253480556889102,
+      "grad_norm": 1.427612957312725,
+      "learning_rate": 8.994890550512152e-06,
+      "loss": 0.6782,
+      "step": 1485
+    },
+    {
+      "epoch": 1.4263082093134902,
+      "grad_norm": 1.2421048458874155,
+      "learning_rate": 8.992875121617224e-06,
+      "loss": 0.6714,
+      "step": 1486
+    },
+    {
+      "epoch": 1.42726836293807,
+      "grad_norm": 1.1475684942018722,
+      "learning_rate": 8.990857900363337e-06,
+      "loss": 0.5341,
+      "step": 1487
+    },
+    {
+      "epoch": 1.42822851656265,
+      "grad_norm": 0.7789219831356208,
+      "learning_rate": 8.988838887655997e-06,
+      "loss": 0.6485,
+      "step": 1488
+    },
+    {
+      "epoch": 1.42918867018723,
+      "grad_norm": 0.9022544631477947,
+      "learning_rate": 8.98681808440152e-06,
+      "loss": 0.4919,
+      "step": 1489
+    },
+    {
+      "epoch": 1.4301488238118099,
+      "grad_norm": 1.0487182980001104,
+      "learning_rate": 8.984795491507021e-06,
+      "loss": 0.5778,
+      "step": 1490
+    },
+    {
+      "epoch": 1.4311089774363899,
+      "grad_norm": 0.7765690777775737,
+      "learning_rate": 8.982771109880421e-06,
+      "loss": 0.6641,
+      "step": 1491
+    },
+    {
+      "epoch": 1.43206913106097,
+      "grad_norm": 0.9770150771741994,
+      "learning_rate": 8.980744940430443e-06,
+      "loss": 0.5373,
+      "step": 1492
+    },
+    {
+      "epoch": 1.4330292846855497,
+      "grad_norm": 1.0842856980753282,
+      "learning_rate": 8.978716984066613e-06,
+      "loss": 0.6931,
+      "step": 1493
+    },
+    {
+      "epoch": 1.4339894383101295,
+      "grad_norm": 1.5905291871795488,
+      "learning_rate": 8.976687241699258e-06,
+      "loss": 0.6763,
+      "step": 1494
+    },
+    {
+      "epoch": 1.4349495919347095,
+      "grad_norm": 1.1291160885943465,
+      "learning_rate": 8.97465571423951e-06,
+      "loss": 0.7141,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4359097455592895,
+      "grad_norm": 0.8847319400098621,
+      "learning_rate": 8.972622402599298e-06,
+      "loss": 0.6199,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4368698991838693,
+      "grad_norm": 1.0193428575785228,
+      "learning_rate": 8.970587307691357e-06,
+      "loss": 0.5403,
+      "step": 1497
+    },
+    {
+      "epoch": 1.4378300528084493,
+      "grad_norm": 1.0872663796559396,
+      "learning_rate": 8.968550430429212e-06,
+      "loss": 0.5434,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4387902064330294,
+      "grad_norm": 1.2406773363844592,
+      "learning_rate": 8.966511771727205e-06,
+      "loss": 0.5879,
+      "step": 1499
+    },
+    {
+      "epoch": 1.4397503600576091,
+      "grad_norm": 0.8156024523464949,
+      "learning_rate": 8.96447133250046e-06,
+      "loss": 0.6312,
+      "step": 1500
+    },
+    {
+      "epoch": 1.4407105136821892,
+      "grad_norm": 1.257440132329098,
+      "learning_rate": 8.962429113664913e-06,
+      "loss": 0.6966,
+      "step": 1501
+    },
+    {
+      "epoch": 1.4416706673067692,
+      "grad_norm": 1.0628200522245184,
+      "learning_rate": 8.960385116137293e-06,
+      "loss": 0.6327,
+      "step": 1502
+    },
+    {
+      "epoch": 1.442630820931349,
+      "grad_norm": 1.1596260122624005,
+      "learning_rate": 8.958339340835128e-06,
+      "loss": 0.5703,
+      "step": 1503
+    },
+    {
+      "epoch": 1.443590974555929,
+      "grad_norm": 1.1298747586240705,
+      "learning_rate": 8.956291788676746e-06,
+      "loss": 0.5657,
+      "step": 1504
+    },
+    {
+      "epoch": 1.444551128180509,
+      "grad_norm": 0.8397355051213713,
+      "learning_rate": 8.95424246058127e-06,
+      "loss": 0.6697,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4455112818050888,
+      "grad_norm": 1.0161743613809866,
+      "learning_rate": 8.952191357468621e-06,
+      "loss": 0.5619,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4464714354296688,
+      "grad_norm": 1.2106464122559442,
+      "learning_rate": 8.950138480259519e-06,
+      "loss": 0.57,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4474315890542486,
+      "grad_norm": 1.1265109245805303,
+      "learning_rate": 8.948083829875478e-06,
+      "loss": 0.5467,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4483917426788286,
+      "grad_norm": 1.0933628766801322,
+      "learning_rate": 8.946027407238809e-06,
+      "loss": 0.588,
+      "step": 1509
+    },
+    {
+      "epoch": 1.4493518963034084,
+      "grad_norm": 1.104099886204487,
+      "learning_rate": 8.943969213272616e-06,
+      "loss": 0.6103,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4503120499279885,
+      "grad_norm": 0.9561757908407287,
+      "learning_rate": 8.941909248900804e-06,
+      "loss": 0.5385,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4512722035525685,
+      "grad_norm": 1.1169073689054023,
+      "learning_rate": 8.939847515048066e-06,
+      "loss": 0.6164,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4522323571771483,
+      "grad_norm": 0.8551643979571723,
+      "learning_rate": 8.937784012639892e-06,
+      "loss": 0.676,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4531925108017283,
+      "grad_norm": 1.0042682995889094,
+      "learning_rate": 8.935718742602568e-06,
+      "loss": 0.6287,
+      "step": 1514
+    },
+    {
+      "epoch": 1.4541526644263083,
+      "grad_norm": 1.0614362268942157,
+      "learning_rate": 8.933651705863172e-06,
+      "loss": 0.7373,
+      "step": 1515
+    },
+    {
+      "epoch": 1.455112818050888,
+      "grad_norm": 1.3401565218668898,
+      "learning_rate": 8.931582903349571e-06,
+      "loss": 0.5156,
+      "step": 1516
+    },
+    {
+      "epoch": 1.4560729716754681,
+      "grad_norm": 1.2341123257068967,
+      "learning_rate": 8.929512335990434e-06,
+      "loss": 0.7821,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4570331253000481,
+      "grad_norm": 1.7078266511969085,
+      "learning_rate": 8.92744000471521e-06,
+      "loss": 0.6796,
+      "step": 1518
+    },
+    {
+      "epoch": 1.457993278924628,
+      "grad_norm": 1.4104645213806324,
+      "learning_rate": 8.925365910454153e-06,
+      "loss": 0.5957,
+      "step": 1519
+    },
+    {
+      "epoch": 1.458953432549208,
+      "grad_norm": 1.172908486399178,
+      "learning_rate": 8.923290054138298e-06,
+      "loss": 0.5026,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4599135861737877,
+      "grad_norm": 0.9886504176722553,
+      "learning_rate": 8.921212436699476e-06,
+      "loss": 0.5457,
+      "step": 1521
+    },
+    {
+      "epoch": 1.4608737397983678,
+      "grad_norm": 0.7259194922521678,
+      "learning_rate": 8.919133059070307e-06,
+      "loss": 0.5767,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4618338934229476,
+      "grad_norm": 1.1329123474845246,
+      "learning_rate": 8.9170519221842e-06,
+      "loss": 0.6652,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4627940470475276,
+      "grad_norm": 1.3825900560877011,
+      "learning_rate": 8.914969026975354e-06,
+      "loss": 0.65,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4637542006721076,
+      "grad_norm": 1.6252305372983962,
+      "learning_rate": 8.912884374378763e-06,
+      "loss": 0.6646,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4647143542966874,
+      "grad_norm": 1.1724088736710376,
+      "learning_rate": 8.910797965330202e-06,
+      "loss": 0.5923,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4656745079212674,
+      "grad_norm": 1.3940002698365594,
+      "learning_rate": 8.908709800766237e-06,
+      "loss": 0.73,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4666346615458474,
+      "grad_norm": 1.2639384312093593,
+      "learning_rate": 8.906619881624225e-06,
+      "loss": 0.6578,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4675948151704272,
+      "grad_norm": 1.0940793340621946,
+      "learning_rate": 8.90452820884231e-06,
+      "loss": 0.607,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4685549687950072,
+      "grad_norm": 1.0542543562205764,
+      "learning_rate": 8.902434783359417e-06,
+      "loss": 0.6936,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4695151224195873,
+      "grad_norm": 1.1879881235521863,
+      "learning_rate": 8.900339606115264e-06,
+      "loss": 0.6887,
+      "step": 1531
+    },
+    {
+      "epoch": 1.470475276044167,
+      "grad_norm": 0.8757796270384739,
+      "learning_rate": 8.898242678050354e-06,
+      "loss": 0.7135,
+      "step": 1532
+    },
+    {
+      "epoch": 1.471435429668747,
+      "grad_norm": 1.0409790443247378,
+      "learning_rate": 8.896144000105979e-06,
+      "loss": 0.5364,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4723955832933269,
+      "grad_norm": 0.8076482116173296,
+      "learning_rate": 8.894043573224207e-06,
+      "loss": 0.6703,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4733557369179069,
+      "grad_norm": 1.0625925178859534,
+      "learning_rate": 8.8919413983479e-06,
+      "loss": 0.5318,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4743158905424867,
+      "grad_norm": 1.049571449075552,
+      "learning_rate": 8.889837476420703e-06,
+      "loss": 0.6438,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4752760441670667,
+      "grad_norm": 0.8197834071354191,
+      "learning_rate": 8.887731808387043e-06,
+      "loss": 0.7928,
+      "step": 1537
+    },
+    {
+      "epoch": 1.4762361977916467,
+      "grad_norm": 1.0166060432466146,
+      "learning_rate": 8.885624395192131e-06,
+      "loss": 0.7061,
+      "step": 1538
+    },
+    {
+      "epoch": 1.4771963514162265,
+      "grad_norm": 1.129472818971423,
+      "learning_rate": 8.883515237781964e-06,
+      "loss": 0.6651,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4781565050408065,
+      "grad_norm": 0.9573464426762843,
+      "learning_rate": 8.881404337103316e-06,
+      "loss": 0.7077,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4791166586653866,
+      "grad_norm": 1.1004021355840239,
+      "learning_rate": 8.879291694103754e-06,
+      "loss": 0.5878,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4800768122899663,
+      "grad_norm": 1.178159407699785,
+      "learning_rate": 8.877177309731618e-06,
+      "loss": 0.5575,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4810369659145464,
+      "grad_norm": 1.0159381485089358,
+      "learning_rate": 8.875061184936031e-06,
+      "loss": 0.6366,
+      "step": 1543
+    },
+    {
+      "epoch": 1.4819971195391264,
+      "grad_norm": 0.8997652197523106,
+      "learning_rate": 8.872943320666901e-06,
+      "loss": 0.6194,
+      "step": 1544
+    },
+    {
+      "epoch": 1.4829572731637062,
+      "grad_norm": 0.8337447335571384,
+      "learning_rate": 8.870823717874912e-06,
+      "loss": 0.4955,
+      "step": 1545
+    },
+    {
+      "epoch": 1.4839174267882862,
+      "grad_norm": 1.3692228718413766,
+      "learning_rate": 8.868702377511535e-06,
+      "loss": 0.7163,
+      "step": 1546
+    },
+    {
+      "epoch": 1.484877580412866,
+      "grad_norm": 1.4991294593929196,
+      "learning_rate": 8.866579300529014e-06,
+      "loss": 0.4715,
+      "step": 1547
+    },
+    {
+      "epoch": 1.485837734037446,
+      "grad_norm": 1.1141143253404242,
+      "learning_rate": 8.864454487880377e-06,
+      "loss": 0.6575,
+      "step": 1548
+    },
+    {
+      "epoch": 1.4867978876620258,
+      "grad_norm": 1.356163506582758,
+      "learning_rate": 8.862327940519429e-06,
+      "loss": 0.6527,
+      "step": 1549
+    },
+    {
+      "epoch": 1.4877580412866058,
+      "grad_norm": 0.7735328461625155,
+      "learning_rate": 8.860199659400754e-06,
+      "loss": 0.6868,
+      "step": 1550
+    },
+    {
+      "epoch": 1.4887181949111858,
+      "grad_norm": 1.0901216355933054,
+      "learning_rate": 8.858069645479715e-06,
+      "loss": 0.6636,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4896783485357656,
+      "grad_norm": 0.9660448042567134,
+      "learning_rate": 8.855937899712454e-06,
+      "loss": 0.5188,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4906385021603457,
+      "grad_norm": 1.0414626727490273,
+      "learning_rate": 8.853804423055887e-06,
+      "loss": 0.6597,
+      "step": 1553
+    },
+    {
+      "epoch": 1.4915986557849257,
+      "grad_norm": 1.1575737580267278,
+      "learning_rate": 8.851669216467708e-06,
+      "loss": 0.6408,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4925588094095055,
+      "grad_norm": 0.9906915429431391,
+      "learning_rate": 8.849532280906391e-06,
+      "loss": 0.5295,
+      "step": 1555
+    },
+    {
+      "epoch": 1.4935189630340855,
+      "grad_norm": 1.3905846032117894,
+      "learning_rate": 8.847393617331182e-06,
+      "loss": 0.5678,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4944791166586655,
+      "grad_norm": 1.2418429156286361,
+      "learning_rate": 8.845253226702103e-06,
+      "loss": 0.5972,
+      "step": 1557
+    },
+    {
+      "epoch": 1.4954392702832453,
+      "grad_norm": 1.1443760223671664,
+      "learning_rate": 8.843111109979957e-06,
+      "loss": 0.6809,
+      "step": 1558
+    },
+    {
+      "epoch": 1.4963994239078253,
+      "grad_norm": 1.2741808025232688,
+      "learning_rate": 8.840967268126313e-06,
+      "loss": 0.6569,
+      "step": 1559
+    },
+    {
+      "epoch": 1.4973595775324051,
+      "grad_norm": 1.410259091742139,
+      "learning_rate": 8.83882170210352e-06,
+      "loss": 0.6657,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4983197311569851,
+      "grad_norm": 1.228627765664867,
+      "learning_rate": 8.836674412874699e-06,
+      "loss": 0.6022,
+      "step": 1561
+    },
+    {
+      "epoch": 1.499279884781565,
+      "grad_norm": 1.1911640645100297,
+      "learning_rate": 8.834525401403747e-06,
+      "loss": 0.6037,
+      "step": 1562
+    },
+    {
+      "epoch": 1.500240038406145,
+      "grad_norm": 1.193801561305446,
+      "learning_rate": 8.83237466865533e-06,
+      "loss": 0.6554,
+      "step": 1563
+    },
+    {
+      "epoch": 1.501200192030725,
+      "grad_norm": 1.1818358561874878,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.6398,
+      "step": 1564
+    },
+    {
+      "epoch": 1.5021603456553048,
+      "grad_norm": 0.8670851130447548,
+      "learning_rate": 8.828068043188641e-06,
+      "loss": 0.6182,
+      "step": 1565
+    },
+    {
+      "epoch": 1.5031204992798848,
+      "grad_norm": 0.9407499345905682,
+      "learning_rate": 8.825912152403568e-06,
+      "loss": 0.6049,
+      "step": 1566
+    },
+    {
+      "epoch": 1.5040806529044648,
+      "grad_norm": 1.0151430781400177,
+      "learning_rate": 8.823754544207426e-06,
+      "loss": 0.6779,
+      "step": 1567
+    },
+    {
+      "epoch": 1.5050408065290446,
+      "grad_norm": 2.185154987340034,
+      "learning_rate": 8.821595219568741e-06,
+      "loss": 0.7579,
+      "step": 1568
+    },
+    {
+      "epoch": 1.5060009601536246,
+      "grad_norm": 1.6014949715673676,
+      "learning_rate": 8.819434179456814e-06,
+      "loss": 0.6781,
+      "step": 1569
+    },
+    {
+      "epoch": 1.5069611137782046,
+      "grad_norm": 1.1293765394233934,
+      "learning_rate": 8.817271424841711e-06,
+      "loss": 0.7799,
+      "step": 1570
+    },
+    {
+      "epoch": 1.5079212674027844,
+      "grad_norm": 1.4630651688147316,
+      "learning_rate": 8.81510695669427e-06,
+      "loss": 0.6272,
+      "step": 1571
+    },
+    {
+      "epoch": 1.5088814210273642,
+      "grad_norm": 1.0806000078130642,
+      "learning_rate": 8.812940775986098e-06,
+      "loss": 0.5766,
+      "step": 1572
+    },
+    {
+      "epoch": 1.5098415746519445,
+      "grad_norm": 1.0636465241656676,
+      "learning_rate": 8.810772883689567e-06,
+      "loss": 0.5887,
+      "step": 1573
+    },
+    {
+      "epoch": 1.5108017282765243,
+      "grad_norm": 1.0826398042376562,
+      "learning_rate": 8.808603280777826e-06,
+      "loss": 0.6777,
+      "step": 1574
+    },
+    {
+      "epoch": 1.511761881901104,
+      "grad_norm": 1.0211975863239149,
+      "learning_rate": 8.806431968224784e-06,
+      "loss": 0.4892,
+      "step": 1575
+    },
+    {
+      "epoch": 1.512722035525684,
+      "grad_norm": 0.9734221712990038,
+      "learning_rate": 8.80425894700512e-06,
+      "loss": 0.676,
+      "step": 1576
+    },
+    {
+      "epoch": 1.513682189150264,
+      "grad_norm": 1.1483419353910045,
+      "learning_rate": 8.802084218094278e-06,
+      "loss": 0.6756,
+      "step": 1577
+    },
+    {
+      "epoch": 1.5146423427748439,
+      "grad_norm": 0.9317491709337571,
+      "learning_rate": 8.799907782468473e-06,
+      "loss": 0.5977,
+      "step": 1578
+    },
+    {
+      "epoch": 1.515602496399424,
+      "grad_norm": 0.960517889530444,
+      "learning_rate": 8.797729641104687e-06,
+      "loss": 0.5713,
+      "step": 1579
+    },
+    {
+      "epoch": 1.516562650024004,
+      "grad_norm": 0.9671139854015403,
+      "learning_rate": 8.795549794980655e-06,
+      "loss": 0.7245,
+      "step": 1580
+    },
+    {
+      "epoch": 1.5175228036485837,
+      "grad_norm": 1.3137409695984128,
+      "learning_rate": 8.793368245074896e-06,
+      "loss": 0.7528,
+      "step": 1581
+    },
+    {
+      "epoch": 1.5184829572731637,
+      "grad_norm": 1.0429450014620933,
+      "learning_rate": 8.791184992366678e-06,
+      "loss": 0.6675,
+      "step": 1582
+    },
+    {
+      "epoch": 1.5194431108977438,
+      "grad_norm": 1.2760113502583366,
+      "learning_rate": 8.789000037836045e-06,
+      "loss": 0.6094,
+      "step": 1583
+    },
+    {
+      "epoch": 1.5204032645223235,
+      "grad_norm": 1.1925925207183954,
+      "learning_rate": 8.786813382463796e-06,
+      "loss": 0.5015,
+      "step": 1584
+    },
+    {
+      "epoch": 1.5213634181469033,
+      "grad_norm": 0.8226805823705247,
+      "learning_rate": 8.7846250272315e-06,
+      "loss": 0.6676,
+      "step": 1585
+    },
+    {
+      "epoch": 1.5223235717714836,
+      "grad_norm": 0.8228915234369152,
+      "learning_rate": 8.782434973121481e-06,
+      "loss": 0.5947,
+      "step": 1586
+    },
+    {
+      "epoch": 1.5232837253960634,
+      "grad_norm": 0.9574007111776064,
+      "learning_rate": 8.780243221116838e-06,
+      "loss": 0.7298,
+      "step": 1587
+    },
+    {
+      "epoch": 1.5242438790206432,
+      "grad_norm": 1.0767250417525245,
+      "learning_rate": 8.77804977220142e-06,
+      "loss": 0.7061,
+      "step": 1588
+    },
+    {
+      "epoch": 1.5252040326452232,
+      "grad_norm": 1.0975411528156236,
+      "learning_rate": 8.775854627359844e-06,
+      "loss": 0.5675,
+      "step": 1589
+    },
+    {
+      "epoch": 1.5261641862698032,
+      "grad_norm": 1.0705241678902788,
+      "learning_rate": 8.773657787577488e-06,
+      "loss": 0.6927,
+      "step": 1590
+    },
+    {
+      "epoch": 1.527124339894383,
+      "grad_norm": 1.2543988040625176,
+      "learning_rate": 8.771459253840489e-06,
+      "loss": 0.6857,
+      "step": 1591
+    },
+    {
+      "epoch": 1.528084493518963,
+      "grad_norm": 1.0609016838340781,
+      "learning_rate": 8.769259027135746e-06,
+      "loss": 0.6519,
+      "step": 1592
+    },
+    {
+      "epoch": 1.529044647143543,
+      "grad_norm": 0.8704882959786878,
+      "learning_rate": 8.767057108450919e-06,
+      "loss": 0.531,
+      "step": 1593
+    },
+    {
+      "epoch": 1.5300048007681228,
+      "grad_norm": 0.9282481560389485,
+      "learning_rate": 8.76485349877442e-06,
+      "loss": 0.7366,
+      "step": 1594
+    },
+    {
+      "epoch": 1.5309649543927029,
+      "grad_norm": 0.9511335341437126,
+      "learning_rate": 8.76264819909543e-06,
+      "loss": 0.6666,
+      "step": 1595
+    },
+    {
+      "epoch": 1.5319251080172829,
+      "grad_norm": 1.1111787893102145,
+      "learning_rate": 8.760441210403886e-06,
+      "loss": 0.6375,
+      "step": 1596
+    },
+    {
+      "epoch": 1.5328852616418627,
+      "grad_norm": 1.092816915711456,
+      "learning_rate": 8.758232533690479e-06,
+      "loss": 0.6397,
+      "step": 1597
+    },
+    {
+      "epoch": 1.5338454152664425,
+      "grad_norm": 1.3155672256646667,
+      "learning_rate": 8.75602216994666e-06,
+      "loss": 0.6269,
+      "step": 1598
+    },
+    {
+      "epoch": 1.5348055688910227,
+      "grad_norm": 0.741803062736798,
+      "learning_rate": 8.753810120164639e-06,
+      "loss": 0.6418,
+      "step": 1599
+    },
+    {
+      "epoch": 1.5357657225156025,
+      "grad_norm": 1.1873303304018221,
+      "learning_rate": 8.75159638533738e-06,
+      "loss": 0.5432,
+      "step": 1600
+    },
+    {
+      "epoch": 1.5367258761401823,
+      "grad_norm": 0.7668290863461923,
+      "learning_rate": 8.749380966458608e-06,
+      "loss": 0.7538,
+      "step": 1601
+    },
+    {
+      "epoch": 1.5376860297647623,
+      "grad_norm": 1.3027080236914383,
+      "learning_rate": 8.747163864522796e-06,
+      "loss": 0.6736,
+      "step": 1602
+    },
+    {
+      "epoch": 1.5386461833893423,
+      "grad_norm": 0.646484719298794,
+      "learning_rate": 8.744945080525182e-06,
+      "loss": 0.6115,
+      "step": 1603
+    },
+    {
+      "epoch": 1.5396063370139221,
+      "grad_norm": 1.5003456905289576,
+      "learning_rate": 8.742724615461753e-06,
+      "loss": 0.6875,
+      "step": 1604
+    },
+    {
+      "epoch": 1.5405664906385022,
+      "grad_norm": 1.4529237102280468,
+      "learning_rate": 8.740502470329251e-06,
+      "loss": 0.7543,
+      "step": 1605
+    },
+    {
+      "epoch": 1.5415266442630822,
+      "grad_norm": 1.264249597177811,
+      "learning_rate": 8.738278646125174e-06,
+      "loss": 0.7174,
+      "step": 1606
+    },
+    {
+      "epoch": 1.542486797887662,
+      "grad_norm": 0.8006931681927969,
+      "learning_rate": 8.736053143847774e-06,
+      "loss": 0.5752,
+      "step": 1607
+    },
+    {
+      "epoch": 1.543446951512242,
+      "grad_norm": 0.9612503521794223,
+      "learning_rate": 8.733825964496052e-06,
+      "loss": 0.6369,
+      "step": 1608
+    },
+    {
+      "epoch": 1.544407105136822,
+      "grad_norm": 1.2263827521809219,
+      "learning_rate": 8.731597109069768e-06,
+      "loss": 0.6944,
+      "step": 1609
+    },
+    {
+      "epoch": 1.5453672587614018,
+      "grad_norm": 0.8117605700550553,
+      "learning_rate": 8.72936657856943e-06,
+      "loss": 0.6205,
+      "step": 1610
+    },
+    {
+      "epoch": 1.5463274123859816,
+      "grad_norm": 1.0844304296657585,
+      "learning_rate": 8.727134373996298e-06,
+      "loss": 0.6762,
+      "step": 1611
+    },
+    {
+      "epoch": 1.5472875660105618,
+      "grad_norm": 1.0356413997768932,
+      "learning_rate": 8.724900496352388e-06,
+      "loss": 0.5406,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5482477196351416,
+      "grad_norm": 1.2889216426140757,
+      "learning_rate": 8.722664946640463e-06,
+      "loss": 0.5896,
+      "step": 1613
+    },
+    {
+      "epoch": 1.5492078732597214,
+      "grad_norm": 1.1650345873037744,
+      "learning_rate": 8.720427725864035e-06,
+      "loss": 0.5814,
+      "step": 1614
+    },
+    {
+      "epoch": 1.5501680268843014,
+      "grad_norm": 0.7366014480030797,
+      "learning_rate": 8.71818883502737e-06,
+      "loss": 0.6081,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5511281805088815,
+      "grad_norm": 1.1318867172011045,
+      "learning_rate": 8.715948275135482e-06,
+      "loss": 0.6559,
+      "step": 1616
+    },
+    {
+      "epoch": 1.5520883341334613,
+      "grad_norm": 0.9487112907304925,
+      "learning_rate": 8.713706047194135e-06,
+      "loss": 0.6027,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5530484877580413,
+      "grad_norm": 1.0802687230047274,
+      "learning_rate": 8.711462152209843e-06,
+      "loss": 0.8259,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5540086413826213,
+      "grad_norm": 1.6731252047990353,
+      "learning_rate": 8.709216591189861e-06,
+      "loss": 0.75,
+      "step": 1619
+    },
+    {
+      "epoch": 1.554968795007201,
+      "grad_norm": 1.1383448741414959,
+      "learning_rate": 8.706969365142202e-06,
+      "loss": 0.6404,
+      "step": 1620
+    },
+    {
+      "epoch": 1.555928948631781,
+      "grad_norm": 1.1778185492412427,
+      "learning_rate": 8.70472047507562e-06,
+      "loss": 0.6958,
+      "step": 1621
+    },
+    {
+      "epoch": 1.5568891022563611,
+      "grad_norm": 0.9686264191189811,
+      "learning_rate": 8.702469921999617e-06,
+      "loss": 0.6271,
+      "step": 1622
+    },
+    {
+      "epoch": 1.557849255880941,
+      "grad_norm": 1.397611275414117,
+      "learning_rate": 8.700217706924445e-06,
+      "loss": 0.7544,
+      "step": 1623
+    },
+    {
+      "epoch": 1.558809409505521,
+      "grad_norm": 1.4513596363263108,
+      "learning_rate": 8.697963830861095e-06,
+      "loss": 0.7178,
+      "step": 1624
+    },
+    {
+      "epoch": 1.559769563130101,
+      "grad_norm": 1.1690765029827235,
+      "learning_rate": 8.695708294821314e-06,
+      "loss": 0.659,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5607297167546808,
+      "grad_norm": 1.0258868567164887,
+      "learning_rate": 8.693451099817583e-06,
+      "loss": 0.693,
+      "step": 1626
+    },
+    {
+      "epoch": 1.5616898703792605,
+      "grad_norm": 1.2127126511703938,
+      "learning_rate": 8.691192246863133e-06,
+      "loss": 0.676,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5626500240038406,
+      "grad_norm": 1.4851939462454826,
+      "learning_rate": 8.688931736971941e-06,
+      "loss": 0.5758,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5636101776284206,
+      "grad_norm": 1.0634247458009611,
+      "learning_rate": 8.686669571158724e-06,
+      "loss": 0.4995,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5645703312530004,
+      "grad_norm": 0.8394074207994456,
+      "learning_rate": 8.684405750438944e-06,
+      "loss": 0.5563,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5655304848775804,
+      "grad_norm": 1.1468296996400023,
+      "learning_rate": 8.682140275828809e-06,
+      "loss": 0.7286,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5664906385021604,
+      "grad_norm": 1.0289285466404867,
+      "learning_rate": 8.679873148345262e-06,
+      "loss": 0.5079,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5674507921267402,
+      "grad_norm": 0.9468257845941952,
+      "learning_rate": 8.677604369005996e-06,
+      "loss": 0.6046,
+      "step": 1633
+    },
+    {
+      "epoch": 1.5684109457513202,
+      "grad_norm": 1.1697711097966919,
+      "learning_rate": 8.67533393882944e-06,
+      "loss": 0.7152,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5693710993759002,
+      "grad_norm": 1.4658046671031377,
+      "learning_rate": 8.673061858834766e-06,
+      "loss": 0.7187,
+      "step": 1635
+    },
+    {
+      "epoch": 1.57033125300048,
+      "grad_norm": 1.3430348470894165,
+      "learning_rate": 8.670788130041886e-06,
+      "loss": 0.7524,
+      "step": 1636
+    },
+    {
+      "epoch": 1.57129140662506,
+      "grad_norm": 1.0160045975668386,
+      "learning_rate": 8.668512753471453e-06,
+      "loss": 0.5755,
+      "step": 1637
+    },
+    {
+      "epoch": 1.57225156024964,
+      "grad_norm": 1.1038876177114614,
+      "learning_rate": 8.666235730144858e-06,
+      "loss": 0.7345,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5732117138742199,
+      "grad_norm": 1.1054129476807606,
+      "learning_rate": 8.663957061084234e-06,
+      "loss": 0.6143,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5741718674987997,
+      "grad_norm": 0.9453810347799845,
+      "learning_rate": 8.661676747312453e-06,
+      "loss": 0.569,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5751320211233797,
+      "grad_norm": 0.9163203128093458,
+      "learning_rate": 8.65939478985312e-06,
+      "loss": 0.6403,
+      "step": 1641
+    },
+    {
+      "epoch": 1.5760921747479597,
+      "grad_norm": 0.9042041854556812,
+      "learning_rate": 8.657111189730583e-06,
+      "loss": 0.5737,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5770523283725395,
+      "grad_norm": 0.9211637353686093,
+      "learning_rate": 8.654825947969924e-06,
+      "loss": 0.6265,
+      "step": 1643
+    },
+    {
+      "epoch": 1.5780124819971195,
+      "grad_norm": 0.8703238211166408,
+      "learning_rate": 8.652539065596966e-06,
+      "loss": 0.5899,
+      "step": 1644
+    },
+    {
+      "epoch": 1.5789726356216995,
+      "grad_norm": 1.2076903241189283,
+      "learning_rate": 8.650250543638264e-06,
+      "loss": 0.7573,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5799327892462793,
+      "grad_norm": 1.2989796355785819,
+      "learning_rate": 8.647960383121113e-06,
+      "loss": 0.4868,
+      "step": 1646
+    },
+    {
+      "epoch": 1.5808929428708594,
+      "grad_norm": 0.8735979859576927,
+      "learning_rate": 8.64566858507354e-06,
+      "loss": 0.5502,
+      "step": 1647
+    },
+    {
+      "epoch": 1.5818530964954394,
+      "grad_norm": 0.8546304127362417,
+      "learning_rate": 8.643375150524308e-06,
+      "loss": 0.5565,
+      "step": 1648
+    },
+    {
+      "epoch": 1.5828132501200192,
+      "grad_norm": 1.0235829348311989,
+      "learning_rate": 8.641080080502919e-06,
+      "loss": 0.6458,
+      "step": 1649
+    },
+    {
+      "epoch": 1.5837734037445992,
+      "grad_norm": 1.1038066712654366,
+      "learning_rate": 8.6387833760396e-06,
+      "loss": 0.6618,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5847335573691792,
+      "grad_norm": 1.2922891293982486,
+      "learning_rate": 8.636485038165323e-06,
+      "loss": 0.6367,
+      "step": 1651
+    },
+    {
+      "epoch": 1.585693710993759,
+      "grad_norm": 1.03589005937051,
+      "learning_rate": 8.634185067911781e-06,
+      "loss": 0.6252,
+      "step": 1652
+    },
+    {
+      "epoch": 1.5866538646183388,
+      "grad_norm": 1.1252711021383222,
+      "learning_rate": 8.631883466311412e-06,
+      "loss": 0.6921,
+      "step": 1653
+    },
+    {
+      "epoch": 1.5876140182429188,
+      "grad_norm": 0.8074113897518285,
+      "learning_rate": 8.629580234397377e-06,
+      "loss": 0.6712,
+      "step": 1654
+    },
+    {
+      "epoch": 1.5885741718674988,
+      "grad_norm": 1.098319269114691,
+      "learning_rate": 8.627275373203572e-06,
+      "loss": 0.6199,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5895343254920786,
+      "grad_norm": 0.902340765050979,
+      "learning_rate": 8.624968883764626e-06,
+      "loss": 0.6737,
+      "step": 1656
+    },
+    {
+      "epoch": 1.5904944791166586,
+      "grad_norm": 0.9334608234803287,
+      "learning_rate": 8.622660767115897e-06,
+      "loss": 0.6133,
+      "step": 1657
+    },
+    {
+      "epoch": 1.5914546327412387,
+      "grad_norm": 1.4876461904454708,
+      "learning_rate": 8.620351024293475e-06,
+      "loss": 0.7382,
+      "step": 1658
+    },
+    {
+      "epoch": 1.5924147863658185,
+      "grad_norm": 1.207322463040773,
+      "learning_rate": 8.618039656334173e-06,
+      "loss": 0.5967,
+      "step": 1659
+    },
+    {
+      "epoch": 1.5933749399903985,
+      "grad_norm": 0.9301430109977181,
+      "learning_rate": 8.615726664275547e-06,
+      "loss": 0.5669,
+      "step": 1660
+    },
+    {
+      "epoch": 1.5943350936149785,
+      "grad_norm": 0.9535514138575241,
+      "learning_rate": 8.613412049155872e-06,
+      "loss": 0.6078,
+      "step": 1661
+    },
+    {
+      "epoch": 1.5952952472395583,
+      "grad_norm": 1.4079098575610178,
+      "learning_rate": 8.611095812014155e-06,
+      "loss": 0.7434,
+      "step": 1662
+    },
+    {
+      "epoch": 1.5962554008641383,
+      "grad_norm": 1.1057978602184975,
+      "learning_rate": 8.608777953890125e-06,
+      "loss": 0.5356,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5972155544887183,
+      "grad_norm": 0.7716019001281486,
+      "learning_rate": 8.60645847582425e-06,
+      "loss": 0.5902,
+      "step": 1664
+    },
+    {
+      "epoch": 1.5981757081132981,
+      "grad_norm": 0.9718399381375142,
+      "learning_rate": 8.604137378857713e-06,
+      "loss": 0.6267,
+      "step": 1665
+    },
+    {
+      "epoch": 1.599135861737878,
+      "grad_norm": 0.9107669288703225,
+      "learning_rate": 8.601814664032434e-06,
+      "loss": 0.6825,
+      "step": 1666
+    },
+    {
+      "epoch": 1.6000960153624582,
+      "grad_norm": 1.0621926560828503,
+      "learning_rate": 8.599490332391054e-06,
+      "loss": 0.7198,
+      "step": 1667
+    },
+    {
+      "epoch": 1.601056168987038,
+      "grad_norm": 1.1535861334047675,
+      "learning_rate": 8.59716438497694e-06,
+      "loss": 0.6236,
+      "step": 1668
+    },
+    {
+      "epoch": 1.6020163226116177,
+      "grad_norm": 0.7516776372352658,
+      "learning_rate": 8.594836822834185e-06,
+      "loss": 0.6028,
+      "step": 1669
+    },
+    {
+      "epoch": 1.6029764762361978,
+      "grad_norm": 1.1685643580367917,
+      "learning_rate": 8.592507647007606e-06,
+      "loss": 0.6472,
+      "step": 1670
+    },
+    {
+      "epoch": 1.6039366298607778,
+      "grad_norm": 1.0980991426900863,
+      "learning_rate": 8.590176858542748e-06,
+      "loss": 0.6724,
+      "step": 1671
+    },
+    {
+      "epoch": 1.6048967834853576,
+      "grad_norm": 1.0716280689989421,
+      "learning_rate": 8.587844458485874e-06,
+      "loss": 0.7001,
+      "step": 1672
+    },
+    {
+      "epoch": 1.6058569371099376,
+      "grad_norm": 1.2080004720156137,
+      "learning_rate": 8.585510447883975e-06,
+      "loss": 0.657,
+      "step": 1673
+    },
+    {
+      "epoch": 1.6068170907345176,
+      "grad_norm": 1.089672816112207,
+      "learning_rate": 8.583174827784762e-06,
+      "loss": 0.5745,
+      "step": 1674
+    },
+    {
+      "epoch": 1.6077772443590974,
+      "grad_norm": 0.8344681189241184,
+      "learning_rate": 8.580837599236673e-06,
+      "loss": 0.5936,
+      "step": 1675
+    },
+    {
+      "epoch": 1.6087373979836774,
+      "grad_norm": 1.0357626272752878,
+      "learning_rate": 8.57849876328886e-06,
+      "loss": 0.6799,
+      "step": 1676
+    },
+    {
+      "epoch": 1.6096975516082574,
+      "grad_norm": 1.005742651907251,
+      "learning_rate": 8.576158320991205e-06,
+      "loss": 0.62,
+      "step": 1677
+    },
+    {
+      "epoch": 1.6106577052328372,
+      "grad_norm": 0.9722385404833888,
+      "learning_rate": 8.573816273394308e-06,
+      "loss": 0.6483,
+      "step": 1678
+    },
+    {
+      "epoch": 1.611617858857417,
+      "grad_norm": 1.2115310207437309,
+      "learning_rate": 8.571472621549488e-06,
+      "loss": 0.6395,
+      "step": 1679
+    },
+    {
+      "epoch": 1.6125780124819973,
+      "grad_norm": 1.0063716252451547,
+      "learning_rate": 8.569127366508782e-06,
+      "loss": 0.6235,
+      "step": 1680
+    },
+    {
+      "epoch": 1.613538166106577,
+      "grad_norm": 0.9894620552857323,
+      "learning_rate": 8.566780509324956e-06,
+      "loss": 0.5872,
+      "step": 1681
+    },
+    {
+      "epoch": 1.6144983197311569,
+      "grad_norm": 0.9067993442019877,
+      "learning_rate": 8.564432051051485e-06,
+      "loss": 0.6886,
+      "step": 1682
+    },
+    {
+      "epoch": 1.615458473355737,
+      "grad_norm": 1.361695050418203,
+      "learning_rate": 8.562081992742569e-06,
+      "loss": 0.64,
+      "step": 1683
+    },
+    {
+      "epoch": 1.616418626980317,
+      "grad_norm": 0.9111982814820835,
+      "learning_rate": 8.559730335453122e-06,
+      "loss": 0.5832,
+      "step": 1684
+    },
+    {
+      "epoch": 1.6173787806048967,
+      "grad_norm": 1.1180812700141665,
+      "learning_rate": 8.557377080238778e-06,
+      "loss": 0.5716,
+      "step": 1685
+    },
+    {
+      "epoch": 1.6183389342294767,
+      "grad_norm": 1.0120829591301022,
+      "learning_rate": 8.55502222815589e-06,
+      "loss": 0.4687,
+      "step": 1686
+    },
+    {
+      "epoch": 1.6192990878540567,
+      "grad_norm": 0.9703184027207251,
+      "learning_rate": 8.552665780261526e-06,
+      "loss": 0.6809,
+      "step": 1687
+    },
+    {
+      "epoch": 1.6202592414786365,
+      "grad_norm": 1.1399371500972935,
+      "learning_rate": 8.550307737613468e-06,
+      "loss": 0.65,
+      "step": 1688
+    },
+    {
+      "epoch": 1.6212193951032166,
+      "grad_norm": 1.2564973647208353,
+      "learning_rate": 8.547948101270215e-06,
+      "loss": 0.6402,
+      "step": 1689
+    },
+    {
+      "epoch": 1.6221795487277966,
+      "grad_norm": 0.810702579081814,
+      "learning_rate": 8.545586872290987e-06,
+      "loss": 0.5705,
+      "step": 1690
+    },
+    {
+      "epoch": 1.6231397023523764,
+      "grad_norm": 1.3279957089503416,
+      "learning_rate": 8.543224051735714e-06,
+      "loss": 0.7017,
+      "step": 1691
+    },
+    {
+      "epoch": 1.6240998559769562,
+      "grad_norm": 1.1568430995470347,
+      "learning_rate": 8.540859640665036e-06,
+      "loss": 0.7889,
+      "step": 1692
+    },
+    {
+      "epoch": 1.6250600096015364,
+      "grad_norm": 1.1895042285769468,
+      "learning_rate": 8.538493640140317e-06,
+      "loss": 0.649,
+      "step": 1693
+    },
+    {
+      "epoch": 1.6260201632261162,
+      "grad_norm": 0.9908405450035387,
+      "learning_rate": 8.536126051223627e-06,
+      "loss": 0.5093,
+      "step": 1694
+    },
+    {
+      "epoch": 1.626980316850696,
+      "grad_norm": 1.1436806207978791,
+      "learning_rate": 8.533756874977751e-06,
+      "loss": 0.6795,
+      "step": 1695
+    },
+    {
+      "epoch": 1.627940470475276,
+      "grad_norm": 1.0001606823595681,
+      "learning_rate": 8.53138611246619e-06,
+      "loss": 0.7585,
+      "step": 1696
+    },
+    {
+      "epoch": 1.628900624099856,
+      "grad_norm": 1.1612341599585667,
+      "learning_rate": 8.529013764753147e-06,
+      "loss": 0.6908,
+      "step": 1697
+    },
+    {
+      "epoch": 1.6298607777244358,
+      "grad_norm": 1.0257237359423448,
+      "learning_rate": 8.526639832903552e-06,
+      "loss": 0.6067,
+      "step": 1698
+    },
+    {
+      "epoch": 1.6308209313490158,
+      "grad_norm": 0.943695610556328,
+      "learning_rate": 8.524264317983032e-06,
+      "loss": 0.5724,
+      "step": 1699
+    },
+    {
+      "epoch": 1.6317810849735959,
+      "grad_norm": 1.1001793125055708,
+      "learning_rate": 8.521887221057932e-06,
+      "loss": 0.5564,
+      "step": 1700
+    },
+    {
+      "epoch": 1.6327412385981757,
+      "grad_norm": 0.7967117986230224,
+      "learning_rate": 8.519508543195304e-06,
+      "loss": 0.4922,
+      "step": 1701
+    },
+    {
+      "epoch": 1.6337013922227557,
+      "grad_norm": 0.9404742155404143,
+      "learning_rate": 8.517128285462914e-06,
+      "loss": 0.6039,
+      "step": 1702
+    },
+    {
+      "epoch": 1.6346615458473357,
+      "grad_norm": 1.1292581101586436,
+      "learning_rate": 8.514746448929235e-06,
+      "loss": 0.6944,
+      "step": 1703
+    },
+    {
+      "epoch": 1.6356216994719155,
+      "grad_norm": 1.4004869256437205,
+      "learning_rate": 8.512363034663441e-06,
+      "loss": 0.7079,
+      "step": 1704
+    },
+    {
+      "epoch": 1.6365818530964953,
+      "grad_norm": 1.3431801226281752,
+      "learning_rate": 8.50997804373543e-06,
+      "loss": 0.7172,
+      "step": 1705
+    },
+    {
+      "epoch": 1.6375420067210755,
+      "grad_norm": 1.3314862190720576,
+      "learning_rate": 8.507591477215793e-06,
+      "loss": 0.6085,
+      "step": 1706
+    },
+    {
+      "epoch": 1.6385021603456553,
+      "grad_norm": 1.209629508540758,
+      "learning_rate": 8.505203336175836e-06,
+      "loss": 0.7252,
+      "step": 1707
+    },
+    {
+      "epoch": 1.6394623139702351,
+      "grad_norm": 1.274399903410045,
+      "learning_rate": 8.502813621687568e-06,
+      "loss": 0.4603,
+      "step": 1708
+    },
+    {
+      "epoch": 1.6404224675948151,
+      "grad_norm": 1.408366215341856,
+      "learning_rate": 8.50042233482371e-06,
+      "loss": 0.6636,
+      "step": 1709
+    },
+    {
+      "epoch": 1.6413826212193952,
+      "grad_norm": 0.8719579855297336,
+      "learning_rate": 8.498029476657686e-06,
+      "loss": 0.5953,
+      "step": 1710
+    },
+    {
+      "epoch": 1.642342774843975,
+      "grad_norm": 1.049429410346072,
+      "learning_rate": 8.495635048263617e-06,
+      "loss": 0.5743,
+      "step": 1711
+    },
+    {
+      "epoch": 1.643302928468555,
+      "grad_norm": 1.0268285567030226,
+      "learning_rate": 8.493239050716344e-06,
+      "loss": 0.6271,
+      "step": 1712
+    },
+    {
+      "epoch": 1.644263082093135,
+      "grad_norm": 1.3780276128217048,
+      "learning_rate": 8.4908414850914e-06,
+      "loss": 0.5648,
+      "step": 1713
+    },
+    {
+      "epoch": 1.6452232357177148,
+      "grad_norm": 1.1025995482917144,
+      "learning_rate": 8.488442352465029e-06,
+      "loss": 0.5872,
+      "step": 1714
+    },
+    {
+      "epoch": 1.6461833893422948,
+      "grad_norm": 1.2177032533446264,
+      "learning_rate": 8.486041653914177e-06,
+      "loss": 0.6434,
+      "step": 1715
+    },
+    {
+      "epoch": 1.6471435429668748,
+      "grad_norm": 1.318212060341539,
+      "learning_rate": 8.483639390516488e-06,
+      "loss": 0.7588,
+      "step": 1716
+    },
+    {
+      "epoch": 1.6481036965914546,
+      "grad_norm": 1.4281094927704476,
+      "learning_rate": 8.481235563350316e-06,
+      "loss": 0.6115,
+      "step": 1717
+    },
+    {
+      "epoch": 1.6490638502160344,
+      "grad_norm": 1.1792419894081392,
+      "learning_rate": 8.478830173494712e-06,
+      "loss": 0.6358,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6500240038406147,
+      "grad_norm": 0.9943172186490088,
+      "learning_rate": 8.47642322202943e-06,
+      "loss": 0.5389,
+      "step": 1719
+    },
+    {
+      "epoch": 1.6509841574651944,
+      "grad_norm": 1.0148518234658925,
+      "learning_rate": 8.474014710034923e-06,
+      "loss": 0.7093,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6519443110897742,
+      "grad_norm": 1.1948035124293634,
+      "learning_rate": 8.471604638592348e-06,
+      "loss": 0.7758,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6529044647143543,
+      "grad_norm": 0.9681052074623742,
+      "learning_rate": 8.469193008783562e-06,
+      "loss": 0.6116,
+      "step": 1722
+    },
+    {
+      "epoch": 1.6538646183389343,
+      "grad_norm": 1.3128560663557998,
+      "learning_rate": 8.466779821691117e-06,
+      "loss": 0.6443,
+      "step": 1723
+    },
+    {
+      "epoch": 1.654824771963514,
+      "grad_norm": 0.8640570945968897,
+      "learning_rate": 8.464365078398269e-06,
+      "loss": 0.6419,
+      "step": 1724
+    },
+    {
+      "epoch": 1.655784925588094,
+      "grad_norm": 1.4007162676256382,
+      "learning_rate": 8.461948779988967e-06,
+      "loss": 0.6014,
+      "step": 1725
+    },
+    {
+      "epoch": 1.656745079212674,
+      "grad_norm": 1.294619488237341,
+      "learning_rate": 8.459530927547864e-06,
+      "loss": 0.6103,
+      "step": 1726
+    },
+    {
+      "epoch": 1.657705232837254,
+      "grad_norm": 0.9582533270010303,
+      "learning_rate": 8.457111522160309e-06,
+      "loss": 0.679,
+      "step": 1727
+    },
+    {
+      "epoch": 1.658665386461834,
+      "grad_norm": 1.2081739113840666,
+      "learning_rate": 8.454690564912347e-06,
+      "loss": 0.6485,
+      "step": 1728
+    },
+    {
+      "epoch": 1.659625540086414,
+      "grad_norm": 0.9562788530236775,
+      "learning_rate": 8.452268056890717e-06,
+      "loss": 0.7311,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6605856937109937,
+      "grad_norm": 1.0530110294051391,
+      "learning_rate": 8.44984399918286e-06,
+      "loss": 0.6905,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6615458473355735,
+      "grad_norm": 0.8980922663700222,
+      "learning_rate": 8.447418392876907e-06,
+      "loss": 0.7196,
+      "step": 1731
+    },
+    {
+      "epoch": 1.6625060009601538,
+      "grad_norm": 1.2779398677085188,
+      "learning_rate": 8.44499123906169e-06,
+      "loss": 0.699,
+      "step": 1732
+    },
+    {
+      "epoch": 1.6634661545847336,
+      "grad_norm": 1.3388477077082197,
+      "learning_rate": 8.44256253882673e-06,
+      "loss": 0.5993,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6644263082093134,
+      "grad_norm": 0.9604175271967059,
+      "learning_rate": 8.440132293262246e-06,
+      "loss": 0.6443,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6653864618338934,
+      "grad_norm": 1.059165325720594,
+      "learning_rate": 8.437700503459149e-06,
+      "loss": 0.6622,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6663466154584734,
+      "grad_norm": 1.2847355817366322,
+      "learning_rate": 8.435267170509044e-06,
+      "loss": 0.6376,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6673067690830532,
+      "grad_norm": 0.9486013461835123,
+      "learning_rate": 8.432832295504224e-06,
+      "loss": 0.6135,
+      "step": 1737
+    },
+    {
+      "epoch": 1.6682669227076332,
+      "grad_norm": 1.1217559786960296,
+      "learning_rate": 8.430395879537684e-06,
+      "loss": 0.7307,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6692270763322132,
+      "grad_norm": 1.166539193180166,
+      "learning_rate": 8.4279579237031e-06,
+      "loss": 0.6745,
+      "step": 1739
+    },
+    {
+      "epoch": 1.670187229956793,
+      "grad_norm": 1.2944973539291245,
+      "learning_rate": 8.425518429094848e-06,
+      "loss": 0.6738,
+      "step": 1740
+    },
+    {
+      "epoch": 1.671147383581373,
+      "grad_norm": 1.7769472382713831,
+      "learning_rate": 8.423077396807991e-06,
+      "loss": 0.72,
+      "step": 1741
+    },
+    {
+      "epoch": 1.672107537205953,
+      "grad_norm": 0.9923753385483782,
+      "learning_rate": 8.42063482793828e-06,
+      "loss": 0.6447,
+      "step": 1742
+    },
+    {
+      "epoch": 1.6730676908305329,
+      "grad_norm": 1.5206413904288867,
+      "learning_rate": 8.41819072358216e-06,
+      "loss": 0.7335,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6740278444551127,
+      "grad_norm": 1.2603132686679719,
+      "learning_rate": 8.415745084836763e-06,
+      "loss": 0.7357,
+      "step": 1744
+    },
+    {
+      "epoch": 1.674987998079693,
+      "grad_norm": 1.3263524233400652,
+      "learning_rate": 8.413297912799909e-06,
+      "loss": 0.603,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6759481517042727,
+      "grad_norm": 1.034102796940537,
+      "learning_rate": 8.410849208570108e-06,
+      "loss": 0.5277,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6769083053288525,
+      "grad_norm": 0.9802629714566773,
+      "learning_rate": 8.408398973246557e-06,
+      "loss": 0.6366,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6778684589534325,
+      "grad_norm": 1.7367971970710725,
+      "learning_rate": 8.405947207929142e-06,
+      "loss": 0.6344,
+      "step": 1748
+    },
+    {
+      "epoch": 1.6788286125780125,
+      "grad_norm": 1.0468229139336518,
+      "learning_rate": 8.40349391371843e-06,
+      "loss": 0.6003,
+      "step": 1749
+    },
+    {
+      "epoch": 1.6797887662025923,
+      "grad_norm": 0.8799505709158508,
+      "learning_rate": 8.401039091715685e-06,
+      "loss": 0.7298,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6807489198271723,
+      "grad_norm": 1.0134659723179875,
+      "learning_rate": 8.398582743022845e-06,
+      "loss": 0.6389,
+      "step": 1751
+    },
+    {
+      "epoch": 1.6817090734517524,
+      "grad_norm": 0.9872492043726284,
+      "learning_rate": 8.39612486874254e-06,
+      "loss": 0.5881,
+      "step": 1752
+    },
+    {
+      "epoch": 1.6826692270763322,
+      "grad_norm": 1.1250605266987366,
+      "learning_rate": 8.393665469978086e-06,
+      "loss": 0.6378,
+      "step": 1753
+    },
+    {
+      "epoch": 1.6836293807009122,
+      "grad_norm": 1.0626430971346574,
+      "learning_rate": 8.391204547833478e-06,
+      "loss": 0.7311,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6845895343254922,
+      "grad_norm": 1.1428411752670604,
+      "learning_rate": 8.388742103413397e-06,
+      "loss": 0.613,
+      "step": 1755
+    },
+    {
+      "epoch": 1.685549687950072,
+      "grad_norm": 1.064563945141614,
+      "learning_rate": 8.38627813782321e-06,
+      "loss": 0.6215,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6865098415746518,
+      "grad_norm": 1.3895186464336546,
+      "learning_rate": 8.383812652168966e-06,
+      "loss": 0.6832,
+      "step": 1757
+    },
+    {
+      "epoch": 1.687469995199232,
+      "grad_norm": 1.0500015031401468,
+      "learning_rate": 8.38134564755739e-06,
+      "loss": 0.6395,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6884301488238118,
+      "grad_norm": 1.0701915658229737,
+      "learning_rate": 8.378877125095901e-06,
+      "loss": 0.6869,
+      "step": 1759
+    },
+    {
+      "epoch": 1.6893903024483916,
+      "grad_norm": 0.9883890260744455,
+      "learning_rate": 8.376407085892586e-06,
+      "loss": 0.627,
+      "step": 1760
+    },
+    {
+      "epoch": 1.6903504560729716,
+      "grad_norm": 1.204954413250014,
+      "learning_rate": 8.373935531056222e-06,
+      "loss": 0.6609,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6913106096975516,
+      "grad_norm": 0.9014119447216241,
+      "learning_rate": 8.371462461696267e-06,
+      "loss": 0.6924,
+      "step": 1762
+    },
+    {
+      "epoch": 1.6922707633221314,
+      "grad_norm": 0.9454956924843029,
+      "learning_rate": 8.36898787892285e-06,
+      "loss": 0.4615,
+      "step": 1763
+    },
+    {
+      "epoch": 1.6932309169467115,
+      "grad_norm": 1.016954677498344,
+      "learning_rate": 8.366511783846785e-06,
+      "loss": 0.5375,
+      "step": 1764
+    },
+    {
+      "epoch": 1.6941910705712915,
+      "grad_norm": 1.4344533155048056,
+      "learning_rate": 8.364034177579568e-06,
+      "loss": 0.7729,
+      "step": 1765
+    },
+    {
+      "epoch": 1.6951512241958713,
+      "grad_norm": 0.9825308632016129,
+      "learning_rate": 8.361555061233367e-06,
+      "loss": 0.5923,
+      "step": 1766
+    },
+    {
+      "epoch": 1.6961113778204513,
+      "grad_norm": 1.044828001221576,
+      "learning_rate": 8.359074435921032e-06,
+      "loss": 0.581,
+      "step": 1767
+    },
+    {
+      "epoch": 1.6970715314450313,
+      "grad_norm": 1.162410608452558,
+      "learning_rate": 8.35659230275609e-06,
+      "loss": 0.5913,
+      "step": 1768
+    },
+    {
+      "epoch": 1.698031685069611,
+      "grad_norm": 1.235270867502264,
+      "learning_rate": 8.354108662852743e-06,
+      "loss": 0.6416,
+      "step": 1769
+    },
+    {
+      "epoch": 1.6989918386941911,
+      "grad_norm": 1.166953971280625,
+      "learning_rate": 8.351623517325872e-06,
+      "loss": 0.5994,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6999519923187711,
+      "grad_norm": 1.0380648883740493,
+      "learning_rate": 8.349136867291029e-06,
+      "loss": 0.6621,
+      "step": 1771
+    },
+    {
+      "epoch": 1.700912145943351,
+      "grad_norm": 1.1119353863991024,
+      "learning_rate": 8.346648713864447e-06,
+      "loss": 0.6469,
+      "step": 1772
+    },
+    {
+      "epoch": 1.7018722995679307,
+      "grad_norm": 0.896899098530902,
+      "learning_rate": 8.344159058163032e-06,
+      "loss": 0.6084,
+      "step": 1773
+    },
+    {
+      "epoch": 1.7028324531925108,
+      "grad_norm": 0.8633910130054333,
+      "learning_rate": 8.341667901304362e-06,
+      "loss": 0.5014,
+      "step": 1774
+    },
+    {
+      "epoch": 1.7037926068170908,
+      "grad_norm": 0.8323822273701,
+      "learning_rate": 8.339175244406693e-06,
+      "loss": 0.5515,
+      "step": 1775
+    },
+    {
+      "epoch": 1.7047527604416706,
+      "grad_norm": 1.0272487976042384,
+      "learning_rate": 8.336681088588949e-06,
+      "loss": 0.6712,
+      "step": 1776
+    },
+    {
+      "epoch": 1.7057129140662506,
+      "grad_norm": 1.3287844681034346,
+      "learning_rate": 8.33418543497073e-06,
+      "loss": 0.7486,
+      "step": 1777
+    },
+    {
+      "epoch": 1.7066730676908306,
+      "grad_norm": 1.071009262146182,
+      "learning_rate": 8.331688284672313e-06,
+      "loss": 0.5426,
+      "step": 1778
+    },
+    {
+      "epoch": 1.7076332213154104,
+      "grad_norm": 0.9979771404488532,
+      "learning_rate": 8.329189638814637e-06,
+      "loss": 0.5924,
+      "step": 1779
+    },
+    {
+      "epoch": 1.7085933749399904,
+      "grad_norm": 0.9845955447724891,
+      "learning_rate": 8.32668949851932e-06,
+      "loss": 0.5748,
+      "step": 1780
+    },
+    {
+      "epoch": 1.7095535285645704,
+      "grad_norm": 1.038187251919438,
+      "learning_rate": 8.324187864908646e-06,
+      "loss": 0.5973,
+      "step": 1781
+    },
+    {
+      "epoch": 1.7105136821891502,
+      "grad_norm": 0.9283230295023462,
+      "learning_rate": 8.321684739105573e-06,
+      "loss": 0.5842,
+      "step": 1782
+    },
+    {
+      "epoch": 1.7114738358137302,
+      "grad_norm": 0.9610776326476227,
+      "learning_rate": 8.319180122233729e-06,
+      "loss": 0.5744,
+      "step": 1783
+    },
+    {
+      "epoch": 1.7124339894383103,
+      "grad_norm": 1.1848026680157284,
+      "learning_rate": 8.316674015417407e-06,
+      "loss": 0.5415,
+      "step": 1784
+    },
+    {
+      "epoch": 1.71339414306289,
+      "grad_norm": 0.8383489208310458,
+      "learning_rate": 8.314166419781574e-06,
+      "loss": 0.5809,
+      "step": 1785
+    },
+    {
+      "epoch": 1.7143542966874699,
+      "grad_norm": 0.8902371551760121,
+      "learning_rate": 8.311657336451862e-06,
+      "loss": 0.6163,
+      "step": 1786
+    },
+    {
+      "epoch": 1.7153144503120499,
+      "grad_norm": 1.0486580496373912,
+      "learning_rate": 8.30914676655457e-06,
+      "loss": 0.6104,
+      "step": 1787
+    },
+    {
+      "epoch": 1.71627460393663,
+      "grad_norm": 1.171465563574018,
+      "learning_rate": 8.306634711216668e-06,
+      "loss": 0.702,
+      "step": 1788
+    },
+    {
+      "epoch": 1.7172347575612097,
+      "grad_norm": 1.1521729735887885,
+      "learning_rate": 8.304121171565788e-06,
+      "loss": 0.572,
+      "step": 1789
+    },
+    {
+      "epoch": 1.7181949111857897,
+      "grad_norm": 0.8221298578072059,
+      "learning_rate": 8.301606148730237e-06,
+      "loss": 0.5707,
+      "step": 1790
+    },
+    {
+      "epoch": 1.7191550648103697,
+      "grad_norm": 1.1029215524544493,
+      "learning_rate": 8.299089643838976e-06,
+      "loss": 0.6609,
+      "step": 1791
+    },
+    {
+      "epoch": 1.7201152184349495,
+      "grad_norm": 1.4561379954869085,
+      "learning_rate": 8.296571658021638e-06,
+      "loss": 0.6613,
+      "step": 1792
+    },
+    {
+      "epoch": 1.7210753720595295,
+      "grad_norm": 1.2738669594132566,
+      "learning_rate": 8.294052192408522e-06,
+      "loss": 0.792,
+      "step": 1793
+    },
+    {
+      "epoch": 1.7220355256841096,
+      "grad_norm": 1.4798740905642442,
+      "learning_rate": 8.291531248130589e-06,
+      "loss": 0.6698,
+      "step": 1794
+    },
+    {
+      "epoch": 1.7229956793086894,
+      "grad_norm": 1.134409326137547,
+      "learning_rate": 8.289008826319463e-06,
+      "loss": 0.6577,
+      "step": 1795
+    },
+    {
+      "epoch": 1.7239558329332694,
+      "grad_norm": 0.9551706414818583,
+      "learning_rate": 8.286484928107431e-06,
+      "loss": 0.6724,
+      "step": 1796
+    },
+    {
+      "epoch": 1.7249159865578494,
+      "grad_norm": 1.2374840247458698,
+      "learning_rate": 8.283959554627448e-06,
+      "loss": 0.7348,
+      "step": 1797
+    },
+    {
+      "epoch": 1.7258761401824292,
+      "grad_norm": 1.3416378545918735,
+      "learning_rate": 8.281432707013123e-06,
+      "loss": 0.7342,
+      "step": 1798
+    },
+    {
+      "epoch": 1.726836293807009,
+      "grad_norm": 1.227198660378129,
+      "learning_rate": 8.278904386398733e-06,
+      "loss": 0.697,
+      "step": 1799
+    },
+    {
+      "epoch": 1.727796447431589,
+      "grad_norm": 0.8602654088414831,
+      "learning_rate": 8.276374593919213e-06,
+      "loss": 0.5971,
+      "step": 1800
+    },
+    {
+      "epoch": 1.728756601056169,
+      "grad_norm": 0.8383919774027291,
+      "learning_rate": 8.27384333071016e-06,
+      "loss": 0.6112,
+      "step": 1801
+    },
+    {
+      "epoch": 1.7297167546807488,
+      "grad_norm": 1.0016028936180106,
+      "learning_rate": 8.27131059790783e-06,
+      "loss": 0.5832,
+      "step": 1802
+    },
+    {
+      "epoch": 1.7306769083053288,
+      "grad_norm": 1.1733238954383638,
+      "learning_rate": 8.268776396649145e-06,
+      "loss": 0.793,
+      "step": 1803
+    },
+    {
+      "epoch": 1.7316370619299088,
+      "grad_norm": 1.2035517212362503,
+      "learning_rate": 8.266240728071673e-06,
+      "loss": 0.6498,
+      "step": 1804
+    },
+    {
+      "epoch": 1.7325972155544886,
+      "grad_norm": 1.0897666129294943,
+      "learning_rate": 8.263703593313654e-06,
+      "loss": 0.6667,
+      "step": 1805
+    },
+    {
+      "epoch": 1.7335573691790687,
+      "grad_norm": 0.9073661107039916,
+      "learning_rate": 8.261164993513978e-06,
+      "loss": 0.6347,
+      "step": 1806
+    },
+    {
+      "epoch": 1.7345175228036487,
+      "grad_norm": 1.5648320656747918,
+      "learning_rate": 8.258624929812198e-06,
+      "loss": 0.7912,
+      "step": 1807
+    },
+    {
+      "epoch": 1.7354776764282285,
+      "grad_norm": 1.3532627183223036,
+      "learning_rate": 8.256083403348518e-06,
+      "loss": 0.6484,
+      "step": 1808
+    },
+    {
+      "epoch": 1.7364378300528085,
+      "grad_norm": 0.9287022732318587,
+      "learning_rate": 8.253540415263805e-06,
+      "loss": 0.5194,
+      "step": 1809
+    },
+    {
+      "epoch": 1.7373979836773885,
+      "grad_norm": 1.0796822687805636,
+      "learning_rate": 8.250995966699577e-06,
+      "loss": 0.5494,
+      "step": 1810
+    },
+    {
+      "epoch": 1.7383581373019683,
+      "grad_norm": 1.1107009166158903,
+      "learning_rate": 8.248450058798012e-06,
+      "loss": 0.5783,
+      "step": 1811
+    },
+    {
+      "epoch": 1.739318290926548,
+      "grad_norm": 0.9604962537986486,
+      "learning_rate": 8.245902692701939e-06,
+      "loss": 0.5204,
+      "step": 1812
+    },
+    {
+      "epoch": 1.7402784445511283,
+      "grad_norm": 0.8985142428195022,
+      "learning_rate": 8.243353869554845e-06,
+      "loss": 0.6438,
+      "step": 1813
+    },
+    {
+      "epoch": 1.7412385981757081,
+      "grad_norm": 0.900420965577512,
+      "learning_rate": 8.240803590500865e-06,
+      "loss": 0.6015,
+      "step": 1814
+    },
+    {
+      "epoch": 1.742198751800288,
+      "grad_norm": 1.229040739833296,
+      "learning_rate": 8.238251856684801e-06,
+      "loss": 0.6532,
+      "step": 1815
+    },
+    {
+      "epoch": 1.743158905424868,
+      "grad_norm": 0.9596600579734443,
+      "learning_rate": 8.23569866925209e-06,
+      "loss": 0.6131,
+      "step": 1816
+    },
+    {
+      "epoch": 1.744119059049448,
+      "grad_norm": 0.9098868199427298,
+      "learning_rate": 8.233144029348834e-06,
+      "loss": 0.5849,
+      "step": 1817
+    },
+    {
+      "epoch": 1.7450792126740278,
+      "grad_norm": 0.7450651835506069,
+      "learning_rate": 8.230587938121783e-06,
+      "loss": 0.6278,
+      "step": 1818
+    },
+    {
+      "epoch": 1.7460393662986078,
+      "grad_norm": 1.3584641927417642,
+      "learning_rate": 8.228030396718342e-06,
+      "loss": 0.8177,
+      "step": 1819
+    },
+    {
+      "epoch": 1.7469995199231878,
+      "grad_norm": 1.1138215385227221,
+      "learning_rate": 8.22547140628656e-06,
+      "loss": 0.6034,
+      "step": 1820
+    },
+    {
+      "epoch": 1.7479596735477676,
+      "grad_norm": 1.0849675016586429,
+      "learning_rate": 8.222910967975143e-06,
+      "loss": 0.6782,
+      "step": 1821
+    },
+    {
+      "epoch": 1.7489198271723476,
+      "grad_norm": 1.0631193133434942,
+      "learning_rate": 8.220349082933444e-06,
+      "loss": 0.6223,
+      "step": 1822
+    },
+    {
+      "epoch": 1.7498799807969276,
+      "grad_norm": 1.0135727168930355,
+      "learning_rate": 8.217785752311464e-06,
+      "loss": 0.5909,
+      "step": 1823
+    },
+    {
+      "epoch": 1.7508401344215074,
+      "grad_norm": 0.7561661716159543,
+      "learning_rate": 8.215220977259855e-06,
+      "loss": 0.6372,
+      "step": 1824
+    },
+    {
+      "epoch": 1.7518002880460872,
+      "grad_norm": 0.9250441764475402,
+      "learning_rate": 8.21265475892992e-06,
+      "loss": 0.6655,
+      "step": 1825
+    },
+    {
+      "epoch": 1.7527604416706675,
+      "grad_norm": 1.2282270838391987,
+      "learning_rate": 8.210087098473605e-06,
+      "loss": 0.653,
+      "step": 1826
+    },
+    {
+      "epoch": 1.7537205952952473,
+      "grad_norm": 1.1097200783241623,
+      "learning_rate": 8.207517997043504e-06,
+      "loss": 0.689,
+      "step": 1827
+    },
+    {
+      "epoch": 1.754680748919827,
+      "grad_norm": 0.9590947969068119,
+      "learning_rate": 8.204947455792863e-06,
+      "loss": 0.6778,
+      "step": 1828
+    },
+    {
+      "epoch": 1.755640902544407,
+      "grad_norm": 1.0631355802999023,
+      "learning_rate": 8.202375475875566e-06,
+      "loss": 0.5967,
+      "step": 1829
+    },
+    {
+      "epoch": 1.756601056168987,
+      "grad_norm": 1.0972744735808178,
+      "learning_rate": 8.19980205844615e-06,
+      "loss": 0.6279,
+      "step": 1830
+    },
+    {
+      "epoch": 1.757561209793567,
+      "grad_norm": 0.9125316595551967,
+      "learning_rate": 8.197227204659795e-06,
+      "loss": 0.6852,
+      "step": 1831
+    },
+    {
+      "epoch": 1.758521363418147,
+      "grad_norm": 1.0202579993835212,
+      "learning_rate": 8.194650915672323e-06,
+      "loss": 0.5599,
+      "step": 1832
+    },
+    {
+      "epoch": 1.759481517042727,
+      "grad_norm": 0.9641028651047512,
+      "learning_rate": 8.192073192640204e-06,
+      "loss": 0.6607,
+      "step": 1833
+    },
+    {
+      "epoch": 1.7604416706673067,
+      "grad_norm": 1.0321264396840382,
+      "learning_rate": 8.18949403672055e-06,
+      "loss": 0.685,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7614018242918867,
+      "grad_norm": 1.1142882750756766,
+      "learning_rate": 8.186913449071117e-06,
+      "loss": 0.6774,
+      "step": 1835
+    },
+    {
+      "epoch": 1.7623619779164668,
+      "grad_norm": 1.2196631969025054,
+      "learning_rate": 8.184331430850304e-06,
+      "loss": 0.5104,
+      "step": 1836
+    },
+    {
+      "epoch": 1.7633221315410466,
+      "grad_norm": 1.021990399884995,
+      "learning_rate": 8.181747983217148e-06,
+      "loss": 0.591,
+      "step": 1837
+    },
+    {
+      "epoch": 1.7642822851656264,
+      "grad_norm": 0.9859026254377197,
+      "learning_rate": 8.179163107331336e-06,
+      "loss": 0.5823,
+      "step": 1838
+    },
+    {
+      "epoch": 1.7652424387902066,
+      "grad_norm": 0.8971928754930129,
+      "learning_rate": 8.176576804353186e-06,
+      "loss": 0.5824,
+      "step": 1839
+    },
+    {
+      "epoch": 1.7662025924147864,
+      "grad_norm": 0.8651992267585531,
+      "learning_rate": 8.173989075443666e-06,
+      "loss": 0.575,
+      "step": 1840
+    },
+    {
+      "epoch": 1.7671627460393662,
+      "grad_norm": 1.12415934868087,
+      "learning_rate": 8.171399921764377e-06,
+      "loss": 0.502,
+      "step": 1841
+    },
+    {
+      "epoch": 1.7681228996639462,
+      "grad_norm": 1.2742387716529824,
+      "learning_rate": 8.168809344477564e-06,
+      "loss": 0.7826,
+      "step": 1842
+    },
+    {
+      "epoch": 1.7690830532885262,
+      "grad_norm": 1.4484109927599673,
+      "learning_rate": 8.16621734474611e-06,
+      "loss": 0.5709,
+      "step": 1843
+    },
+    {
+      "epoch": 1.770043206913106,
+      "grad_norm": 1.2696285065207322,
+      "learning_rate": 8.163623923733533e-06,
+      "loss": 0.6374,
+      "step": 1844
+    },
+    {
+      "epoch": 1.771003360537686,
+      "grad_norm": 1.2535412969858437,
+      "learning_rate": 8.161029082603994e-06,
+      "loss": 0.6868,
+      "step": 1845
+    },
+    {
+      "epoch": 1.771963514162266,
+      "grad_norm": 1.3855775261414205,
+      "learning_rate": 8.158432822522291e-06,
+      "loss": 0.6984,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7729236677868458,
+      "grad_norm": 0.8498398072770822,
+      "learning_rate": 8.155835144653856e-06,
+      "loss": 0.626,
+      "step": 1847
+    },
+    {
+      "epoch": 1.7738838214114259,
+      "grad_norm": 1.219824573506507,
+      "learning_rate": 8.153236050164757e-06,
+      "loss": 0.5783,
+      "step": 1848
+    },
+    {
+      "epoch": 1.7748439750360059,
+      "grad_norm": 0.9608903408185807,
+      "learning_rate": 8.1506355402217e-06,
+      "loss": 0.7243,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7758041286605857,
+      "grad_norm": 1.0144995350899024,
+      "learning_rate": 8.14803361599203e-06,
+      "loss": 0.6871,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7767642822851655,
+      "grad_norm": 1.3313755262778169,
+      "learning_rate": 8.14543027864372e-06,
+      "loss": 0.586,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7777244359097457,
+      "grad_norm": 0.9957132992143335,
+      "learning_rate": 8.14282552934538e-06,
+      "loss": 0.6559,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7786845895343255,
+      "grad_norm": 1.2923986777784746,
+      "learning_rate": 8.140219369266255e-06,
+      "loss": 0.6073,
+      "step": 1853
+    },
+    {
+      "epoch": 1.7796447431589053,
+      "grad_norm": 0.9127845535992563,
+      "learning_rate": 8.137611799576222e-06,
+      "loss": 0.5435,
+      "step": 1854
+    },
+    {
+      "epoch": 1.7806048967834853,
+      "grad_norm": 0.929215519990613,
+      "learning_rate": 8.135002821445793e-06,
+      "loss": 0.6305,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7815650504080653,
+      "grad_norm": 1.0143603358268427,
+      "learning_rate": 8.132392436046107e-06,
+      "loss": 0.6905,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7825252040326451,
+      "grad_norm": 0.9559284383420343,
+      "learning_rate": 8.129780644548938e-06,
+      "loss": 0.6128,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7834853576572252,
+      "grad_norm": 0.9519699589571047,
+      "learning_rate": 8.127167448126698e-06,
+      "loss": 0.635,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7844455112818052,
+      "grad_norm": 1.2331502386379996,
+      "learning_rate": 8.124552847952416e-06,
+      "loss": 0.5966,
+      "step": 1859
+    },
+    {
+      "epoch": 1.785405664906385,
+      "grad_norm": 1.1994027600373456,
+      "learning_rate": 8.121936845199759e-06,
+      "loss": 0.7028,
+      "step": 1860
+    },
+    {
+      "epoch": 1.786365818530965,
+      "grad_norm": 1.4147779911257676,
+      "learning_rate": 8.119319441043024e-06,
+      "loss": 0.6477,
+      "step": 1861
+    },
+    {
+      "epoch": 1.787325972155545,
+      "grad_norm": 0.8490520314350598,
+      "learning_rate": 8.116700636657137e-06,
+      "loss": 0.5516,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7882861257801248,
+      "grad_norm": 1.1696054113032124,
+      "learning_rate": 8.11408043321765e-06,
+      "loss": 0.7455,
+      "step": 1863
+    },
+    {
+      "epoch": 1.7892462794047046,
+      "grad_norm": 1.225511448354425,
+      "learning_rate": 8.111458831900743e-06,
+      "loss": 0.5417,
+      "step": 1864
+    },
+    {
+      "epoch": 1.7902064330292848,
+      "grad_norm": 0.8514978364567153,
+      "learning_rate": 8.108835833883227e-06,
+      "loss": 0.6946,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7911665866538646,
+      "grad_norm": 1.0733158336064204,
+      "learning_rate": 8.106211440342535e-06,
+      "loss": 0.6104,
+      "step": 1866
+    },
+    {
+      "epoch": 1.7921267402784444,
+      "grad_norm": 0.859932043516501,
+      "learning_rate": 8.103585652456732e-06,
+      "loss": 0.6925,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7930868939030244,
+      "grad_norm": 0.7924816855317768,
+      "learning_rate": 8.100958471404506e-06,
+      "loss": 0.4791,
+      "step": 1868
+    },
+    {
+      "epoch": 1.7940470475276045,
+      "grad_norm": 0.9026770361495339,
+      "learning_rate": 8.098329898365168e-06,
+      "loss": 0.5668,
+      "step": 1869
+    },
+    {
+      "epoch": 1.7950072011521843,
+      "grad_norm": 0.7783217855735335,
+      "learning_rate": 8.095699934518659e-06,
+      "loss": 0.6691,
+      "step": 1870
+    },
+    {
+      "epoch": 1.7959673547767643,
+      "grad_norm": 1.3245388261385187,
+      "learning_rate": 8.093068581045538e-06,
+      "loss": 0.6839,
+      "step": 1871
+    },
+    {
+      "epoch": 1.7969275084013443,
+      "grad_norm": 0.9264206387795253,
+      "learning_rate": 8.090435839126998e-06,
+      "loss": 0.6947,
+      "step": 1872
+    },
+    {
+      "epoch": 1.797887662025924,
+      "grad_norm": 1.015192823602121,
+      "learning_rate": 8.087801709944842e-06,
+      "loss": 0.63,
+      "step": 1873
+    },
+    {
+      "epoch": 1.7988478156505041,
+      "grad_norm": 1.2166516771098983,
+      "learning_rate": 8.085166194681508e-06,
+      "loss": 0.6834,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7998079692750841,
+      "grad_norm": 1.4029832180085575,
+      "learning_rate": 8.082529294520046e-06,
+      "loss": 0.7395,
+      "step": 1875
+    },
+    {
+      "epoch": 1.800768122899664,
+      "grad_norm": 1.0670024630500639,
+      "learning_rate": 8.079891010644134e-06,
+      "loss": 0.7066,
+      "step": 1876
+    },
+    {
+      "epoch": 1.8017282765242437,
+      "grad_norm": 0.9591006276036653,
+      "learning_rate": 8.077251344238068e-06,
+      "loss": 0.5438,
+      "step": 1877
+    },
+    {
+      "epoch": 1.802688430148824,
+      "grad_norm": 0.807560347256751,
+      "learning_rate": 8.074610296486771e-06,
+      "loss": 0.5772,
+      "step": 1878
+    },
+    {
+      "epoch": 1.8036485837734038,
+      "grad_norm": 1.1301704136208068,
+      "learning_rate": 8.071967868575775e-06,
+      "loss": 0.5224,
+      "step": 1879
+    },
+    {
+      "epoch": 1.8046087373979836,
+      "grad_norm": 0.8090738491171423,
+      "learning_rate": 8.069324061691242e-06,
+      "loss": 0.5858,
+      "step": 1880
+    },
+    {
+      "epoch": 1.8055688910225636,
+      "grad_norm": 1.0707391627369216,
+      "learning_rate": 8.066678877019945e-06,
+      "loss": 0.7289,
+      "step": 1881
+    },
+    {
+      "epoch": 1.8065290446471436,
+      "grad_norm": 0.9828589858485237,
+      "learning_rate": 8.064032315749283e-06,
+      "loss": 0.6256,
+      "step": 1882
+    },
+    {
+      "epoch": 1.8074891982717234,
+      "grad_norm": 0.9962401337718808,
+      "learning_rate": 8.061384379067266e-06,
+      "loss": 0.6766,
+      "step": 1883
+    },
+    {
+      "epoch": 1.8084493518963034,
+      "grad_norm": 0.8138116443056042,
+      "learning_rate": 8.058735068162527e-06,
+      "loss": 0.6071,
+      "step": 1884
+    },
+    {
+      "epoch": 1.8094095055208834,
+      "grad_norm": 0.8179160915657997,
+      "learning_rate": 8.05608438422431e-06,
+      "loss": 0.7092,
+      "step": 1885
+    },
+    {
+      "epoch": 1.8103696591454632,
+      "grad_norm": 1.3163333231081986,
+      "learning_rate": 8.053432328442483e-06,
+      "loss": 0.6915,
+      "step": 1886
+    },
+    {
+      "epoch": 1.8113298127700432,
+      "grad_norm": 1.3579706080157559,
+      "learning_rate": 8.05077890200752e-06,
+      "loss": 0.6606,
+      "step": 1887
+    },
+    {
+      "epoch": 1.8122899663946233,
+      "grad_norm": 1.1776867295102547,
+      "learning_rate": 8.04812410611052e-06,
+      "loss": 0.6988,
+      "step": 1888
+    },
+    {
+      "epoch": 1.813250120019203,
+      "grad_norm": 1.2589174346369656,
+      "learning_rate": 8.045467941943193e-06,
+      "loss": 0.5862,
+      "step": 1889
+    },
+    {
+      "epoch": 1.8142102736437828,
+      "grad_norm": 1.2265270873184735,
+      "learning_rate": 8.042810410697861e-06,
+      "loss": 0.6179,
+      "step": 1890
+    },
+    {
+      "epoch": 1.815170427268363,
+      "grad_norm": 1.0133059812027188,
+      "learning_rate": 8.04015151356746e-06,
+      "loss": 0.6033,
+      "step": 1891
+    },
+    {
+      "epoch": 1.8161305808929429,
+      "grad_norm": 0.8791820381095383,
+      "learning_rate": 8.037491251745543e-06,
+      "loss": 0.7375,
+      "step": 1892
+    },
+    {
+      "epoch": 1.8170907345175227,
+      "grad_norm": 1.040871201272267,
+      "learning_rate": 8.034829626426273e-06,
+      "loss": 0.6059,
+      "step": 1893
+    },
+    {
+      "epoch": 1.8180508881421027,
+      "grad_norm": 1.2459478903863541,
+      "learning_rate": 8.032166638804422e-06,
+      "loss": 0.7627,
+      "step": 1894
+    },
+    {
+      "epoch": 1.8190110417666827,
+      "grad_norm": 1.1669011277696746,
+      "learning_rate": 8.02950229007538e-06,
+      "loss": 0.6767,
+      "step": 1895
+    },
+    {
+      "epoch": 1.8199711953912625,
+      "grad_norm": 1.116550200147148,
+      "learning_rate": 8.026836581435142e-06,
+      "loss": 0.6242,
+      "step": 1896
+    },
+    {
+      "epoch": 1.8209313490158425,
+      "grad_norm": 1.005814348522918,
+      "learning_rate": 8.02416951408032e-06,
+      "loss": 0.586,
+      "step": 1897
+    },
+    {
+      "epoch": 1.8218915026404225,
+      "grad_norm": 1.3295399752112438,
+      "learning_rate": 8.02150108920813e-06,
+      "loss": 0.6137,
+      "step": 1898
+    },
+    {
+      "epoch": 1.8228516562650023,
+      "grad_norm": 1.3126930311763862,
+      "learning_rate": 8.018831308016398e-06,
+      "loss": 0.7176,
+      "step": 1899
+    },
+    {
+      "epoch": 1.8238118098895824,
+      "grad_norm": 1.4757208629612801,
+      "learning_rate": 8.01616017170356e-06,
+      "loss": 0.6723,
+      "step": 1900
+    },
+    {
+      "epoch": 1.8247719635141624,
+      "grad_norm": 0.7801099357841741,
+      "learning_rate": 8.013487681468663e-06,
+      "loss": 0.5786,
+      "step": 1901
+    },
+    {
+      "epoch": 1.8257321171387422,
+      "grad_norm": 1.0134564029272144,
+      "learning_rate": 8.010813838511357e-06,
+      "loss": 0.6652,
+      "step": 1902
+    },
+    {
+      "epoch": 1.826692270763322,
+      "grad_norm": 0.8964154794897466,
+      "learning_rate": 8.008138644031906e-06,
+      "loss": 0.4801,
+      "step": 1903
+    },
+    {
+      "epoch": 1.8276524243879022,
+      "grad_norm": 0.9900836513820425,
+      "learning_rate": 8.005462099231171e-06,
+      "loss": 0.6187,
+      "step": 1904
+    },
+    {
+      "epoch": 1.828612578012482,
+      "grad_norm": 1.0219218374611196,
+      "learning_rate": 8.002784205310626e-06,
+      "loss": 0.4844,
+      "step": 1905
+    },
+    {
+      "epoch": 1.8295727316370618,
+      "grad_norm": 0.9937049664117276,
+      "learning_rate": 8.00010496347235e-06,
+      "loss": 0.81,
+      "step": 1906
+    },
+    {
+      "epoch": 1.8305328852616418,
+      "grad_norm": 1.3696912850171175,
+      "learning_rate": 7.997424374919024e-06,
+      "loss": 0.6047,
+      "step": 1907
+    },
+    {
+      "epoch": 1.8314930388862218,
+      "grad_norm": 0.8695888682231789,
+      "learning_rate": 7.994742440853937e-06,
+      "loss": 0.5507,
+      "step": 1908
+    },
+    {
+      "epoch": 1.8324531925108016,
+      "grad_norm": 0.8029912452351989,
+      "learning_rate": 7.992059162480983e-06,
+      "loss": 0.7144,
+      "step": 1909
+    },
+    {
+      "epoch": 1.8334133461353816,
+      "grad_norm": 1.024031315206989,
+      "learning_rate": 7.98937454100465e-06,
+      "loss": 0.7038,
+      "step": 1910
+    },
+    {
+      "epoch": 1.8343734997599617,
+      "grad_norm": 0.9379701292046752,
+      "learning_rate": 7.986688577630042e-06,
+      "loss": 0.6368,
+      "step": 1911
+    },
+    {
+      "epoch": 1.8353336533845415,
+      "grad_norm": 1.1626921746350636,
+      "learning_rate": 7.984001273562858e-06,
+      "loss": 0.4876,
+      "step": 1912
+    },
+    {
+      "epoch": 1.8362938070091215,
+      "grad_norm": 0.765093029375882,
+      "learning_rate": 7.981312630009398e-06,
+      "loss": 0.5535,
+      "step": 1913
+    },
+    {
+      "epoch": 1.8372539606337015,
+      "grad_norm": 1.493058303836218,
+      "learning_rate": 7.97862264817657e-06,
+      "loss": 0.6126,
+      "step": 1914
+    },
+    {
+      "epoch": 1.8382141142582813,
+      "grad_norm": 1.0474886899563316,
+      "learning_rate": 7.975931329271874e-06,
+      "loss": 0.627,
+      "step": 1915
+    },
+    {
+      "epoch": 1.839174267882861,
+      "grad_norm": 1.0060827978459788,
+      "learning_rate": 7.973238674503412e-06,
+      "loss": 0.5583,
+      "step": 1916
+    },
+    {
+      "epoch": 1.8401344215074413,
+      "grad_norm": 0.8485255878794656,
+      "learning_rate": 7.970544685079895e-06,
+      "loss": 0.6075,
+      "step": 1917
+    },
+    {
+      "epoch": 1.8410945751320211,
+      "grad_norm": 1.0948252496214814,
+      "learning_rate": 7.96784936221062e-06,
+      "loss": 0.5305,
+      "step": 1918
+    },
+    {
+      "epoch": 1.842054728756601,
+      "grad_norm": 1.1354876526540547,
+      "learning_rate": 7.965152707105491e-06,
+      "loss": 0.5774,
+      "step": 1919
+    },
+    {
+      "epoch": 1.843014882381181,
+      "grad_norm": 1.2082984729614092,
+      "learning_rate": 7.962454720975008e-06,
+      "loss": 0.694,
+      "step": 1920
+    },
+    {
+      "epoch": 1.843975036005761,
+      "grad_norm": 1.1873456218694367,
+      "learning_rate": 7.959755405030269e-06,
+      "loss": 0.6698,
+      "step": 1921
+    },
+    {
+      "epoch": 1.8449351896303408,
+      "grad_norm": 1.1345266713800688,
+      "learning_rate": 7.957054760482964e-06,
+      "loss": 0.6462,
+      "step": 1922
+    },
+    {
+      "epoch": 1.8458953432549208,
+      "grad_norm": 0.9554105867510718,
+      "learning_rate": 7.954352788545388e-06,
+      "loss": 0.5843,
+      "step": 1923
+    },
+    {
+      "epoch": 1.8468554968795008,
+      "grad_norm": 0.9468802876993183,
+      "learning_rate": 7.951649490430424e-06,
+      "loss": 0.6401,
+      "step": 1924
+    },
+    {
+      "epoch": 1.8478156505040806,
+      "grad_norm": 1.0926269329377654,
+      "learning_rate": 7.948944867351556e-06,
+      "loss": 0.6098,
+      "step": 1925
+    },
+    {
+      "epoch": 1.8487758041286606,
+      "grad_norm": 0.8654462228441753,
+      "learning_rate": 7.946238920522862e-06,
+      "loss": 0.7336,
+      "step": 1926
+    },
+    {
+      "epoch": 1.8497359577532406,
+      "grad_norm": 0.8394409652705445,
+      "learning_rate": 7.943531651159007e-06,
+      "loss": 0.6228,
+      "step": 1927
+    },
+    {
+      "epoch": 1.8506961113778204,
+      "grad_norm": 0.8727248717468625,
+      "learning_rate": 7.940823060475257e-06,
+      "loss": 0.6318,
+      "step": 1928
+    },
+    {
+      "epoch": 1.8516562650024004,
+      "grad_norm": 1.0237836414075832,
+      "learning_rate": 7.938113149687474e-06,
+      "loss": 0.6455,
+      "step": 1929
+    },
+    {
+      "epoch": 1.8526164186269805,
+      "grad_norm": 1.598106761224662,
+      "learning_rate": 7.935401920012105e-06,
+      "loss": 0.6924,
+      "step": 1930
+    },
+    {
+      "epoch": 1.8535765722515603,
+      "grad_norm": 1.154091511674192,
+      "learning_rate": 7.932689372666192e-06,
+      "loss": 0.6076,
+      "step": 1931
+    },
+    {
+      "epoch": 1.85453672587614,
+      "grad_norm": 0.9933962172027189,
+      "learning_rate": 7.929975508867364e-06,
+      "loss": 0.6653,
+      "step": 1932
+    },
+    {
+      "epoch": 1.85549687950072,
+      "grad_norm": 1.0289186292742585,
+      "learning_rate": 7.927260329833854e-06,
+      "loss": 0.6823,
+      "step": 1933
+    },
+    {
+      "epoch": 1.8564570331253,
+      "grad_norm": 0.9757375555786784,
+      "learning_rate": 7.924543836784471e-06,
+      "loss": 0.5639,
+      "step": 1934
+    },
+    {
+      "epoch": 1.8574171867498799,
+      "grad_norm": 0.9313246628917996,
+      "learning_rate": 7.921826030938623e-06,
+      "loss": 0.6496,
+      "step": 1935
+    },
+    {
+      "epoch": 1.85837734037446,
+      "grad_norm": 0.9904979867772578,
+      "learning_rate": 7.919106913516299e-06,
+      "loss": 0.6132,
+      "step": 1936
+    },
+    {
+      "epoch": 1.85933749399904,
+      "grad_norm": 1.3040673337594413,
+      "learning_rate": 7.916386485738089e-06,
+      "loss": 0.6787,
+      "step": 1937
+    },
+    {
+      "epoch": 1.8602976476236197,
+      "grad_norm": 1.0202838458263193,
+      "learning_rate": 7.913664748825156e-06,
+      "loss": 0.6105,
+      "step": 1938
+    },
+    {
+      "epoch": 1.8612578012481997,
+      "grad_norm": 1.137407655299372,
+      "learning_rate": 7.910941703999265e-06,
+      "loss": 0.6515,
+      "step": 1939
+    },
+    {
+      "epoch": 1.8622179548727797,
+      "grad_norm": 1.1799016075204445,
+      "learning_rate": 7.908217352482759e-06,
+      "loss": 0.7167,
+      "step": 1940
+    },
+    {
+      "epoch": 1.8631781084973595,
+      "grad_norm": 1.376572934679239,
+      "learning_rate": 7.905491695498571e-06,
+      "loss": 0.6602,
+      "step": 1941
+    },
+    {
+      "epoch": 1.8641382621219396,
+      "grad_norm": 0.9024551367490704,
+      "learning_rate": 7.902764734270219e-06,
+      "loss": 0.6101,
+      "step": 1942
+    },
+    {
+      "epoch": 1.8650984157465196,
+      "grad_norm": 1.2623165248007306,
+      "learning_rate": 7.900036470021805e-06,
+      "loss": 0.613,
+      "step": 1943
+    },
+    {
+      "epoch": 1.8660585693710994,
+      "grad_norm": 0.8908534019026104,
+      "learning_rate": 7.89730690397802e-06,
+      "loss": 0.7345,
+      "step": 1944
+    },
+    {
+      "epoch": 1.8670187229956792,
+      "grad_norm": 1.2219964995820927,
+      "learning_rate": 7.894576037364136e-06,
+      "loss": 0.6142,
+      "step": 1945
+    },
+    {
+      "epoch": 1.8679788766202592,
+      "grad_norm": 1.1701283331410564,
+      "learning_rate": 7.89184387140601e-06,
+      "loss": 0.7137,
+      "step": 1946
+    },
+    {
+      "epoch": 1.8689390302448392,
+      "grad_norm": 1.1780638161310644,
+      "learning_rate": 7.889110407330083e-06,
+      "loss": 0.6452,
+      "step": 1947
+    },
+    {
+      "epoch": 1.869899183869419,
+      "grad_norm": 0.8909626695365555,
+      "learning_rate": 7.886375646363378e-06,
+      "loss": 0.6545,
+      "step": 1948
+    },
+    {
+      "epoch": 1.870859337493999,
+      "grad_norm": 1.1229937869782511,
+      "learning_rate": 7.883639589733498e-06,
+      "loss": 0.7138,
+      "step": 1949
+    },
+    {
+      "epoch": 1.871819491118579,
+      "grad_norm": 0.7759493102845842,
+      "learning_rate": 7.880902238668631e-06,
+      "loss": 0.6015,
+      "step": 1950
+    },
+    {
+      "epoch": 1.8727796447431588,
+      "grad_norm": 1.098330681816144,
+      "learning_rate": 7.878163594397543e-06,
+      "loss": 0.7063,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8737397983677389,
+      "grad_norm": 1.1044481021724823,
+      "learning_rate": 7.875423658149583e-06,
+      "loss": 0.6902,
+      "step": 1952
+    },
+    {
+      "epoch": 1.8746999519923189,
+      "grad_norm": 1.2663156630113597,
+      "learning_rate": 7.87268243115468e-06,
+      "loss": 0.7063,
+      "step": 1953
+    },
+    {
+      "epoch": 1.8756601056168987,
+      "grad_norm": 1.4520521258078687,
+      "learning_rate": 7.869939914643343e-06,
+      "loss": 0.7229,
+      "step": 1954
+    },
+    {
+      "epoch": 1.8766202592414787,
+      "grad_norm": 1.0617459167629775,
+      "learning_rate": 7.867196109846653e-06,
+      "loss": 0.6476,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8775804128660587,
+      "grad_norm": 1.3958866985961658,
+      "learning_rate": 7.86445101799628e-06,
+      "loss": 0.6608,
+      "step": 1956
+    },
+    {
+      "epoch": 1.8785405664906385,
+      "grad_norm": 1.2206399228833131,
+      "learning_rate": 7.861704640324466e-06,
+      "loss": 0.5386,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8795007201152183,
+      "grad_norm": 1.0046968552784716,
+      "learning_rate": 7.858956978064025e-06,
+      "loss": 0.5919,
+      "step": 1958
+    },
+    {
+      "epoch": 1.8804608737397985,
+      "grad_norm": 1.0151619535695873,
+      "learning_rate": 7.85620803244836e-06,
+      "loss": 0.6173,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8814210273643783,
+      "grad_norm": 1.0207747206779128,
+      "learning_rate": 7.85345780471144e-06,
+      "loss": 0.6329,
+      "step": 1960
+    },
+    {
+      "epoch": 1.8823811809889581,
+      "grad_norm": 0.9326902851316377,
+      "learning_rate": 7.850706296087811e-06,
+      "loss": 0.6369,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8833413346135381,
+      "grad_norm": 1.0361948400957086,
+      "learning_rate": 7.8479535078126e-06,
+      "loss": 0.5498,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8843014882381182,
+      "grad_norm": 0.957960449142528,
+      "learning_rate": 7.8451994411215e-06,
+      "loss": 0.6353,
+      "step": 1963
+    },
+    {
+      "epoch": 1.885261641862698,
+      "grad_norm": 0.9019753967112145,
+      "learning_rate": 7.842444097250787e-06,
+      "loss": 0.6288,
+      "step": 1964
+    },
+    {
+      "epoch": 1.886221795487278,
+      "grad_norm": 1.0750655440090873,
+      "learning_rate": 7.839687477437304e-06,
+      "loss": 0.5878,
+      "step": 1965
+    },
+    {
+      "epoch": 1.887181949111858,
+      "grad_norm": 0.9259874258322465,
+      "learning_rate": 7.836929582918464e-06,
+      "loss": 0.6749,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8881421027364378,
+      "grad_norm": 1.147776615043257,
+      "learning_rate": 7.834170414932259e-06,
+      "loss": 0.5769,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8891022563610178,
+      "grad_norm": 0.8621825258564246,
+      "learning_rate": 7.831409974717252e-06,
+      "loss": 0.5565,
+      "step": 1968
+    },
+    {
+      "epoch": 1.8900624099855978,
+      "grad_norm": 1.2248913855444168,
+      "learning_rate": 7.828648263512573e-06,
+      "loss": 0.6344,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8910225636101776,
+      "grad_norm": 1.2907510720676452,
+      "learning_rate": 7.825885282557926e-06,
+      "loss": 0.6224,
+      "step": 1970
+    },
+    {
+      "epoch": 1.8919827172347574,
+      "grad_norm": 1.1862880222849024,
+      "learning_rate": 7.823121033093581e-06,
+      "loss": 0.6892,
+      "step": 1971
+    },
+    {
+      "epoch": 1.8929428708593377,
+      "grad_norm": 1.002890596821769,
+      "learning_rate": 7.820355516360383e-06,
+      "loss": 0.6472,
+      "step": 1972
+    },
+    {
+      "epoch": 1.8939030244839175,
+      "grad_norm": 1.2047448111603778,
+      "learning_rate": 7.817588733599742e-06,
+      "loss": 0.6551,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8948631781084972,
+      "grad_norm": 0.9525275129416088,
+      "learning_rate": 7.814820686053634e-06,
+      "loss": 0.6269,
+      "step": 1974
+    },
+    {
+      "epoch": 1.8958233317330773,
+      "grad_norm": 0.6923440757128753,
+      "learning_rate": 7.812051374964611e-06,
+      "loss": 0.5931,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8967834853576573,
+      "grad_norm": 0.8917493964077245,
+      "learning_rate": 7.809280801575784e-06,
+      "loss": 0.6596,
+      "step": 1976
+    },
+    {
+      "epoch": 1.897743638982237,
+      "grad_norm": 1.044122606482162,
+      "learning_rate": 7.806508967130838e-06,
+      "loss": 0.6629,
+      "step": 1977
+    },
+    {
+      "epoch": 1.898703792606817,
+      "grad_norm": 0.8740264354924745,
+      "learning_rate": 7.803735872874012e-06,
+      "loss": 0.6139,
+      "step": 1978
+    },
+    {
+      "epoch": 1.8996639462313971,
+      "grad_norm": 1.0016841226101536,
+      "learning_rate": 7.800961520050125e-06,
+      "loss": 0.6834,
+      "step": 1979
+    },
+    {
+      "epoch": 1.900624099855977,
+      "grad_norm": 0.8857095140329992,
+      "learning_rate": 7.798185909904552e-06,
+      "loss": 0.6602,
+      "step": 1980
+    },
+    {
+      "epoch": 1.901584253480557,
+      "grad_norm": 0.9550896040999265,
+      "learning_rate": 7.795409043683237e-06,
+      "loss": 0.6896,
+      "step": 1981
+    },
+    {
+      "epoch": 1.902544407105137,
+      "grad_norm": 0.8851191160717644,
+      "learning_rate": 7.792630922632682e-06,
+      "loss": 0.5623,
+      "step": 1982
+    },
+    {
+      "epoch": 1.9035045607297167,
+      "grad_norm": 1.1502009947200451,
+      "learning_rate": 7.789851547999957e-06,
+      "loss": 0.7963,
+      "step": 1983
+    },
+    {
+      "epoch": 1.9044647143542965,
+      "grad_norm": 1.1733042609983582,
+      "learning_rate": 7.787070921032693e-06,
+      "loss": 0.647,
+      "step": 1984
+    },
+    {
+      "epoch": 1.9054248679788768,
+      "grad_norm": 0.9294392569689884,
+      "learning_rate": 7.784289042979085e-06,
+      "loss": 0.6825,
+      "step": 1985
+    },
+    {
+      "epoch": 1.9063850216034566,
+      "grad_norm": 0.9934211403486944,
+      "learning_rate": 7.781505915087891e-06,
+      "loss": 0.6155,
+      "step": 1986
+    },
+    {
+      "epoch": 1.9073451752280364,
+      "grad_norm": 1.1918964161217966,
+      "learning_rate": 7.778721538608422e-06,
+      "loss": 0.6664,
+      "step": 1987
+    },
+    {
+      "epoch": 1.9083053288526164,
+      "grad_norm": 1.0426428053082084,
+      "learning_rate": 7.775935914790557e-06,
+      "loss": 0.6234,
+      "step": 1988
+    },
+    {
+      "epoch": 1.9092654824771964,
+      "grad_norm": 1.0510493862237833,
+      "learning_rate": 7.773149044884732e-06,
+      "loss": 0.7403,
+      "step": 1989
+    },
+    {
+      "epoch": 1.9102256361017762,
+      "grad_norm": 1.5115286509769619,
+      "learning_rate": 7.770360930141946e-06,
+      "loss": 0.6588,
+      "step": 1990
+    },
+    {
+      "epoch": 1.9111857897263562,
+      "grad_norm": 1.1302363335567547,
+      "learning_rate": 7.767571571813752e-06,
+      "loss": 0.5889,
+      "step": 1991
+    },
+    {
+      "epoch": 1.9121459433509362,
+      "grad_norm": 0.8397706236392392,
+      "learning_rate": 7.764780971152262e-06,
+      "loss": 0.5496,
+      "step": 1992
+    },
+    {
+      "epoch": 1.913106096975516,
+      "grad_norm": 1.0688534481923766,
+      "learning_rate": 7.761989129410149e-06,
+      "loss": 0.6942,
+      "step": 1993
+    },
+    {
+      "epoch": 1.914066250600096,
+      "grad_norm": 1.1145804131889687,
+      "learning_rate": 7.759196047840639e-06,
+      "loss": 0.7286,
+      "step": 1994
+    },
+    {
+      "epoch": 1.915026404224676,
+      "grad_norm": 1.238021302805448,
+      "learning_rate": 7.756401727697518e-06,
+      "loss": 0.6105,
+      "step": 1995
+    },
+    {
+      "epoch": 1.9159865578492559,
+      "grad_norm": 1.1728378204701007,
+      "learning_rate": 7.753606170235124e-06,
+      "loss": 0.6221,
+      "step": 1996
+    },
+    {
+      "epoch": 1.9169467114738357,
+      "grad_norm": 1.083706528150401,
+      "learning_rate": 7.750809376708357e-06,
+      "loss": 0.6819,
+      "step": 1997
+    },
+    {
+      "epoch": 1.917906865098416,
+      "grad_norm": 1.1110884491237523,
+      "learning_rate": 7.748011348372663e-06,
+      "loss": 0.6413,
+      "step": 1998
+    },
+    {
+      "epoch": 1.9188670187229957,
+      "grad_norm": 1.0215471235052087,
+      "learning_rate": 7.74521208648405e-06,
+      "loss": 0.719,
+      "step": 1999
+    },
+    {
+      "epoch": 1.9198271723475755,
+      "grad_norm": 0.9261412164580689,
+      "learning_rate": 7.742411592299077e-06,
+      "loss": 0.6255,
+      "step": 2000
+    },
+    {
+      "epoch": 1.9207873259721555,
+      "grad_norm": 0.872463961667252,
+      "learning_rate": 7.739609867074856e-06,
+      "loss": 0.634,
+      "step": 2001
+    },
+    {
+      "epoch": 1.9217474795967355,
+      "grad_norm": 1.0863303089243082,
+      "learning_rate": 7.736806912069048e-06,
+      "loss": 0.6633,
+      "step": 2002
+    },
+    {
+      "epoch": 1.9227076332213153,
+      "grad_norm": 0.8139746210348665,
+      "learning_rate": 7.734002728539873e-06,
+      "loss": 0.6113,
+      "step": 2003
+    },
+    {
+      "epoch": 1.9236677868458953,
+      "grad_norm": 0.8817300854256722,
+      "learning_rate": 7.731197317746099e-06,
+      "loss": 0.5994,
+      "step": 2004
+    },
+    {
+      "epoch": 1.9246279404704754,
+      "grad_norm": 1.136222412651614,
+      "learning_rate": 7.728390680947043e-06,
+      "loss": 0.655,
+      "step": 2005
+    },
+    {
+      "epoch": 1.9255880940950552,
+      "grad_norm": 1.2325785352151544,
+      "learning_rate": 7.725582819402575e-06,
+      "loss": 0.6905,
+      "step": 2006
+    },
+    {
+      "epoch": 1.9265482477196352,
+      "grad_norm": 1.0641373207977778,
+      "learning_rate": 7.722773734373114e-06,
+      "loss": 0.587,
+      "step": 2007
+    },
+    {
+      "epoch": 1.9275084013442152,
+      "grad_norm": 1.0629082498817422,
+      "learning_rate": 7.71996342711963e-06,
+      "loss": 0.6946,
+      "step": 2008
+    },
+    {
+      "epoch": 1.928468554968795,
+      "grad_norm": 1.5321764093757542,
+      "learning_rate": 7.717151898903636e-06,
+      "loss": 0.596,
+      "step": 2009
+    },
+    {
+      "epoch": 1.9294287085933748,
+      "grad_norm": 1.2130629515284264,
+      "learning_rate": 7.714339150987202e-06,
+      "loss": 0.6919,
+      "step": 2010
+    },
+    {
+      "epoch": 1.930388862217955,
+      "grad_norm": 1.1116261332670805,
+      "learning_rate": 7.711525184632936e-06,
+      "loss": 0.7311,
+      "step": 2011
+    },
+    {
+      "epoch": 1.9313490158425348,
+      "grad_norm": 1.1953661542991707,
+      "learning_rate": 7.708710001103998e-06,
+      "loss": 0.6029,
+      "step": 2012
+    },
+    {
+      "epoch": 1.9323091694671146,
+      "grad_norm": 1.497513151133857,
+      "learning_rate": 7.705893601664099e-06,
+      "loss": 0.6562,
+      "step": 2013
+    },
+    {
+      "epoch": 1.9332693230916946,
+      "grad_norm": 1.2799348499546876,
+      "learning_rate": 7.703075987577483e-06,
+      "loss": 0.7839,
+      "step": 2014
+    },
+    {
+      "epoch": 1.9342294767162747,
+      "grad_norm": 1.525526033893843,
+      "learning_rate": 7.700257160108953e-06,
+      "loss": 0.6587,
+      "step": 2015
+    },
+    {
+      "epoch": 1.9351896303408544,
+      "grad_norm": 1.1700523940769612,
+      "learning_rate": 7.697437120523845e-06,
+      "loss": 0.7259,
+      "step": 2016
+    },
+    {
+      "epoch": 1.9361497839654345,
+      "grad_norm": 1.0531375077014808,
+      "learning_rate": 7.694615870088051e-06,
+      "loss": 0.7215,
+      "step": 2017
+    },
+    {
+      "epoch": 1.9371099375900145,
+      "grad_norm": 1.3969434971840897,
+      "learning_rate": 7.691793410067993e-06,
+      "loss": 0.5855,
+      "step": 2018
+    },
+    {
+      "epoch": 1.9380700912145943,
+      "grad_norm": 1.3604813865106966,
+      "learning_rate": 7.68896974173065e-06,
+      "loss": 0.6194,
+      "step": 2019
+    },
+    {
+      "epoch": 1.9390302448391743,
+      "grad_norm": 0.9952031977741248,
+      "learning_rate": 7.686144866343533e-06,
+      "loss": 0.5427,
+      "step": 2020
+    },
+    {
+      "epoch": 1.9399903984637543,
+      "grad_norm": 0.812314858579823,
+      "learning_rate": 7.683318785174698e-06,
+      "loss": 0.5197,
+      "step": 2021
+    },
+    {
+      "epoch": 1.9409505520883341,
+      "grad_norm": 0.95987119699267,
+      "learning_rate": 7.680491499492743e-06,
+      "loss": 0.6596,
+      "step": 2022
+    },
+    {
+      "epoch": 1.941910705712914,
+      "grad_norm": 1.2319589747848858,
+      "learning_rate": 7.677663010566807e-06,
+      "loss": 0.6821,
+      "step": 2023
+    },
+    {
+      "epoch": 1.9428708593374941,
+      "grad_norm": 0.6606150963033354,
+      "learning_rate": 7.674833319666568e-06,
+      "loss": 0.6038,
+      "step": 2024
+    },
+    {
+      "epoch": 1.943831012962074,
+      "grad_norm": 1.200520352178802,
+      "learning_rate": 7.672002428062245e-06,
+      "loss": 0.653,
+      "step": 2025
+    },
+    {
+      "epoch": 1.9447911665866537,
+      "grad_norm": 0.7192438014804379,
+      "learning_rate": 7.669170337024593e-06,
+      "loss": 0.4693,
+      "step": 2026
+    },
+    {
+      "epoch": 1.9457513202112338,
+      "grad_norm": 1.4486619176506674,
+      "learning_rate": 7.66633704782491e-06,
+      "loss": 0.7564,
+      "step": 2027
+    },
+    {
+      "epoch": 1.9467114738358138,
+      "grad_norm": 1.2242450439191879,
+      "learning_rate": 7.663502561735028e-06,
+      "loss": 0.733,
+      "step": 2028
+    },
+    {
+      "epoch": 1.9476716274603936,
+      "grad_norm": 1.1407736193961824,
+      "learning_rate": 7.660666880027316e-06,
+      "loss": 0.593,
+      "step": 2029
+    },
+    {
+      "epoch": 1.9486317810849736,
+      "grad_norm": 1.1271456340334174,
+      "learning_rate": 7.657830003974684e-06,
+      "loss": 0.644,
+      "step": 2030
+    },
+    {
+      "epoch": 1.9495919347095536,
+      "grad_norm": 0.846049078907581,
+      "learning_rate": 7.654991934850573e-06,
+      "loss": 0.5331,
+      "step": 2031
+    },
+    {
+      "epoch": 1.9505520883341334,
+      "grad_norm": 0.7872449106149692,
+      "learning_rate": 7.652152673928964e-06,
+      "loss": 0.6157,
+      "step": 2032
+    },
+    {
+      "epoch": 1.9515122419587134,
+      "grad_norm": 0.9912770009275691,
+      "learning_rate": 7.649312222484367e-06,
+      "loss": 0.6688,
+      "step": 2033
+    },
+    {
+      "epoch": 1.9524723955832934,
+      "grad_norm": 1.0527641311770262,
+      "learning_rate": 7.646470581791834e-06,
+      "loss": 0.7184,
+      "step": 2034
+    },
+    {
+      "epoch": 1.9534325492078732,
+      "grad_norm": 0.9104283901825249,
+      "learning_rate": 7.643627753126946e-06,
+      "loss": 0.6,
+      "step": 2035
+    },
+    {
+      "epoch": 1.954392702832453,
+      "grad_norm": 1.2765136056418498,
+      "learning_rate": 7.640783737765817e-06,
+      "loss": 0.7672,
+      "step": 2036
+    },
+    {
+      "epoch": 1.9553528564570333,
+      "grad_norm": 1.4204717911379219,
+      "learning_rate": 7.6379385369851e-06,
+      "loss": 0.6158,
+      "step": 2037
+    },
+    {
+      "epoch": 1.956313010081613,
+      "grad_norm": 1.9379913067353052,
+      "learning_rate": 7.63509215206197e-06,
+      "loss": 0.7316,
+      "step": 2038
+    },
+    {
+      "epoch": 1.9572731637061929,
+      "grad_norm": 1.1268648862052064,
+      "learning_rate": 7.63224458427414e-06,
+      "loss": 0.4365,
+      "step": 2039
+    },
+    {
+      "epoch": 1.9582333173307729,
+      "grad_norm": 0.7974068444813239,
+      "learning_rate": 7.629395834899852e-06,
+      "loss": 0.477,
+      "step": 2040
+    },
+    {
+      "epoch": 1.959193470955353,
+      "grad_norm": 1.1109805277309888,
+      "learning_rate": 7.626545905217884e-06,
+      "loss": 0.6397,
+      "step": 2041
+    },
+    {
+      "epoch": 1.9601536245799327,
+      "grad_norm": 1.2398850876268863,
+      "learning_rate": 7.6236947965075325e-06,
+      "loss": 0.6679,
+      "step": 2042
+    },
+    {
+      "epoch": 1.9611137782045127,
+      "grad_norm": 1.1737119231817035,
+      "learning_rate": 7.6208425100486334e-06,
+      "loss": 0.6392,
+      "step": 2043
+    },
+    {
+      "epoch": 1.9620739318290927,
+      "grad_norm": 0.8854080472389517,
+      "learning_rate": 7.617989047121548e-06,
+      "loss": 0.6596,
+      "step": 2044
+    },
+    {
+      "epoch": 1.9630340854536725,
+      "grad_norm": 0.7676659103912914,
+      "learning_rate": 7.6151344090071624e-06,
+      "loss": 0.5967,
+      "step": 2045
+    },
+    {
+      "epoch": 1.9639942390782525,
+      "grad_norm": 1.0291718099826375,
+      "learning_rate": 7.612278596986895e-06,
+      "loss": 0.7092,
+      "step": 2046
+    },
+    {
+      "epoch": 1.9649543927028326,
+      "grad_norm": 0.7766136208698822,
+      "learning_rate": 7.609421612342689e-06,
+      "loss": 0.6606,
+      "step": 2047
+    },
+    {
+      "epoch": 1.9659145463274124,
+      "grad_norm": 1.1861269837590105,
+      "learning_rate": 7.606563456357016e-06,
+      "loss": 0.7037,
+      "step": 2048
+    },
+    {
+      "epoch": 1.9668746999519922,
+      "grad_norm": 1.0408178029162052,
+      "learning_rate": 7.603704130312868e-06,
+      "loss": 0.7429,
+      "step": 2049
+    },
+    {
+      "epoch": 1.9678348535765724,
+      "grad_norm": 1.1869576752680744,
+      "learning_rate": 7.6008436354937695e-06,
+      "loss": 0.7273,
+      "step": 2050
+    },
+    {
+      "epoch": 1.9687950072011522,
+      "grad_norm": 1.3051495633501524,
+      "learning_rate": 7.5979819731837635e-06,
+      "loss": 0.6918,
+      "step": 2051
+    },
+    {
+      "epoch": 1.969755160825732,
+      "grad_norm": 0.7596205378022328,
+      "learning_rate": 7.595119144667421e-06,
+      "loss": 0.5797,
+      "step": 2052
+    },
+    {
+      "epoch": 1.970715314450312,
+      "grad_norm": 0.9586448051063569,
+      "learning_rate": 7.5922551512298345e-06,
+      "loss": 0.689,
+      "step": 2053
+    },
+    {
+      "epoch": 1.971675468074892,
+      "grad_norm": 0.8801070575115794,
+      "learning_rate": 7.589389994156622e-06,
+      "loss": 0.6686,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9726356216994718,
+      "grad_norm": 1.3509221888327454,
+      "learning_rate": 7.586523674733918e-06,
+      "loss": 0.6191,
+      "step": 2055
+    },
+    {
+      "epoch": 1.9735957753240518,
+      "grad_norm": 1.1120138619462479,
+      "learning_rate": 7.583656194248386e-06,
+      "loss": 0.5759,
+      "step": 2056
+    },
+    {
+      "epoch": 1.9745559289486319,
+      "grad_norm": 1.0233499758935727,
+      "learning_rate": 7.5807875539872054e-06,
+      "loss": 0.5934,
+      "step": 2057
+    },
+    {
+      "epoch": 1.9755160825732117,
+      "grad_norm": 1.3666746387565665,
+      "learning_rate": 7.577917755238079e-06,
+      "loss": 0.7105,
+      "step": 2058
+    },
+    {
+      "epoch": 1.9764762361977917,
+      "grad_norm": 1.1636696265207904,
+      "learning_rate": 7.57504679928923e-06,
+      "loss": 0.7389,
+      "step": 2059
+    },
+    {
+      "epoch": 1.9774363898223717,
+      "grad_norm": 1.1833938285444674,
+      "learning_rate": 7.572174687429395e-06,
+      "loss": 0.6764,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9783965434469515,
+      "grad_norm": 1.0399387517184737,
+      "learning_rate": 7.569301420947841e-06,
+      "loss": 0.5481,
+      "step": 2061
+    },
+    {
+      "epoch": 1.9793566970715313,
+      "grad_norm": 0.9375153941817806,
+      "learning_rate": 7.566427001134342e-06,
+      "loss": 0.7324,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9803168506961115,
+      "grad_norm": 1.19747434705813,
+      "learning_rate": 7.563551429279198e-06,
+      "loss": 0.6257,
+      "step": 2063
+    },
+    {
+      "epoch": 1.9812770043206913,
+      "grad_norm": 1.3612075834492465,
+      "learning_rate": 7.560674706673219e-06,
+      "loss": 0.5541,
+      "step": 2064
+    },
+    {
+      "epoch": 1.982237157945271,
+      "grad_norm": 1.2460446469104085,
+      "learning_rate": 7.557796834607736e-06,
+      "loss": 0.7148,
+      "step": 2065
+    },
+    {
+      "epoch": 1.9831973115698511,
+      "grad_norm": 1.2505377581381347,
+      "learning_rate": 7.554917814374595e-06,
+      "loss": 0.6051,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9841574651944311,
+      "grad_norm": 1.3178231047232203,
+      "learning_rate": 7.552037647266157e-06,
+      "loss": 0.666,
+      "step": 2067
+    },
+    {
+      "epoch": 1.985117618819011,
+      "grad_norm": 1.2053322598268592,
+      "learning_rate": 7.5491563345753e-06,
+      "loss": 0.6736,
+      "step": 2068
+    },
+    {
+      "epoch": 1.986077772443591,
+      "grad_norm": 1.5118119390957352,
+      "learning_rate": 7.546273877595413e-06,
+      "loss": 0.6847,
+      "step": 2069
+    },
+    {
+      "epoch": 1.987037926068171,
+      "grad_norm": 1.0329897029908128,
+      "learning_rate": 7.5433902776204015e-06,
+      "loss": 0.6205,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9879980796927508,
+      "grad_norm": 0.9974569099985744,
+      "learning_rate": 7.54050553594468e-06,
+      "loss": 0.5332,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9889582333173308,
+      "grad_norm": 1.0690421060675295,
+      "learning_rate": 7.537619653863182e-06,
+      "loss": 0.7159,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9899183869419108,
+      "grad_norm": 1.1632058889940493,
+      "learning_rate": 7.5347326326713454e-06,
+      "loss": 0.5751,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9908785405664906,
+      "grad_norm": 1.422020334434976,
+      "learning_rate": 7.531844473665125e-06,
+      "loss": 0.7107,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9918386941910706,
+      "grad_norm": 1.0079638983911854,
+      "learning_rate": 7.528955178140985e-06,
+      "loss": 0.5734,
+      "step": 2075
+    },
+    {
+      "epoch": 1.9927988478156506,
+      "grad_norm": 0.9123186568870082,
+      "learning_rate": 7.526064747395898e-06,
+      "loss": 0.6097,
+      "step": 2076
+    },
+    {
+      "epoch": 1.9937590014402304,
+      "grad_norm": 1.0309543923066218,
+      "learning_rate": 7.523173182727348e-06,
+      "loss": 0.5464,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9947191550648102,
+      "grad_norm": 1.0854712934998054,
+      "learning_rate": 7.520280485433327e-06,
+      "loss": 0.7086,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9956793086893903,
+      "grad_norm": 1.006296613540927,
+      "learning_rate": 7.517386656812338e-06,
+      "loss": 0.6084,
+      "step": 2079
+    },
+    {
+      "epoch": 1.9966394623139703,
+      "grad_norm": 0.915483611749352,
+      "learning_rate": 7.514491698163387e-06,
+      "loss": 0.6173,
+      "step": 2080
+    },
+    {
+      "epoch": 1.99759961593855,
+      "grad_norm": 1.0737464838300075,
+      "learning_rate": 7.511595610785991e-06,
+      "loss": 0.6215,
+      "step": 2081
+    },
+    {
+      "epoch": 1.99855976956313,
+      "grad_norm": 1.0588984534223536,
+      "learning_rate": 7.508698395980173e-06,
+      "loss": 0.7455,
+      "step": 2082
+    },
+    {
+      "epoch": 1.99951992318771,
+      "grad_norm": 1.0259188689456935,
+      "learning_rate": 7.5058000550464615e-06,
+      "loss": 0.5611,
+      "step": 2083
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3486689052941767,
+      "learning_rate": 7.502900589285891e-06,
+      "loss": 0.5107,
+      "step": 2084
+    },
+    {
+      "epoch": 2.00096015362458,
+      "grad_norm": 0.9242603330168586,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.54,
+      "step": 2085
+    },
+    {
+      "epoch": 2.00192030724916,
+      "grad_norm": 0.9543728385159,
+      "learning_rate": 7.497098288490834e-06,
+      "loss": 0.4942,
+      "step": 2086
+    },
+    {
+      "epoch": 2.00288046087374,
+      "grad_norm": 1.4659028048986025,
+      "learning_rate": 7.494195456060938e-06,
+      "loss": 0.5756,
+      "step": 2087
+    },
+    {
+      "epoch": 2.0038406144983196,
+      "grad_norm": 1.2257178724297157,
+      "learning_rate": 7.491291504013363e-06,
+      "loss": 0.4658,
+      "step": 2088
+    },
+    {
+      "epoch": 2.0048007681229,
+      "grad_norm": 0.8867977842650157,
+      "learning_rate": 7.488386433651662e-06,
+      "loss": 0.5053,
+      "step": 2089
+    },
+    {
+      "epoch": 2.0057609217474797,
+      "grad_norm": 1.192674325910438,
+      "learning_rate": 7.4854802462798905e-06,
+      "loss": 0.5525,
+      "step": 2090
+    },
+    {
+      "epoch": 2.0067210753720595,
+      "grad_norm": 1.0367888174874489,
+      "learning_rate": 7.482572943202604e-06,
+      "loss": 0.6436,
+      "step": 2091
+    },
+    {
+      "epoch": 2.0076812289966393,
+      "grad_norm": 1.0340421819235337,
+      "learning_rate": 7.4796645257248625e-06,
+      "loss": 0.5068,
+      "step": 2092
+    },
+    {
+      "epoch": 2.0086413826212195,
+      "grad_norm": 0.8516771088470378,
+      "learning_rate": 7.47675499515222e-06,
+      "loss": 0.4932,
+      "step": 2093
+    },
+    {
+      "epoch": 2.0096015362457993,
+      "grad_norm": 1.9015516331041107,
+      "learning_rate": 7.473844352790735e-06,
+      "loss": 0.6114,
+      "step": 2094
+    },
+    {
+      "epoch": 2.010561689870379,
+      "grad_norm": 1.2434236142855897,
+      "learning_rate": 7.470932599946965e-06,
+      "loss": 0.5542,
+      "step": 2095
+    },
+    {
+      "epoch": 2.0115218434949593,
+      "grad_norm": 1.1912469569660604,
+      "learning_rate": 7.468019737927963e-06,
+      "loss": 0.6125,
+      "step": 2096
+    },
+    {
+      "epoch": 2.012481997119539,
+      "grad_norm": 1.1075309095334798,
+      "learning_rate": 7.465105768041283e-06,
+      "loss": 0.5807,
+      "step": 2097
+    },
+    {
+      "epoch": 2.013442150744119,
+      "grad_norm": 1.3920065240518684,
+      "learning_rate": 7.462190691594973e-06,
+      "loss": 0.644,
+      "step": 2098
+    },
+    {
+      "epoch": 2.014402304368699,
+      "grad_norm": 1.0387681847196268,
+      "learning_rate": 7.459274509897583e-06,
+      "loss": 0.5815,
+      "step": 2099
+    },
+    {
+      "epoch": 2.015362457993279,
+      "grad_norm": 1.0086737416782097,
+      "learning_rate": 7.45635722425815e-06,
+      "loss": 0.6361,
+      "step": 2100
+    },
+    {
+      "epoch": 2.0163226116178588,
+      "grad_norm": 1.5131229782622566,
+      "learning_rate": 7.45343883598622e-06,
+      "loss": 0.4398,
+      "step": 2101
+    },
+    {
+      "epoch": 2.017282765242439,
+      "grad_norm": 0.9217517932823873,
+      "learning_rate": 7.45051934639182e-06,
+      "loss": 0.4969,
+      "step": 2102
+    },
+    {
+      "epoch": 2.018242918867019,
+      "grad_norm": 1.0006289807747306,
+      "learning_rate": 7.447598756785482e-06,
+      "loss": 0.5403,
+      "step": 2103
+    },
+    {
+      "epoch": 2.0192030724915986,
+      "grad_norm": 1.034626184859905,
+      "learning_rate": 7.444677068478223e-06,
+      "loss": 0.5091,
+      "step": 2104
+    },
+    {
+      "epoch": 2.0201632261161784,
+      "grad_norm": 0.9478952916646003,
+      "learning_rate": 7.441754282781561e-06,
+      "loss": 0.6067,
+      "step": 2105
+    },
+    {
+      "epoch": 2.0211233797407586,
+      "grad_norm": 1.1611452501373463,
+      "learning_rate": 7.438830401007504e-06,
+      "loss": 0.5924,
+      "step": 2106
+    },
+    {
+      "epoch": 2.0220835333653384,
+      "grad_norm": 1.316256258593302,
+      "learning_rate": 7.435905424468548e-06,
+      "loss": 0.5202,
+      "step": 2107
+    },
+    {
+      "epoch": 2.023043686989918,
+      "grad_norm": 1.23591737148139,
+      "learning_rate": 7.432979354477686e-06,
+      "loss": 0.5226,
+      "step": 2108
+    },
+    {
+      "epoch": 2.0240038406144985,
+      "grad_norm": 1.134488568665678,
+      "learning_rate": 7.430052192348398e-06,
+      "loss": 0.5593,
+      "step": 2109
+    },
+    {
+      "epoch": 2.0249639942390782,
+      "grad_norm": 0.873863500954897,
+      "learning_rate": 7.427123939394658e-06,
+      "loss": 0.5202,
+      "step": 2110
+    },
+    {
+      "epoch": 2.025924147863658,
+      "grad_norm": 1.5403396675611922,
+      "learning_rate": 7.424194596930925e-06,
+      "loss": 0.5958,
+      "step": 2111
+    },
+    {
+      "epoch": 2.0268843014882383,
+      "grad_norm": 1.4791655228119358,
+      "learning_rate": 7.421264166272152e-06,
+      "loss": 0.5972,
+      "step": 2112
+    },
+    {
+      "epoch": 2.027844455112818,
+      "grad_norm": 1.3053642943908599,
+      "learning_rate": 7.418332648733774e-06,
+      "loss": 0.5313,
+      "step": 2113
+    },
+    {
+      "epoch": 2.028804608737398,
+      "grad_norm": 1.3286714063074645,
+      "learning_rate": 7.415400045631723e-06,
+      "loss": 0.612,
+      "step": 2114
+    },
+    {
+      "epoch": 2.029764762361978,
+      "grad_norm": 0.9569340901030102,
+      "learning_rate": 7.41246635828241e-06,
+      "loss": 0.4132,
+      "step": 2115
+    },
+    {
+      "epoch": 2.030724915986558,
+      "grad_norm": 1.1873003684933803,
+      "learning_rate": 7.4095315880027365e-06,
+      "loss": 0.5841,
+      "step": 2116
+    },
+    {
+      "epoch": 2.0316850696111377,
+      "grad_norm": 1.5116393108346937,
+      "learning_rate": 7.406595736110089e-06,
+      "loss": 0.5357,
+      "step": 2117
+    },
+    {
+      "epoch": 2.0326452232357175,
+      "grad_norm": 1.0182481052961394,
+      "learning_rate": 7.4036588039223415e-06,
+      "loss": 0.6686,
+      "step": 2118
+    },
+    {
+      "epoch": 2.0336053768602977,
+      "grad_norm": 0.8425290443373554,
+      "learning_rate": 7.400720792757851e-06,
+      "loss": 0.5016,
+      "step": 2119
+    },
+    {
+      "epoch": 2.0345655304848775,
+      "grad_norm": 1.477362353438116,
+      "learning_rate": 7.397781703935458e-06,
+      "loss": 0.5692,
+      "step": 2120
+    },
+    {
+      "epoch": 2.0355256841094573,
+      "grad_norm": 1.4325155153609799,
+      "learning_rate": 7.39484153877449e-06,
+      "loss": 0.5293,
+      "step": 2121
+    },
+    {
+      "epoch": 2.0364858377340376,
+      "grad_norm": 0.9282735950305444,
+      "learning_rate": 7.391900298594752e-06,
+      "loss": 0.4984,
+      "step": 2122
+    },
+    {
+      "epoch": 2.0374459913586174,
+      "grad_norm": 0.9532270139954646,
+      "learning_rate": 7.388957984716541e-06,
+      "loss": 0.5398,
+      "step": 2123
+    },
+    {
+      "epoch": 2.038406144983197,
+      "grad_norm": 0.8762355601311758,
+      "learning_rate": 7.386014598460626e-06,
+      "loss": 0.5812,
+      "step": 2124
+    },
+    {
+      "epoch": 2.0393662986077774,
+      "grad_norm": 1.0739822627421551,
+      "learning_rate": 7.383070141148263e-06,
+      "loss": 0.5636,
+      "step": 2125
+    },
+    {
+      "epoch": 2.040326452232357,
+      "grad_norm": 1.2489651649480997,
+      "learning_rate": 7.380124614101185e-06,
+      "loss": 0.5317,
+      "step": 2126
+    },
+    {
+      "epoch": 2.041286605856937,
+      "grad_norm": 1.2927524784687077,
+      "learning_rate": 7.377178018641613e-06,
+      "loss": 0.5812,
+      "step": 2127
+    },
+    {
+      "epoch": 2.0422467594815172,
+      "grad_norm": 1.3946462494710277,
+      "learning_rate": 7.374230356092238e-06,
+      "loss": 0.5615,
+      "step": 2128
+    },
+    {
+      "epoch": 2.043206913106097,
+      "grad_norm": 0.9822202806600208,
+      "learning_rate": 7.371281627776237e-06,
+      "loss": 0.6693,
+      "step": 2129
+    },
+    {
+      "epoch": 2.044167066730677,
+      "grad_norm": 1.2864927131454644,
+      "learning_rate": 7.368331835017257e-06,
+      "loss": 0.5295,
+      "step": 2130
+    },
+    {
+      "epoch": 2.0451272203552566,
+      "grad_norm": 1.2331212956816664,
+      "learning_rate": 7.365380979139435e-06,
+      "loss": 0.6415,
+      "step": 2131
+    },
+    {
+      "epoch": 2.046087373979837,
+      "grad_norm": 1.1555702611744292,
+      "learning_rate": 7.362429061467377e-06,
+      "loss": 0.6135,
+      "step": 2132
+    },
+    {
+      "epoch": 2.0470475276044167,
+      "grad_norm": 1.0250907289031153,
+      "learning_rate": 7.359476083326164e-06,
+      "loss": 0.6169,
+      "step": 2133
+    },
+    {
+      "epoch": 2.0480076812289965,
+      "grad_norm": 1.0358471386936456,
+      "learning_rate": 7.35652204604136e-06,
+      "loss": 0.5128,
+      "step": 2134
+    },
+    {
+      "epoch": 2.0489678348535767,
+      "grad_norm": 1.0637155576382715,
+      "learning_rate": 7.353566950938997e-06,
+      "loss": 0.4894,
+      "step": 2135
+    },
+    {
+      "epoch": 2.0499279884781565,
+      "grad_norm": 1.5028422968427364,
+      "learning_rate": 7.35061079934559e-06,
+      "loss": 0.573,
+      "step": 2136
+    },
+    {
+      "epoch": 2.0508881421027363,
+      "grad_norm": 1.1093624220431402,
+      "learning_rate": 7.347653592588122e-06,
+      "loss": 0.5491,
+      "step": 2137
+    },
+    {
+      "epoch": 2.0518482957273165,
+      "grad_norm": 1.2646723948358842,
+      "learning_rate": 7.344695331994052e-06,
+      "loss": 0.4999,
+      "step": 2138
+    },
+    {
+      "epoch": 2.0528084493518963,
+      "grad_norm": 1.030349237091044,
+      "learning_rate": 7.341736018891308e-06,
+      "loss": 0.511,
+      "step": 2139
+    },
+    {
+      "epoch": 2.053768602976476,
+      "grad_norm": 1.1360884059780991,
+      "learning_rate": 7.338775654608299e-06,
+      "loss": 0.544,
+      "step": 2140
+    },
+    {
+      "epoch": 2.0547287566010564,
+      "grad_norm": 1.048952709164682,
+      "learning_rate": 7.335814240473899e-06,
+      "loss": 0.6217,
+      "step": 2141
+    },
+    {
+      "epoch": 2.055688910225636,
+      "grad_norm": 1.2728010363199347,
+      "learning_rate": 7.332851777817453e-06,
+      "loss": 0.4647,
+      "step": 2142
+    },
+    {
+      "epoch": 2.056649063850216,
+      "grad_norm": 1.157030872007594,
+      "learning_rate": 7.3298882679687836e-06,
+      "loss": 0.4886,
+      "step": 2143
+    },
+    {
+      "epoch": 2.0576092174747957,
+      "grad_norm": 0.9473979824746046,
+      "learning_rate": 7.326923712258173e-06,
+      "loss": 0.5894,
+      "step": 2144
+    },
+    {
+      "epoch": 2.058569371099376,
+      "grad_norm": 0.91389585244967,
+      "learning_rate": 7.323958112016383e-06,
+      "loss": 0.482,
+      "step": 2145
+    },
+    {
+      "epoch": 2.059529524723956,
+      "grad_norm": 0.9679054569476475,
+      "learning_rate": 7.320991468574637e-06,
+      "loss": 0.5391,
+      "step": 2146
+    },
+    {
+      "epoch": 2.0604896783485356,
+      "grad_norm": 1.3249677389529415,
+      "learning_rate": 7.318023783264633e-06,
+      "loss": 0.5262,
+      "step": 2147
+    },
+    {
+      "epoch": 2.061449831973116,
+      "grad_norm": 0.9183488536355273,
+      "learning_rate": 7.315055057418528e-06,
+      "loss": 0.4921,
+      "step": 2148
+    },
+    {
+      "epoch": 2.0624099855976956,
+      "grad_norm": 1.4264562988423015,
+      "learning_rate": 7.312085292368957e-06,
+      "loss": 0.4443,
+      "step": 2149
+    },
+    {
+      "epoch": 2.0633701392222754,
+      "grad_norm": 1.1406047321067216,
+      "learning_rate": 7.309114489449012e-06,
+      "loss": 0.5625,
+      "step": 2150
+    },
+    {
+      "epoch": 2.0643302928468557,
+      "grad_norm": 1.422593860911789,
+      "learning_rate": 7.306142649992256e-06,
+      "loss": 0.5264,
+      "step": 2151
+    },
+    {
+      "epoch": 2.0652904464714354,
+      "grad_norm": 1.0953180106941673,
+      "learning_rate": 7.3031697753327145e-06,
+      "loss": 0.5759,
+      "step": 2152
+    },
+    {
+      "epoch": 2.0662506000960152,
+      "grad_norm": 0.8941745426562419,
+      "learning_rate": 7.300195866804881e-06,
+      "loss": 0.4539,
+      "step": 2153
+    },
+    {
+      "epoch": 2.0672107537205955,
+      "grad_norm": 1.2100475289024843,
+      "learning_rate": 7.2972209257437125e-06,
+      "loss": 0.6848,
+      "step": 2154
+    },
+    {
+      "epoch": 2.0681709073451753,
+      "grad_norm": 1.3674083245206652,
+      "learning_rate": 7.2942449534846234e-06,
+      "loss": 0.5534,
+      "step": 2155
+    },
+    {
+      "epoch": 2.069131060969755,
+      "grad_norm": 0.9586407848894423,
+      "learning_rate": 7.291267951363501e-06,
+      "loss": 0.49,
+      "step": 2156
+    },
+    {
+      "epoch": 2.070091214594335,
+      "grad_norm": 1.0061110896808698,
+      "learning_rate": 7.288289920716686e-06,
+      "loss": 0.4975,
+      "step": 2157
+    },
+    {
+      "epoch": 2.071051368218915,
+      "grad_norm": 1.0432840568946489,
+      "learning_rate": 7.285310862880987e-06,
+      "loss": 0.575,
+      "step": 2158
+    },
+    {
+      "epoch": 2.072011521843495,
+      "grad_norm": 0.9872238049121338,
+      "learning_rate": 7.28233077919367e-06,
+      "loss": 0.5424,
+      "step": 2159
+    },
+    {
+      "epoch": 2.0729716754680747,
+      "grad_norm": 1.790221412169585,
+      "learning_rate": 7.279349670992464e-06,
+      "loss": 0.541,
+      "step": 2160
+    },
+    {
+      "epoch": 2.073931829092655,
+      "grad_norm": 1.032360706897286,
+      "learning_rate": 7.2763675396155545e-06,
+      "loss": 0.56,
+      "step": 2161
+    },
+    {
+      "epoch": 2.0748919827172347,
+      "grad_norm": 0.824803012243617,
+      "learning_rate": 7.27338438640159e-06,
+      "loss": 0.525,
+      "step": 2162
+    },
+    {
+      "epoch": 2.0758521363418145,
+      "grad_norm": 1.1790315047925064,
+      "learning_rate": 7.270400212689675e-06,
+      "loss": 0.577,
+      "step": 2163
+    },
+    {
+      "epoch": 2.0768122899663948,
+      "grad_norm": 1.2404278132966462,
+      "learning_rate": 7.2674150198193735e-06,
+      "loss": 0.5341,
+      "step": 2164
+    },
+    {
+      "epoch": 2.0777724435909746,
+      "grad_norm": 1.022275468924025,
+      "learning_rate": 7.264428809130708e-06,
+      "loss": 0.6559,
+      "step": 2165
+    },
+    {
+      "epoch": 2.0787325972155544,
+      "grad_norm": 1.2774007022863516,
+      "learning_rate": 7.2614415819641535e-06,
+      "loss": 0.5772,
+      "step": 2166
+    },
+    {
+      "epoch": 2.0796927508401346,
+      "grad_norm": 1.2454958358304937,
+      "learning_rate": 7.258453339660648e-06,
+      "loss": 0.6805,
+      "step": 2167
+    },
+    {
+      "epoch": 2.0806529044647144,
+      "grad_norm": 0.9254926320094997,
+      "learning_rate": 7.25546408356158e-06,
+      "loss": 0.5528,
+      "step": 2168
+    },
+    {
+      "epoch": 2.081613058089294,
+      "grad_norm": 1.1372835025449755,
+      "learning_rate": 7.252473815008794e-06,
+      "loss": 0.4636,
+      "step": 2169
+    },
+    {
+      "epoch": 2.082573211713874,
+      "grad_norm": 0.9132542974941551,
+      "learning_rate": 7.24948253534459e-06,
+      "loss": 0.4813,
+      "step": 2170
+    },
+    {
+      "epoch": 2.0835333653384542,
+      "grad_norm": 1.1585164013087945,
+      "learning_rate": 7.246490245911722e-06,
+      "loss": 0.4817,
+      "step": 2171
+    },
+    {
+      "epoch": 2.084493518963034,
+      "grad_norm": 0.9130120744811105,
+      "learning_rate": 7.243496948053394e-06,
+      "loss": 0.5988,
+      "step": 2172
+    },
+    {
+      "epoch": 2.085453672587614,
+      "grad_norm": 1.1587018142316117,
+      "learning_rate": 7.240502643113271e-06,
+      "loss": 0.664,
+      "step": 2173
+    },
+    {
+      "epoch": 2.086413826212194,
+      "grad_norm": 0.9006150195164618,
+      "learning_rate": 7.237507332435459e-06,
+      "loss": 0.5423,
+      "step": 2174
+    },
+    {
+      "epoch": 2.087373979836774,
+      "grad_norm": 0.946251970852323,
+      "learning_rate": 7.234511017364524e-06,
+      "loss": 0.6016,
+      "step": 2175
+    },
+    {
+      "epoch": 2.0883341334613537,
+      "grad_norm": 1.242912712579443,
+      "learning_rate": 7.2315136992454794e-06,
+      "loss": 0.6251,
+      "step": 2176
+    },
+    {
+      "epoch": 2.089294287085934,
+      "grad_norm": 1.1224268015584455,
+      "learning_rate": 7.228515379423787e-06,
+      "loss": 0.6026,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0902544407105137,
+      "grad_norm": 0.8180908380728011,
+      "learning_rate": 7.225516059245365e-06,
+      "loss": 0.5581,
+      "step": 2178
+    },
+    {
+      "epoch": 2.0912145943350935,
+      "grad_norm": 0.9144789064207813,
+      "learning_rate": 7.222515740056571e-06,
+      "loss": 0.5401,
+      "step": 2179
+    },
+    {
+      "epoch": 2.0921747479596737,
+      "grad_norm": 1.2387776357857734,
+      "learning_rate": 7.219514423204219e-06,
+      "loss": 0.5254,
+      "step": 2180
+    },
+    {
+      "epoch": 2.0931349015842535,
+      "grad_norm": 1.5793470381166037,
+      "learning_rate": 7.216512110035567e-06,
+      "loss": 0.5215,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0940950552088333,
+      "grad_norm": 1.0212283069024395,
+      "learning_rate": 7.213508801898325e-06,
+      "loss": 0.4813,
+      "step": 2182
+    },
+    {
+      "epoch": 2.095055208833413,
+      "grad_norm": 1.1890346255264337,
+      "learning_rate": 7.21050450014064e-06,
+      "loss": 0.5749,
+      "step": 2183
+    },
+    {
+      "epoch": 2.0960153624579934,
+      "grad_norm": 1.5029802892697517,
+      "learning_rate": 7.2074992061111145e-06,
+      "loss": 0.5671,
+      "step": 2184
+    },
+    {
+      "epoch": 2.096975516082573,
+      "grad_norm": 0.929036317951188,
+      "learning_rate": 7.204492921158794e-06,
+      "loss": 0.5416,
+      "step": 2185
+    },
+    {
+      "epoch": 2.097935669707153,
+      "grad_norm": 1.1339110148519138,
+      "learning_rate": 7.201485646633164e-06,
+      "loss": 0.6285,
+      "step": 2186
+    },
+    {
+      "epoch": 2.098895823331733,
+      "grad_norm": 0.8938875483158,
+      "learning_rate": 7.198477383884161e-06,
+      "loss": 0.4489,
+      "step": 2187
+    },
+    {
+      "epoch": 2.099855976956313,
+      "grad_norm": 1.1838577139933362,
+      "learning_rate": 7.195468134262159e-06,
+      "loss": 0.5581,
+      "step": 2188
+    },
+    {
+      "epoch": 2.100816130580893,
+      "grad_norm": 0.8025569757175272,
+      "learning_rate": 7.1924578991179815e-06,
+      "loss": 0.5391,
+      "step": 2189
+    },
+    {
+      "epoch": 2.101776284205473,
+      "grad_norm": 1.5927228117090462,
+      "learning_rate": 7.189446679802889e-06,
+      "loss": 0.5613,
+      "step": 2190
+    },
+    {
+      "epoch": 2.102736437830053,
+      "grad_norm": 1.0842887995533879,
+      "learning_rate": 7.186434477668586e-06,
+      "loss": 0.5342,
+      "step": 2191
+    },
+    {
+      "epoch": 2.1036965914546326,
+      "grad_norm": 1.3835446087699197,
+      "learning_rate": 7.1834212940672155e-06,
+      "loss": 0.6185,
+      "step": 2192
+    },
+    {
+      "epoch": 2.104656745079213,
+      "grad_norm": 1.1413603502415013,
+      "learning_rate": 7.180407130351368e-06,
+      "loss": 0.4431,
+      "step": 2193
+    },
+    {
+      "epoch": 2.1056168987037926,
+      "grad_norm": 1.0203641308253522,
+      "learning_rate": 7.177391987874065e-06,
+      "loss": 0.548,
+      "step": 2194
+    },
+    {
+      "epoch": 2.1065770523283724,
+      "grad_norm": 1.1363311509920815,
+      "learning_rate": 7.174375867988776e-06,
+      "loss": 0.5308,
+      "step": 2195
+    },
+    {
+      "epoch": 2.1075372059529527,
+      "grad_norm": 0.9567673445544885,
+      "learning_rate": 7.171358772049402e-06,
+      "loss": 0.5575,
+      "step": 2196
+    },
+    {
+      "epoch": 2.1084973595775325,
+      "grad_norm": 0.9923700778138494,
+      "learning_rate": 7.168340701410284e-06,
+      "loss": 0.6697,
+      "step": 2197
+    },
+    {
+      "epoch": 2.1094575132021123,
+      "grad_norm": 1.0364082103638248,
+      "learning_rate": 7.165321657426205e-06,
+      "loss": 0.5625,
+      "step": 2198
+    },
+    {
+      "epoch": 2.110417666826692,
+      "grad_norm": 0.9024784229244153,
+      "learning_rate": 7.1623016414523795e-06,
+      "loss": 0.6424,
+      "step": 2199
+    },
+    {
+      "epoch": 2.1113778204512723,
+      "grad_norm": 1.367149463303392,
+      "learning_rate": 7.159280654844462e-06,
+      "loss": 0.5617,
+      "step": 2200
+    },
+    {
+      "epoch": 2.112337974075852,
+      "grad_norm": 1.1617838082001823,
+      "learning_rate": 7.156258698958537e-06,
+      "loss": 0.554,
+      "step": 2201
+    },
+    {
+      "epoch": 2.113298127700432,
+      "grad_norm": 1.021449634098613,
+      "learning_rate": 7.153235775151134e-06,
+      "loss": 0.5342,
+      "step": 2202
+    },
+    {
+      "epoch": 2.114258281325012,
+      "grad_norm": 0.8308406626044873,
+      "learning_rate": 7.150211884779208e-06,
+      "loss": 0.5278,
+      "step": 2203
+    },
+    {
+      "epoch": 2.115218434949592,
+      "grad_norm": 0.9307394949034711,
+      "learning_rate": 7.147187029200151e-06,
+      "loss": 0.6127,
+      "step": 2204
+    },
+    {
+      "epoch": 2.1161785885741717,
+      "grad_norm": 1.8052285647276194,
+      "learning_rate": 7.144161209771788e-06,
+      "loss": 0.4938,
+      "step": 2205
+    },
+    {
+      "epoch": 2.117138742198752,
+      "grad_norm": 0.9771174668616823,
+      "learning_rate": 7.141134427852378e-06,
+      "loss": 0.5277,
+      "step": 2206
+    },
+    {
+      "epoch": 2.1180988958233318,
+      "grad_norm": 1.0292956633743262,
+      "learning_rate": 7.138106684800612e-06,
+      "loss": 0.5094,
+      "step": 2207
+    },
+    {
+      "epoch": 2.1190590494479116,
+      "grad_norm": 1.2451525712883664,
+      "learning_rate": 7.1350779819756085e-06,
+      "loss": 0.5448,
+      "step": 2208
+    },
+    {
+      "epoch": 2.120019203072492,
+      "grad_norm": 0.9661193881293042,
+      "learning_rate": 7.132048320736924e-06,
+      "loss": 0.5537,
+      "step": 2209
+    },
+    {
+      "epoch": 2.1209793566970716,
+      "grad_norm": 1.3789163777407878,
+      "learning_rate": 7.129017702444536e-06,
+      "loss": 0.611,
+      "step": 2210
+    },
+    {
+      "epoch": 2.1219395103216514,
+      "grad_norm": 1.5546841461614513,
+      "learning_rate": 7.12598612845886e-06,
+      "loss": 0.6078,
+      "step": 2211
+    },
+    {
+      "epoch": 2.122899663946231,
+      "grad_norm": 1.0926028952817535,
+      "learning_rate": 7.1229536001407376e-06,
+      "loss": 0.4314,
+      "step": 2212
+    },
+    {
+      "epoch": 2.1238598175708114,
+      "grad_norm": 1.0646968709677729,
+      "learning_rate": 7.119920118851438e-06,
+      "loss": 0.4824,
+      "step": 2213
+    },
+    {
+      "epoch": 2.1248199711953912,
+      "grad_norm": 1.3513451104794474,
+      "learning_rate": 7.116885685952655e-06,
+      "loss": 0.469,
+      "step": 2214
+    },
+    {
+      "epoch": 2.125780124819971,
+      "grad_norm": 0.9639076857555183,
+      "learning_rate": 7.113850302806518e-06,
+      "loss": 0.5212,
+      "step": 2215
+    },
+    {
+      "epoch": 2.1267402784445513,
+      "grad_norm": 1.0092878405418004,
+      "learning_rate": 7.110813970775574e-06,
+      "loss": 0.5365,
+      "step": 2216
+    },
+    {
+      "epoch": 2.127700432069131,
+      "grad_norm": 1.2257022771483324,
+      "learning_rate": 7.107776691222803e-06,
+      "loss": 0.4947,
+      "step": 2217
+    },
+    {
+      "epoch": 2.128660585693711,
+      "grad_norm": 1.0281895426486987,
+      "learning_rate": 7.104738465511606e-06,
+      "loss": 0.6361,
+      "step": 2218
+    },
+    {
+      "epoch": 2.129620739318291,
+      "grad_norm": 1.268479130840763,
+      "learning_rate": 7.1016992950058104e-06,
+      "loss": 0.5749,
+      "step": 2219
+    },
+    {
+      "epoch": 2.130580892942871,
+      "grad_norm": 1.0817101074711388,
+      "learning_rate": 7.098659181069667e-06,
+      "loss": 0.5998,
+      "step": 2220
+    },
+    {
+      "epoch": 2.1315410465674507,
+      "grad_norm": 0.9419880894216737,
+      "learning_rate": 7.0956181250678514e-06,
+      "loss": 0.5361,
+      "step": 2221
+    },
+    {
+      "epoch": 2.132501200192031,
+      "grad_norm": 1.1560205008626543,
+      "learning_rate": 7.09257612836546e-06,
+      "loss": 0.5673,
+      "step": 2222
+    },
+    {
+      "epoch": 2.1334613538166107,
+      "grad_norm": 1.2057235149267305,
+      "learning_rate": 7.089533192328011e-06,
+      "loss": 0.4456,
+      "step": 2223
+    },
+    {
+      "epoch": 2.1344215074411905,
+      "grad_norm": 1.2052989405818355,
+      "learning_rate": 7.08648931832145e-06,
+      "loss": 0.6141,
+      "step": 2224
+    },
+    {
+      "epoch": 2.1353816610657703,
+      "grad_norm": 1.0478638695612665,
+      "learning_rate": 7.0834445077121374e-06,
+      "loss": 0.6443,
+      "step": 2225
+    },
+    {
+      "epoch": 2.1363418146903506,
+      "grad_norm": 1.0239613188518975,
+      "learning_rate": 7.080398761866856e-06,
+      "loss": 0.5032,
+      "step": 2226
+    },
+    {
+      "epoch": 2.1373019683149304,
+      "grad_norm": 1.018760226884325,
+      "learning_rate": 7.077352082152809e-06,
+      "loss": 0.607,
+      "step": 2227
+    },
+    {
+      "epoch": 2.13826212193951,
+      "grad_norm": 1.148025927195372,
+      "learning_rate": 7.0743044699376165e-06,
+      "loss": 0.5498,
+      "step": 2228
+    },
+    {
+      "epoch": 2.1392222755640904,
+      "grad_norm": 0.8200781721139206,
+      "learning_rate": 7.0712559265893214e-06,
+      "loss": 0.5648,
+      "step": 2229
+    },
+    {
+      "epoch": 2.14018242918867,
+      "grad_norm": 1.1389520733011873,
+      "learning_rate": 7.068206453476381e-06,
+      "loss": 0.5307,
+      "step": 2230
+    },
+    {
+      "epoch": 2.14114258281325,
+      "grad_norm": 0.9912886676294732,
+      "learning_rate": 7.065156051967672e-06,
+      "loss": 0.4595,
+      "step": 2231
+    },
+    {
+      "epoch": 2.1421027364378302,
+      "grad_norm": 1.458899560172927,
+      "learning_rate": 7.062104723432485e-06,
+      "loss": 0.6631,
+      "step": 2232
+    },
+    {
+      "epoch": 2.14306289006241,
+      "grad_norm": 0.8813816859162074,
+      "learning_rate": 7.059052469240533e-06,
+      "loss": 0.4875,
+      "step": 2233
+    },
+    {
+      "epoch": 2.14402304368699,
+      "grad_norm": 0.9422304069243908,
+      "learning_rate": 7.0559992907619346e-06,
+      "loss": 0.4818,
+      "step": 2234
+    },
+    {
+      "epoch": 2.14498319731157,
+      "grad_norm": 1.0433863270739594,
+      "learning_rate": 7.0529451893672356e-06,
+      "loss": 0.4883,
+      "step": 2235
+    },
+    {
+      "epoch": 2.14594335093615,
+      "grad_norm": 1.083463354491641,
+      "learning_rate": 7.0498901664273825e-06,
+      "loss": 0.5466,
+      "step": 2236
+    },
+    {
+      "epoch": 2.1469035045607296,
+      "grad_norm": 0.9168061171424753,
+      "learning_rate": 7.046834223313746e-06,
+      "loss": 0.416,
+      "step": 2237
+    },
+    {
+      "epoch": 2.1478636581853094,
+      "grad_norm": 1.0983573939844062,
+      "learning_rate": 7.043777361398108e-06,
+      "loss": 0.5345,
+      "step": 2238
+    },
+    {
+      "epoch": 2.1488238118098897,
+      "grad_norm": 1.1026006301523767,
+      "learning_rate": 7.040719582052659e-06,
+      "loss": 0.5849,
+      "step": 2239
+    },
+    {
+      "epoch": 2.1497839654344695,
+      "grad_norm": 1.1229041477293433,
+      "learning_rate": 7.037660886650003e-06,
+      "loss": 0.608,
+      "step": 2240
+    },
+    {
+      "epoch": 2.1507441190590493,
+      "grad_norm": 1.0282597731998386,
+      "learning_rate": 7.034601276563155e-06,
+      "loss": 0.5216,
+      "step": 2241
+    },
+    {
+      "epoch": 2.1517042726836295,
+      "grad_norm": 1.2262513647641144,
+      "learning_rate": 7.031540753165545e-06,
+      "loss": 0.5375,
+      "step": 2242
+    },
+    {
+      "epoch": 2.1526644263082093,
+      "grad_norm": 1.3657286880174617,
+      "learning_rate": 7.028479317831006e-06,
+      "loss": 0.5714,
+      "step": 2243
+    },
+    {
+      "epoch": 2.153624579932789,
+      "grad_norm": 1.0830473500612943,
+      "learning_rate": 7.0254169719337864e-06,
+      "loss": 0.552,
+      "step": 2244
+    },
+    {
+      "epoch": 2.1545847335573693,
+      "grad_norm": 1.4461561056588648,
+      "learning_rate": 7.022353716848535e-06,
+      "loss": 0.4827,
+      "step": 2245
+    },
+    {
+      "epoch": 2.155544887181949,
+      "grad_norm": 1.057096879420475,
+      "learning_rate": 7.0192895539503205e-06,
+      "loss": 0.4658,
+      "step": 2246
+    },
+    {
+      "epoch": 2.156505040806529,
+      "grad_norm": 1.3215393080024684,
+      "learning_rate": 7.016224484614609e-06,
+      "loss": 0.6065,
+      "step": 2247
+    },
+    {
+      "epoch": 2.157465194431109,
+      "grad_norm": 1.2754535852244573,
+      "learning_rate": 7.01315851021728e-06,
+      "loss": 0.4785,
+      "step": 2248
+    },
+    {
+      "epoch": 2.158425348055689,
+      "grad_norm": 0.8719319849964379,
+      "learning_rate": 7.010091632134612e-06,
+      "loss": 0.5544,
+      "step": 2249
+    },
+    {
+      "epoch": 2.1593855016802688,
+      "grad_norm": 1.2593155434928438,
+      "learning_rate": 7.0070238517433e-06,
+      "loss": 0.5334,
+      "step": 2250
+    },
+    {
+      "epoch": 2.160345655304849,
+      "grad_norm": 1.1330654046107738,
+      "learning_rate": 7.003955170420434e-06,
+      "loss": 0.5208,
+      "step": 2251
+    },
+    {
+      "epoch": 2.161305808929429,
+      "grad_norm": 1.2484710985903849,
+      "learning_rate": 7.000885589543513e-06,
+      "loss": 0.5937,
+      "step": 2252
+    },
+    {
+      "epoch": 2.1622659625540086,
+      "grad_norm": 0.9836315272716349,
+      "learning_rate": 6.99781511049044e-06,
+      "loss": 0.5836,
+      "step": 2253
+    },
+    {
+      "epoch": 2.1632261161785884,
+      "grad_norm": 0.8452358278737337,
+      "learning_rate": 6.994743734639517e-06,
+      "loss": 0.5031,
+      "step": 2254
+    },
+    {
+      "epoch": 2.1641862698031686,
+      "grad_norm": 1.1668919313717263,
+      "learning_rate": 6.991671463369457e-06,
+      "loss": 0.6261,
+      "step": 2255
+    },
+    {
+      "epoch": 2.1651464234277484,
+      "grad_norm": 0.9413696122477315,
+      "learning_rate": 6.988598298059366e-06,
+      "loss": 0.531,
+      "step": 2256
+    },
+    {
+      "epoch": 2.1661065770523282,
+      "grad_norm": 0.9420458908509529,
+      "learning_rate": 6.985524240088756e-06,
+      "loss": 0.4476,
+      "step": 2257
+    },
+    {
+      "epoch": 2.1670667306769085,
+      "grad_norm": 1.1758684457243866,
+      "learning_rate": 6.982449290837538e-06,
+      "loss": 0.517,
+      "step": 2258
+    },
+    {
+      "epoch": 2.1680268843014883,
+      "grad_norm": 1.7800386777188943,
+      "learning_rate": 6.979373451686025e-06,
+      "loss": 0.6417,
+      "step": 2259
+    },
+    {
+      "epoch": 2.168987037926068,
+      "grad_norm": 1.4337750025305862,
+      "learning_rate": 6.976296724014927e-06,
+      "loss": 0.5546,
+      "step": 2260
+    },
+    {
+      "epoch": 2.1699471915506483,
+      "grad_norm": 1.3786218860218389,
+      "learning_rate": 6.973219109205356e-06,
+      "loss": 0.5388,
+      "step": 2261
+    },
+    {
+      "epoch": 2.170907345175228,
+      "grad_norm": 1.0618258194646955,
+      "learning_rate": 6.970140608638818e-06,
+      "loss": 0.5076,
+      "step": 2262
+    },
+    {
+      "epoch": 2.171867498799808,
+      "grad_norm": 1.7338047296555672,
+      "learning_rate": 6.967061223697223e-06,
+      "loss": 0.4912,
+      "step": 2263
+    },
+    {
+      "epoch": 2.172827652424388,
+      "grad_norm": 0.8916519090513646,
+      "learning_rate": 6.963980955762869e-06,
+      "loss": 0.5516,
+      "step": 2264
+    },
+    {
+      "epoch": 2.173787806048968,
+      "grad_norm": 0.9422127017638235,
+      "learning_rate": 6.960899806218458e-06,
+      "loss": 0.4112,
+      "step": 2265
+    },
+    {
+      "epoch": 2.1747479596735477,
+      "grad_norm": 1.2627279572143573,
+      "learning_rate": 6.957817776447087e-06,
+      "loss": 0.6057,
+      "step": 2266
+    },
+    {
+      "epoch": 2.1757081132981275,
+      "grad_norm": 0.9280773130916784,
+      "learning_rate": 6.954734867832241e-06,
+      "loss": 0.5171,
+      "step": 2267
+    },
+    {
+      "epoch": 2.1766682669227078,
+      "grad_norm": 0.9905577683589822,
+      "learning_rate": 6.9516510817578105e-06,
+      "loss": 0.5717,
+      "step": 2268
+    },
+    {
+      "epoch": 2.1776284205472876,
+      "grad_norm": 1.1751454090366942,
+      "learning_rate": 6.948566419608069e-06,
+      "loss": 0.5614,
+      "step": 2269
+    },
+    {
+      "epoch": 2.1785885741718674,
+      "grad_norm": 1.2412026855284037,
+      "learning_rate": 6.945480882767694e-06,
+      "loss": 0.6331,
+      "step": 2270
+    },
+    {
+      "epoch": 2.1795487277964476,
+      "grad_norm": 0.9185685217686413,
+      "learning_rate": 6.942394472621744e-06,
+      "loss": 0.5177,
+      "step": 2271
+    },
+    {
+      "epoch": 2.1805088814210274,
+      "grad_norm": 1.434092355643127,
+      "learning_rate": 6.939307190555679e-06,
+      "loss": 0.6101,
+      "step": 2272
+    },
+    {
+      "epoch": 2.181469035045607,
+      "grad_norm": 1.1546664904412804,
+      "learning_rate": 6.936219037955347e-06,
+      "loss": 0.5022,
+      "step": 2273
+    },
+    {
+      "epoch": 2.1824291886701874,
+      "grad_norm": 1.3881411890858537,
+      "learning_rate": 6.9331300162069846e-06,
+      "loss": 0.6449,
+      "step": 2274
+    },
+    {
+      "epoch": 2.183389342294767,
+      "grad_norm": 1.4231626563775166,
+      "learning_rate": 6.930040126697224e-06,
+      "loss": 0.5352,
+      "step": 2275
+    },
+    {
+      "epoch": 2.184349495919347,
+      "grad_norm": 0.9643207536001666,
+      "learning_rate": 6.926949370813079e-06,
+      "loss": 0.5814,
+      "step": 2276
+    },
+    {
+      "epoch": 2.1853096495439273,
+      "grad_norm": 0.8894548405312629,
+      "learning_rate": 6.92385774994196e-06,
+      "loss": 0.5097,
+      "step": 2277
+    },
+    {
+      "epoch": 2.186269803168507,
+      "grad_norm": 1.3618572655327714,
+      "learning_rate": 6.920765265471663e-06,
+      "loss": 0.5298,
+      "step": 2278
+    },
+    {
+      "epoch": 2.187229956793087,
+      "grad_norm": 1.077037347090665,
+      "learning_rate": 6.91767191879037e-06,
+      "loss": 0.4841,
+      "step": 2279
+    },
+    {
+      "epoch": 2.1881901104176666,
+      "grad_norm": 1.187289731607155,
+      "learning_rate": 6.9145777112866495e-06,
+      "loss": 0.5638,
+      "step": 2280
+    },
+    {
+      "epoch": 2.189150264042247,
+      "grad_norm": 1.3509930182156265,
+      "learning_rate": 6.911482644349462e-06,
+      "loss": 0.5707,
+      "step": 2281
+    },
+    {
+      "epoch": 2.1901104176668267,
+      "grad_norm": 0.9079890206437792,
+      "learning_rate": 6.908386719368147e-06,
+      "loss": 0.4809,
+      "step": 2282
+    },
+    {
+      "epoch": 2.1910705712914065,
+      "grad_norm": 1.3158411640553533,
+      "learning_rate": 6.9052899377324345e-06,
+      "loss": 0.552,
+      "step": 2283
+    },
+    {
+      "epoch": 2.1920307249159867,
+      "grad_norm": 1.1945920307842048,
+      "learning_rate": 6.902192300832435e-06,
+      "loss": 0.5206,
+      "step": 2284
+    },
+    {
+      "epoch": 2.1929908785405665,
+      "grad_norm": 1.1161698542709764,
+      "learning_rate": 6.899093810058643e-06,
+      "loss": 0.5349,
+      "step": 2285
+    },
+    {
+      "epoch": 2.1939510321651463,
+      "grad_norm": 1.0129646662385823,
+      "learning_rate": 6.895994466801943e-06,
+      "loss": 0.4909,
+      "step": 2286
+    },
+    {
+      "epoch": 2.1949111857897265,
+      "grad_norm": 0.785720546066982,
+      "learning_rate": 6.892894272453592e-06,
+      "loss": 0.5562,
+      "step": 2287
+    },
+    {
+      "epoch": 2.1958713394143063,
+      "grad_norm": 1.176481379742249,
+      "learning_rate": 6.889793228405237e-06,
+      "loss": 0.487,
+      "step": 2288
+    },
+    {
+      "epoch": 2.196831493038886,
+      "grad_norm": 1.2659944494299649,
+      "learning_rate": 6.8866913360489005e-06,
+      "loss": 0.5603,
+      "step": 2289
+    },
+    {
+      "epoch": 2.1977916466634664,
+      "grad_norm": 0.891824396482085,
+      "learning_rate": 6.883588596776994e-06,
+      "loss": 0.4518,
+      "step": 2290
+    },
+    {
+      "epoch": 2.198751800288046,
+      "grad_norm": 0.9718378046907263,
+      "learning_rate": 6.8804850119822986e-06,
+      "loss": 0.5012,
+      "step": 2291
+    },
+    {
+      "epoch": 2.199711953912626,
+      "grad_norm": 1.1515253226218582,
+      "learning_rate": 6.8773805830579845e-06,
+      "loss": 0.6426,
+      "step": 2292
+    },
+    {
+      "epoch": 2.2006721075372058,
+      "grad_norm": 1.0808812915933304,
+      "learning_rate": 6.8742753113975925e-06,
+      "loss": 0.4578,
+      "step": 2293
+    },
+    {
+      "epoch": 2.201632261161786,
+      "grad_norm": 1.1587103593406312,
+      "learning_rate": 6.871169198395049e-06,
+      "loss": 0.5647,
+      "step": 2294
+    },
+    {
+      "epoch": 2.202592414786366,
+      "grad_norm": 1.2830368134483428,
+      "learning_rate": 6.868062245444655e-06,
+      "loss": 0.5116,
+      "step": 2295
+    },
+    {
+      "epoch": 2.2035525684109456,
+      "grad_norm": 0.9888273438380815,
+      "learning_rate": 6.864954453941086e-06,
+      "loss": 0.5891,
+      "step": 2296
+    },
+    {
+      "epoch": 2.204512722035526,
+      "grad_norm": 1.0910465773902722,
+      "learning_rate": 6.861845825279399e-06,
+      "loss": 0.5684,
+      "step": 2297
+    },
+    {
+      "epoch": 2.2054728756601056,
+      "grad_norm": 0.9836998262220984,
+      "learning_rate": 6.85873636085502e-06,
+      "loss": 0.6049,
+      "step": 2298
+    },
+    {
+      "epoch": 2.2064330292846854,
+      "grad_norm": 1.143683722460904,
+      "learning_rate": 6.855626062063761e-06,
+      "loss": 0.5415,
+      "step": 2299
+    },
+    {
+      "epoch": 2.2073931829092657,
+      "grad_norm": 1.3627628991871574,
+      "learning_rate": 6.852514930301797e-06,
+      "loss": 0.4894,
+      "step": 2300
+    },
+    {
+      "epoch": 2.2083533365338455,
+      "grad_norm": 0.9646215360960887,
+      "learning_rate": 6.8494029669656845e-06,
+      "loss": 0.5391,
+      "step": 2301
+    },
+    {
+      "epoch": 2.2093134901584253,
+      "grad_norm": 1.6614106845749312,
+      "learning_rate": 6.846290173452348e-06,
+      "loss": 0.5892,
+      "step": 2302
+    },
+    {
+      "epoch": 2.2102736437830055,
+      "grad_norm": 0.7681870339356103,
+      "learning_rate": 6.843176551159087e-06,
+      "loss": 0.551,
+      "step": 2303
+    },
+    {
+      "epoch": 2.2112337974075853,
+      "grad_norm": 1.688322695234949,
+      "learning_rate": 6.8400621014835775e-06,
+      "loss": 0.6805,
+      "step": 2304
+    },
+    {
+      "epoch": 2.212193951032165,
+      "grad_norm": 1.0924474130289121,
+      "learning_rate": 6.836946825823858e-06,
+      "loss": 0.3479,
+      "step": 2305
+    },
+    {
+      "epoch": 2.213154104656745,
+      "grad_norm": 1.0545213214412013,
+      "learning_rate": 6.833830725578346e-06,
+      "loss": 0.5024,
+      "step": 2306
+    },
+    {
+      "epoch": 2.214114258281325,
+      "grad_norm": 0.9771886643473012,
+      "learning_rate": 6.83071380214582e-06,
+      "loss": 0.5355,
+      "step": 2307
+    },
+    {
+      "epoch": 2.215074411905905,
+      "grad_norm": 1.3549067705037805,
+      "learning_rate": 6.827596056925441e-06,
+      "loss": 0.641,
+      "step": 2308
+    },
+    {
+      "epoch": 2.2160345655304847,
+      "grad_norm": 0.9694903981259282,
+      "learning_rate": 6.824477491316725e-06,
+      "loss": 0.6004,
+      "step": 2309
+    },
+    {
+      "epoch": 2.216994719155065,
+      "grad_norm": 0.9874739474800195,
+      "learning_rate": 6.821358106719569e-06,
+      "loss": 0.5204,
+      "step": 2310
+    },
+    {
+      "epoch": 2.2179548727796448,
+      "grad_norm": 1.510713052530886,
+      "learning_rate": 6.818237904534226e-06,
+      "loss": 0.5127,
+      "step": 2311
+    },
+    {
+      "epoch": 2.2189150264042246,
+      "grad_norm": 1.2595518403245647,
+      "learning_rate": 6.815116886161324e-06,
+      "loss": 0.6615,
+      "step": 2312
+    },
+    {
+      "epoch": 2.219875180028805,
+      "grad_norm": 1.2551386659866655,
+      "learning_rate": 6.811995053001854e-06,
+      "loss": 0.5522,
+      "step": 2313
+    },
+    {
+      "epoch": 2.2208353336533846,
+      "grad_norm": 1.1456539086225281,
+      "learning_rate": 6.808872406457175e-06,
+      "loss": 0.4721,
+      "step": 2314
+    },
+    {
+      "epoch": 2.2217954872779644,
+      "grad_norm": 0.8045166359873522,
+      "learning_rate": 6.8057489479290085e-06,
+      "loss": 0.5141,
+      "step": 2315
+    },
+    {
+      "epoch": 2.2227556409025446,
+      "grad_norm": 1.3517950602434772,
+      "learning_rate": 6.802624678819441e-06,
+      "loss": 0.6937,
+      "step": 2316
+    },
+    {
+      "epoch": 2.2237157945271244,
+      "grad_norm": 1.1857114818615182,
+      "learning_rate": 6.799499600530925e-06,
+      "loss": 0.56,
+      "step": 2317
+    },
+    {
+      "epoch": 2.224675948151704,
+      "grad_norm": 1.2652105327532868,
+      "learning_rate": 6.796373714466272e-06,
+      "loss": 0.4745,
+      "step": 2318
+    },
+    {
+      "epoch": 2.225636101776284,
+      "grad_norm": 1.0637397227409495,
+      "learning_rate": 6.793247022028663e-06,
+      "loss": 0.5657,
+      "step": 2319
+    },
+    {
+      "epoch": 2.2265962554008643,
+      "grad_norm": 1.2460238834022297,
+      "learning_rate": 6.790119524621632e-06,
+      "loss": 0.4977,
+      "step": 2320
+    },
+    {
+      "epoch": 2.227556409025444,
+      "grad_norm": 1.2740604894255427,
+      "learning_rate": 6.786991223649084e-06,
+      "loss": 0.5146,
+      "step": 2321
+    },
+    {
+      "epoch": 2.228516562650024,
+      "grad_norm": 0.926237153900135,
+      "learning_rate": 6.783862120515277e-06,
+      "loss": 0.5681,
+      "step": 2322
+    },
+    {
+      "epoch": 2.229476716274604,
+      "grad_norm": 0.9988330432333267,
+      "learning_rate": 6.780732216624833e-06,
+      "loss": 0.3972,
+      "step": 2323
+    },
+    {
+      "epoch": 2.230436869899184,
+      "grad_norm": 1.095885002537761,
+      "learning_rate": 6.77760151338273e-06,
+      "loss": 0.6618,
+      "step": 2324
+    },
+    {
+      "epoch": 2.2313970235237637,
+      "grad_norm": 0.955222234964677,
+      "learning_rate": 6.7744700121943095e-06,
+      "loss": 0.548,
+      "step": 2325
+    },
+    {
+      "epoch": 2.232357177148344,
+      "grad_norm": 1.1476624773336968,
+      "learning_rate": 6.77133771446527e-06,
+      "loss": 0.5159,
+      "step": 2326
+    },
+    {
+      "epoch": 2.2333173307729237,
+      "grad_norm": 0.8123273760209445,
+      "learning_rate": 6.768204621601665e-06,
+      "loss": 0.528,
+      "step": 2327
+    },
+    {
+      "epoch": 2.2342774843975035,
+      "grad_norm": 1.6314501191901696,
+      "learning_rate": 6.765070735009906e-06,
+      "loss": 0.6033,
+      "step": 2328
+    },
+    {
+      "epoch": 2.2352376380220838,
+      "grad_norm": 1.5105126997235994,
+      "learning_rate": 6.761936056096761e-06,
+      "loss": 0.5298,
+      "step": 2329
+    },
+    {
+      "epoch": 2.2361977916466635,
+      "grad_norm": 1.193808886296744,
+      "learning_rate": 6.758800586269355e-06,
+      "loss": 0.5199,
+      "step": 2330
+    },
+    {
+      "epoch": 2.2371579452712433,
+      "grad_norm": 1.0487225146728896,
+      "learning_rate": 6.755664326935167e-06,
+      "loss": 0.6136,
+      "step": 2331
+    },
+    {
+      "epoch": 2.238118098895823,
+      "grad_norm": 1.362824827408439,
+      "learning_rate": 6.752527279502029e-06,
+      "loss": 0.6011,
+      "step": 2332
+    },
+    {
+      "epoch": 2.2390782525204034,
+      "grad_norm": 0.9389073331635173,
+      "learning_rate": 6.749389445378127e-06,
+      "loss": 0.5052,
+      "step": 2333
+    },
+    {
+      "epoch": 2.240038406144983,
+      "grad_norm": 1.0526327734300798,
+      "learning_rate": 6.746250825972005e-06,
+      "loss": 0.5006,
+      "step": 2334
+    },
+    {
+      "epoch": 2.240998559769563,
+      "grad_norm": 1.1827824553700668,
+      "learning_rate": 6.743111422692553e-06,
+      "loss": 0.6073,
+      "step": 2335
+    },
+    {
+      "epoch": 2.241958713394143,
+      "grad_norm": 1.51578979380233,
+      "learning_rate": 6.739971236949017e-06,
+      "loss": 0.5195,
+      "step": 2336
+    },
+    {
+      "epoch": 2.242918867018723,
+      "grad_norm": 1.125227461717384,
+      "learning_rate": 6.736830270150992e-06,
+      "loss": 0.4413,
+      "step": 2337
+    },
+    {
+      "epoch": 2.243879020643303,
+      "grad_norm": 1.042673510287957,
+      "learning_rate": 6.733688523708422e-06,
+      "loss": 0.6148,
+      "step": 2338
+    },
+    {
+      "epoch": 2.244839174267883,
+      "grad_norm": 1.4466183633336012,
+      "learning_rate": 6.7305459990316094e-06,
+      "loss": 0.6003,
+      "step": 2339
+    },
+    {
+      "epoch": 2.245799327892463,
+      "grad_norm": 1.0357649916251936,
+      "learning_rate": 6.727402697531193e-06,
+      "loss": 0.5981,
+      "step": 2340
+    },
+    {
+      "epoch": 2.2467594815170426,
+      "grad_norm": 0.9929925479732605,
+      "learning_rate": 6.7242586206181735e-06,
+      "loss": 0.6478,
+      "step": 2341
+    },
+    {
+      "epoch": 2.247719635141623,
+      "grad_norm": 0.8004720229019504,
+      "learning_rate": 6.721113769703887e-06,
+      "loss": 0.5089,
+      "step": 2342
+    },
+    {
+      "epoch": 2.2486797887662027,
+      "grad_norm": 1.0056538203059768,
+      "learning_rate": 6.71796814620003e-06,
+      "loss": 0.5418,
+      "step": 2343
+    },
+    {
+      "epoch": 2.2496399423907825,
+      "grad_norm": 1.1260843247375578,
+      "learning_rate": 6.714821751518635e-06,
+      "loss": 0.5194,
+      "step": 2344
+    },
+    {
+      "epoch": 2.2506000960153623,
+      "grad_norm": 0.9516273323340717,
+      "learning_rate": 6.7116745870720875e-06,
+      "loss": 0.5228,
+      "step": 2345
+    },
+    {
+      "epoch": 2.2515602496399425,
+      "grad_norm": 1.456476808458998,
+      "learning_rate": 6.708526654273113e-06,
+      "loss": 0.6069,
+      "step": 2346
+    },
+    {
+      "epoch": 2.2525204032645223,
+      "grad_norm": 1.3072688806574018,
+      "learning_rate": 6.7053779545347874e-06,
+      "loss": 0.5497,
+      "step": 2347
+    },
+    {
+      "epoch": 2.253480556889102,
+      "grad_norm": 1.048250496505981,
+      "learning_rate": 6.702228489270528e-06,
+      "loss": 0.4946,
+      "step": 2348
+    },
+    {
+      "epoch": 2.2544407105136823,
+      "grad_norm": 1.0496690320472093,
+      "learning_rate": 6.699078259894094e-06,
+      "loss": 0.5388,
+      "step": 2349
+    },
+    {
+      "epoch": 2.255400864138262,
+      "grad_norm": 0.9688487220489557,
+      "learning_rate": 6.695927267819593e-06,
+      "loss": 0.5984,
+      "step": 2350
+    },
+    {
+      "epoch": 2.256361017762842,
+      "grad_norm": 1.2185141354022164,
+      "learning_rate": 6.692775514461468e-06,
+      "loss": 0.5056,
+      "step": 2351
+    },
+    {
+      "epoch": 2.257321171387422,
+      "grad_norm": 1.0614292358503847,
+      "learning_rate": 6.689623001234511e-06,
+      "loss": 0.571,
+      "step": 2352
+    },
+    {
+      "epoch": 2.258281325012002,
+      "grad_norm": 1.1296621276111845,
+      "learning_rate": 6.686469729553849e-06,
+      "loss": 0.5724,
+      "step": 2353
+    },
+    {
+      "epoch": 2.2592414786365818,
+      "grad_norm": 0.8582511322406426,
+      "learning_rate": 6.683315700834953e-06,
+      "loss": 0.4999,
+      "step": 2354
+    },
+    {
+      "epoch": 2.260201632261162,
+      "grad_norm": 0.8613287945064796,
+      "learning_rate": 6.680160916493631e-06,
+      "loss": 0.4628,
+      "step": 2355
+    },
+    {
+      "epoch": 2.261161785885742,
+      "grad_norm": 1.483542890527988,
+      "learning_rate": 6.677005377946037e-06,
+      "loss": 0.4807,
+      "step": 2356
+    },
+    {
+      "epoch": 2.2621219395103216,
+      "grad_norm": 1.0556765585600885,
+      "learning_rate": 6.6738490866086515e-06,
+      "loss": 0.63,
+      "step": 2357
+    },
+    {
+      "epoch": 2.2630820931349014,
+      "grad_norm": 1.0456996598851618,
+      "learning_rate": 6.670692043898306e-06,
+      "loss": 0.5289,
+      "step": 2358
+    },
+    {
+      "epoch": 2.2640422467594816,
+      "grad_norm": 1.3551152692604027,
+      "learning_rate": 6.66753425123216e-06,
+      "loss": 0.5914,
+      "step": 2359
+    },
+    {
+      "epoch": 2.2650024003840614,
+      "grad_norm": 0.9504059413020357,
+      "learning_rate": 6.664375710027713e-06,
+      "loss": 0.5807,
+      "step": 2360
+    },
+    {
+      "epoch": 2.265962554008641,
+      "grad_norm": 0.9733497390231736,
+      "learning_rate": 6.661216421702801e-06,
+      "loss": 0.5759,
+      "step": 2361
+    },
+    {
+      "epoch": 2.2669227076332215,
+      "grad_norm": 1.239651371353343,
+      "learning_rate": 6.6580563876755945e-06,
+      "loss": 0.5163,
+      "step": 2362
+    },
+    {
+      "epoch": 2.2678828612578013,
+      "grad_norm": 0.9195396850869483,
+      "learning_rate": 6.6548956093646e-06,
+      "loss": 0.5418,
+      "step": 2363
+    },
+    {
+      "epoch": 2.268843014882381,
+      "grad_norm": 1.1034369651300087,
+      "learning_rate": 6.651734088188655e-06,
+      "loss": 0.5811,
+      "step": 2364
+    },
+    {
+      "epoch": 2.2698031685069613,
+      "grad_norm": 1.3877249971980943,
+      "learning_rate": 6.648571825566936e-06,
+      "loss": 0.5501,
+      "step": 2365
+    },
+    {
+      "epoch": 2.270763322131541,
+      "grad_norm": 1.1901368203660332,
+      "learning_rate": 6.645408822918945e-06,
+      "loss": 0.5802,
+      "step": 2366
+    },
+    {
+      "epoch": 2.271723475756121,
+      "grad_norm": 1.7551834561249287,
+      "learning_rate": 6.642245081664524e-06,
+      "loss": 0.6671,
+      "step": 2367
+    },
+    {
+      "epoch": 2.272683629380701,
+      "grad_norm": 1.2420794293486392,
+      "learning_rate": 6.6390806032238375e-06,
+      "loss": 0.5046,
+      "step": 2368
+    },
+    {
+      "epoch": 2.273643783005281,
+      "grad_norm": 0.9272166116224568,
+      "learning_rate": 6.63591538901739e-06,
+      "loss": 0.5252,
+      "step": 2369
+    },
+    {
+      "epoch": 2.2746039366298607,
+      "grad_norm": 1.0236145872195985,
+      "learning_rate": 6.632749440466013e-06,
+      "loss": 0.6363,
+      "step": 2370
+    },
+    {
+      "epoch": 2.2755640902544405,
+      "grad_norm": 0.9268058857786272,
+      "learning_rate": 6.629582758990864e-06,
+      "loss": 0.558,
+      "step": 2371
+    },
+    {
+      "epoch": 2.2765242438790207,
+      "grad_norm": 0.8661224125161839,
+      "learning_rate": 6.626415346013434e-06,
+      "loss": 0.5373,
+      "step": 2372
+    },
+    {
+      "epoch": 2.2774843975036005,
+      "grad_norm": 1.1503416267820565,
+      "learning_rate": 6.623247202955538e-06,
+      "loss": 0.6417,
+      "step": 2373
+    },
+    {
+      "epoch": 2.2784445511281803,
+      "grad_norm": 0.9631774310954262,
+      "learning_rate": 6.620078331239328e-06,
+      "loss": 0.6008,
+      "step": 2374
+    },
+    {
+      "epoch": 2.2794047047527606,
+      "grad_norm": 1.0572988947020006,
+      "learning_rate": 6.61690873228727e-06,
+      "loss": 0.544,
+      "step": 2375
+    },
+    {
+      "epoch": 2.2803648583773404,
+      "grad_norm": 1.6109399609121728,
+      "learning_rate": 6.613738407522169e-06,
+      "loss": 0.6017,
+      "step": 2376
+    },
+    {
+      "epoch": 2.28132501200192,
+      "grad_norm": 0.9386008298486445,
+      "learning_rate": 6.610567358367143e-06,
+      "loss": 0.5316,
+      "step": 2377
+    },
+    {
+      "epoch": 2.2822851656265004,
+      "grad_norm": 1.2291257971075966,
+      "learning_rate": 6.607395586245648e-06,
+      "loss": 0.5553,
+      "step": 2378
+    },
+    {
+      "epoch": 2.28324531925108,
+      "grad_norm": 0.9880000748680476,
+      "learning_rate": 6.6042230925814565e-06,
+      "loss": 0.5706,
+      "step": 2379
+    },
+    {
+      "epoch": 2.28420547287566,
+      "grad_norm": 1.3882384375731867,
+      "learning_rate": 6.601049878798667e-06,
+      "loss": 0.6662,
+      "step": 2380
+    },
+    {
+      "epoch": 2.2851656265002402,
+      "grad_norm": 1.0024878661703003,
+      "learning_rate": 6.597875946321701e-06,
+      "loss": 0.5747,
+      "step": 2381
+    },
+    {
+      "epoch": 2.28612578012482,
+      "grad_norm": 0.9871022506826898,
+      "learning_rate": 6.594701296575303e-06,
+      "loss": 0.5533,
+      "step": 2382
+    },
+    {
+      "epoch": 2.2870859337494,
+      "grad_norm": 1.0405673527467023,
+      "learning_rate": 6.5915259309845404e-06,
+      "loss": 0.5858,
+      "step": 2383
+    },
+    {
+      "epoch": 2.2880460873739796,
+      "grad_norm": 1.1935101241961101,
+      "learning_rate": 6.588349850974798e-06,
+      "loss": 0.4727,
+      "step": 2384
+    },
+    {
+      "epoch": 2.28900624099856,
+      "grad_norm": 1.6937651826408868,
+      "learning_rate": 6.585173057971787e-06,
+      "loss": 0.6207,
+      "step": 2385
+    },
+    {
+      "epoch": 2.2899663946231397,
+      "grad_norm": 0.9273325222250683,
+      "learning_rate": 6.5819955534015335e-06,
+      "loss": 0.5046,
+      "step": 2386
+    },
+    {
+      "epoch": 2.2909265482477195,
+      "grad_norm": 1.0349028239541116,
+      "learning_rate": 6.578817338690389e-06,
+      "loss": 0.5695,
+      "step": 2387
+    },
+    {
+      "epoch": 2.2918867018722997,
+      "grad_norm": 1.0067078089408212,
+      "learning_rate": 6.575638415265016e-06,
+      "loss": 0.5692,
+      "step": 2388
+    },
+    {
+      "epoch": 2.2928468554968795,
+      "grad_norm": 1.4122462213753166,
+      "learning_rate": 6.5724587845524024e-06,
+      "loss": 0.648,
+      "step": 2389
+    },
+    {
+      "epoch": 2.2938070091214593,
+      "grad_norm": 0.9882915938991353,
+      "learning_rate": 6.569278447979848e-06,
+      "loss": 0.6179,
+      "step": 2390
+    },
+    {
+      "epoch": 2.2947671627460395,
+      "grad_norm": 1.3776070344364486,
+      "learning_rate": 6.566097406974972e-06,
+      "loss": 0.5681,
+      "step": 2391
+    },
+    {
+      "epoch": 2.2957273163706193,
+      "grad_norm": 1.0032261079729023,
+      "learning_rate": 6.562915662965713e-06,
+      "loss": 0.567,
+      "step": 2392
+    },
+    {
+      "epoch": 2.296687469995199,
+      "grad_norm": 0.837628254593809,
+      "learning_rate": 6.5597332173803175e-06,
+      "loss": 0.4813,
+      "step": 2393
+    },
+    {
+      "epoch": 2.2976476236197794,
+      "grad_norm": 0.9072512534340048,
+      "learning_rate": 6.556550071647355e-06,
+      "loss": 0.5525,
+      "step": 2394
+    },
+    {
+      "epoch": 2.298607777244359,
+      "grad_norm": 1.1699080465870773,
+      "learning_rate": 6.553366227195702e-06,
+      "loss": 0.5854,
+      "step": 2395
+    },
+    {
+      "epoch": 2.299567930868939,
+      "grad_norm": 0.9777031836326662,
+      "learning_rate": 6.550181685454556e-06,
+      "loss": 0.5771,
+      "step": 2396
+    },
+    {
+      "epoch": 2.3005280844935188,
+      "grad_norm": 0.8933149048160616,
+      "learning_rate": 6.546996447853421e-06,
+      "loss": 0.5101,
+      "step": 2397
+    },
+    {
+      "epoch": 2.301488238118099,
+      "grad_norm": 1.3469317902735265,
+      "learning_rate": 6.5438105158221175e-06,
+      "loss": 0.5216,
+      "step": 2398
+    },
+    {
+      "epoch": 2.302448391742679,
+      "grad_norm": 0.9440025069528666,
+      "learning_rate": 6.5406238907907745e-06,
+      "loss": 0.3806,
+      "step": 2399
+    },
+    {
+      "epoch": 2.3034085453672586,
+      "grad_norm": 1.1338584353480947,
+      "learning_rate": 6.537436574189837e-06,
+      "loss": 0.496,
+      "step": 2400
+    },
+    {
+      "epoch": 2.304368698991839,
+      "grad_norm": 0.8664060336273369,
+      "learning_rate": 6.5342485674500554e-06,
+      "loss": 0.5588,
+      "step": 2401
+    },
+    {
+      "epoch": 2.3053288526164186,
+      "grad_norm": 0.844386738936581,
+      "learning_rate": 6.531059872002492e-06,
+      "loss": 0.5238,
+      "step": 2402
+    },
+    {
+      "epoch": 2.3062890062409984,
+      "grad_norm": 0.97008817337292,
+      "learning_rate": 6.527870489278519e-06,
+      "loss": 0.4965,
+      "step": 2403
+    },
+    {
+      "epoch": 2.3072491598655787,
+      "grad_norm": 1.2113093411636833,
+      "learning_rate": 6.524680420709813e-06,
+      "loss": 0.5232,
+      "step": 2404
+    },
+    {
+      "epoch": 2.3082093134901585,
+      "grad_norm": 0.7933762415126152,
+      "learning_rate": 6.521489667728365e-06,
+      "loss": 0.5103,
+      "step": 2405
+    },
+    {
+      "epoch": 2.3091694671147383,
+      "grad_norm": 1.6040011514959773,
+      "learning_rate": 6.51829823176647e-06,
+      "loss": 0.4945,
+      "step": 2406
+    },
+    {
+      "epoch": 2.3101296207393185,
+      "grad_norm": 1.1872680952204704,
+      "learning_rate": 6.5151061142567285e-06,
+      "loss": 0.5818,
+      "step": 2407
+    },
+    {
+      "epoch": 2.3110897743638983,
+      "grad_norm": 1.1867275559034063,
+      "learning_rate": 6.511913316632046e-06,
+      "loss": 0.4382,
+      "step": 2408
+    },
+    {
+      "epoch": 2.312049927988478,
+      "grad_norm": 1.0241798518777094,
+      "learning_rate": 6.508719840325639e-06,
+      "loss": 0.5529,
+      "step": 2409
+    },
+    {
+      "epoch": 2.313010081613058,
+      "grad_norm": 0.932597269090835,
+      "learning_rate": 6.505525686771021e-06,
+      "loss": 0.4931,
+      "step": 2410
+    },
+    {
+      "epoch": 2.313970235237638,
+      "grad_norm": 2.0244549254616926,
+      "learning_rate": 6.502330857402018e-06,
+      "loss": 0.4747,
+      "step": 2411
+    },
+    {
+      "epoch": 2.314930388862218,
+      "grad_norm": 0.9489041896092459,
+      "learning_rate": 6.49913535365275e-06,
+      "loss": 0.5403,
+      "step": 2412
+    },
+    {
+      "epoch": 2.3158905424867977,
+      "grad_norm": 0.9132995165306347,
+      "learning_rate": 6.495939176957648e-06,
+      "loss": 0.524,
+      "step": 2413
+    },
+    {
+      "epoch": 2.316850696111378,
+      "grad_norm": 1.1110988605125405,
+      "learning_rate": 6.49274232875144e-06,
+      "loss": 0.4239,
+      "step": 2414
+    },
+    {
+      "epoch": 2.3178108497359577,
+      "grad_norm": 1.1042601415143272,
+      "learning_rate": 6.489544810469157e-06,
+      "loss": 0.4548,
+      "step": 2415
+    },
+    {
+      "epoch": 2.3187710033605375,
+      "grad_norm": 0.7788382029560041,
+      "learning_rate": 6.4863466235461316e-06,
+      "loss": 0.5014,
+      "step": 2416
+    },
+    {
+      "epoch": 2.319731156985118,
+      "grad_norm": 0.8844795604785618,
+      "learning_rate": 6.483147769417992e-06,
+      "loss": 0.5911,
+      "step": 2417
+    },
+    {
+      "epoch": 2.3206913106096976,
+      "grad_norm": 1.155745030904022,
+      "learning_rate": 6.479948249520676e-06,
+      "loss": 0.5965,
+      "step": 2418
+    },
+    {
+      "epoch": 2.3216514642342774,
+      "grad_norm": 0.8708622921604449,
+      "learning_rate": 6.476748065290409e-06,
+      "loss": 0.5494,
+      "step": 2419
+    },
+    {
+      "epoch": 2.3226116178588576,
+      "grad_norm": 1.0298916457963645,
+      "learning_rate": 6.473547218163723e-06,
+      "loss": 0.5817,
+      "step": 2420
+    },
+    {
+      "epoch": 2.3235717714834374,
+      "grad_norm": 0.9776650358768887,
+      "learning_rate": 6.4703457095774395e-06,
+      "loss": 0.5358,
+      "step": 2421
+    },
+    {
+      "epoch": 2.324531925108017,
+      "grad_norm": 1.080675144348416,
+      "learning_rate": 6.467143540968684e-06,
+      "loss": 0.4904,
+      "step": 2422
+    },
+    {
+      "epoch": 2.325492078732597,
+      "grad_norm": 1.230767202332983,
+      "learning_rate": 6.4639407137748774e-06,
+      "loss": 0.5198,
+      "step": 2423
+    },
+    {
+      "epoch": 2.3264522323571772,
+      "grad_norm": 1.030541708041631,
+      "learning_rate": 6.4607372294337335e-06,
+      "loss": 0.4811,
+      "step": 2424
+    },
+    {
+      "epoch": 2.327412385981757,
+      "grad_norm": 1.0798731215135828,
+      "learning_rate": 6.457533089383261e-06,
+      "loss": 0.6193,
+      "step": 2425
+    },
+    {
+      "epoch": 2.328372539606337,
+      "grad_norm": 1.1990530071734347,
+      "learning_rate": 6.454328295061764e-06,
+      "loss": 0.5993,
+      "step": 2426
+    },
+    {
+      "epoch": 2.329332693230917,
+      "grad_norm": 1.0319319277157921,
+      "learning_rate": 6.451122847907843e-06,
+      "loss": 0.5686,
+      "step": 2427
+    },
+    {
+      "epoch": 2.330292846855497,
+      "grad_norm": 0.929150769455459,
+      "learning_rate": 6.447916749360386e-06,
+      "loss": 0.5257,
+      "step": 2428
+    },
+    {
+      "epoch": 2.3312530004800767,
+      "grad_norm": 0.8378201535255333,
+      "learning_rate": 6.444710000858578e-06,
+      "loss": 0.6194,
+      "step": 2429
+    },
+    {
+      "epoch": 2.332213154104657,
+      "grad_norm": 0.9679996392554938,
+      "learning_rate": 6.441502603841892e-06,
+      "loss": 0.5243,
+      "step": 2430
+    },
+    {
+      "epoch": 2.3331733077292367,
+      "grad_norm": 1.132744381814498,
+      "learning_rate": 6.438294559750098e-06,
+      "loss": 0.5697,
+      "step": 2431
+    },
+    {
+      "epoch": 2.3341334613538165,
+      "grad_norm": 0.9314276331579016,
+      "learning_rate": 6.43508587002325e-06,
+      "loss": 0.4477,
+      "step": 2432
+    },
+    {
+      "epoch": 2.3350936149783967,
+      "grad_norm": 1.5226059113858328,
+      "learning_rate": 6.431876536101695e-06,
+      "loss": 0.6278,
+      "step": 2433
+    },
+    {
+      "epoch": 2.3360537686029765,
+      "grad_norm": 1.2881405764305072,
+      "learning_rate": 6.42866655942607e-06,
+      "loss": 0.5379,
+      "step": 2434
+    },
+    {
+      "epoch": 2.3370139222275563,
+      "grad_norm": 1.12727054271157,
+      "learning_rate": 6.425455941437297e-06,
+      "loss": 0.5785,
+      "step": 2435
+    },
+    {
+      "epoch": 2.337974075852136,
+      "grad_norm": 1.723016354144707,
+      "learning_rate": 6.422244683576589e-06,
+      "loss": 0.4627,
+      "step": 2436
+    },
+    {
+      "epoch": 2.3389342294767164,
+      "grad_norm": 1.142417894256119,
+      "learning_rate": 6.419032787285446e-06,
+      "loss": 0.5908,
+      "step": 2437
+    },
+    {
+      "epoch": 2.339894383101296,
+      "grad_norm": 1.8797402056376855,
+      "learning_rate": 6.4158202540056536e-06,
+      "loss": 0.6451,
+      "step": 2438
+    },
+    {
+      "epoch": 2.340854536725876,
+      "grad_norm": 1.2870211657494477,
+      "learning_rate": 6.412607085179282e-06,
+      "loss": 0.5784,
+      "step": 2439
+    },
+    {
+      "epoch": 2.341814690350456,
+      "grad_norm": 1.159389786483229,
+      "learning_rate": 6.409393282248692e-06,
+      "loss": 0.5527,
+      "step": 2440
+    },
+    {
+      "epoch": 2.342774843975036,
+      "grad_norm": 1.0038200173284817,
+      "learning_rate": 6.406178846656521e-06,
+      "loss": 0.4619,
+      "step": 2441
+    },
+    {
+      "epoch": 2.343734997599616,
+      "grad_norm": 1.2176155016950183,
+      "learning_rate": 6.402963779845699e-06,
+      "loss": 0.5757,
+      "step": 2442
+    },
+    {
+      "epoch": 2.344695151224196,
+      "grad_norm": 1.442166780259645,
+      "learning_rate": 6.399748083259432e-06,
+      "loss": 0.6336,
+      "step": 2443
+    },
+    {
+      "epoch": 2.345655304848776,
+      "grad_norm": 0.9343353686932061,
+      "learning_rate": 6.3965317583412135e-06,
+      "loss": 0.4759,
+      "step": 2444
+    },
+    {
+      "epoch": 2.3466154584733556,
+      "grad_norm": 1.2702910517169863,
+      "learning_rate": 6.393314806534815e-06,
+      "loss": 0.5686,
+      "step": 2445
+    },
+    {
+      "epoch": 2.347575612097936,
+      "grad_norm": 1.2536723128102305,
+      "learning_rate": 6.390097229284296e-06,
+      "loss": 0.4995,
+      "step": 2446
+    },
+    {
+      "epoch": 2.3485357657225157,
+      "grad_norm": 1.0470844232843697,
+      "learning_rate": 6.386879028033988e-06,
+      "loss": 0.5053,
+      "step": 2447
+    },
+    {
+      "epoch": 2.3494959193470955,
+      "grad_norm": 1.1722093956745143,
+      "learning_rate": 6.3836602042285105e-06,
+      "loss": 0.4668,
+      "step": 2448
+    },
+    {
+      "epoch": 2.3504560729716752,
+      "grad_norm": 0.944788504592577,
+      "learning_rate": 6.3804407593127585e-06,
+      "loss": 0.5594,
+      "step": 2449
+    },
+    {
+      "epoch": 2.3514162265962555,
+      "grad_norm": 1.090347727493859,
+      "learning_rate": 6.377220694731904e-06,
+      "loss": 0.5469,
+      "step": 2450
+    },
+    {
+      "epoch": 2.3523763802208353,
+      "grad_norm": 1.1260538228003005,
+      "learning_rate": 6.374000011931403e-06,
+      "loss": 0.5433,
+      "step": 2451
+    },
+    {
+      "epoch": 2.353336533845415,
+      "grad_norm": 1.0542951287411089,
+      "learning_rate": 6.370778712356981e-06,
+      "loss": 0.4382,
+      "step": 2452
+    },
+    {
+      "epoch": 2.3542966874699953,
+      "grad_norm": 0.9019927253450573,
+      "learning_rate": 6.3675567974546506e-06,
+      "loss": 0.546,
+      "step": 2453
+    },
+    {
+      "epoch": 2.355256841094575,
+      "grad_norm": 1.4738986979211055,
+      "learning_rate": 6.364334268670691e-06,
+      "loss": 0.6841,
+      "step": 2454
+    },
+    {
+      "epoch": 2.356216994719155,
+      "grad_norm": 1.3022796833262742,
+      "learning_rate": 6.361111127451664e-06,
+      "loss": 0.5596,
+      "step": 2455
+    },
+    {
+      "epoch": 2.357177148343735,
+      "grad_norm": 1.0840341756830607,
+      "learning_rate": 6.357887375244398e-06,
+      "loss": 0.4968,
+      "step": 2456
+    },
+    {
+      "epoch": 2.358137301968315,
+      "grad_norm": 1.2951952187839153,
+      "learning_rate": 6.354663013496006e-06,
+      "loss": 0.4638,
+      "step": 2457
+    },
+    {
+      "epoch": 2.3590974555928947,
+      "grad_norm": 1.1111086015730918,
+      "learning_rate": 6.3514380436538695e-06,
+      "loss": 0.5432,
+      "step": 2458
+    },
+    {
+      "epoch": 2.360057609217475,
+      "grad_norm": 1.188571308975247,
+      "learning_rate": 6.348212467165639e-06,
+      "loss": 0.4349,
+      "step": 2459
+    },
+    {
+      "epoch": 2.361017762842055,
+      "grad_norm": 0.9545985278672484,
+      "learning_rate": 6.344986285479244e-06,
+      "loss": 0.5683,
+      "step": 2460
+    },
+    {
+      "epoch": 2.3619779164666346,
+      "grad_norm": 1.6278258187837187,
+      "learning_rate": 6.341759500042882e-06,
+      "loss": 0.6329,
+      "step": 2461
+    },
+    {
+      "epoch": 2.3629380700912144,
+      "grad_norm": 1.022302684447469,
+      "learning_rate": 6.338532112305025e-06,
+      "loss": 0.4961,
+      "step": 2462
+    },
+    {
+      "epoch": 2.3638982237157946,
+      "grad_norm": 1.107779000074782,
+      "learning_rate": 6.3353041237144096e-06,
+      "loss": 0.6454,
+      "step": 2463
+    },
+    {
+      "epoch": 2.3648583773403744,
+      "grad_norm": 1.1517936843117076,
+      "learning_rate": 6.332075535720047e-06,
+      "loss": 0.5464,
+      "step": 2464
+    },
+    {
+      "epoch": 2.365818530964954,
+      "grad_norm": 1.0677799092338685,
+      "learning_rate": 6.328846349771215e-06,
+      "loss": 0.4648,
+      "step": 2465
+    },
+    {
+      "epoch": 2.3667786845895344,
+      "grad_norm": 1.6285500755462932,
+      "learning_rate": 6.325616567317461e-06,
+      "loss": 0.5713,
+      "step": 2466
+    },
+    {
+      "epoch": 2.3677388382141142,
+      "grad_norm": 1.136262753647289,
+      "learning_rate": 6.3223861898086e-06,
+      "loss": 0.6101,
+      "step": 2467
+    },
+    {
+      "epoch": 2.368698991838694,
+      "grad_norm": 0.90489576314916,
+      "learning_rate": 6.319155218694716e-06,
+      "loss": 0.5386,
+      "step": 2468
+    },
+    {
+      "epoch": 2.3696591454632743,
+      "grad_norm": 1.878512432031467,
+      "learning_rate": 6.315923655426152e-06,
+      "loss": 0.6031,
+      "step": 2469
+    },
+    {
+      "epoch": 2.370619299087854,
+      "grad_norm": 1.1158148731899769,
+      "learning_rate": 6.312691501453527e-06,
+      "loss": 0.462,
+      "step": 2470
+    },
+    {
+      "epoch": 2.371579452712434,
+      "grad_norm": 0.9969182900346529,
+      "learning_rate": 6.30945875822772e-06,
+      "loss": 0.6176,
+      "step": 2471
+    },
+    {
+      "epoch": 2.372539606337014,
+      "grad_norm": 0.8883371107710113,
+      "learning_rate": 6.306225427199873e-06,
+      "loss": 0.5047,
+      "step": 2472
+    },
+    {
+      "epoch": 2.373499759961594,
+      "grad_norm": 1.011686097060432,
+      "learning_rate": 6.302991509821396e-06,
+      "loss": 0.5202,
+      "step": 2473
+    },
+    {
+      "epoch": 2.3744599135861737,
+      "grad_norm": 0.9917933465194929,
+      "learning_rate": 6.299757007543954e-06,
+      "loss": 0.5423,
+      "step": 2474
+    },
+    {
+      "epoch": 2.3754200672107535,
+      "grad_norm": 0.867445761455369,
+      "learning_rate": 6.296521921819489e-06,
+      "loss": 0.4968,
+      "step": 2475
+    },
+    {
+      "epoch": 2.3763802208353337,
+      "grad_norm": 1.226080037659591,
+      "learning_rate": 6.29328625410019e-06,
+      "loss": 0.6016,
+      "step": 2476
+    },
+    {
+      "epoch": 2.3773403744599135,
+      "grad_norm": 0.8714179357262549,
+      "learning_rate": 6.290050005838519e-06,
+      "loss": 0.5684,
+      "step": 2477
+    },
+    {
+      "epoch": 2.3783005280844933,
+      "grad_norm": 0.9008505419061984,
+      "learning_rate": 6.286813178487188e-06,
+      "loss": 0.6102,
+      "step": 2478
+    },
+    {
+      "epoch": 2.3792606817090736,
+      "grad_norm": 1.1543272859182996,
+      "learning_rate": 6.283575773499176e-06,
+      "loss": 0.4259,
+      "step": 2479
+    },
+    {
+      "epoch": 2.3802208353336534,
+      "grad_norm": 1.250841848975828,
+      "learning_rate": 6.280337792327722e-06,
+      "loss": 0.5412,
+      "step": 2480
+    },
+    {
+      "epoch": 2.381180988958233,
+      "grad_norm": 1.1754961010375546,
+      "learning_rate": 6.277099236426317e-06,
+      "loss": 0.4733,
+      "step": 2481
+    },
+    {
+      "epoch": 2.3821411425828134,
+      "grad_norm": 1.5148238383477088,
+      "learning_rate": 6.273860107248718e-06,
+      "loss": 0.5905,
+      "step": 2482
+    },
+    {
+      "epoch": 2.383101296207393,
+      "grad_norm": 1.1762920142876214,
+      "learning_rate": 6.270620406248931e-06,
+      "loss": 0.5957,
+      "step": 2483
+    },
+    {
+      "epoch": 2.384061449831973,
+      "grad_norm": 1.049323319918193,
+      "learning_rate": 6.267380134881229e-06,
+      "loss": 0.603,
+      "step": 2484
+    },
+    {
+      "epoch": 2.3850216034565532,
+      "grad_norm": 0.9600595231048848,
+      "learning_rate": 6.26413929460013e-06,
+      "loss": 0.5481,
+      "step": 2485
+    },
+    {
+      "epoch": 2.385981757081133,
+      "grad_norm": 0.8949485585188345,
+      "learning_rate": 6.260897886860415e-06,
+      "loss": 0.5511,
+      "step": 2486
+    },
+    {
+      "epoch": 2.386941910705713,
+      "grad_norm": 0.8341641848405127,
+      "learning_rate": 6.2576559131171174e-06,
+      "loss": 0.561,
+      "step": 2487
+    },
+    {
+      "epoch": 2.3879020643302926,
+      "grad_norm": 0.8356952968329642,
+      "learning_rate": 6.254413374825523e-06,
+      "loss": 0.512,
+      "step": 2488
+    },
+    {
+      "epoch": 2.388862217954873,
+      "grad_norm": 1.2912356680160988,
+      "learning_rate": 6.251170273441174e-06,
+      "loss": 0.501,
+      "step": 2489
+    },
+    {
+      "epoch": 2.3898223715794527,
+      "grad_norm": 1.3687753415351693,
+      "learning_rate": 6.247926610419865e-06,
+      "loss": 0.5907,
+      "step": 2490
+    },
+    {
+      "epoch": 2.390782525204033,
+      "grad_norm": 1.741973792817936,
+      "learning_rate": 6.244682387217638e-06,
+      "loss": 0.5909,
+      "step": 2491
+    },
+    {
+      "epoch": 2.3917426788286127,
+      "grad_norm": 1.733414899449277,
+      "learning_rate": 6.241437605290792e-06,
+      "loss": 0.638,
+      "step": 2492
+    },
+    {
+      "epoch": 2.3927028324531925,
+      "grad_norm": 1.098017920644056,
+      "learning_rate": 6.238192266095877e-06,
+      "loss": 0.5966,
+      "step": 2493
+    },
+    {
+      "epoch": 2.3936629860777723,
+      "grad_norm": 1.858505879300705,
+      "learning_rate": 6.234946371089688e-06,
+      "loss": 0.5663,
+      "step": 2494
+    },
+    {
+      "epoch": 2.3946231397023525,
+      "grad_norm": 0.9586710525031326,
+      "learning_rate": 6.231699921729275e-06,
+      "loss": 0.5443,
+      "step": 2495
+    },
+    {
+      "epoch": 2.3955832933269323,
+      "grad_norm": 1.2256198791441315,
+      "learning_rate": 6.228452919471933e-06,
+      "loss": 0.62,
+      "step": 2496
+    },
+    {
+      "epoch": 2.396543446951512,
+      "grad_norm": 0.9852276786729444,
+      "learning_rate": 6.225205365775206e-06,
+      "loss": 0.4704,
+      "step": 2497
+    },
+    {
+      "epoch": 2.3975036005760924,
+      "grad_norm": 0.8635032986917627,
+      "learning_rate": 6.221957262096887e-06,
+      "loss": 0.5254,
+      "step": 2498
+    },
+    {
+      "epoch": 2.398463754200672,
+      "grad_norm": 1.1244557413706344,
+      "learning_rate": 6.218708609895017e-06,
+      "loss": 0.6046,
+      "step": 2499
+    },
+    {
+      "epoch": 2.399423907825252,
+      "grad_norm": 1.146043243064041,
+      "learning_rate": 6.2154594106278775e-06,
+      "loss": 0.5711,
+      "step": 2500
+    },
+    {
+      "epoch": 2.4003840614498317,
+      "grad_norm": 0.9653797913238615,
+      "learning_rate": 6.212209665754001e-06,
+      "loss": 0.54,
+      "step": 2501
+    },
+    {
+      "epoch": 2.401344215074412,
+      "grad_norm": 0.9576143723527571,
+      "learning_rate": 6.2089593767321645e-06,
+      "loss": 0.4741,
+      "step": 2502
+    },
+    {
+      "epoch": 2.4023043686989918,
+      "grad_norm": 1.1649218757253375,
+      "learning_rate": 6.2057085450213864e-06,
+      "loss": 0.5723,
+      "step": 2503
+    },
+    {
+      "epoch": 2.403264522323572,
+      "grad_norm": 0.9175682966509715,
+      "learning_rate": 6.202457172080931e-06,
+      "loss": 0.6017,
+      "step": 2504
+    },
+    {
+      "epoch": 2.404224675948152,
+      "grad_norm": 1.2241928486652216,
+      "learning_rate": 6.199205259370304e-06,
+      "loss": 0.5167,
+      "step": 2505
+    },
+    {
+      "epoch": 2.4051848295727316,
+      "grad_norm": 0.93584799869271,
+      "learning_rate": 6.195952808349255e-06,
+      "loss": 0.4923,
+      "step": 2506
+    },
+    {
+      "epoch": 2.4061449831973114,
+      "grad_norm": 0.7834795850296479,
+      "learning_rate": 6.192699820477775e-06,
+      "loss": 0.4509,
+      "step": 2507
+    },
+    {
+      "epoch": 2.4071051368218916,
+      "grad_norm": 1.9248511223100664,
+      "learning_rate": 6.189446297216094e-06,
+      "loss": 0.7094,
+      "step": 2508
+    },
+    {
+      "epoch": 2.4080652904464714,
+      "grad_norm": 1.055263058174764,
+      "learning_rate": 6.186192240024684e-06,
+      "loss": 0.5797,
+      "step": 2509
+    },
+    {
+      "epoch": 2.4090254440710512,
+      "grad_norm": 1.451596012988708,
+      "learning_rate": 6.182937650364258e-06,
+      "loss": 0.5341,
+      "step": 2510
+    },
+    {
+      "epoch": 2.4099855976956315,
+      "grad_norm": 1.0779510051470163,
+      "learning_rate": 6.179682529695766e-06,
+      "loss": 0.5495,
+      "step": 2511
+    },
+    {
+      "epoch": 2.4109457513202113,
+      "grad_norm": 1.0889028245993762,
+      "learning_rate": 6.176426879480396e-06,
+      "loss": 0.5927,
+      "step": 2512
+    },
+    {
+      "epoch": 2.411905904944791,
+      "grad_norm": 1.4630636365034198,
+      "learning_rate": 6.173170701179576e-06,
+      "loss": 0.6391,
+      "step": 2513
+    },
+    {
+      "epoch": 2.412866058569371,
+      "grad_norm": 1.110996876012569,
+      "learning_rate": 6.169913996254967e-06,
+      "loss": 0.4694,
+      "step": 2514
+    },
+    {
+      "epoch": 2.413826212193951,
+      "grad_norm": 1.3124809791345493,
+      "learning_rate": 6.166656766168473e-06,
+      "loss": 0.5996,
+      "step": 2515
+    },
+    {
+      "epoch": 2.414786365818531,
+      "grad_norm": 0.9981468821518614,
+      "learning_rate": 6.1633990123822265e-06,
+      "loss": 0.4888,
+      "step": 2516
+    },
+    {
+      "epoch": 2.415746519443111,
+      "grad_norm": 0.8868689938749357,
+      "learning_rate": 6.1601407363586e-06,
+      "loss": 0.4886,
+      "step": 2517
+    },
+    {
+      "epoch": 2.416706673067691,
+      "grad_norm": 0.9276247128822482,
+      "learning_rate": 6.156881939560196e-06,
+      "loss": 0.479,
+      "step": 2518
+    },
+    {
+      "epoch": 2.4176668266922707,
+      "grad_norm": 1.4470991332967298,
+      "learning_rate": 6.153622623449858e-06,
+      "loss": 0.5096,
+      "step": 2519
+    },
+    {
+      "epoch": 2.4186269803168505,
+      "grad_norm": 1.3237570866236188,
+      "learning_rate": 6.150362789490654e-06,
+      "loss": 0.473,
+      "step": 2520
+    },
+    {
+      "epoch": 2.4195871339414308,
+      "grad_norm": 1.7682334211933788,
+      "learning_rate": 6.147102439145892e-06,
+      "loss": 0.6328,
+      "step": 2521
+    },
+    {
+      "epoch": 2.4205472875660106,
+      "grad_norm": 0.7815698215631511,
+      "learning_rate": 6.143841573879107e-06,
+      "loss": 0.4841,
+      "step": 2522
+    },
+    {
+      "epoch": 2.4215074411905904,
+      "grad_norm": 1.19672195307083,
+      "learning_rate": 6.140580195154065e-06,
+      "loss": 0.5615,
+      "step": 2523
+    },
+    {
+      "epoch": 2.4224675948151706,
+      "grad_norm": 0.9681133678481987,
+      "learning_rate": 6.137318304434768e-06,
+      "loss": 0.5036,
+      "step": 2524
+    },
+    {
+      "epoch": 2.4234277484397504,
+      "grad_norm": 1.0484371141077236,
+      "learning_rate": 6.1340559031854385e-06,
+      "loss": 0.6357,
+      "step": 2525
+    },
+    {
+      "epoch": 2.42438790206433,
+      "grad_norm": 0.9071284605464971,
+      "learning_rate": 6.130792992870539e-06,
+      "loss": 0.6447,
+      "step": 2526
+    },
+    {
+      "epoch": 2.42534805568891,
+      "grad_norm": 1.4903328940675329,
+      "learning_rate": 6.127529574954749e-06,
+      "loss": 0.5664,
+      "step": 2527
+    },
+    {
+      "epoch": 2.4263082093134902,
+      "grad_norm": 0.7112508735080384,
+      "learning_rate": 6.124265650902989e-06,
+      "loss": 0.3958,
+      "step": 2528
+    },
+    {
+      "epoch": 2.42726836293807,
+      "grad_norm": 0.876598457400479,
+      "learning_rate": 6.121001222180394e-06,
+      "loss": 0.5359,
+      "step": 2529
+    },
+    {
+      "epoch": 2.4282285165626503,
+      "grad_norm": 1.838248783503213,
+      "learning_rate": 6.117736290252335e-06,
+      "loss": 0.4881,
+      "step": 2530
+    },
+    {
+      "epoch": 2.42918867018723,
+      "grad_norm": 0.8244178851212894,
+      "learning_rate": 6.1144708565844026e-06,
+      "loss": 0.5553,
+      "step": 2531
+    },
+    {
+      "epoch": 2.43014882381181,
+      "grad_norm": 1.370833039667388,
+      "learning_rate": 6.111204922642414e-06,
+      "loss": 0.5652,
+      "step": 2532
+    },
+    {
+      "epoch": 2.4311089774363897,
+      "grad_norm": 1.0224639269077314,
+      "learning_rate": 6.107938489892417e-06,
+      "loss": 0.6292,
+      "step": 2533
+    },
+    {
+      "epoch": 2.43206913106097,
+      "grad_norm": 1.4056638544422844,
+      "learning_rate": 6.104671559800672e-06,
+      "loss": 0.6494,
+      "step": 2534
+    },
+    {
+      "epoch": 2.4330292846855497,
+      "grad_norm": 1.279030589080096,
+      "learning_rate": 6.101404133833674e-06,
+      "loss": 0.5567,
+      "step": 2535
+    },
+    {
+      "epoch": 2.4339894383101295,
+      "grad_norm": 0.965369013712244,
+      "learning_rate": 6.098136213458131e-06,
+      "loss": 0.5479,
+      "step": 2536
+    },
+    {
+      "epoch": 2.4349495919347097,
+      "grad_norm": 0.811936788515548,
+      "learning_rate": 6.0948678001409835e-06,
+      "loss": 0.4838,
+      "step": 2537
+    },
+    {
+      "epoch": 2.4359097455592895,
+      "grad_norm": 1.4508632995724344,
+      "learning_rate": 6.091598895349382e-06,
+      "loss": 0.6178,
+      "step": 2538
+    },
+    {
+      "epoch": 2.4368698991838693,
+      "grad_norm": 1.315776079609651,
+      "learning_rate": 6.088329500550706e-06,
+      "loss": 0.551,
+      "step": 2539
+    },
+    {
+      "epoch": 2.437830052808449,
+      "grad_norm": 1.0309376843616966,
+      "learning_rate": 6.08505961721255e-06,
+      "loss": 0.5907,
+      "step": 2540
+    },
+    {
+      "epoch": 2.4387902064330294,
+      "grad_norm": 0.9774921339196672,
+      "learning_rate": 6.081789246802731e-06,
+      "loss": 0.5253,
+      "step": 2541
+    },
+    {
+      "epoch": 2.439750360057609,
+      "grad_norm": 1.8226000020033482,
+      "learning_rate": 6.078518390789282e-06,
+      "loss": 0.4373,
+      "step": 2542
+    },
+    {
+      "epoch": 2.4407105136821894,
+      "grad_norm": 1.0582251492749066,
+      "learning_rate": 6.075247050640458e-06,
+      "loss": 0.5234,
+      "step": 2543
+    },
+    {
+      "epoch": 2.441670667306769,
+      "grad_norm": 1.7260864444229784,
+      "learning_rate": 6.071975227824724e-06,
+      "loss": 0.6115,
+      "step": 2544
+    },
+    {
+      "epoch": 2.442630820931349,
+      "grad_norm": 1.7230141368781013,
+      "learning_rate": 6.068702923810771e-06,
+      "loss": 0.547,
+      "step": 2545
+    },
+    {
+      "epoch": 2.4435909745559288,
+      "grad_norm": 1.0337078967488165,
+      "learning_rate": 6.0654301400675e-06,
+      "loss": 0.5322,
+      "step": 2546
+    },
+    {
+      "epoch": 2.444551128180509,
+      "grad_norm": 1.077894068675427,
+      "learning_rate": 6.062156878064026e-06,
+      "loss": 0.5091,
+      "step": 2547
+    },
+    {
+      "epoch": 2.445511281805089,
+      "grad_norm": 1.3279128101668733,
+      "learning_rate": 6.058883139269685e-06,
+      "loss": 0.5047,
+      "step": 2548
+    },
+    {
+      "epoch": 2.4464714354296686,
+      "grad_norm": 2.147066360747897,
+      "learning_rate": 6.055608925154019e-06,
+      "loss": 0.4498,
+      "step": 2549
+    },
+    {
+      "epoch": 2.447431589054249,
+      "grad_norm": 1.3374049080761576,
+      "learning_rate": 6.052334237186792e-06,
+      "loss": 0.5532,
+      "step": 2550
+    },
+    {
+      "epoch": 2.4483917426788286,
+      "grad_norm": 1.0663473328404744,
+      "learning_rate": 6.049059076837974e-06,
+      "loss": 0.7161,
+      "step": 2551
+    },
+    {
+      "epoch": 2.4493518963034084,
+      "grad_norm": 1.0044549382114532,
+      "learning_rate": 6.045783445577752e-06,
+      "loss": 0.522,
+      "step": 2552
+    },
+    {
+      "epoch": 2.4503120499279882,
+      "grad_norm": 1.3742670890197568,
+      "learning_rate": 6.042507344876518e-06,
+      "loss": 0.5938,
+      "step": 2553
+    },
+    {
+      "epoch": 2.4512722035525685,
+      "grad_norm": 0.910650313970339,
+      "learning_rate": 6.0392307762048816e-06,
+      "loss": 0.5379,
+      "step": 2554
+    },
+    {
+      "epoch": 2.4522323571771483,
+      "grad_norm": 0.9627230310579845,
+      "learning_rate": 6.03595374103366e-06,
+      "loss": 0.5392,
+      "step": 2555
+    },
+    {
+      "epoch": 2.4531925108017285,
+      "grad_norm": 1.145458285959235,
+      "learning_rate": 6.032676240833876e-06,
+      "loss": 0.5202,
+      "step": 2556
+    },
+    {
+      "epoch": 2.4541526644263083,
+      "grad_norm": 0.8140750529777154,
+      "learning_rate": 6.029398277076766e-06,
+      "loss": 0.5469,
+      "step": 2557
+    },
+    {
+      "epoch": 2.455112818050888,
+      "grad_norm": 1.0405936049835796,
+      "learning_rate": 6.0261198512337725e-06,
+      "loss": 0.4157,
+      "step": 2558
+    },
+    {
+      "epoch": 2.456072971675468,
+      "grad_norm": 0.8788759787462748,
+      "learning_rate": 6.022840964776548e-06,
+      "loss": 0.5172,
+      "step": 2559
+    },
+    {
+      "epoch": 2.457033125300048,
+      "grad_norm": 1.0732883769248467,
+      "learning_rate": 6.019561619176948e-06,
+      "loss": 0.5356,
+      "step": 2560
+    },
+    {
+      "epoch": 2.457993278924628,
+      "grad_norm": 0.8906641941924691,
+      "learning_rate": 6.016281815907036e-06,
+      "loss": 0.5221,
+      "step": 2561
+    },
+    {
+      "epoch": 2.4589534325492077,
+      "grad_norm": 1.2324033430189187,
+      "learning_rate": 6.013001556439079e-06,
+      "loss": 0.4163,
+      "step": 2562
+    },
+    {
+      "epoch": 2.459913586173788,
+      "grad_norm": 1.2455497806231217,
+      "learning_rate": 6.0097208422455526e-06,
+      "loss": 0.5585,
+      "step": 2563
+    },
+    {
+      "epoch": 2.4608737397983678,
+      "grad_norm": 1.0533865252806114,
+      "learning_rate": 6.006439674799132e-06,
+      "loss": 0.5587,
+      "step": 2564
+    },
+    {
+      "epoch": 2.4618338934229476,
+      "grad_norm": 0.8556710253638442,
+      "learning_rate": 6.0031580555727005e-06,
+      "loss": 0.505,
+      "step": 2565
+    },
+    {
+      "epoch": 2.4627940470475274,
+      "grad_norm": 1.0939080771517722,
+      "learning_rate": 5.999875986039341e-06,
+      "loss": 0.5696,
+      "step": 2566
+    },
+    {
+      "epoch": 2.4637542006721076,
+      "grad_norm": 0.938244140579834,
+      "learning_rate": 5.996593467672339e-06,
+      "loss": 0.5682,
+      "step": 2567
+    },
+    {
+      "epoch": 2.4647143542966874,
+      "grad_norm": 1.0243723639883222,
+      "learning_rate": 5.993310501945184e-06,
+      "loss": 0.5133,
+      "step": 2568
+    },
+    {
+      "epoch": 2.4656745079212676,
+      "grad_norm": 1.039600891886981,
+      "learning_rate": 5.990027090331562e-06,
+      "loss": 0.6911,
+      "step": 2569
+    },
+    {
+      "epoch": 2.4666346615458474,
+      "grad_norm": 0.8806020071668524,
+      "learning_rate": 5.986743234305361e-06,
+      "loss": 0.5951,
+      "step": 2570
+    },
+    {
+      "epoch": 2.4675948151704272,
+      "grad_norm": 0.9458631120569398,
+      "learning_rate": 5.983458935340669e-06,
+      "loss": 0.5176,
+      "step": 2571
+    },
+    {
+      "epoch": 2.468554968795007,
+      "grad_norm": 1.6442454513402067,
+      "learning_rate": 5.980174194911776e-06,
+      "loss": 0.5958,
+      "step": 2572
+    },
+    {
+      "epoch": 2.4695151224195873,
+      "grad_norm": 1.4319293390390937,
+      "learning_rate": 5.9768890144931626e-06,
+      "loss": 0.5127,
+      "step": 2573
+    },
+    {
+      "epoch": 2.470475276044167,
+      "grad_norm": 1.8969459507515172,
+      "learning_rate": 5.973603395559514e-06,
+      "loss": 0.5942,
+      "step": 2574
+    },
+    {
+      "epoch": 2.471435429668747,
+      "grad_norm": 0.9224176575825475,
+      "learning_rate": 5.970317339585707e-06,
+      "loss": 0.5498,
+      "step": 2575
+    },
+    {
+      "epoch": 2.472395583293327,
+      "grad_norm": 1.1947498550400801,
+      "learning_rate": 5.96703084804682e-06,
+      "loss": 0.4773,
+      "step": 2576
+    },
+    {
+      "epoch": 2.473355736917907,
+      "grad_norm": 1.004578430059494,
+      "learning_rate": 5.963743922418122e-06,
+      "loss": 0.583,
+      "step": 2577
+    },
+    {
+      "epoch": 2.4743158905424867,
+      "grad_norm": 1.357725894295687,
+      "learning_rate": 5.960456564175079e-06,
+      "loss": 0.5136,
+      "step": 2578
+    },
+    {
+      "epoch": 2.475276044167067,
+      "grad_norm": 0.966777258485813,
+      "learning_rate": 5.9571687747933505e-06,
+      "loss": 0.6178,
+      "step": 2579
+    },
+    {
+      "epoch": 2.4762361977916467,
+      "grad_norm": 0.971768856681797,
+      "learning_rate": 5.9538805557487896e-06,
+      "loss": 0.6749,
+      "step": 2580
+    },
+    {
+      "epoch": 2.4771963514162265,
+      "grad_norm": 1.194876344310327,
+      "learning_rate": 5.950591908517446e-06,
+      "loss": 0.5473,
+      "step": 2581
+    },
+    {
+      "epoch": 2.4781565050408068,
+      "grad_norm": 1.3229983395220837,
+      "learning_rate": 5.9473028345755535e-06,
+      "loss": 0.5612,
+      "step": 2582
+    },
+    {
+      "epoch": 2.4791166586653866,
+      "grad_norm": 1.0317895502097822,
+      "learning_rate": 5.944013335399546e-06,
+      "loss": 0.6091,
+      "step": 2583
+    },
+    {
+      "epoch": 2.4800768122899663,
+      "grad_norm": 1.0414457476888133,
+      "learning_rate": 5.940723412466044e-06,
+      "loss": 0.5305,
+      "step": 2584
+    },
+    {
+      "epoch": 2.481036965914546,
+      "grad_norm": 1.3840832496866207,
+      "learning_rate": 5.937433067251857e-06,
+      "loss": 0.5803,
+      "step": 2585
+    },
+    {
+      "epoch": 2.4819971195391264,
+      "grad_norm": 0.9928133197922491,
+      "learning_rate": 5.934142301233986e-06,
+      "loss": 0.618,
+      "step": 2586
+    },
+    {
+      "epoch": 2.482957273163706,
+      "grad_norm": 1.0766586909207183,
+      "learning_rate": 5.9308511158896235e-06,
+      "loss": 0.5838,
+      "step": 2587
+    },
+    {
+      "epoch": 2.483917426788286,
+      "grad_norm": 1.1825859995947656,
+      "learning_rate": 5.927559512696143e-06,
+      "loss": 0.6283,
+      "step": 2588
+    },
+    {
+      "epoch": 2.484877580412866,
+      "grad_norm": 1.032415138872871,
+      "learning_rate": 5.924267493131113e-06,
+      "loss": 0.6022,
+      "step": 2589
+    },
+    {
+      "epoch": 2.485837734037446,
+      "grad_norm": 0.9628596424680305,
+      "learning_rate": 5.9209750586722855e-06,
+      "loss": 0.6158,
+      "step": 2590
+    },
+    {
+      "epoch": 2.486797887662026,
+      "grad_norm": 1.2887726983995444,
+      "learning_rate": 5.917682210797598e-06,
+      "loss": 0.6021,
+      "step": 2591
+    },
+    {
+      "epoch": 2.487758041286606,
+      "grad_norm": 1.0204898747641036,
+      "learning_rate": 5.9143889509851785e-06,
+      "loss": 0.4879,
+      "step": 2592
+    },
+    {
+      "epoch": 2.488718194911186,
+      "grad_norm": 1.151434491288258,
+      "learning_rate": 5.91109528071333e-06,
+      "loss": 0.5356,
+      "step": 2593
+    },
+    {
+      "epoch": 2.4896783485357656,
+      "grad_norm": 0.9842095217042718,
+      "learning_rate": 5.9078012014605505e-06,
+      "loss": 0.4471,
+      "step": 2594
+    },
+    {
+      "epoch": 2.490638502160346,
+      "grad_norm": 1.33150273835374,
+      "learning_rate": 5.904506714705514e-06,
+      "loss": 0.6534,
+      "step": 2595
+    },
+    {
+      "epoch": 2.4915986557849257,
+      "grad_norm": 1.7258705689035156,
+      "learning_rate": 5.901211821927084e-06,
+      "loss": 0.6171,
+      "step": 2596
+    },
+    {
+      "epoch": 2.4925588094095055,
+      "grad_norm": 1.2626862822791978,
+      "learning_rate": 5.897916524604299e-06,
+      "loss": 0.5844,
+      "step": 2597
+    },
+    {
+      "epoch": 2.4935189630340853,
+      "grad_norm": 0.9250204182091345,
+      "learning_rate": 5.894620824216385e-06,
+      "loss": 0.5661,
+      "step": 2598
+    },
+    {
+      "epoch": 2.4944791166586655,
+      "grad_norm": 1.1826866287673776,
+      "learning_rate": 5.8913247222427464e-06,
+      "loss": 0.5676,
+      "step": 2599
+    },
+    {
+      "epoch": 2.4954392702832453,
+      "grad_norm": 0.9766514675342373,
+      "learning_rate": 5.888028220162967e-06,
+      "loss": 0.5587,
+      "step": 2600
+    },
+    {
+      "epoch": 2.496399423907825,
+      "grad_norm": 0.8732797696155599,
+      "learning_rate": 5.884731319456812e-06,
+      "loss": 0.5236,
+      "step": 2601
+    },
+    {
+      "epoch": 2.4973595775324053,
+      "grad_norm": 0.8430315332928399,
+      "learning_rate": 5.881434021604224e-06,
+      "loss": 0.3869,
+      "step": 2602
+    },
+    {
+      "epoch": 2.498319731156985,
+      "grad_norm": 1.274877961681524,
+      "learning_rate": 5.878136328085328e-06,
+      "loss": 0.6309,
+      "step": 2603
+    },
+    {
+      "epoch": 2.499279884781565,
+      "grad_norm": 1.8930945096618759,
+      "learning_rate": 5.874838240380419e-06,
+      "loss": 0.5417,
+      "step": 2604
+    },
+    {
+      "epoch": 2.5002400384061447,
+      "grad_norm": 1.1010418480785087,
+      "learning_rate": 5.871539759969978e-06,
+      "loss": 0.4826,
+      "step": 2605
+    },
+    {
+      "epoch": 2.501200192030725,
+      "grad_norm": 1.88062980130338,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.585,
+      "step": 2606
+    },
+    {
+      "epoch": 2.5021603456553048,
+      "grad_norm": 1.212461006479849,
+      "learning_rate": 5.864941626955274e-06,
+      "loss": 0.5101,
+      "step": 2607
+    },
+    {
+      "epoch": 2.503120499279885,
+      "grad_norm": 0.8668789311085789,
+      "learning_rate": 5.861641977312844e-06,
+      "loss": 0.5702,
+      "step": 2608
+    },
+    {
+      "epoch": 2.504080652904465,
+      "grad_norm": 0.8667080794531593,
+      "learning_rate": 5.858341940888543e-06,
+      "loss": 0.5367,
+      "step": 2609
+    },
+    {
+      "epoch": 2.5050408065290446,
+      "grad_norm": 1.07778753805212,
+      "learning_rate": 5.855041519163718e-06,
+      "loss": 0.5217,
+      "step": 2610
+    },
+    {
+      "epoch": 2.5060009601536244,
+      "grad_norm": 1.1078478028218475,
+      "learning_rate": 5.851740713619896e-06,
+      "loss": 0.521,
+      "step": 2611
+    },
+    {
+      "epoch": 2.5069611137782046,
+      "grad_norm": 1.179715917210379,
+      "learning_rate": 5.848439525738772e-06,
+      "loss": 0.5676,
+      "step": 2612
+    },
+    {
+      "epoch": 2.5079212674027844,
+      "grad_norm": 1.1097176015372312,
+      "learning_rate": 5.845137957002214e-06,
+      "loss": 0.6292,
+      "step": 2613
+    },
+    {
+      "epoch": 2.5088814210273642,
+      "grad_norm": 1.2984048738403013,
+      "learning_rate": 5.841836008892262e-06,
+      "loss": 0.5074,
+      "step": 2614
+    },
+    {
+      "epoch": 2.5098415746519445,
+      "grad_norm": 1.0559915309308197,
+      "learning_rate": 5.838533682891123e-06,
+      "loss": 0.5888,
+      "step": 2615
+    },
+    {
+      "epoch": 2.5108017282765243,
+      "grad_norm": 0.9964536318540524,
+      "learning_rate": 5.835230980481177e-06,
+      "loss": 0.5694,
+      "step": 2616
+    },
+    {
+      "epoch": 2.511761881901104,
+      "grad_norm": 1.1724431371098636,
+      "learning_rate": 5.831927903144972e-06,
+      "loss": 0.507,
+      "step": 2617
+    },
+    {
+      "epoch": 2.512722035525684,
+      "grad_norm": 0.9083953835098307,
+      "learning_rate": 5.828624452365226e-06,
+      "loss": 0.5548,
+      "step": 2618
+    },
+    {
+      "epoch": 2.513682189150264,
+      "grad_norm": 1.5918065837595812,
+      "learning_rate": 5.82532062962482e-06,
+      "loss": 0.5391,
+      "step": 2619
+    },
+    {
+      "epoch": 2.514642342774844,
+      "grad_norm": 1.010645399404615,
+      "learning_rate": 5.822016436406805e-06,
+      "loss": 0.5103,
+      "step": 2620
+    },
+    {
+      "epoch": 2.515602496399424,
+      "grad_norm": 1.3509602518688228,
+      "learning_rate": 5.818711874194398e-06,
+      "loss": 0.5193,
+      "step": 2621
+    },
+    {
+      "epoch": 2.516562650024004,
+      "grad_norm": 1.2127030501075124,
+      "learning_rate": 5.8154069444709825e-06,
+      "loss": 0.5876,
+      "step": 2622
+    },
+    {
+      "epoch": 2.5175228036485837,
+      "grad_norm": 1.0025237235571185,
+      "learning_rate": 5.8121016487201074e-06,
+      "loss": 0.4904,
+      "step": 2623
+    },
+    {
+      "epoch": 2.5184829572731635,
+      "grad_norm": 1.0354864302247444,
+      "learning_rate": 5.80879598842548e-06,
+      "loss": 0.6199,
+      "step": 2624
+    },
+    {
+      "epoch": 2.5194431108977438,
+      "grad_norm": 1.5814691353189114,
+      "learning_rate": 5.805489965070983e-06,
+      "loss": 0.5325,
+      "step": 2625
+    },
+    {
+      "epoch": 2.5204032645223235,
+      "grad_norm": 0.8799344319638772,
+      "learning_rate": 5.8021835801406475e-06,
+      "loss": 0.5243,
+      "step": 2626
+    },
+    {
+      "epoch": 2.5213634181469033,
+      "grad_norm": 0.909422512011236,
+      "learning_rate": 5.798876835118681e-06,
+      "loss": 0.5372,
+      "step": 2627
+    },
+    {
+      "epoch": 2.5223235717714836,
+      "grad_norm": 1.1460977214328045,
+      "learning_rate": 5.79556973148944e-06,
+      "loss": 0.481,
+      "step": 2628
+    },
+    {
+      "epoch": 2.5232837253960634,
+      "grad_norm": 1.06360120832438,
+      "learning_rate": 5.792262270737451e-06,
+      "loss": 0.5626,
+      "step": 2629
+    },
+    {
+      "epoch": 2.524243879020643,
+      "grad_norm": 1.0066875885568904,
+      "learning_rate": 5.788954454347397e-06,
+      "loss": 0.5557,
+      "step": 2630
+    },
+    {
+      "epoch": 2.525204032645223,
+      "grad_norm": 1.2447816180455138,
+      "learning_rate": 5.7856462838041235e-06,
+      "loss": 0.5445,
+      "step": 2631
+    },
+    {
+      "epoch": 2.526164186269803,
+      "grad_norm": 1.4941631381271145,
+      "learning_rate": 5.782337760592631e-06,
+      "loss": 0.5912,
+      "step": 2632
+    },
+    {
+      "epoch": 2.527124339894383,
+      "grad_norm": 0.8837403230313842,
+      "learning_rate": 5.7790288861980795e-06,
+      "loss": 0.5031,
+      "step": 2633
+    },
+    {
+      "epoch": 2.5280844935189632,
+      "grad_norm": 1.0322148290442799,
+      "learning_rate": 5.77571966210579e-06,
+      "loss": 0.5291,
+      "step": 2634
+    },
+    {
+      "epoch": 2.529044647143543,
+      "grad_norm": 1.074461969605371,
+      "learning_rate": 5.772410089801234e-06,
+      "loss": 0.5288,
+      "step": 2635
+    },
+    {
+      "epoch": 2.530004800768123,
+      "grad_norm": 1.2540822198794377,
+      "learning_rate": 5.769100170770048e-06,
+      "loss": 0.3869,
+      "step": 2636
+    },
+    {
+      "epoch": 2.5309649543927026,
+      "grad_norm": 1.2917790473998079,
+      "learning_rate": 5.765789906498015e-06,
+      "loss": 0.4972,
+      "step": 2637
+    },
+    {
+      "epoch": 2.531925108017283,
+      "grad_norm": 1.185457765621514,
+      "learning_rate": 5.762479298471078e-06,
+      "loss": 0.6066,
+      "step": 2638
+    },
+    {
+      "epoch": 2.5328852616418627,
+      "grad_norm": 1.615534221090781,
+      "learning_rate": 5.759168348175335e-06,
+      "loss": 0.6056,
+      "step": 2639
+    },
+    {
+      "epoch": 2.5338454152664425,
+      "grad_norm": 1.039914366142861,
+      "learning_rate": 5.755857057097036e-06,
+      "loss": 0.624,
+      "step": 2640
+    },
+    {
+      "epoch": 2.5348055688910227,
+      "grad_norm": 1.2310248807706055,
+      "learning_rate": 5.752545426722583e-06,
+      "loss": 0.5919,
+      "step": 2641
+    },
+    {
+      "epoch": 2.5357657225156025,
+      "grad_norm": 0.963716619513092,
+      "learning_rate": 5.7492334585385325e-06,
+      "loss": 0.5855,
+      "step": 2642
+    },
+    {
+      "epoch": 2.5367258761401823,
+      "grad_norm": 0.8762076343673909,
+      "learning_rate": 5.745921154031593e-06,
+      "loss": 0.5786,
+      "step": 2643
+    },
+    {
+      "epoch": 2.537686029764762,
+      "grad_norm": 0.7901722509181774,
+      "learning_rate": 5.742608514688619e-06,
+      "loss": 0.4986,
+      "step": 2644
+    },
+    {
+      "epoch": 2.5386461833893423,
+      "grad_norm": 0.9899570633713458,
+      "learning_rate": 5.739295541996622e-06,
+      "loss": 0.5513,
+      "step": 2645
+    },
+    {
+      "epoch": 2.539606337013922,
+      "grad_norm": 1.1860857708819632,
+      "learning_rate": 5.7359822374427565e-06,
+      "loss": 0.4888,
+      "step": 2646
+    },
+    {
+      "epoch": 2.5405664906385024,
+      "grad_norm": 0.9829635047820741,
+      "learning_rate": 5.732668602514334e-06,
+      "loss": 0.4516,
+      "step": 2647
+    },
+    {
+      "epoch": 2.541526644263082,
+      "grad_norm": 0.857473761716564,
+      "learning_rate": 5.729354638698806e-06,
+      "loss": 0.4112,
+      "step": 2648
+    },
+    {
+      "epoch": 2.542486797887662,
+      "grad_norm": 0.8539632466434419,
+      "learning_rate": 5.726040347483779e-06,
+      "loss": 0.5202,
+      "step": 2649
+    },
+    {
+      "epoch": 2.5434469515122418,
+      "grad_norm": 1.349597198191044,
+      "learning_rate": 5.722725730357001e-06,
+      "loss": 0.6785,
+      "step": 2650
+    },
+    {
+      "epoch": 2.544407105136822,
+      "grad_norm": 1.476308254477948,
+      "learning_rate": 5.719410788806366e-06,
+      "loss": 0.4867,
+      "step": 2651
+    },
+    {
+      "epoch": 2.545367258761402,
+      "grad_norm": 1.1364375560304625,
+      "learning_rate": 5.716095524319919e-06,
+      "loss": 0.4572,
+      "step": 2652
+    },
+    {
+      "epoch": 2.5463274123859816,
+      "grad_norm": 0.9292151349949764,
+      "learning_rate": 5.712779938385846e-06,
+      "loss": 0.5275,
+      "step": 2653
+    },
+    {
+      "epoch": 2.547287566010562,
+      "grad_norm": 1.0603683681510738,
+      "learning_rate": 5.709464032492478e-06,
+      "loss": 0.6118,
+      "step": 2654
+    },
+    {
+      "epoch": 2.5482477196351416,
+      "grad_norm": 1.0695585278526674,
+      "learning_rate": 5.706147808128288e-06,
+      "loss": 0.541,
+      "step": 2655
+    },
+    {
+      "epoch": 2.5492078732597214,
+      "grad_norm": 1.1713880900693274,
+      "learning_rate": 5.702831266781898e-06,
+      "loss": 0.5174,
+      "step": 2656
+    },
+    {
+      "epoch": 2.550168026884301,
+      "grad_norm": 1.0340443847712604,
+      "learning_rate": 5.699514409942064e-06,
+      "loss": 0.4663,
+      "step": 2657
+    },
+    {
+      "epoch": 2.5511281805088815,
+      "grad_norm": 1.6840399739400633,
+      "learning_rate": 5.69619723909769e-06,
+      "loss": 0.5886,
+      "step": 2658
+    },
+    {
+      "epoch": 2.5520883341334613,
+      "grad_norm": 1.5517188707837135,
+      "learning_rate": 5.6928797557378145e-06,
+      "loss": 0.493,
+      "step": 2659
+    },
+    {
+      "epoch": 2.5530484877580415,
+      "grad_norm": 1.01998223291034,
+      "learning_rate": 5.689561961351624e-06,
+      "loss": 0.506,
+      "step": 2660
+    },
+    {
+      "epoch": 2.5540086413826213,
+      "grad_norm": 1.2733355383158487,
+      "learning_rate": 5.68624385742844e-06,
+      "loss": 0.568,
+      "step": 2661
+    },
+    {
+      "epoch": 2.554968795007201,
+      "grad_norm": 1.3483186301470194,
+      "learning_rate": 5.682925445457726e-06,
+      "loss": 0.4972,
+      "step": 2662
+    },
+    {
+      "epoch": 2.555928948631781,
+      "grad_norm": 0.9104335531564698,
+      "learning_rate": 5.679606726929076e-06,
+      "loss": 0.5291,
+      "step": 2663
+    },
+    {
+      "epoch": 2.556889102256361,
+      "grad_norm": 1.1588795293210066,
+      "learning_rate": 5.676287703332234e-06,
+      "loss": 0.553,
+      "step": 2664
+    },
+    {
+      "epoch": 2.557849255880941,
+      "grad_norm": 1.0427251508783315,
+      "learning_rate": 5.672968376157071e-06,
+      "loss": 0.4908,
+      "step": 2665
+    },
+    {
+      "epoch": 2.558809409505521,
+      "grad_norm": 0.943102809941431,
+      "learning_rate": 5.6696487468935955e-06,
+      "loss": 0.5077,
+      "step": 2666
+    },
+    {
+      "epoch": 2.559769563130101,
+      "grad_norm": 1.0191375722161495,
+      "learning_rate": 5.666328817031958e-06,
+      "loss": 0.5781,
+      "step": 2667
+    },
+    {
+      "epoch": 2.5607297167546808,
+      "grad_norm": 1.0860080564510013,
+      "learning_rate": 5.663008588062435e-06,
+      "loss": 0.5606,
+      "step": 2668
+    },
+    {
+      "epoch": 2.5616898703792605,
+      "grad_norm": 1.6069535272566735,
+      "learning_rate": 5.659688061475445e-06,
+      "loss": 0.6415,
+      "step": 2669
+    },
+    {
+      "epoch": 2.5626500240038403,
+      "grad_norm": 0.9241312991066435,
+      "learning_rate": 5.656367238761536e-06,
+      "loss": 0.5017,
+      "step": 2670
+    },
+    {
+      "epoch": 2.5636101776284206,
+      "grad_norm": 1.1079602817747738,
+      "learning_rate": 5.653046121411389e-06,
+      "loss": 0.5436,
+      "step": 2671
+    },
+    {
+      "epoch": 2.5645703312530004,
+      "grad_norm": 1.2295516632519532,
+      "learning_rate": 5.649724710915819e-06,
+      "loss": 0.5775,
+      "step": 2672
+    },
+    {
+      "epoch": 2.5655304848775806,
+      "grad_norm": 0.8842835928651863,
+      "learning_rate": 5.646403008765772e-06,
+      "loss": 0.5541,
+      "step": 2673
+    },
+    {
+      "epoch": 2.5664906385021604,
+      "grad_norm": 0.815281761560292,
+      "learning_rate": 5.643081016452323e-06,
+      "loss": 0.482,
+      "step": 2674
+    },
+    {
+      "epoch": 2.56745079212674,
+      "grad_norm": 1.029634729715191,
+      "learning_rate": 5.639758735466682e-06,
+      "loss": 0.5659,
+      "step": 2675
+    },
+    {
+      "epoch": 2.56841094575132,
+      "grad_norm": 1.1812805553945471,
+      "learning_rate": 5.636436167300181e-06,
+      "loss": 0.6244,
+      "step": 2676
+    },
+    {
+      "epoch": 2.5693710993759002,
+      "grad_norm": 0.9321973409793013,
+      "learning_rate": 5.633113313444288e-06,
+      "loss": 0.6212,
+      "step": 2677
+    },
+    {
+      "epoch": 2.57033125300048,
+      "grad_norm": 1.1961374002922658,
+      "learning_rate": 5.629790175390598e-06,
+      "loss": 0.5985,
+      "step": 2678
+    },
+    {
+      "epoch": 2.5712914066250603,
+      "grad_norm": 1.5530231817597748,
+      "learning_rate": 5.626466754630829e-06,
+      "loss": 0.5804,
+      "step": 2679
+    },
+    {
+      "epoch": 2.57225156024964,
+      "grad_norm": 1.2801498988049118,
+      "learning_rate": 5.623143052656831e-06,
+      "loss": 0.593,
+      "step": 2680
+    },
+    {
+      "epoch": 2.57321171387422,
+      "grad_norm": 0.8825201031543539,
+      "learning_rate": 5.619819070960577e-06,
+      "loss": 0.5012,
+      "step": 2681
+    },
+    {
+      "epoch": 2.5741718674987997,
+      "grad_norm": 1.1200609022324564,
+      "learning_rate": 5.616494811034168e-06,
+      "loss": 0.5668,
+      "step": 2682
+    },
+    {
+      "epoch": 2.5751320211233795,
+      "grad_norm": 1.004337461449322,
+      "learning_rate": 5.613170274369827e-06,
+      "loss": 0.4849,
+      "step": 2683
+    },
+    {
+      "epoch": 2.5760921747479597,
+      "grad_norm": 0.8727360467170426,
+      "learning_rate": 5.609845462459907e-06,
+      "loss": 0.5333,
+      "step": 2684
+    },
+    {
+      "epoch": 2.5770523283725395,
+      "grad_norm": 1.0406545366923414,
+      "learning_rate": 5.606520376796875e-06,
+      "loss": 0.5216,
+      "step": 2685
+    },
+    {
+      "epoch": 2.5780124819971197,
+      "grad_norm": 0.919124011318109,
+      "learning_rate": 5.603195018873329e-06,
+      "loss": 0.6206,
+      "step": 2686
+    },
+    {
+      "epoch": 2.5789726356216995,
+      "grad_norm": 1.4852216859352825,
+      "learning_rate": 5.59986939018199e-06,
+      "loss": 0.5242,
+      "step": 2687
+    },
+    {
+      "epoch": 2.5799327892462793,
+      "grad_norm": 1.2000340248357038,
+      "learning_rate": 5.596543492215691e-06,
+      "loss": 0.6142,
+      "step": 2688
+    },
+    {
+      "epoch": 2.580892942870859,
+      "grad_norm": 1.2157078489152757,
+      "learning_rate": 5.5932173264673975e-06,
+      "loss": 0.5207,
+      "step": 2689
+    },
+    {
+      "epoch": 2.5818530964954394,
+      "grad_norm": 1.226909166384669,
+      "learning_rate": 5.589890894430186e-06,
+      "loss": 0.5099,
+      "step": 2690
+    },
+    {
+      "epoch": 2.582813250120019,
+      "grad_norm": 1.38801630395136,
+      "learning_rate": 5.586564197597256e-06,
+      "loss": 0.5845,
+      "step": 2691
+    },
+    {
+      "epoch": 2.5837734037445994,
+      "grad_norm": 1.2007598474081262,
+      "learning_rate": 5.583237237461929e-06,
+      "loss": 0.5644,
+      "step": 2692
+    },
+    {
+      "epoch": 2.584733557369179,
+      "grad_norm": 1.3318445225092201,
+      "learning_rate": 5.579910015517642e-06,
+      "loss": 0.5989,
+      "step": 2693
+    },
+    {
+      "epoch": 2.585693710993759,
+      "grad_norm": 1.336662008737483,
+      "learning_rate": 5.576582533257946e-06,
+      "loss": 0.5732,
+      "step": 2694
+    },
+    {
+      "epoch": 2.586653864618339,
+      "grad_norm": 0.9636159574045997,
+      "learning_rate": 5.573254792176515e-06,
+      "loss": 0.6207,
+      "step": 2695
+    },
+    {
+      "epoch": 2.5876140182429186,
+      "grad_norm": 1.214235939611955,
+      "learning_rate": 5.569926793767136e-06,
+      "loss": 0.5635,
+      "step": 2696
+    },
+    {
+      "epoch": 2.588574171867499,
+      "grad_norm": 1.061638704288488,
+      "learning_rate": 5.566598539523711e-06,
+      "loss": 0.6396,
+      "step": 2697
+    },
+    {
+      "epoch": 2.5895343254920786,
+      "grad_norm": 1.477231542756541,
+      "learning_rate": 5.563270030940258e-06,
+      "loss": 0.4549,
+      "step": 2698
+    },
+    {
+      "epoch": 2.590494479116659,
+      "grad_norm": 1.6126428115809528,
+      "learning_rate": 5.559941269510908e-06,
+      "loss": 0.6009,
+      "step": 2699
+    },
+    {
+      "epoch": 2.5914546327412387,
+      "grad_norm": 0.863195874077329,
+      "learning_rate": 5.556612256729909e-06,
+      "loss": 0.5583,
+      "step": 2700
+    },
+    {
+      "epoch": 2.5924147863658185,
+      "grad_norm": 0.96510297442427,
+      "learning_rate": 5.553282994091617e-06,
+      "loss": 0.612,
+      "step": 2701
+    },
+    {
+      "epoch": 2.5933749399903983,
+      "grad_norm": 1.3476599241059,
+      "learning_rate": 5.549953483090506e-06,
+      "loss": 0.5237,
+      "step": 2702
+    },
+    {
+      "epoch": 2.5943350936149785,
+      "grad_norm": 0.8232905487050111,
+      "learning_rate": 5.546623725221152e-06,
+      "loss": 0.4715,
+      "step": 2703
+    },
+    {
+      "epoch": 2.5952952472395583,
+      "grad_norm": 1.817626592285635,
+      "learning_rate": 5.543293721978251e-06,
+      "loss": 0.6081,
+      "step": 2704
+    },
+    {
+      "epoch": 2.5962554008641385,
+      "grad_norm": 0.9796305099231157,
+      "learning_rate": 5.539963474856607e-06,
+      "loss": 0.502,
+      "step": 2705
+    },
+    {
+      "epoch": 2.5972155544887183,
+      "grad_norm": 1.1016208610029354,
+      "learning_rate": 5.536632985351132e-06,
+      "loss": 0.5806,
+      "step": 2706
+    },
+    {
+      "epoch": 2.598175708113298,
+      "grad_norm": 1.1112838847468167,
+      "learning_rate": 5.533302254956843e-06,
+      "loss": 0.5873,
+      "step": 2707
+    },
+    {
+      "epoch": 2.599135861737878,
+      "grad_norm": 1.3974448758581155,
+      "learning_rate": 5.529971285168873e-06,
+      "loss": 0.5397,
+      "step": 2708
+    },
+    {
+      "epoch": 2.600096015362458,
+      "grad_norm": 1.102305687285218,
+      "learning_rate": 5.526640077482462e-06,
+      "loss": 0.5855,
+      "step": 2709
+    },
+    {
+      "epoch": 2.601056168987038,
+      "grad_norm": 1.0834105086173802,
+      "learning_rate": 5.523308633392945e-06,
+      "loss": 0.5582,
+      "step": 2710
+    },
+    {
+      "epoch": 2.6020163226116177,
+      "grad_norm": 1.1138895731472624,
+      "learning_rate": 5.519976954395779e-06,
+      "loss": 0.6421,
+      "step": 2711
+    },
+    {
+      "epoch": 2.602976476236198,
+      "grad_norm": 1.0725641102844308,
+      "learning_rate": 5.516645041986516e-06,
+      "loss": 0.4871,
+      "step": 2712
+    },
+    {
+      "epoch": 2.603936629860778,
+      "grad_norm": 1.1949801296467149,
+      "learning_rate": 5.513312897660815e-06,
+      "loss": 0.5309,
+      "step": 2713
+    },
+    {
+      "epoch": 2.6048967834853576,
+      "grad_norm": 0.7610137344554979,
+      "learning_rate": 5.509980522914442e-06,
+      "loss": 0.5354,
+      "step": 2714
+    },
+    {
+      "epoch": 2.6058569371099374,
+      "grad_norm": 1.1752312404192111,
+      "learning_rate": 5.506647919243265e-06,
+      "loss": 0.6004,
+      "step": 2715
+    },
+    {
+      "epoch": 2.6068170907345176,
+      "grad_norm": 1.617690143178605,
+      "learning_rate": 5.503315088143251e-06,
+      "loss": 0.5618,
+      "step": 2716
+    },
+    {
+      "epoch": 2.6077772443590974,
+      "grad_norm": 1.057282458640301,
+      "learning_rate": 5.499982031110475e-06,
+      "loss": 0.6788,
+      "step": 2717
+    },
+    {
+      "epoch": 2.6087373979836777,
+      "grad_norm": 1.0834313330038319,
+      "learning_rate": 5.496648749641109e-06,
+      "loss": 0.5725,
+      "step": 2718
+    },
+    {
+      "epoch": 2.6096975516082574,
+      "grad_norm": 0.8414741095327913,
+      "learning_rate": 5.493315245231429e-06,
+      "loss": 0.5333,
+      "step": 2719
+    },
+    {
+      "epoch": 2.6106577052328372,
+      "grad_norm": 2.0680047470240686,
+      "learning_rate": 5.489981519377808e-06,
+      "loss": 0.6605,
+      "step": 2720
+    },
+    {
+      "epoch": 2.611617858857417,
+      "grad_norm": 1.375141648361915,
+      "learning_rate": 5.486647573576721e-06,
+      "loss": 0.5187,
+      "step": 2721
+    },
+    {
+      "epoch": 2.6125780124819973,
+      "grad_norm": 1.0374484919564513,
+      "learning_rate": 5.483313409324741e-06,
+      "loss": 0.5978,
+      "step": 2722
+    },
+    {
+      "epoch": 2.613538166106577,
+      "grad_norm": 0.8794906694859634,
+      "learning_rate": 5.479979028118536e-06,
+      "loss": 0.4158,
+      "step": 2723
+    },
+    {
+      "epoch": 2.614498319731157,
+      "grad_norm": 0.9556415312439337,
+      "learning_rate": 5.476644431454878e-06,
+      "loss": 0.5909,
+      "step": 2724
+    },
+    {
+      "epoch": 2.615458473355737,
+      "grad_norm": 1.042097717657121,
+      "learning_rate": 5.473309620830628e-06,
+      "loss": 0.4337,
+      "step": 2725
+    },
+    {
+      "epoch": 2.616418626980317,
+      "grad_norm": 1.0235818012601186,
+      "learning_rate": 5.469974597742749e-06,
+      "loss": 0.5137,
+      "step": 2726
+    },
+    {
+      "epoch": 2.6173787806048967,
+      "grad_norm": 0.8434336986130978,
+      "learning_rate": 5.4666393636882965e-06,
+      "loss": 0.5856,
+      "step": 2727
+    },
+    {
+      "epoch": 2.6183389342294765,
+      "grad_norm": 1.0289269282223525,
+      "learning_rate": 5.463303920164423e-06,
+      "loss": 0.5726,
+      "step": 2728
+    },
+    {
+      "epoch": 2.6192990878540567,
+      "grad_norm": 0.876350909586272,
+      "learning_rate": 5.459968268668372e-06,
+      "loss": 0.4906,
+      "step": 2729
+    },
+    {
+      "epoch": 2.6202592414786365,
+      "grad_norm": 1.2840493491555898,
+      "learning_rate": 5.456632410697481e-06,
+      "loss": 0.5594,
+      "step": 2730
+    },
+    {
+      "epoch": 2.6212193951032168,
+      "grad_norm": 1.0127773694122186,
+      "learning_rate": 5.453296347749183e-06,
+      "loss": 0.5075,
+      "step": 2731
+    },
+    {
+      "epoch": 2.6221795487277966,
+      "grad_norm": 1.3860567146301412,
+      "learning_rate": 5.449960081321e-06,
+      "loss": 0.4575,
+      "step": 2732
+    },
+    {
+      "epoch": 2.6231397023523764,
+      "grad_norm": 1.0199533453696517,
+      "learning_rate": 5.446623612910545e-06,
+      "loss": 0.5225,
+      "step": 2733
+    },
+    {
+      "epoch": 2.624099855976956,
+      "grad_norm": 0.8269796783584771,
+      "learning_rate": 5.443286944015525e-06,
+      "loss": 0.5097,
+      "step": 2734
+    },
+    {
+      "epoch": 2.6250600096015364,
+      "grad_norm": 1.0664631979932955,
+      "learning_rate": 5.439950076133734e-06,
+      "loss": 0.5606,
+      "step": 2735
+    },
+    {
+      "epoch": 2.626020163226116,
+      "grad_norm": 0.8895570701010829,
+      "learning_rate": 5.4366130107630566e-06,
+      "loss": 0.4576,
+      "step": 2736
+    },
+    {
+      "epoch": 2.626980316850696,
+      "grad_norm": 0.9217817865861233,
+      "learning_rate": 5.4332757494014675e-06,
+      "loss": 0.5129,
+      "step": 2737
+    },
+    {
+      "epoch": 2.6279404704752762,
+      "grad_norm": 1.066177295205498,
+      "learning_rate": 5.429938293547024e-06,
+      "loss": 0.5681,
+      "step": 2738
+    },
+    {
+      "epoch": 2.628900624099856,
+      "grad_norm": 1.0606628802778508,
+      "learning_rate": 5.426600644697877e-06,
+      "loss": 0.5056,
+      "step": 2739
+    },
+    {
+      "epoch": 2.629860777724436,
+      "grad_norm": 1.0503053541421365,
+      "learning_rate": 5.423262804352262e-06,
+      "loss": 0.4709,
+      "step": 2740
+    },
+    {
+      "epoch": 2.6308209313490156,
+      "grad_norm": 1.8386042944937735,
+      "learning_rate": 5.419924774008496e-06,
+      "loss": 0.601,
+      "step": 2741
+    },
+    {
+      "epoch": 2.631781084973596,
+      "grad_norm": 1.082547521247347,
+      "learning_rate": 5.416586555164989e-06,
+      "loss": 0.5412,
+      "step": 2742
+    },
+    {
+      "epoch": 2.6327412385981757,
+      "grad_norm": 0.8814343852429971,
+      "learning_rate": 5.4132481493202295e-06,
+      "loss": 0.586,
+      "step": 2743
+    },
+    {
+      "epoch": 2.633701392222756,
+      "grad_norm": 1.1250651363324953,
+      "learning_rate": 5.409909557972796e-06,
+      "loss": 0.5744,
+      "step": 2744
+    },
+    {
+      "epoch": 2.6346615458473357,
+      "grad_norm": 0.8687291175550564,
+      "learning_rate": 5.406570782621341e-06,
+      "loss": 0.558,
+      "step": 2745
+    },
+    {
+      "epoch": 2.6356216994719155,
+      "grad_norm": 1.030358641675026,
+      "learning_rate": 5.403231824764609e-06,
+      "loss": 0.5893,
+      "step": 2746
+    },
+    {
+      "epoch": 2.6365818530964953,
+      "grad_norm": 1.4855280963727258,
+      "learning_rate": 5.39989268590142e-06,
+      "loss": 0.6443,
+      "step": 2747
+    },
+    {
+      "epoch": 2.6375420067210755,
+      "grad_norm": 1.353291787264611,
+      "learning_rate": 5.396553367530679e-06,
+      "loss": 0.5846,
+      "step": 2748
+    },
+    {
+      "epoch": 2.6385021603456553,
+      "grad_norm": 0.8241599883739923,
+      "learning_rate": 5.39321387115137e-06,
+      "loss": 0.6277,
+      "step": 2749
+    },
+    {
+      "epoch": 2.639462313970235,
+      "grad_norm": 0.9429568535593399,
+      "learning_rate": 5.389874198262559e-06,
+      "loss": 0.4257,
+      "step": 2750
+    },
+    {
+      "epoch": 2.6404224675948154,
+      "grad_norm": 1.2544147684571567,
+      "learning_rate": 5.386534350363387e-06,
+      "loss": 0.4699,
+      "step": 2751
+    },
+    {
+      "epoch": 2.641382621219395,
+      "grad_norm": 0.9451996316895992,
+      "learning_rate": 5.3831943289530765e-06,
+      "loss": 0.4887,
+      "step": 2752
+    },
+    {
+      "epoch": 2.642342774843975,
+      "grad_norm": 1.1899392309777594,
+      "learning_rate": 5.379854135530929e-06,
+      "loss": 0.5271,
+      "step": 2753
+    },
+    {
+      "epoch": 2.6433029284685547,
+      "grad_norm": 0.8425542526503982,
+      "learning_rate": 5.376513771596319e-06,
+      "loss": 0.5593,
+      "step": 2754
+    },
+    {
+      "epoch": 2.644263082093135,
+      "grad_norm": 1.3338829011787532,
+      "learning_rate": 5.373173238648704e-06,
+      "loss": 0.5741,
+      "step": 2755
+    },
+    {
+      "epoch": 2.645223235717715,
+      "grad_norm": 0.9471815196604269,
+      "learning_rate": 5.369832538187609e-06,
+      "loss": 0.5136,
+      "step": 2756
+    },
+    {
+      "epoch": 2.646183389342295,
+      "grad_norm": 1.1035308295953958,
+      "learning_rate": 5.366491671712641e-06,
+      "loss": 0.54,
+      "step": 2757
+    },
+    {
+      "epoch": 2.647143542966875,
+      "grad_norm": 1.290015711595849,
+      "learning_rate": 5.363150640723479e-06,
+      "loss": 0.4995,
+      "step": 2758
+    },
+    {
+      "epoch": 2.6481036965914546,
+      "grad_norm": 1.4030864072237978,
+      "learning_rate": 5.359809446719878e-06,
+      "loss": 0.6412,
+      "step": 2759
+    },
+    {
+      "epoch": 2.6490638502160344,
+      "grad_norm": 1.3374177423017348,
+      "learning_rate": 5.356468091201661e-06,
+      "loss": 0.6014,
+      "step": 2760
+    },
+    {
+      "epoch": 2.6500240038406147,
+      "grad_norm": 1.011465602607862,
+      "learning_rate": 5.353126575668728e-06,
+      "loss": 0.4744,
+      "step": 2761
+    },
+    {
+      "epoch": 2.6509841574651944,
+      "grad_norm": 0.9607418937883007,
+      "learning_rate": 5.3497849016210515e-06,
+      "loss": 0.6151,
+      "step": 2762
+    },
+    {
+      "epoch": 2.6519443110897742,
+      "grad_norm": 0.7651995220803354,
+      "learning_rate": 5.3464430705586715e-06,
+      "loss": 0.5129,
+      "step": 2763
+    },
+    {
+      "epoch": 2.6529044647143545,
+      "grad_norm": 1.0016321728616067,
+      "learning_rate": 5.3431010839817e-06,
+      "loss": 0.5879,
+      "step": 2764
+    },
+    {
+      "epoch": 2.6538646183389343,
+      "grad_norm": 1.0984388087436279,
+      "learning_rate": 5.339758943390319e-06,
+      "loss": 0.6534,
+      "step": 2765
+    },
+    {
+      "epoch": 2.654824771963514,
+      "grad_norm": 0.9621442488908168,
+      "learning_rate": 5.3364166502847796e-06,
+      "loss": 0.5664,
+      "step": 2766
+    },
+    {
+      "epoch": 2.655784925588094,
+      "grad_norm": 0.8468291580428136,
+      "learning_rate": 5.3330742061654014e-06,
+      "loss": 0.4852,
+      "step": 2767
+    },
+    {
+      "epoch": 2.656745079212674,
+      "grad_norm": 1.2985751049055718,
+      "learning_rate": 5.329731612532575e-06,
+      "loss": 0.5849,
+      "step": 2768
+    },
+    {
+      "epoch": 2.657705232837254,
+      "grad_norm": 0.8763111580394592,
+      "learning_rate": 5.32638887088675e-06,
+      "loss": 0.5138,
+      "step": 2769
+    },
+    {
+      "epoch": 2.658665386461834,
+      "grad_norm": 0.7339367242929759,
+      "learning_rate": 5.323045982728451e-06,
+      "loss": 0.5476,
+      "step": 2770
+    },
+    {
+      "epoch": 2.659625540086414,
+      "grad_norm": 1.2811757059202413,
+      "learning_rate": 5.319702949558263e-06,
+      "loss": 0.5656,
+      "step": 2771
+    },
+    {
+      "epoch": 2.6605856937109937,
+      "grad_norm": 1.5681436780587878,
+      "learning_rate": 5.316359772876839e-06,
+      "loss": 0.493,
+      "step": 2772
+    },
+    {
+      "epoch": 2.6615458473355735,
+      "grad_norm": 1.347679937660324,
+      "learning_rate": 5.313016454184894e-06,
+      "loss": 0.585,
+      "step": 2773
+    },
+    {
+      "epoch": 2.6625060009601538,
+      "grad_norm": 0.8893616509634205,
+      "learning_rate": 5.309672994983209e-06,
+      "loss": 0.5743,
+      "step": 2774
+    },
+    {
+      "epoch": 2.6634661545847336,
+      "grad_norm": 1.5476080539515944,
+      "learning_rate": 5.306329396772627e-06,
+      "loss": 0.5673,
+      "step": 2775
+    },
+    {
+      "epoch": 2.6644263082093134,
+      "grad_norm": 1.3915506222380327,
+      "learning_rate": 5.302985661054053e-06,
+      "loss": 0.5582,
+      "step": 2776
+    },
+    {
+      "epoch": 2.6653864618338936,
+      "grad_norm": 0.8927995632650635,
+      "learning_rate": 5.2996417893284556e-06,
+      "loss": 0.519,
+      "step": 2777
+    },
+    {
+      "epoch": 2.6663466154584734,
+      "grad_norm": 1.3392126472571715,
+      "learning_rate": 5.296297783096861e-06,
+      "loss": 0.515,
+      "step": 2778
+    },
+    {
+      "epoch": 2.667306769083053,
+      "grad_norm": 1.0587509975921972,
+      "learning_rate": 5.292953643860358e-06,
+      "loss": 0.6577,
+      "step": 2779
+    },
+    {
+      "epoch": 2.668266922707633,
+      "grad_norm": 0.8777896031068066,
+      "learning_rate": 5.289609373120096e-06,
+      "loss": 0.6479,
+      "step": 2780
+    },
+    {
+      "epoch": 2.6692270763322132,
+      "grad_norm": 1.3657735621927407,
+      "learning_rate": 5.286264972377283e-06,
+      "loss": 0.5274,
+      "step": 2781
+    },
+    {
+      "epoch": 2.670187229956793,
+      "grad_norm": 1.1016616615171386,
+      "learning_rate": 5.282920443133184e-06,
+      "loss": 0.545,
+      "step": 2782
+    },
+    {
+      "epoch": 2.6711473835813733,
+      "grad_norm": 1.4640582077517341,
+      "learning_rate": 5.279575786889121e-06,
+      "loss": 0.6541,
+      "step": 2783
+    },
+    {
+      "epoch": 2.672107537205953,
+      "grad_norm": 0.9447135546130015,
+      "learning_rate": 5.2762310051464775e-06,
+      "loss": 0.6143,
+      "step": 2784
+    },
+    {
+      "epoch": 2.673067690830533,
+      "grad_norm": 0.7081120383251607,
+      "learning_rate": 5.272886099406688e-06,
+      "loss": 0.4984,
+      "step": 2785
+    },
+    {
+      "epoch": 2.6740278444551127,
+      "grad_norm": 1.2125439642491755,
+      "learning_rate": 5.269541071171246e-06,
+      "loss": 0.6023,
+      "step": 2786
+    },
+    {
+      "epoch": 2.674987998079693,
+      "grad_norm": 0.8922467247793752,
+      "learning_rate": 5.266195921941697e-06,
+      "loss": 0.4071,
+      "step": 2787
+    },
+    {
+      "epoch": 2.6759481517042727,
+      "grad_norm": 0.9275700548723981,
+      "learning_rate": 5.262850653219643e-06,
+      "loss": 0.5377,
+      "step": 2788
+    },
+    {
+      "epoch": 2.6769083053288525,
+      "grad_norm": 1.726446147407602,
+      "learning_rate": 5.259505266506742e-06,
+      "loss": 0.5053,
+      "step": 2789
+    },
+    {
+      "epoch": 2.6778684589534327,
+      "grad_norm": 1.3977026317966972,
+      "learning_rate": 5.256159763304703e-06,
+      "loss": 0.5421,
+      "step": 2790
+    },
+    {
+      "epoch": 2.6788286125780125,
+      "grad_norm": 1.1183935336200055,
+      "learning_rate": 5.252814145115281e-06,
+      "loss": 0.4734,
+      "step": 2791
+    },
+    {
+      "epoch": 2.6797887662025923,
+      "grad_norm": 1.0158485746568071,
+      "learning_rate": 5.249468413440293e-06,
+      "loss": 0.4863,
+      "step": 2792
+    },
+    {
+      "epoch": 2.680748919827172,
+      "grad_norm": 0.8616929124320747,
+      "learning_rate": 5.246122569781601e-06,
+      "loss": 0.5978,
+      "step": 2793
+    },
+    {
+      "epoch": 2.6817090734517524,
+      "grad_norm": 0.9340691499609243,
+      "learning_rate": 5.242776615641119e-06,
+      "loss": 0.5256,
+      "step": 2794
+    },
+    {
+      "epoch": 2.682669227076332,
+      "grad_norm": 1.0959314695399902,
+      "learning_rate": 5.2394305525208076e-06,
+      "loss": 0.6211,
+      "step": 2795
+    },
+    {
+      "epoch": 2.6836293807009124,
+      "grad_norm": 0.9167429208775514,
+      "learning_rate": 5.236084381922681e-06,
+      "loss": 0.5164,
+      "step": 2796
+    },
+    {
+      "epoch": 2.684589534325492,
+      "grad_norm": 1.26657483456402,
+      "learning_rate": 5.232738105348801e-06,
+      "loss": 0.6207,
+      "step": 2797
+    },
+    {
+      "epoch": 2.685549687950072,
+      "grad_norm": 1.302400802244244,
+      "learning_rate": 5.229391724301271e-06,
+      "loss": 0.6118,
+      "step": 2798
+    },
+    {
+      "epoch": 2.686509841574652,
+      "grad_norm": 0.9671848101889771,
+      "learning_rate": 5.22604524028225e-06,
+      "loss": 0.5496,
+      "step": 2799
+    },
+    {
+      "epoch": 2.687469995199232,
+      "grad_norm": 1.3124969202474654,
+      "learning_rate": 5.222698654793936e-06,
+      "loss": 0.4981,
+      "step": 2800
+    },
+    {
+      "epoch": 2.688430148823812,
+      "grad_norm": 1.0602400209247174,
+      "learning_rate": 5.219351969338576e-06,
+      "loss": 0.5206,
+      "step": 2801
+    },
+    {
+      "epoch": 2.6893903024483916,
+      "grad_norm": 1.266528249280877,
+      "learning_rate": 5.2160051854184616e-06,
+      "loss": 0.4813,
+      "step": 2802
+    },
+    {
+      "epoch": 2.690350456072972,
+      "grad_norm": 1.6180762661837926,
+      "learning_rate": 5.212658304535929e-06,
+      "loss": 0.5331,
+      "step": 2803
+    },
+    {
+      "epoch": 2.6913106096975516,
+      "grad_norm": 0.9858838394310673,
+      "learning_rate": 5.209311328193356e-06,
+      "loss": 0.5161,
+      "step": 2804
+    },
+    {
+      "epoch": 2.6922707633221314,
+      "grad_norm": 0.8941238520646599,
+      "learning_rate": 5.205964257893166e-06,
+      "loss": 0.5927,
+      "step": 2805
+    },
+    {
+      "epoch": 2.6932309169467112,
+      "grad_norm": 1.4707676253801614,
+      "learning_rate": 5.202617095137823e-06,
+      "loss": 0.5158,
+      "step": 2806
+    },
+    {
+      "epoch": 2.6941910705712915,
+      "grad_norm": 0.9200661719503983,
+      "learning_rate": 5.199269841429831e-06,
+      "loss": 0.5481,
+      "step": 2807
+    },
+    {
+      "epoch": 2.6951512241958713,
+      "grad_norm": 0.907416561880925,
+      "learning_rate": 5.195922498271737e-06,
+      "loss": 0.5722,
+      "step": 2808
+    },
+    {
+      "epoch": 2.6961113778204515,
+      "grad_norm": 1.8798059389301354,
+      "learning_rate": 5.192575067166126e-06,
+      "loss": 0.6103,
+      "step": 2809
+    },
+    {
+      "epoch": 2.6970715314450313,
+      "grad_norm": 1.423299903141147,
+      "learning_rate": 5.189227549615629e-06,
+      "loss": 0.6826,
+      "step": 2810
+    },
+    {
+      "epoch": 2.698031685069611,
+      "grad_norm": 0.851764037938656,
+      "learning_rate": 5.185879947122906e-06,
+      "loss": 0.6435,
+      "step": 2811
+    },
+    {
+      "epoch": 2.698991838694191,
+      "grad_norm": 1.1148566470561174,
+      "learning_rate": 5.182532261190663e-06,
+      "loss": 0.6071,
+      "step": 2812
+    },
+    {
+      "epoch": 2.699951992318771,
+      "grad_norm": 0.8721542840393476,
+      "learning_rate": 5.179184493321637e-06,
+      "loss": 0.5941,
+      "step": 2813
+    },
+    {
+      "epoch": 2.700912145943351,
+      "grad_norm": 0.9319223293987254,
+      "learning_rate": 5.175836645018608e-06,
+      "loss": 0.6197,
+      "step": 2814
+    },
+    {
+      "epoch": 2.7018722995679307,
+      "grad_norm": 0.9165388015037748,
+      "learning_rate": 5.1724887177843894e-06,
+      "loss": 0.6133,
+      "step": 2815
+    },
+    {
+      "epoch": 2.702832453192511,
+      "grad_norm": 1.0465203044552438,
+      "learning_rate": 5.1691407131218296e-06,
+      "loss": 0.5232,
+      "step": 2816
+    },
+    {
+      "epoch": 2.7037926068170908,
+      "grad_norm": 1.1803840783202921,
+      "learning_rate": 5.165792632533811e-06,
+      "loss": 0.5574,
+      "step": 2817
+    },
+    {
+      "epoch": 2.7047527604416706,
+      "grad_norm": 0.8704643622364252,
+      "learning_rate": 5.162444477523251e-06,
+      "loss": 0.4348,
+      "step": 2818
+    },
+    {
+      "epoch": 2.7057129140662504,
+      "grad_norm": 0.9190152242703277,
+      "learning_rate": 5.159096249593104e-06,
+      "loss": 0.5118,
+      "step": 2819
+    },
+    {
+      "epoch": 2.7066730676908306,
+      "grad_norm": 1.0719406102416777,
+      "learning_rate": 5.1557479502463495e-06,
+      "loss": 0.5721,
+      "step": 2820
+    },
+    {
+      "epoch": 2.7076332213154104,
+      "grad_norm": 0.9439114402850619,
+      "learning_rate": 5.152399580986005e-06,
+      "loss": 0.5997,
+      "step": 2821
+    },
+    {
+      "epoch": 2.7085933749399906,
+      "grad_norm": 0.9071815706578216,
+      "learning_rate": 5.149051143315118e-06,
+      "loss": 0.4983,
+      "step": 2822
+    },
+    {
+      "epoch": 2.7095535285645704,
+      "grad_norm": 0.9696598987887864,
+      "learning_rate": 5.1457026387367655e-06,
+      "loss": 0.5385,
+      "step": 2823
+    },
+    {
+      "epoch": 2.7105136821891502,
+      "grad_norm": 0.9638273619431147,
+      "learning_rate": 5.142354068754057e-06,
+      "loss": 0.5157,
+      "step": 2824
+    },
+    {
+      "epoch": 2.71147383581373,
+      "grad_norm": 0.8736657470189363,
+      "learning_rate": 5.139005434870128e-06,
+      "loss": 0.543,
+      "step": 2825
+    },
+    {
+      "epoch": 2.7124339894383103,
+      "grad_norm": 0.8985746559178751,
+      "learning_rate": 5.135656738588144e-06,
+      "loss": 0.5126,
+      "step": 2826
+    },
+    {
+      "epoch": 2.71339414306289,
+      "grad_norm": 0.8339644668159699,
+      "learning_rate": 5.132307981411301e-06,
+      "loss": 0.5605,
+      "step": 2827
+    },
+    {
+      "epoch": 2.71435429668747,
+      "grad_norm": 0.9584927474740805,
+      "learning_rate": 5.128959164842821e-06,
+      "loss": 0.6624,
+      "step": 2828
+    },
+    {
+      "epoch": 2.71531445031205,
+      "grad_norm": 0.9832834120811093,
+      "learning_rate": 5.125610290385948e-06,
+      "loss": 0.5626,
+      "step": 2829
+    },
+    {
+      "epoch": 2.71627460393663,
+      "grad_norm": 1.4531130506929584,
+      "learning_rate": 5.1222613595439605e-06,
+      "loss": 0.5188,
+      "step": 2830
+    },
+    {
+      "epoch": 2.7172347575612097,
+      "grad_norm": 1.3354914155952573,
+      "learning_rate": 5.118912373820154e-06,
+      "loss": 0.5278,
+      "step": 2831
+    },
+    {
+      "epoch": 2.7181949111857895,
+      "grad_norm": 1.5115030214565903,
+      "learning_rate": 5.1155633347178545e-06,
+      "loss": 0.6736,
+      "step": 2832
+    },
+    {
+      "epoch": 2.7191550648103697,
+      "grad_norm": 1.108017215305917,
+      "learning_rate": 5.1122142437404085e-06,
+      "loss": 0.5684,
+      "step": 2833
+    },
+    {
+      "epoch": 2.7201152184349495,
+      "grad_norm": 0.8892713533041173,
+      "learning_rate": 5.1088651023911905e-06,
+      "loss": 0.441,
+      "step": 2834
+    },
+    {
+      "epoch": 2.7210753720595298,
+      "grad_norm": 1.1567874326646286,
+      "learning_rate": 5.10551591217359e-06,
+      "loss": 0.5614,
+      "step": 2835
+    },
+    {
+      "epoch": 2.7220355256841096,
+      "grad_norm": 1.2315998773842656,
+      "learning_rate": 5.102166674591024e-06,
+      "loss": 0.5762,
+      "step": 2836
+    },
+    {
+      "epoch": 2.7229956793086894,
+      "grad_norm": 1.0734260610627955,
+      "learning_rate": 5.09881739114693e-06,
+      "loss": 0.4065,
+      "step": 2837
+    },
+    {
+      "epoch": 2.723955832933269,
+      "grad_norm": 1.1504831323701001,
+      "learning_rate": 5.095468063344767e-06,
+      "loss": 0.4678,
+      "step": 2838
+    },
+    {
+      "epoch": 2.7249159865578494,
+      "grad_norm": 1.339344806209807,
+      "learning_rate": 5.09211869268801e-06,
+      "loss": 0.5219,
+      "step": 2839
+    },
+    {
+      "epoch": 2.725876140182429,
+      "grad_norm": 1.1804555966763617,
+      "learning_rate": 5.088769280680156e-06,
+      "loss": 0.4341,
+      "step": 2840
+    },
+    {
+      "epoch": 2.726836293807009,
+      "grad_norm": 1.296624678563868,
+      "learning_rate": 5.08541982882472e-06,
+      "loss": 0.6106,
+      "step": 2841
+    },
+    {
+      "epoch": 2.727796447431589,
+      "grad_norm": 1.2549957277373123,
+      "learning_rate": 5.082070338625238e-06,
+      "loss": 0.5008,
+      "step": 2842
+    },
+    {
+      "epoch": 2.728756601056169,
+      "grad_norm": 0.9794688150434735,
+      "learning_rate": 5.078720811585257e-06,
+      "loss": 0.5464,
+      "step": 2843
+    },
+    {
+      "epoch": 2.729716754680749,
+      "grad_norm": 0.7783284828697912,
+      "learning_rate": 5.0753712492083455e-06,
+      "loss": 0.5024,
+      "step": 2844
+    },
+    {
+      "epoch": 2.7306769083053286,
+      "grad_norm": 1.3035420271573768,
+      "learning_rate": 5.072021652998086e-06,
+      "loss": 0.6407,
+      "step": 2845
+    },
+    {
+      "epoch": 2.731637061929909,
+      "grad_norm": 1.0125724172958173,
+      "learning_rate": 5.068672024458076e-06,
+      "loss": 0.5838,
+      "step": 2846
+    },
+    {
+      "epoch": 2.7325972155544886,
+      "grad_norm": 1.2426701036049863,
+      "learning_rate": 5.065322365091928e-06,
+      "loss": 0.6571,
+      "step": 2847
+    },
+    {
+      "epoch": 2.733557369179069,
+      "grad_norm": 1.3012295328835128,
+      "learning_rate": 5.061972676403269e-06,
+      "loss": 0.4668,
+      "step": 2848
+    },
+    {
+      "epoch": 2.7345175228036487,
+      "grad_norm": 0.8865773452230334,
+      "learning_rate": 5.058622959895738e-06,
+      "loss": 0.6179,
+      "step": 2849
+    },
+    {
+      "epoch": 2.7354776764282285,
+      "grad_norm": 1.0885644191112207,
+      "learning_rate": 5.055273217072988e-06,
+      "loss": 0.4369,
+      "step": 2850
+    },
+    {
+      "epoch": 2.7364378300528083,
+      "grad_norm": 0.9165856036911711,
+      "learning_rate": 5.051923449438679e-06,
+      "loss": 0.5733,
+      "step": 2851
+    },
+    {
+      "epoch": 2.7373979836773885,
+      "grad_norm": 0.9540132292687737,
+      "learning_rate": 5.048573658496492e-06,
+      "loss": 0.5949,
+      "step": 2852
+    },
+    {
+      "epoch": 2.7383581373019683,
+      "grad_norm": 0.8366834093479427,
+      "learning_rate": 5.0452238457501065e-06,
+      "loss": 0.4407,
+      "step": 2853
+    },
+    {
+      "epoch": 2.739318290926548,
+      "grad_norm": 0.8971465876763657,
+      "learning_rate": 5.0418740127032204e-06,
+      "loss": 0.5203,
+      "step": 2854
+    },
+    {
+      "epoch": 2.7402784445511283,
+      "grad_norm": 1.3695778579771662,
+      "learning_rate": 5.038524160859538e-06,
+      "loss": 0.5721,
+      "step": 2855
+    },
+    {
+      "epoch": 2.741238598175708,
+      "grad_norm": 1.3018734808593488,
+      "learning_rate": 5.035174291722772e-06,
+      "loss": 0.5383,
+      "step": 2856
+    },
+    {
+      "epoch": 2.742198751800288,
+      "grad_norm": 0.9550679487013927,
+      "learning_rate": 5.031824406796642e-06,
+      "loss": 0.5882,
+      "step": 2857
+    },
+    {
+      "epoch": 2.7431589054248677,
+      "grad_norm": 0.8927485043459621,
+      "learning_rate": 5.028474507584875e-06,
+      "loss": 0.6261,
+      "step": 2858
+    },
+    {
+      "epoch": 2.744119059049448,
+      "grad_norm": 0.7893565117198948,
+      "learning_rate": 5.025124595591206e-06,
+      "loss": 0.5053,
+      "step": 2859
+    },
+    {
+      "epoch": 2.7450792126740278,
+      "grad_norm": 1.0459122033902766,
+      "learning_rate": 5.0217746723193775e-06,
+      "loss": 0.5781,
+      "step": 2860
+    },
+    {
+      "epoch": 2.746039366298608,
+      "grad_norm": 1.0907240317149534,
+      "learning_rate": 5.018424739273129e-06,
+      "loss": 0.6002,
+      "step": 2861
+    },
+    {
+      "epoch": 2.746999519923188,
+      "grad_norm": 1.0419964460467377,
+      "learning_rate": 5.015074797956212e-06,
+      "loss": 0.4895,
+      "step": 2862
+    },
+    {
+      "epoch": 2.7479596735477676,
+      "grad_norm": 1.216878499598488,
+      "learning_rate": 5.01172484987238e-06,
+      "loss": 0.454,
+      "step": 2863
+    },
+    {
+      "epoch": 2.7489198271723474,
+      "grad_norm": 1.0576687649963499,
+      "learning_rate": 5.008374896525387e-06,
+      "loss": 0.509,
+      "step": 2864
+    },
+    {
+      "epoch": 2.7498799807969276,
+      "grad_norm": 1.8347390514096837,
+      "learning_rate": 5.0050249394189945e-06,
+      "loss": 0.6098,
+      "step": 2865
+    },
+    {
+      "epoch": 2.7508401344215074,
+      "grad_norm": 1.405774031281213,
+      "learning_rate": 5.001674980056959e-06,
+      "loss": 0.5591,
+      "step": 2866
+    },
+    {
+      "epoch": 2.7518002880460872,
+      "grad_norm": 1.1151900662766834,
+      "learning_rate": 4.998325019943044e-06,
+      "loss": 0.6112,
+      "step": 2867
+    },
+    {
+      "epoch": 2.7527604416706675,
+      "grad_norm": 0.9580892640928749,
+      "learning_rate": 4.994975060581007e-06,
+      "loss": 0.5234,
+      "step": 2868
+    },
+    {
+      "epoch": 2.7537205952952473,
+      "grad_norm": 0.9864450350398929,
+      "learning_rate": 4.991625103474614e-06,
+      "loss": 0.6642,
+      "step": 2869
+    },
+    {
+      "epoch": 2.754680748919827,
+      "grad_norm": 1.3753482467077893,
+      "learning_rate": 4.988275150127622e-06,
+      "loss": 0.6233,
+      "step": 2870
+    },
+    {
+      "epoch": 2.755640902544407,
+      "grad_norm": 1.0859931168811305,
+      "learning_rate": 4.984925202043789e-06,
+      "loss": 0.521,
+      "step": 2871
+    },
+    {
+      "epoch": 2.756601056168987,
+      "grad_norm": 1.0229236000477513,
+      "learning_rate": 4.981575260726873e-06,
+      "loss": 0.5044,
+      "step": 2872
+    },
+    {
+      "epoch": 2.757561209793567,
+      "grad_norm": 0.7735495482332387,
+      "learning_rate": 4.978225327680625e-06,
+      "loss": 0.4529,
+      "step": 2873
+    },
+    {
+      "epoch": 2.758521363418147,
+      "grad_norm": 0.9809142385638872,
+      "learning_rate": 4.974875404408794e-06,
+      "loss": 0.6003,
+      "step": 2874
+    },
+    {
+      "epoch": 2.759481517042727,
+      "grad_norm": 0.9247342259426234,
+      "learning_rate": 4.971525492415126e-06,
+      "loss": 0.5366,
+      "step": 2875
+    },
+    {
+      "epoch": 2.7604416706673067,
+      "grad_norm": 1.9299835428669425,
+      "learning_rate": 4.96817559320336e-06,
+      "loss": 0.4561,
+      "step": 2876
+    },
+    {
+      "epoch": 2.7614018242918865,
+      "grad_norm": 0.8949534381152232,
+      "learning_rate": 4.9648257082772305e-06,
+      "loss": 0.5095,
+      "step": 2877
+    },
+    {
+      "epoch": 2.7623619779164668,
+      "grad_norm": 1.1198384601344433,
+      "learning_rate": 4.9614758391404634e-06,
+      "loss": 0.5142,
+      "step": 2878
+    },
+    {
+      "epoch": 2.7633221315410466,
+      "grad_norm": 0.9360147241694083,
+      "learning_rate": 4.9581259872967795e-06,
+      "loss": 0.6007,
+      "step": 2879
+    },
+    {
+      "epoch": 2.7642822851656264,
+      "grad_norm": 0.9027492001719692,
+      "learning_rate": 4.954776154249896e-06,
+      "loss": 0.5609,
+      "step": 2880
+    },
+    {
+      "epoch": 2.7652424387902066,
+      "grad_norm": 1.267557537425477,
+      "learning_rate": 4.95142634150351e-06,
+      "loss": 0.6081,
+      "step": 2881
+    },
+    {
+      "epoch": 2.7662025924147864,
+      "grad_norm": 0.902476589047353,
+      "learning_rate": 4.94807655056132e-06,
+      "loss": 0.4766,
+      "step": 2882
+    },
+    {
+      "epoch": 2.767162746039366,
+      "grad_norm": 1.136589337895067,
+      "learning_rate": 4.944726782927014e-06,
+      "loss": 0.4529,
+      "step": 2883
+    },
+    {
+      "epoch": 2.768122899663946,
+      "grad_norm": 1.327656014793884,
+      "learning_rate": 4.9413770401042625e-06,
+      "loss": 0.6217,
+      "step": 2884
+    },
+    {
+      "epoch": 2.769083053288526,
+      "grad_norm": 1.5131516110491452,
+      "learning_rate": 4.938027323596732e-06,
+      "loss": 0.5914,
+      "step": 2885
+    },
+    {
+      "epoch": 2.770043206913106,
+      "grad_norm": 1.1522831309033787,
+      "learning_rate": 4.934677634908073e-06,
+      "loss": 0.5679,
+      "step": 2886
+    },
+    {
+      "epoch": 2.7710033605376863,
+      "grad_norm": 1.4848204252131276,
+      "learning_rate": 4.931327975541925e-06,
+      "loss": 0.4961,
+      "step": 2887
+    },
+    {
+      "epoch": 2.771963514162266,
+      "grad_norm": 1.0322928053227625,
+      "learning_rate": 4.927978347001916e-06,
+      "loss": 0.5455,
+      "step": 2888
+    },
+    {
+      "epoch": 2.772923667786846,
+      "grad_norm": 0.9299388285419378,
+      "learning_rate": 4.924628750791656e-06,
+      "loss": 0.5697,
+      "step": 2889
+    },
+    {
+      "epoch": 2.7738838214114256,
+      "grad_norm": 1.3715596322966592,
+      "learning_rate": 4.9212791884147456e-06,
+      "loss": 0.6172,
+      "step": 2890
+    },
+    {
+      "epoch": 2.774843975036006,
+      "grad_norm": 1.021732103832636,
+      "learning_rate": 4.917929661374764e-06,
+      "loss": 0.5135,
+      "step": 2891
+    },
+    {
+      "epoch": 2.7758041286605857,
+      "grad_norm": 1.2142376265131427,
+      "learning_rate": 4.9145801711752804e-06,
+      "loss": 0.5315,
+      "step": 2892
+    },
+    {
+      "epoch": 2.7767642822851655,
+      "grad_norm": 0.982717420233173,
+      "learning_rate": 4.911230719319846e-06,
+      "loss": 0.5403,
+      "step": 2893
+    },
+    {
+      "epoch": 2.7777244359097457,
+      "grad_norm": 1.1441271424914692,
+      "learning_rate": 4.907881307311992e-06,
+      "loss": 0.5888,
+      "step": 2894
+    },
+    {
+      "epoch": 2.7786845895343255,
+      "grad_norm": 1.109638211799611,
+      "learning_rate": 4.904531936655236e-06,
+      "loss": 0.513,
+      "step": 2895
+    },
+    {
+      "epoch": 2.7796447431589053,
+      "grad_norm": 1.1796956457561363,
+      "learning_rate": 4.901182608853071e-06,
+      "loss": 0.5994,
+      "step": 2896
+    },
+    {
+      "epoch": 2.780604896783485,
+      "grad_norm": 0.9698122948469821,
+      "learning_rate": 4.897833325408977e-06,
+      "loss": 0.4911,
+      "step": 2897
+    },
+    {
+      "epoch": 2.7815650504080653,
+      "grad_norm": 1.1345128740035557,
+      "learning_rate": 4.894484087826413e-06,
+      "loss": 0.5344,
+      "step": 2898
+    },
+    {
+      "epoch": 2.782525204032645,
+      "grad_norm": 1.1458155224917117,
+      "learning_rate": 4.891134897608811e-06,
+      "loss": 0.5364,
+      "step": 2899
+    },
+    {
+      "epoch": 2.7834853576572254,
+      "grad_norm": 0.9397418319561159,
+      "learning_rate": 4.887785756259591e-06,
+      "loss": 0.5709,
+      "step": 2900
+    },
+    {
+      "epoch": 2.784445511281805,
+      "grad_norm": 1.0170311583518241,
+      "learning_rate": 4.884436665282146e-06,
+      "loss": 0.5443,
+      "step": 2901
+    },
+    {
+      "epoch": 2.785405664906385,
+      "grad_norm": 1.2170631253268918,
+      "learning_rate": 4.881087626179847e-06,
+      "loss": 0.6374,
+      "step": 2902
+    },
+    {
+      "epoch": 2.7863658185309648,
+      "grad_norm": 1.5775108093947912,
+      "learning_rate": 4.877738640456042e-06,
+      "loss": 0.6124,
+      "step": 2903
+    },
+    {
+      "epoch": 2.787325972155545,
+      "grad_norm": 1.3087798900178886,
+      "learning_rate": 4.874389709614053e-06,
+      "loss": 0.5946,
+      "step": 2904
+    },
+    {
+      "epoch": 2.788286125780125,
+      "grad_norm": 0.9347091809549669,
+      "learning_rate": 4.871040835157181e-06,
+      "loss": 0.6385,
+      "step": 2905
+    },
+    {
+      "epoch": 2.7892462794047046,
+      "grad_norm": 0.7220209203380794,
+      "learning_rate": 4.8676920185887e-06,
+      "loss": 0.508,
+      "step": 2906
+    },
+    {
+      "epoch": 2.790206433029285,
+      "grad_norm": 0.9498971025305757,
+      "learning_rate": 4.8643432614118565e-06,
+      "loss": 0.4853,
+      "step": 2907
+    },
+    {
+      "epoch": 2.7911665866538646,
+      "grad_norm": 1.1940499656509447,
+      "learning_rate": 4.860994565129875e-06,
+      "loss": 0.6684,
+      "step": 2908
+    },
+    {
+      "epoch": 2.7921267402784444,
+      "grad_norm": 0.9148733237251847,
+      "learning_rate": 4.857645931245946e-06,
+      "loss": 0.5631,
+      "step": 2909
+    },
+    {
+      "epoch": 2.7930868939030242,
+      "grad_norm": 1.0164189280076747,
+      "learning_rate": 4.854297361263235e-06,
+      "loss": 0.5831,
+      "step": 2910
+    },
+    {
+      "epoch": 2.7940470475276045,
+      "grad_norm": 1.028809477737421,
+      "learning_rate": 4.850948856684884e-06,
+      "loss": 0.6655,
+      "step": 2911
+    },
+    {
+      "epoch": 2.7950072011521843,
+      "grad_norm": 1.3339558334636454,
+      "learning_rate": 4.8476004190139954e-06,
+      "loss": 0.5883,
+      "step": 2912
+    },
+    {
+      "epoch": 2.7959673547767645,
+      "grad_norm": 1.216316236366803,
+      "learning_rate": 4.844252049753651e-06,
+      "loss": 0.5467,
+      "step": 2913
+    },
+    {
+      "epoch": 2.7969275084013443,
+      "grad_norm": 1.0088969301176822,
+      "learning_rate": 4.840903750406898e-06,
+      "loss": 0.559,
+      "step": 2914
+    },
+    {
+      "epoch": 2.797887662025924,
+      "grad_norm": 1.0206529363082588,
+      "learning_rate": 4.8375555224767486e-06,
+      "loss": 0.5089,
+      "step": 2915
+    },
+    {
+      "epoch": 2.798847815650504,
+      "grad_norm": 1.4111941947397157,
+      "learning_rate": 4.83420736746619e-06,
+      "loss": 0.5706,
+      "step": 2916
+    },
+    {
+      "epoch": 2.799807969275084,
+      "grad_norm": 0.8486448778179049,
+      "learning_rate": 4.830859286878172e-06,
+      "loss": 0.4695,
+      "step": 2917
+    },
+    {
+      "epoch": 2.800768122899664,
+      "grad_norm": 1.2717958003470367,
+      "learning_rate": 4.8275112822156105e-06,
+      "loss": 0.4842,
+      "step": 2918
+    },
+    {
+      "epoch": 2.8017282765242437,
+      "grad_norm": 1.2693866916210295,
+      "learning_rate": 4.824163354981393e-06,
+      "loss": 0.6731,
+      "step": 2919
+    },
+    {
+      "epoch": 2.802688430148824,
+      "grad_norm": 1.639470803075458,
+      "learning_rate": 4.820815506678364e-06,
+      "loss": 0.5737,
+      "step": 2920
+    },
+    {
+      "epoch": 2.8036485837734038,
+      "grad_norm": 0.9229804157692403,
+      "learning_rate": 4.81746773880934e-06,
+      "loss": 0.576,
+      "step": 2921
+    },
+    {
+      "epoch": 2.8046087373979836,
+      "grad_norm": 0.9251750930089446,
+      "learning_rate": 4.8141200528770955e-06,
+      "loss": 0.6212,
+      "step": 2922
+    },
+    {
+      "epoch": 2.8055688910225633,
+      "grad_norm": 1.1494616993252966,
+      "learning_rate": 4.810772450384372e-06,
+      "loss": 0.4777,
+      "step": 2923
+    },
+    {
+      "epoch": 2.8065290446471436,
+      "grad_norm": 0.8265258950149946,
+      "learning_rate": 4.807424932833875e-06,
+      "loss": 0.6193,
+      "step": 2924
+    },
+    {
+      "epoch": 2.8074891982717234,
+      "grad_norm": 0.9044863619052229,
+      "learning_rate": 4.8040775017282644e-06,
+      "loss": 0.5613,
+      "step": 2925
+    },
+    {
+      "epoch": 2.8084493518963036,
+      "grad_norm": 1.158329265336177,
+      "learning_rate": 4.8007301585701726e-06,
+      "loss": 0.4908,
+      "step": 2926
+    },
+    {
+      "epoch": 2.8094095055208834,
+      "grad_norm": 1.0484283022409993,
+      "learning_rate": 4.7973829048621786e-06,
+      "loss": 0.6873,
+      "step": 2927
+    },
+    {
+      "epoch": 2.810369659145463,
+      "grad_norm": 1.0352659403810793,
+      "learning_rate": 4.794035742106834e-06,
+      "loss": 0.5028,
+      "step": 2928
+    },
+    {
+      "epoch": 2.811329812770043,
+      "grad_norm": 0.8197048707055922,
+      "learning_rate": 4.7906886718066445e-06,
+      "loss": 0.5401,
+      "step": 2929
+    },
+    {
+      "epoch": 2.8122899663946233,
+      "grad_norm": 1.5350399626221585,
+      "learning_rate": 4.7873416954640725e-06,
+      "loss": 0.5613,
+      "step": 2930
+    },
+    {
+      "epoch": 2.813250120019203,
+      "grad_norm": 0.887983419444355,
+      "learning_rate": 4.783994814581539e-06,
+      "loss": 0.54,
+      "step": 2931
+    },
+    {
+      "epoch": 2.814210273643783,
+      "grad_norm": 1.0429987583112454,
+      "learning_rate": 4.7806480306614256e-06,
+      "loss": 0.5531,
+      "step": 2932
+    },
+    {
+      "epoch": 2.815170427268363,
+      "grad_norm": 1.1935995907940316,
+      "learning_rate": 4.7773013452060654e-06,
+      "loss": 0.539,
+      "step": 2933
+    },
+    {
+      "epoch": 2.816130580892943,
+      "grad_norm": 1.2647882862564,
+      "learning_rate": 4.773954759717752e-06,
+      "loss": 0.5965,
+      "step": 2934
+    },
+    {
+      "epoch": 2.8170907345175227,
+      "grad_norm": 0.8387385243105556,
+      "learning_rate": 4.7706082756987295e-06,
+      "loss": 0.6562,
+      "step": 2935
+    },
+    {
+      "epoch": 2.8180508881421025,
+      "grad_norm": 1.2832883921377154,
+      "learning_rate": 4.7672618946512e-06,
+      "loss": 0.502,
+      "step": 2936
+    },
+    {
+      "epoch": 2.8190110417666827,
+      "grad_norm": 1.092179843546104,
+      "learning_rate": 4.76391561807732e-06,
+      "loss": 0.6711,
+      "step": 2937
+    },
+    {
+      "epoch": 2.8199711953912625,
+      "grad_norm": 1.0055939141975967,
+      "learning_rate": 4.760569447479193e-06,
+      "loss": 0.4059,
+      "step": 2938
+    },
+    {
+      "epoch": 2.8209313490158427,
+      "grad_norm": 0.9786410157357517,
+      "learning_rate": 4.757223384358884e-06,
+      "loss": 0.5426,
+      "step": 2939
+    },
+    {
+      "epoch": 2.8218915026404225,
+      "grad_norm": 1.0152212772507851,
+      "learning_rate": 4.753877430218401e-06,
+      "loss": 0.4667,
+      "step": 2940
+    },
+    {
+      "epoch": 2.8228516562650023,
+      "grad_norm": 1.0236248565204924,
+      "learning_rate": 4.750531586559707e-06,
+      "loss": 0.5683,
+      "step": 2941
+    },
+    {
+      "epoch": 2.823811809889582,
+      "grad_norm": 0.9447469720712466,
+      "learning_rate": 4.7471858548847214e-06,
+      "loss": 0.547,
+      "step": 2942
+    },
+    {
+      "epoch": 2.8247719635141624,
+      "grad_norm": 0.9256791194856672,
+      "learning_rate": 4.743840236695299e-06,
+      "loss": 0.4015,
+      "step": 2943
+    },
+    {
+      "epoch": 2.825732117138742,
+      "grad_norm": 1.0494974554288614,
+      "learning_rate": 4.740494733493258e-06,
+      "loss": 0.4664,
+      "step": 2944
+    },
+    {
+      "epoch": 2.826692270763322,
+      "grad_norm": 1.9042579098918375,
+      "learning_rate": 4.737149346780358e-06,
+      "loss": 0.6442,
+      "step": 2945
+    },
+    {
+      "epoch": 2.827652424387902,
+      "grad_norm": 1.2222045147265035,
+      "learning_rate": 4.733804078058305e-06,
+      "loss": 0.5163,
+      "step": 2946
+    },
+    {
+      "epoch": 2.828612578012482,
+      "grad_norm": 1.0208873395442013,
+      "learning_rate": 4.730458928828757e-06,
+      "loss": 0.4993,
+      "step": 2947
+    },
+    {
+      "epoch": 2.829572731637062,
+      "grad_norm": 1.0085328539903105,
+      "learning_rate": 4.7271139005933144e-06,
+      "loss": 0.511,
+      "step": 2948
+    },
+    {
+      "epoch": 2.8305328852616416,
+      "grad_norm": 0.9530941990881852,
+      "learning_rate": 4.723768994853523e-06,
+      "loss": 0.5053,
+      "step": 2949
+    },
+    {
+      "epoch": 2.831493038886222,
+      "grad_norm": 1.0544682630381126,
+      "learning_rate": 4.72042421311088e-06,
+      "loss": 0.6032,
+      "step": 2950
+    },
+    {
+      "epoch": 2.8324531925108016,
+      "grad_norm": 2.039841179674737,
+      "learning_rate": 4.717079556866818e-06,
+      "loss": 0.6091,
+      "step": 2951
+    },
+    {
+      "epoch": 2.833413346135382,
+      "grad_norm": 0.7630898576165969,
+      "learning_rate": 4.713735027622719e-06,
+      "loss": 0.4976,
+      "step": 2952
+    },
+    {
+      "epoch": 2.8343734997599617,
+      "grad_norm": 1.1800668687363773,
+      "learning_rate": 4.710390626879906e-06,
+      "loss": 0.4927,
+      "step": 2953
+    },
+    {
+      "epoch": 2.8353336533845415,
+      "grad_norm": 1.1137877328928956,
+      "learning_rate": 4.707046356139643e-06,
+      "loss": 0.5815,
+      "step": 2954
+    },
+    {
+      "epoch": 2.8362938070091213,
+      "grad_norm": 1.324999776891313,
+      "learning_rate": 4.703702216903142e-06,
+      "loss": 0.5604,
+      "step": 2955
+    },
+    {
+      "epoch": 2.8372539606337015,
+      "grad_norm": 1.2021393651999435,
+      "learning_rate": 4.700358210671545e-06,
+      "loss": 0.5078,
+      "step": 2956
+    },
+    {
+      "epoch": 2.8382141142582813,
+      "grad_norm": 0.9841798111047575,
+      "learning_rate": 4.697014338945947e-06,
+      "loss": 0.6838,
+      "step": 2957
+    },
+    {
+      "epoch": 2.839174267882861,
+      "grad_norm": 0.8593640999692991,
+      "learning_rate": 4.693670603227374e-06,
+      "loss": 0.5313,
+      "step": 2958
+    },
+    {
+      "epoch": 2.8401344215074413,
+      "grad_norm": 0.839383051904467,
+      "learning_rate": 4.690327005016792e-06,
+      "loss": 0.5929,
+      "step": 2959
+    },
+    {
+      "epoch": 2.841094575132021,
+      "grad_norm": 1.249979030318789,
+      "learning_rate": 4.686983545815107e-06,
+      "loss": 0.5601,
+      "step": 2960
+    },
+    {
+      "epoch": 2.842054728756601,
+      "grad_norm": 1.1991651360082194,
+      "learning_rate": 4.683640227123162e-06,
+      "loss": 0.5548,
+      "step": 2961
+    },
+    {
+      "epoch": 2.8430148823811807,
+      "grad_norm": 0.9230837658500113,
+      "learning_rate": 4.680297050441737e-06,
+      "loss": 0.5189,
+      "step": 2962
+    },
+    {
+      "epoch": 2.843975036005761,
+      "grad_norm": 1.0786500524276668,
+      "learning_rate": 4.67695401727155e-06,
+      "loss": 0.546,
+      "step": 2963
+    },
+    {
+      "epoch": 2.8449351896303408,
+      "grad_norm": 1.0300845286097802,
+      "learning_rate": 4.6736111291132505e-06,
+      "loss": 0.5902,
+      "step": 2964
+    },
+    {
+      "epoch": 2.845895343254921,
+      "grad_norm": 0.9355354955667121,
+      "learning_rate": 4.670268387467427e-06,
+      "loss": 0.458,
+      "step": 2965
+    },
+    {
+      "epoch": 2.846855496879501,
+      "grad_norm": 1.117434615366541,
+      "learning_rate": 4.666925793834599e-06,
+      "loss": 0.5796,
+      "step": 2966
+    },
+    {
+      "epoch": 2.8478156505040806,
+      "grad_norm": 0.9974172974227485,
+      "learning_rate": 4.663583349715222e-06,
+      "loss": 0.6005,
+      "step": 2967
+    },
+    {
+      "epoch": 2.8487758041286604,
+      "grad_norm": 1.112334304697473,
+      "learning_rate": 4.660241056609685e-06,
+      "loss": 0.4449,
+      "step": 2968
+    },
+    {
+      "epoch": 2.8497359577532406,
+      "grad_norm": 0.8208724186800234,
+      "learning_rate": 4.656898916018302e-06,
+      "loss": 0.5611,
+      "step": 2969
+    },
+    {
+      "epoch": 2.8506961113778204,
+      "grad_norm": 1.0722641318898372,
+      "learning_rate": 4.653556929441332e-06,
+      "loss": 0.6135,
+      "step": 2970
+    },
+    {
+      "epoch": 2.8516562650024007,
+      "grad_norm": 1.0965869899942944,
+      "learning_rate": 4.650215098378949e-06,
+      "loss": 0.609,
+      "step": 2971
+    },
+    {
+      "epoch": 2.8526164186269805,
+      "grad_norm": 1.3223276111201334,
+      "learning_rate": 4.646873424331272e-06,
+      "loss": 0.4943,
+      "step": 2972
+    },
+    {
+      "epoch": 2.8535765722515603,
+      "grad_norm": 0.8175684508979174,
+      "learning_rate": 4.64353190879834e-06,
+      "loss": 0.4866,
+      "step": 2973
+    },
+    {
+      "epoch": 2.85453672587614,
+      "grad_norm": 1.9024552987874916,
+      "learning_rate": 4.640190553280123e-06,
+      "loss": 0.5274,
+      "step": 2974
+    },
+    {
+      "epoch": 2.85549687950072,
+      "grad_norm": 1.5474165271644353,
+      "learning_rate": 4.636849359276521e-06,
+      "loss": 0.6289,
+      "step": 2975
+    },
+    {
+      "epoch": 2.8564570331253,
+      "grad_norm": 1.3727133407513454,
+      "learning_rate": 4.6335083282873605e-06,
+      "loss": 0.4485,
+      "step": 2976
+    },
+    {
+      "epoch": 2.85741718674988,
+      "grad_norm": 1.2338334468601542,
+      "learning_rate": 4.6301674618123924e-06,
+      "loss": 0.577,
+      "step": 2977
+    },
+    {
+      "epoch": 2.85837734037446,
+      "grad_norm": 1.1130501420510492,
+      "learning_rate": 4.626826761351299e-06,
+      "loss": 0.4546,
+      "step": 2978
+    },
+    {
+      "epoch": 2.85933749399904,
+      "grad_norm": 0.7632792686659663,
+      "learning_rate": 4.623486228403682e-06,
+      "loss": 0.4242,
+      "step": 2979
+    },
+    {
+      "epoch": 2.8602976476236197,
+      "grad_norm": 0.8560025752215499,
+      "learning_rate": 4.620145864469073e-06,
+      "loss": 0.4676,
+      "step": 2980
+    },
+    {
+      "epoch": 2.8612578012481995,
+      "grad_norm": 0.8304956239781156,
+      "learning_rate": 4.616805671046925e-06,
+      "loss": 0.5169,
+      "step": 2981
+    },
+    {
+      "epoch": 2.8622179548727797,
+      "grad_norm": 0.8369784073571794,
+      "learning_rate": 4.613465649636615e-06,
+      "loss": 0.6583,
+      "step": 2982
+    },
+    {
+      "epoch": 2.8631781084973595,
+      "grad_norm": 0.9018829233854131,
+      "learning_rate": 4.610125801737443e-06,
+      "loss": 0.4335,
+      "step": 2983
+    },
+    {
+      "epoch": 2.86413826212194,
+      "grad_norm": 1.077563234475322,
+      "learning_rate": 4.6067861288486315e-06,
+      "loss": 0.5521,
+      "step": 2984
+    },
+    {
+      "epoch": 2.8650984157465196,
+      "grad_norm": 0.9451248894082102,
+      "learning_rate": 4.603446632469322e-06,
+      "loss": 0.4489,
+      "step": 2985
+    },
+    {
+      "epoch": 2.8660585693710994,
+      "grad_norm": 0.8653059661890927,
+      "learning_rate": 4.600107314098583e-06,
+      "loss": 0.5393,
+      "step": 2986
+    },
+    {
+      "epoch": 2.867018722995679,
+      "grad_norm": 1.0572417546328456,
+      "learning_rate": 4.596768175235393e-06,
+      "loss": 0.5382,
+      "step": 2987
+    },
+    {
+      "epoch": 2.867978876620259,
+      "grad_norm": 1.019972356032487,
+      "learning_rate": 4.59342921737866e-06,
+      "loss": 0.5814,
+      "step": 2988
+    },
+    {
+      "epoch": 2.868939030244839,
+      "grad_norm": 1.3033043712718195,
+      "learning_rate": 4.590090442027206e-06,
+      "loss": 0.5162,
+      "step": 2989
+    },
+    {
+      "epoch": 2.869899183869419,
+      "grad_norm": 0.8925748328223191,
+      "learning_rate": 4.5867518506797705e-06,
+      "loss": 0.5998,
+      "step": 2990
+    },
+    {
+      "epoch": 2.8708593374939992,
+      "grad_norm": 1.3113807118490572,
+      "learning_rate": 4.583413444835013e-06,
+      "loss": 0.5451,
+      "step": 2991
+    },
+    {
+      "epoch": 2.871819491118579,
+      "grad_norm": 1.2727824784003843,
+      "learning_rate": 4.580075225991505e-06,
+      "loss": 0.5719,
+      "step": 2992
+    },
+    {
+      "epoch": 2.872779644743159,
+      "grad_norm": 1.3164380952076393,
+      "learning_rate": 4.576737195647741e-06,
+      "loss": 0.5968,
+      "step": 2993
+    },
+    {
+      "epoch": 2.8737397983677386,
+      "grad_norm": 0.9221847639444891,
+      "learning_rate": 4.573399355302124e-06,
+      "loss": 0.663,
+      "step": 2994
+    },
+    {
+      "epoch": 2.874699951992319,
+      "grad_norm": 0.916116575610588,
+      "learning_rate": 4.5700617064529774e-06,
+      "loss": 0.5737,
+      "step": 2995
+    },
+    {
+      "epoch": 2.8756601056168987,
+      "grad_norm": 1.4319594670206892,
+      "learning_rate": 4.566724250598535e-06,
+      "loss": 0.6928,
+      "step": 2996
+    },
+    {
+      "epoch": 2.876620259241479,
+      "grad_norm": 1.191313441910245,
+      "learning_rate": 4.563386989236944e-06,
+      "loss": 0.4899,
+      "step": 2997
+    },
+    {
+      "epoch": 2.8775804128660587,
+      "grad_norm": 1.0149931134472947,
+      "learning_rate": 4.560049923866266e-06,
+      "loss": 0.5079,
+      "step": 2998
+    },
+    {
+      "epoch": 2.8785405664906385,
+      "grad_norm": 1.5373731386660563,
+      "learning_rate": 4.556713055984477e-06,
+      "loss": 0.432,
+      "step": 2999
+    },
+    {
+      "epoch": 2.8795007201152183,
+      "grad_norm": 1.7074343779552237,
+      "learning_rate": 4.5533763870894555e-06,
+      "loss": 0.6384,
+      "step": 3000
+    },
+    {
+      "epoch": 2.8804608737397985,
+      "grad_norm": 0.8375219710394645,
+      "learning_rate": 4.5500399186790034e-06,
+      "loss": 0.6383,
+      "step": 3001
+    },
+    {
+      "epoch": 2.8814210273643783,
+      "grad_norm": 1.0632428704884742,
+      "learning_rate": 4.5467036522508185e-06,
+      "loss": 0.5493,
+      "step": 3002
+    },
+    {
+      "epoch": 2.882381180988958,
+      "grad_norm": 1.207052947052255,
+      "learning_rate": 4.54336758930252e-06,
+      "loss": 0.5615,
+      "step": 3003
+    },
+    {
+      "epoch": 2.8833413346135384,
+      "grad_norm": 0.7678019396949012,
+      "learning_rate": 4.54003173133163e-06,
+      "loss": 0.4637,
+      "step": 3004
+    },
+    {
+      "epoch": 2.884301488238118,
+      "grad_norm": 1.1681088860376656,
+      "learning_rate": 4.536696079835578e-06,
+      "loss": 0.6318,
+      "step": 3005
+    },
+    {
+      "epoch": 2.885261641862698,
+      "grad_norm": 1.23080410397212,
+      "learning_rate": 4.5333606363117035e-06,
+      "loss": 0.5864,
+      "step": 3006
+    },
+    {
+      "epoch": 2.8862217954872778,
+      "grad_norm": 1.0049753863707054,
+      "learning_rate": 4.530025402257252e-06,
+      "loss": 0.5366,
+      "step": 3007
+    },
+    {
+      "epoch": 2.887181949111858,
+      "grad_norm": 1.1876107138311136,
+      "learning_rate": 4.526690379169374e-06,
+      "loss": 0.5576,
+      "step": 3008
+    },
+    {
+      "epoch": 2.888142102736438,
+      "grad_norm": 0.9531136818959197,
+      "learning_rate": 4.5233555685451245e-06,
+      "loss": 0.4856,
+      "step": 3009
+    },
+    {
+      "epoch": 2.889102256361018,
+      "grad_norm": 1.2586919211655028,
+      "learning_rate": 4.5200209718814656e-06,
+      "loss": 0.5349,
+      "step": 3010
+    },
+    {
+      "epoch": 2.890062409985598,
+      "grad_norm": 1.87157955692469,
+      "learning_rate": 4.516686590675261e-06,
+      "loss": 0.636,
+      "step": 3011
+    },
+    {
+      "epoch": 2.8910225636101776,
+      "grad_norm": 1.1001353549011164,
+      "learning_rate": 4.513352426423281e-06,
+      "loss": 0.5301,
+      "step": 3012
+    },
+    {
+      "epoch": 2.8919827172347574,
+      "grad_norm": 0.8672854790851329,
+      "learning_rate": 4.510018480622193e-06,
+      "loss": 0.479,
+      "step": 3013
+    },
+    {
+      "epoch": 2.8929428708593377,
+      "grad_norm": 1.069144089238986,
+      "learning_rate": 4.506684754768573e-06,
+      "loss": 0.6274,
+      "step": 3014
+    },
+    {
+      "epoch": 2.8939030244839175,
+      "grad_norm": 1.0047158763513129,
+      "learning_rate": 4.503351250358893e-06,
+      "loss": 0.594,
+      "step": 3015
+    },
+    {
+      "epoch": 2.8948631781084972,
+      "grad_norm": 0.9946312004958189,
+      "learning_rate": 4.500017968889525e-06,
+      "loss": 0.4962,
+      "step": 3016
+    },
+    {
+      "epoch": 2.8958233317330775,
+      "grad_norm": 1.3986911433989682,
+      "learning_rate": 4.4966849118567515e-06,
+      "loss": 0.5117,
+      "step": 3017
+    },
+    {
+      "epoch": 2.8967834853576573,
+      "grad_norm": 0.7769163481046195,
+      "learning_rate": 4.4933520807567365e-06,
+      "loss": 0.5894,
+      "step": 3018
+    },
+    {
+      "epoch": 2.897743638982237,
+      "grad_norm": 1.0431629023014957,
+      "learning_rate": 4.490019477085558e-06,
+      "loss": 0.4432,
+      "step": 3019
+    },
+    {
+      "epoch": 2.898703792606817,
+      "grad_norm": 1.0246612106470556,
+      "learning_rate": 4.4866871023391865e-06,
+      "loss": 0.5405,
+      "step": 3020
+    },
+    {
+      "epoch": 2.899663946231397,
+      "grad_norm": 1.0507082713031515,
+      "learning_rate": 4.483354958013485e-06,
+      "loss": 0.5078,
+      "step": 3021
+    },
+    {
+      "epoch": 2.900624099855977,
+      "grad_norm": 1.8483135265392954,
+      "learning_rate": 4.480023045604223e-06,
+      "loss": 0.5474,
+      "step": 3022
+    },
+    {
+      "epoch": 2.901584253480557,
+      "grad_norm": 0.8890285573880536,
+      "learning_rate": 4.476691366607056e-06,
+      "loss": 0.5069,
+      "step": 3023
+    },
+    {
+      "epoch": 2.902544407105137,
+      "grad_norm": 1.4544176744902257,
+      "learning_rate": 4.473359922517541e-06,
+      "loss": 0.6014,
+      "step": 3024
+    },
+    {
+      "epoch": 2.9035045607297167,
+      "grad_norm": 0.9310882341777574,
+      "learning_rate": 4.4700287148311275e-06,
+      "loss": 0.5207,
+      "step": 3025
+    },
+    {
+      "epoch": 2.9044647143542965,
+      "grad_norm": 0.974213768892102,
+      "learning_rate": 4.466697745043158e-06,
+      "loss": 0.544,
+      "step": 3026
+    },
+    {
+      "epoch": 2.905424867978877,
+      "grad_norm": 1.4873506603586337,
+      "learning_rate": 4.463367014648872e-06,
+      "loss": 0.6402,
+      "step": 3027
+    },
+    {
+      "epoch": 2.9063850216034566,
+      "grad_norm": 1.134407923930257,
+      "learning_rate": 4.460036525143395e-06,
+      "loss": 0.6281,
+      "step": 3028
+    },
+    {
+      "epoch": 2.9073451752280364,
+      "grad_norm": 1.0120977178590904,
+      "learning_rate": 4.4567062780217486e-06,
+      "loss": 0.4844,
+      "step": 3029
+    },
+    {
+      "epoch": 2.9083053288526166,
+      "grad_norm": 1.2791076500871876,
+      "learning_rate": 4.45337627477885e-06,
+      "loss": 0.4925,
+      "step": 3030
+    },
+    {
+      "epoch": 2.9092654824771964,
+      "grad_norm": 1.0668621725583047,
+      "learning_rate": 4.450046516909496e-06,
+      "loss": 0.5593,
+      "step": 3031
+    },
+    {
+      "epoch": 2.910225636101776,
+      "grad_norm": 0.9021891829977514,
+      "learning_rate": 4.446717005908382e-06,
+      "loss": 0.5559,
+      "step": 3032
+    },
+    {
+      "epoch": 2.911185789726356,
+      "grad_norm": 1.0815019586591574,
+      "learning_rate": 4.4433877432700915e-06,
+      "loss": 0.5933,
+      "step": 3033
+    },
+    {
+      "epoch": 2.9121459433509362,
+      "grad_norm": 0.9222059200044922,
+      "learning_rate": 4.440058730489092e-06,
+      "loss": 0.5702,
+      "step": 3034
+    },
+    {
+      "epoch": 2.913106096975516,
+      "grad_norm": 0.9742529311574974,
+      "learning_rate": 4.436729969059745e-06,
+      "loss": 0.4932,
+      "step": 3035
+    },
+    {
+      "epoch": 2.9140662506000963,
+      "grad_norm": 1.2060995834914687,
+      "learning_rate": 4.433401460476291e-06,
+      "loss": 0.5679,
+      "step": 3036
+    },
+    {
+      "epoch": 2.915026404224676,
+      "grad_norm": 1.3623989808235863,
+      "learning_rate": 4.430073206232865e-06,
+      "loss": 0.6511,
+      "step": 3037
+    },
+    {
+      "epoch": 2.915986557849256,
+      "grad_norm": 0.910803855126238,
+      "learning_rate": 4.426745207823486e-06,
+      "loss": 0.4692,
+      "step": 3038
+    },
+    {
+      "epoch": 2.9169467114738357,
+      "grad_norm": 1.0472135165762646,
+      "learning_rate": 4.4234174667420546e-06,
+      "loss": 0.5371,
+      "step": 3039
+    },
+    {
+      "epoch": 2.917906865098416,
+      "grad_norm": 0.8765347699511459,
+      "learning_rate": 4.42008998448236e-06,
+      "loss": 0.5694,
+      "step": 3040
+    },
+    {
+      "epoch": 2.9188670187229957,
+      "grad_norm": 0.8886301221762489,
+      "learning_rate": 4.416762762538072e-06,
+      "loss": 0.562,
+      "step": 3041
+    },
+    {
+      "epoch": 2.9198271723475755,
+      "grad_norm": 0.9065298103819833,
+      "learning_rate": 4.413435802402744e-06,
+      "loss": 0.5103,
+      "step": 3042
+    },
+    {
+      "epoch": 2.9207873259721557,
+      "grad_norm": 1.051560868451231,
+      "learning_rate": 4.410109105569817e-06,
+      "loss": 0.5767,
+      "step": 3043
+    },
+    {
+      "epoch": 2.9217474795967355,
+      "grad_norm": 1.3605211310629468,
+      "learning_rate": 4.406782673532604e-06,
+      "loss": 0.5673,
+      "step": 3044
+    },
+    {
+      "epoch": 2.9227076332213153,
+      "grad_norm": 0.9631342778600814,
+      "learning_rate": 4.403456507784311e-06,
+      "loss": 0.6016,
+      "step": 3045
+    },
+    {
+      "epoch": 2.923667786845895,
+      "grad_norm": 1.5267605914221638,
+      "learning_rate": 4.400130609818012e-06,
+      "loss": 0.6287,
+      "step": 3046
+    },
+    {
+      "epoch": 2.9246279404704754,
+      "grad_norm": 1.2832581809596535,
+      "learning_rate": 4.3968049811266705e-06,
+      "loss": 0.5855,
+      "step": 3047
+    },
+    {
+      "epoch": 2.925588094095055,
+      "grad_norm": 1.1133356273353832,
+      "learning_rate": 4.393479623203126e-06,
+      "loss": 0.5513,
+      "step": 3048
+    },
+    {
+      "epoch": 2.9265482477196354,
+      "grad_norm": 1.3472110867020108,
+      "learning_rate": 4.390154537540095e-06,
+      "loss": 0.5396,
+      "step": 3049
+    },
+    {
+      "epoch": 2.927508401344215,
+      "grad_norm": 0.9741477875318071,
+      "learning_rate": 4.386829725630173e-06,
+      "loss": 0.5727,
+      "step": 3050
+    },
+    {
+      "epoch": 2.928468554968795,
+      "grad_norm": 1.0691734973697014,
+      "learning_rate": 4.383505188965835e-06,
+      "loss": 0.5953,
+      "step": 3051
+    },
+    {
+      "epoch": 2.929428708593375,
+      "grad_norm": 0.9946935210461184,
+      "learning_rate": 4.380180929039424e-06,
+      "loss": 0.6349,
+      "step": 3052
+    },
+    {
+      "epoch": 2.930388862217955,
+      "grad_norm": 0.9572456779873351,
+      "learning_rate": 4.376856947343171e-06,
+      "loss": 0.6243,
+      "step": 3053
+    },
+    {
+      "epoch": 2.931349015842535,
+      "grad_norm": 1.2367312077386956,
+      "learning_rate": 4.3735332453691725e-06,
+      "loss": 0.472,
+      "step": 3054
+    },
+    {
+      "epoch": 2.9323091694671146,
+      "grad_norm": 1.8085708065543822,
+      "learning_rate": 4.370209824609403e-06,
+      "loss": 0.635,
+      "step": 3055
+    },
+    {
+      "epoch": 2.933269323091695,
+      "grad_norm": 0.9167614826946134,
+      "learning_rate": 4.366886686555713e-06,
+      "loss": 0.4998,
+      "step": 3056
+    },
+    {
+      "epoch": 2.9342294767162747,
+      "grad_norm": 1.1843108961133864,
+      "learning_rate": 4.36356383269982e-06,
+      "loss": 0.5833,
+      "step": 3057
+    },
+    {
+      "epoch": 2.9351896303408544,
+      "grad_norm": 1.1558700054712718,
+      "learning_rate": 4.360241264533321e-06,
+      "loss": 0.542,
+      "step": 3058
+    },
+    {
+      "epoch": 2.9361497839654342,
+      "grad_norm": 0.9249557810139207,
+      "learning_rate": 4.356918983547678e-06,
+      "loss": 0.5136,
+      "step": 3059
+    },
+    {
+      "epoch": 2.9371099375900145,
+      "grad_norm": 0.9239812370800711,
+      "learning_rate": 4.353596991234228e-06,
+      "loss": 0.5449,
+      "step": 3060
+    },
+    {
+      "epoch": 2.9380700912145943,
+      "grad_norm": 1.8903715342860052,
+      "learning_rate": 4.3502752890841825e-06,
+      "loss": 0.5592,
+      "step": 3061
+    },
+    {
+      "epoch": 2.9390302448391745,
+      "grad_norm": 1.1373301173933708,
+      "learning_rate": 4.3469538785886115e-06,
+      "loss": 0.4991,
+      "step": 3062
+    },
+    {
+      "epoch": 2.9399903984637543,
+      "grad_norm": 1.3397196120331643,
+      "learning_rate": 4.343632761238465e-06,
+      "loss": 0.4994,
+      "step": 3063
+    },
+    {
+      "epoch": 2.940950552088334,
+      "grad_norm": 1.2104743324760188,
+      "learning_rate": 4.340311938524556e-06,
+      "loss": 0.4654,
+      "step": 3064
+    },
+    {
+      "epoch": 2.941910705712914,
+      "grad_norm": 1.1082009729444584,
+      "learning_rate": 4.336991411937566e-06,
+      "loss": 0.5203,
+      "step": 3065
+    },
+    {
+      "epoch": 2.942870859337494,
+      "grad_norm": 0.9814982846809901,
+      "learning_rate": 4.333671182968045e-06,
+      "loss": 0.5168,
+      "step": 3066
+    },
+    {
+      "epoch": 2.943831012962074,
+      "grad_norm": 0.9877106038691462,
+      "learning_rate": 4.330351253106405e-06,
+      "loss": 0.5403,
+      "step": 3067
+    },
+    {
+      "epoch": 2.9447911665866537,
+      "grad_norm": 1.1985681621184443,
+      "learning_rate": 4.327031623842931e-06,
+      "loss": 0.5349,
+      "step": 3068
+    },
+    {
+      "epoch": 2.945751320211234,
+      "grad_norm": 1.0081926647653574,
+      "learning_rate": 4.323712296667768e-06,
+      "loss": 0.5492,
+      "step": 3069
+    },
+    {
+      "epoch": 2.9467114738358138,
+      "grad_norm": 1.1225174524717747,
+      "learning_rate": 4.320393273070924e-06,
+      "loss": 0.472,
+      "step": 3070
+    },
+    {
+      "epoch": 2.9476716274603936,
+      "grad_norm": 0.9103503434742389,
+      "learning_rate": 4.317074554542277e-06,
+      "loss": 0.5209,
+      "step": 3071
+    },
+    {
+      "epoch": 2.9486317810849734,
+      "grad_norm": 1.0713157935747046,
+      "learning_rate": 4.313756142571561e-06,
+      "loss": 0.5833,
+      "step": 3072
+    },
+    {
+      "epoch": 2.9495919347095536,
+      "grad_norm": 1.4210267935329106,
+      "learning_rate": 4.310438038648376e-06,
+      "loss": 0.5986,
+      "step": 3073
+    },
+    {
+      "epoch": 2.9505520883341334,
+      "grad_norm": 1.0818317997358697,
+      "learning_rate": 4.307120244262189e-06,
+      "loss": 0.5342,
+      "step": 3074
+    },
+    {
+      "epoch": 2.9515122419587136,
+      "grad_norm": 1.038315223312765,
+      "learning_rate": 4.303802760902313e-06,
+      "loss": 0.5476,
+      "step": 3075
+    },
+    {
+      "epoch": 2.9524723955832934,
+      "grad_norm": 0.9581055780089218,
+      "learning_rate": 4.300485590057939e-06,
+      "loss": 0.5294,
+      "step": 3076
+    },
+    {
+      "epoch": 2.9534325492078732,
+      "grad_norm": 0.9610387983102177,
+      "learning_rate": 4.2971687332181035e-06,
+      "loss": 0.5661,
+      "step": 3077
+    },
+    {
+      "epoch": 2.954392702832453,
+      "grad_norm": 1.450956072149878,
+      "learning_rate": 4.293852191871712e-06,
+      "loss": 0.5403,
+      "step": 3078
+    },
+    {
+      "epoch": 2.9553528564570333,
+      "grad_norm": 0.9737547138290833,
+      "learning_rate": 4.290535967507524e-06,
+      "loss": 0.6408,
+      "step": 3079
+    },
+    {
+      "epoch": 2.956313010081613,
+      "grad_norm": 0.9700309566318754,
+      "learning_rate": 4.287220061614155e-06,
+      "loss": 0.6155,
+      "step": 3080
+    },
+    {
+      "epoch": 2.957273163706193,
+      "grad_norm": 1.468628724969963,
+      "learning_rate": 4.283904475680082e-06,
+      "loss": 0.5536,
+      "step": 3081
+    },
+    {
+      "epoch": 2.958233317330773,
+      "grad_norm": 1.2590260462192808,
+      "learning_rate": 4.280589211193635e-06,
+      "loss": 0.5509,
+      "step": 3082
+    },
+    {
+      "epoch": 2.959193470955353,
+      "grad_norm": 1.3289292495566223,
+      "learning_rate": 4.277274269643002e-06,
+      "loss": 0.6351,
+      "step": 3083
+    },
+    {
+      "epoch": 2.9601536245799327,
+      "grad_norm": 1.009392083833032,
+      "learning_rate": 4.273959652516223e-06,
+      "loss": 0.6123,
+      "step": 3084
+    },
+    {
+      "epoch": 2.9611137782045125,
+      "grad_norm": 0.881239926721038,
+      "learning_rate": 4.270645361301195e-06,
+      "loss": 0.629,
+      "step": 3085
+    },
+    {
+      "epoch": 2.9620739318290927,
+      "grad_norm": 1.2398684285711008,
+      "learning_rate": 4.267331397485666e-06,
+      "loss": 0.6218,
+      "step": 3086
+    },
+    {
+      "epoch": 2.9630340854536725,
+      "grad_norm": 1.03551343661382,
+      "learning_rate": 4.264017762557245e-06,
+      "loss": 0.5348,
+      "step": 3087
+    },
+    {
+      "epoch": 2.9639942390782528,
+      "grad_norm": 1.2133383321685893,
+      "learning_rate": 4.260704458003379e-06,
+      "loss": 0.5282,
+      "step": 3088
+    },
+    {
+      "epoch": 2.9649543927028326,
+      "grad_norm": 1.2295496409916837,
+      "learning_rate": 4.257391485311384e-06,
+      "loss": 0.5988,
+      "step": 3089
+    },
+    {
+      "epoch": 2.9659145463274124,
+      "grad_norm": 0.8450511701456784,
+      "learning_rate": 4.254078845968409e-06,
+      "loss": 0.4831,
+      "step": 3090
+    },
+    {
+      "epoch": 2.966874699951992,
+      "grad_norm": 1.1926041598670738,
+      "learning_rate": 4.2507665414614675e-06,
+      "loss": 0.5523,
+      "step": 3091
+    },
+    {
+      "epoch": 2.9678348535765724,
+      "grad_norm": 1.0802042541618795,
+      "learning_rate": 4.247454573277418e-06,
+      "loss": 0.5847,
+      "step": 3092
+    },
+    {
+      "epoch": 2.968795007201152,
+      "grad_norm": 0.8101417404705086,
+      "learning_rate": 4.244142942902965e-06,
+      "loss": 0.557,
+      "step": 3093
+    },
+    {
+      "epoch": 2.969755160825732,
+      "grad_norm": 0.9978787381280483,
+      "learning_rate": 4.240831651824665e-06,
+      "loss": 0.5475,
+      "step": 3094
+    },
+    {
+      "epoch": 2.9707153144503122,
+      "grad_norm": 0.7973446788429128,
+      "learning_rate": 4.237520701528923e-06,
+      "loss": 0.4878,
+      "step": 3095
+    },
+    {
+      "epoch": 2.971675468074892,
+      "grad_norm": 1.5089364654362984,
+      "learning_rate": 4.234210093501987e-06,
+      "loss": 0.5532,
+      "step": 3096
+    },
+    {
+      "epoch": 2.972635621699472,
+      "grad_norm": 1.2666030325705535,
+      "learning_rate": 4.230899829229955e-06,
+      "loss": 0.52,
+      "step": 3097
+    },
+    {
+      "epoch": 2.9735957753240516,
+      "grad_norm": 1.8606649883668172,
+      "learning_rate": 4.2275899101987665e-06,
+      "loss": 0.6,
+      "step": 3098
+    },
+    {
+      "epoch": 2.974555928948632,
+      "grad_norm": 0.9830532061304403,
+      "learning_rate": 4.2242803378942115e-06,
+      "loss": 0.4973,
+      "step": 3099
+    },
+    {
+      "epoch": 2.9755160825732117,
+      "grad_norm": 0.8637325283470042,
+      "learning_rate": 4.220971113801921e-06,
+      "loss": 0.5082,
+      "step": 3100
+    },
+    {
+      "epoch": 2.976476236197792,
+      "grad_norm": 1.0487416355119943,
+      "learning_rate": 4.2176622394073705e-06,
+      "loss": 0.5563,
+      "step": 3101
+    },
+    {
+      "epoch": 2.9774363898223717,
+      "grad_norm": 1.1094095686360002,
+      "learning_rate": 4.214353716195879e-06,
+      "loss": 0.5545,
+      "step": 3102
+    },
+    {
+      "epoch": 2.9783965434469515,
+      "grad_norm": 1.4765371168803285,
+      "learning_rate": 4.211045545652604e-06,
+      "loss": 0.6027,
+      "step": 3103
+    },
+    {
+      "epoch": 2.9793566970715313,
+      "grad_norm": 1.1411382826006238,
+      "learning_rate": 4.207737729262549e-06,
+      "loss": 0.5777,
+      "step": 3104
+    },
+    {
+      "epoch": 2.9803168506961115,
+      "grad_norm": 1.0429061032865632,
+      "learning_rate": 4.2044302685105635e-06,
+      "loss": 0.599,
+      "step": 3105
+    },
+    {
+      "epoch": 2.9812770043206913,
+      "grad_norm": 0.9688763849898634,
+      "learning_rate": 4.201123164881321e-06,
+      "loss": 0.54,
+      "step": 3106
+    },
+    {
+      "epoch": 2.982237157945271,
+      "grad_norm": 0.8287155643391799,
+      "learning_rate": 4.1978164198593524e-06,
+      "loss": 0.5797,
+      "step": 3107
+    },
+    {
+      "epoch": 2.9831973115698514,
+      "grad_norm": 1.0522580344235575,
+      "learning_rate": 4.194510034929019e-06,
+      "loss": 0.6053,
+      "step": 3108
+    },
+    {
+      "epoch": 2.984157465194431,
+      "grad_norm": 1.1055882647196837,
+      "learning_rate": 4.19120401157452e-06,
+      "loss": 0.4796,
+      "step": 3109
+    },
+    {
+      "epoch": 2.985117618819011,
+      "grad_norm": 0.79377984148004,
+      "learning_rate": 4.187898351279895e-06,
+      "loss": 0.3955,
+      "step": 3110
+    },
+    {
+      "epoch": 2.9860777724435907,
+      "grad_norm": 0.9330418861567719,
+      "learning_rate": 4.184593055529018e-06,
+      "loss": 0.4637,
+      "step": 3111
+    },
+    {
+      "epoch": 2.987037926068171,
+      "grad_norm": 1.0066222445936424,
+      "learning_rate": 4.181288125805603e-06,
+      "loss": 0.5016,
+      "step": 3112
+    },
+    {
+      "epoch": 2.9879980796927508,
+      "grad_norm": 0.7760807705242411,
+      "learning_rate": 4.177983563593198e-06,
+      "loss": 0.4345,
+      "step": 3113
+    },
+    {
+      "epoch": 2.988958233317331,
+      "grad_norm": 1.0514174207579965,
+      "learning_rate": 4.174679370375182e-06,
+      "loss": 0.4325,
+      "step": 3114
+    },
+    {
+      "epoch": 2.989918386941911,
+      "grad_norm": 0.9141277819725362,
+      "learning_rate": 4.1713755476347766e-06,
+      "loss": 0.5234,
+      "step": 3115
+    },
+    {
+      "epoch": 2.9908785405664906,
+      "grad_norm": 1.7132038904312563,
+      "learning_rate": 4.168072096855029e-06,
+      "loss": 0.5674,
+      "step": 3116
+    },
+    {
+      "epoch": 2.9918386941910704,
+      "grad_norm": 1.1024169808053594,
+      "learning_rate": 4.164769019518823e-06,
+      "loss": 0.5336,
+      "step": 3117
+    },
+    {
+      "epoch": 2.9927988478156506,
+      "grad_norm": 1.2565038017923282,
+      "learning_rate": 4.16146631710888e-06,
+      "loss": 0.5238,
+      "step": 3118
+    },
+    {
+      "epoch": 2.9937590014402304,
+      "grad_norm": 1.7202871269275235,
+      "learning_rate": 4.15816399110774e-06,
+      "loss": 0.6322,
+      "step": 3119
+    },
+    {
+      "epoch": 2.9947191550648102,
+      "grad_norm": 1.30347203310255,
+      "learning_rate": 4.154862042997789e-06,
+      "loss": 0.6312,
+      "step": 3120
+    },
+    {
+      "epoch": 2.9956793086893905,
+      "grad_norm": 1.2767992360688722,
+      "learning_rate": 4.151560474261229e-06,
+      "loss": 0.4833,
+      "step": 3121
+    },
+    {
+      "epoch": 2.9966394623139703,
+      "grad_norm": 1.5121299471074614,
+      "learning_rate": 4.148259286380105e-06,
+      "loss": 0.6428,
+      "step": 3122
+    },
+    {
+      "epoch": 2.99759961593855,
+      "grad_norm": 0.9268698627588248,
+      "learning_rate": 4.144958480836283e-06,
+      "loss": 0.5043,
+      "step": 3123
+    },
+    {
+      "epoch": 2.99855976956313,
+      "grad_norm": 1.2804185748689254,
+      "learning_rate": 4.141658059111459e-06,
+      "loss": 0.4551,
+      "step": 3124
+    },
+    {
+      "epoch": 2.99951992318771,
+      "grad_norm": 1.406455168534148,
+      "learning_rate": 4.138358022687156e-06,
+      "loss": 0.5526,
+      "step": 3125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.406455168534148,
+      "learning_rate": 4.135058373044728e-06,
+      "loss": 0.4663,
+      "step": 3126
+    },
+    {
+      "epoch": 3.00096015362458,
+      "grad_norm": 1.9239636981940516,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.5365,
+      "step": 3127
+    },
+    {
+      "epoch": 3.00192030724916,
+      "grad_norm": 1.3228733602616092,
+      "learning_rate": 4.128460240030025e-06,
+      "loss": 0.4984,
+      "step": 3128
+    },
+    {
+      "epoch": 3.00288046087374,
+      "grad_norm": 1.0793428323285756,
+      "learning_rate": 4.125161759619582e-06,
+      "loss": 0.4681,
+      "step": 3129
+    },
+    {
+      "epoch": 3.0038406144983196,
+      "grad_norm": 1.3004225045164435,
+      "learning_rate": 4.121863671914672e-06,
+      "loss": 0.4962,
+      "step": 3130
+    },
+    {
+      "epoch": 3.0048007681229,
+      "grad_norm": 1.063124342865363,
+      "learning_rate": 4.118565978395777e-06,
+      "loss": 0.4961,
+      "step": 3131
+    },
+    {
+      "epoch": 3.0057609217474797,
+      "grad_norm": 1.2375039499986231,
+      "learning_rate": 4.115268680543189e-06,
+      "loss": 0.4713,
+      "step": 3132
+    },
+    {
+      "epoch": 3.0067210753720595,
+      "grad_norm": 1.0197191376472794,
+      "learning_rate": 4.111971779837036e-06,
+      "loss": 0.4491,
+      "step": 3133
+    },
+    {
+      "epoch": 3.0076812289966393,
+      "grad_norm": 0.8492726664650959,
+      "learning_rate": 4.108675277757255e-06,
+      "loss": 0.5794,
+      "step": 3134
+    },
+    {
+      "epoch": 3.0086413826212195,
+      "grad_norm": 1.0663777392940654,
+      "learning_rate": 4.105379175783616e-06,
+      "loss": 0.519,
+      "step": 3135
+    },
+    {
+      "epoch": 3.0096015362457993,
+      "grad_norm": 1.3478638059709642,
+      "learning_rate": 4.102083475395702e-06,
+      "loss": 0.4668,
+      "step": 3136
+    },
+    {
+      "epoch": 3.010561689870379,
+      "grad_norm": 1.0797695538189431,
+      "learning_rate": 4.098788178072917e-06,
+      "loss": 0.4843,
+      "step": 3137
+    },
+    {
+      "epoch": 3.0115218434949593,
+      "grad_norm": 1.8100865280275094,
+      "learning_rate": 4.095493285294486e-06,
+      "loss": 0.5063,
+      "step": 3138
+    },
+    {
+      "epoch": 3.012481997119539,
+      "grad_norm": 1.1855532300263525,
+      "learning_rate": 4.092198798539452e-06,
+      "loss": 0.5044,
+      "step": 3139
+    },
+    {
+      "epoch": 3.013442150744119,
+      "grad_norm": 1.1934692504859403,
+      "learning_rate": 4.088904719286671e-06,
+      "loss": 0.3971,
+      "step": 3140
+    },
+    {
+      "epoch": 3.014402304368699,
+      "grad_norm": 1.3325929130702021,
+      "learning_rate": 4.085611049014825e-06,
+      "loss": 0.4535,
+      "step": 3141
+    },
+    {
+      "epoch": 3.015362457993279,
+      "grad_norm": 1.1200829101249894,
+      "learning_rate": 4.082317789202403e-06,
+      "loss": 0.4413,
+      "step": 3142
+    },
+    {
+      "epoch": 3.0163226116178588,
+      "grad_norm": 1.3120401515246576,
+      "learning_rate": 4.079024941327715e-06,
+      "loss": 0.4709,
+      "step": 3143
+    },
+    {
+      "epoch": 3.017282765242439,
+      "grad_norm": 0.9877878500932993,
+      "learning_rate": 4.075732506868889e-06,
+      "loss": 0.5058,
+      "step": 3144
+    },
+    {
+      "epoch": 3.018242918867019,
+      "grad_norm": 1.4300568342369686,
+      "learning_rate": 4.072440487303859e-06,
+      "loss": 0.5296,
+      "step": 3145
+    },
+    {
+      "epoch": 3.0192030724915986,
+      "grad_norm": 1.1336881826643528,
+      "learning_rate": 4.06914888411038e-06,
+      "loss": 0.4567,
+      "step": 3146
+    },
+    {
+      "epoch": 3.0201632261161784,
+      "grad_norm": 1.2940453158406124,
+      "learning_rate": 4.065857698766016e-06,
+      "loss": 0.4519,
+      "step": 3147
+    },
+    {
+      "epoch": 3.0211233797407586,
+      "grad_norm": 1.7038628093050756,
+      "learning_rate": 4.062566932748143e-06,
+      "loss": 0.4981,
+      "step": 3148
+    },
+    {
+      "epoch": 3.0220835333653384,
+      "grad_norm": 1.6262079457565777,
+      "learning_rate": 4.059276587533959e-06,
+      "loss": 0.3239,
+      "step": 3149
+    },
+    {
+      "epoch": 3.023043686989918,
+      "grad_norm": 1.3827984010683527,
+      "learning_rate": 4.0559866646004546e-06,
+      "loss": 0.3996,
+      "step": 3150
+    },
+    {
+      "epoch": 3.0240038406144985,
+      "grad_norm": 1.3727816316241965,
+      "learning_rate": 4.052697165424448e-06,
+      "loss": 0.4812,
+      "step": 3151
+    },
+    {
+      "epoch": 3.0249639942390782,
+      "grad_norm": 1.8550271569220913,
+      "learning_rate": 4.049408091482556e-06,
+      "loss": 0.4482,
+      "step": 3152
+    },
+    {
+      "epoch": 3.025924147863658,
+      "grad_norm": 1.6445361829073784,
+      "learning_rate": 4.046119444251211e-06,
+      "loss": 0.537,
+      "step": 3153
+    },
+    {
+      "epoch": 3.0268843014882383,
+      "grad_norm": 1.1083472073002156,
+      "learning_rate": 4.042831225206651e-06,
+      "loss": 0.5213,
+      "step": 3154
+    },
+    {
+      "epoch": 3.027844455112818,
+      "grad_norm": 0.9762787260620396,
+      "learning_rate": 4.039543435824923e-06,
+      "loss": 0.4877,
+      "step": 3155
+    },
+    {
+      "epoch": 3.028804608737398,
+      "grad_norm": 0.696652041806107,
+      "learning_rate": 4.036256077581879e-06,
+      "loss": 0.4781,
+      "step": 3156
+    },
+    {
+      "epoch": 3.029764762361978,
+      "grad_norm": 1.4782959603246209,
+      "learning_rate": 4.032969151953182e-06,
+      "loss": 0.4845,
+      "step": 3157
+    },
+    {
+      "epoch": 3.030724915986558,
+      "grad_norm": 1.2575536480360427,
+      "learning_rate": 4.029682660414294e-06,
+      "loss": 0.553,
+      "step": 3158
+    },
+    {
+      "epoch": 3.0316850696111377,
+      "grad_norm": 1.0250506698631832,
+      "learning_rate": 4.0263966044404885e-06,
+      "loss": 0.4167,
+      "step": 3159
+    },
+    {
+      "epoch": 3.0326452232357175,
+      "grad_norm": 1.1007246132458646,
+      "learning_rate": 4.023110985506839e-06,
+      "loss": 0.4504,
+      "step": 3160
+    },
+    {
+      "epoch": 3.0336053768602977,
+      "grad_norm": 0.9604262341570069,
+      "learning_rate": 4.019825805088225e-06,
+      "loss": 0.6067,
+      "step": 3161
+    },
+    {
+      "epoch": 3.0345655304848775,
+      "grad_norm": 0.736231017264886,
+      "learning_rate": 4.0165410646593324e-06,
+      "loss": 0.6075,
+      "step": 3162
+    },
+    {
+      "epoch": 3.0355256841094573,
+      "grad_norm": 1.1126325326648259,
+      "learning_rate": 4.01325676569464e-06,
+      "loss": 0.5049,
+      "step": 3163
+    },
+    {
+      "epoch": 3.0364858377340376,
+      "grad_norm": 1.277468939327407,
+      "learning_rate": 4.009972909668442e-06,
+      "loss": 0.366,
+      "step": 3164
+    },
+    {
+      "epoch": 3.0374459913586174,
+      "grad_norm": 1.2387999787478754,
+      "learning_rate": 4.006689498054817e-06,
+      "loss": 0.576,
+      "step": 3165
+    },
+    {
+      "epoch": 3.038406144983197,
+      "grad_norm": 1.0536507323725626,
+      "learning_rate": 4.00340653232766e-06,
+      "loss": 0.5671,
+      "step": 3166
+    },
+    {
+      "epoch": 3.0393662986077774,
+      "grad_norm": 1.124388549333207,
+      "learning_rate": 4.00012401396066e-06,
+      "loss": 0.5529,
+      "step": 3167
+    },
+    {
+      "epoch": 3.040326452232357,
+      "grad_norm": 1.5898668256174617,
+      "learning_rate": 3.9968419444273e-06,
+      "loss": 0.4886,
+      "step": 3168
+    },
+    {
+      "epoch": 3.041286605856937,
+      "grad_norm": 1.3596397519775179,
+      "learning_rate": 3.993560325200869e-06,
+      "loss": 0.4261,
+      "step": 3169
+    },
+    {
+      "epoch": 3.0422467594815172,
+      "grad_norm": 1.173627102680041,
+      "learning_rate": 3.99027915775445e-06,
+      "loss": 0.3541,
+      "step": 3170
+    },
+    {
+      "epoch": 3.043206913106097,
+      "grad_norm": 1.5247708167999623,
+      "learning_rate": 3.986998443560923e-06,
+      "loss": 0.5323,
+      "step": 3171
+    },
+    {
+      "epoch": 3.044167066730677,
+      "grad_norm": 1.3288642811110243,
+      "learning_rate": 3.9837181840929665e-06,
+      "loss": 0.4968,
+      "step": 3172
+    },
+    {
+      "epoch": 3.0451272203552566,
+      "grad_norm": 1.1244635357268893,
+      "learning_rate": 3.980438380823053e-06,
+      "loss": 0.5472,
+      "step": 3173
+    },
+    {
+      "epoch": 3.046087373979837,
+      "grad_norm": 1.1514397821116547,
+      "learning_rate": 3.977159035223451e-06,
+      "loss": 0.4639,
+      "step": 3174
+    },
+    {
+      "epoch": 3.0470475276044167,
+      "grad_norm": 1.3764671192085696,
+      "learning_rate": 3.973880148766228e-06,
+      "loss": 0.5002,
+      "step": 3175
+    },
+    {
+      "epoch": 3.0480076812289965,
+      "grad_norm": 0.9376345105827826,
+      "learning_rate": 3.970601722923234e-06,
+      "loss": 0.4932,
+      "step": 3176
+    },
+    {
+      "epoch": 3.0489678348535767,
+      "grad_norm": 0.9781571016286641,
+      "learning_rate": 3.9673237591661265e-06,
+      "loss": 0.5122,
+      "step": 3177
+    },
+    {
+      "epoch": 3.0499279884781565,
+      "grad_norm": 1.1548070316901065,
+      "learning_rate": 3.964046258966342e-06,
+      "loss": 0.3892,
+      "step": 3178
+    },
+    {
+      "epoch": 3.0508881421027363,
+      "grad_norm": 1.1524801035715257,
+      "learning_rate": 3.960769223795118e-06,
+      "loss": 0.5279,
+      "step": 3179
+    },
+    {
+      "epoch": 3.0518482957273165,
+      "grad_norm": 0.9123047804535923,
+      "learning_rate": 3.957492655123483e-06,
+      "loss": 0.4769,
+      "step": 3180
+    },
+    {
+      "epoch": 3.0528084493518963,
+      "grad_norm": 1.0365703704266618,
+      "learning_rate": 3.954216554422249e-06,
+      "loss": 0.61,
+      "step": 3181
+    },
+    {
+      "epoch": 3.053768602976476,
+      "grad_norm": 1.0829501655121776,
+      "learning_rate": 3.9509409231620264e-06,
+      "loss": 0.5291,
+      "step": 3182
+    },
+    {
+      "epoch": 3.0547287566010564,
+      "grad_norm": 0.9651497491775098,
+      "learning_rate": 3.947665762813209e-06,
+      "loss": 0.4728,
+      "step": 3183
+    },
+    {
+      "epoch": 3.055688910225636,
+      "grad_norm": 1.1138336240051303,
+      "learning_rate": 3.944391074845982e-06,
+      "loss": 0.4506,
+      "step": 3184
+    },
+    {
+      "epoch": 3.056649063850216,
+      "grad_norm": 1.0769846623869541,
+      "learning_rate": 3.941116860730319e-06,
+      "loss": 0.5404,
+      "step": 3185
+    },
+    {
+      "epoch": 3.0576092174747957,
+      "grad_norm": 1.2759802897492079,
+      "learning_rate": 3.9378431219359755e-06,
+      "loss": 0.5389,
+      "step": 3186
+    },
+    {
+      "epoch": 3.058569371099376,
+      "grad_norm": 0.8355257768799682,
+      "learning_rate": 3.934569859932502e-06,
+      "loss": 0.4113,
+      "step": 3187
+    },
+    {
+      "epoch": 3.059529524723956,
+      "grad_norm": 0.7681107528899316,
+      "learning_rate": 3.9312970761892305e-06,
+      "loss": 0.4384,
+      "step": 3188
+    },
+    {
+      "epoch": 3.0604896783485356,
+      "grad_norm": 1.0274639611804495,
+      "learning_rate": 3.9280247721752765e-06,
+      "loss": 0.5275,
+      "step": 3189
+    },
+    {
+      "epoch": 3.061449831973116,
+      "grad_norm": 0.9168806243794024,
+      "learning_rate": 3.924752949359545e-06,
+      "loss": 0.4163,
+      "step": 3190
+    },
+    {
+      "epoch": 3.0624099855976956,
+      "grad_norm": 1.116107818033291,
+      "learning_rate": 3.921481609210719e-06,
+      "loss": 0.3245,
+      "step": 3191
+    },
+    {
+      "epoch": 3.0633701392222754,
+      "grad_norm": 1.3137393963694208,
+      "learning_rate": 3.918210753197269e-06,
+      "loss": 0.3701,
+      "step": 3192
+    },
+    {
+      "epoch": 3.0643302928468557,
+      "grad_norm": 0.8407706817305267,
+      "learning_rate": 3.914940382787452e-06,
+      "loss": 0.4979,
+      "step": 3193
+    },
+    {
+      "epoch": 3.0652904464714354,
+      "grad_norm": 1.598056165872704,
+      "learning_rate": 3.911670499449295e-06,
+      "loss": 0.413,
+      "step": 3194
+    },
+    {
+      "epoch": 3.0662506000960152,
+      "grad_norm": 1.4350338721525306,
+      "learning_rate": 3.908401104650621e-06,
+      "loss": 0.5327,
+      "step": 3195
+    },
+    {
+      "epoch": 3.0672107537205955,
+      "grad_norm": 1.2740972841552856,
+      "learning_rate": 3.905132199859018e-06,
+      "loss": 0.5272,
+      "step": 3196
+    },
+    {
+      "epoch": 3.0681709073451753,
+      "grad_norm": 1.0656897951599038,
+      "learning_rate": 3.901863786541869e-06,
+      "loss": 0.4698,
+      "step": 3197
+    },
+    {
+      "epoch": 3.069131060969755,
+      "grad_norm": 1.2564910897357613,
+      "learning_rate": 3.898595866166329e-06,
+      "loss": 0.4034,
+      "step": 3198
+    },
+    {
+      "epoch": 3.070091214594335,
+      "grad_norm": 0.9616820602484117,
+      "learning_rate": 3.895328440199329e-06,
+      "loss": 0.5009,
+      "step": 3199
+    },
+    {
+      "epoch": 3.071051368218915,
+      "grad_norm": 1.0296404974142745,
+      "learning_rate": 3.892061510107586e-06,
+      "loss": 0.4821,
+      "step": 3200
+    },
+    {
+      "epoch": 3.072011521843495,
+      "grad_norm": 1.1667899184325679,
+      "learning_rate": 3.888795077357587e-06,
+      "loss": 0.4823,
+      "step": 3201
+    },
+    {
+      "epoch": 3.0729716754680747,
+      "grad_norm": 0.9112201502963381,
+      "learning_rate": 3.885529143415599e-06,
+      "loss": 0.4716,
+      "step": 3202
+    },
+    {
+      "epoch": 3.073931829092655,
+      "grad_norm": 1.048266883447334,
+      "learning_rate": 3.882263709747667e-06,
+      "loss": 0.4892,
+      "step": 3203
+    },
+    {
+      "epoch": 3.0748919827172347,
+      "grad_norm": 1.3039531108592672,
+      "learning_rate": 3.878998777819607e-06,
+      "loss": 0.4388,
+      "step": 3204
+    },
+    {
+      "epoch": 3.0758521363418145,
+      "grad_norm": 1.4831555191361303,
+      "learning_rate": 3.8757343490970115e-06,
+      "loss": 0.4733,
+      "step": 3205
+    },
+    {
+      "epoch": 3.0768122899663948,
+      "grad_norm": 1.158334615855496,
+      "learning_rate": 3.872470425045251e-06,
+      "loss": 0.4721,
+      "step": 3206
+    },
+    {
+      "epoch": 3.0777724435909746,
+      "grad_norm": 1.287705739501043,
+      "learning_rate": 3.869207007129463e-06,
+      "loss": 0.4967,
+      "step": 3207
+    },
+    {
+      "epoch": 3.0787325972155544,
+      "grad_norm": 1.232998826458165,
+      "learning_rate": 3.865944096814564e-06,
+      "loss": 0.5472,
+      "step": 3208
+    },
+    {
+      "epoch": 3.0796927508401346,
+      "grad_norm": 1.0185919178748002,
+      "learning_rate": 3.862681695565234e-06,
+      "loss": 0.4131,
+      "step": 3209
+    },
+    {
+      "epoch": 3.0806529044647144,
+      "grad_norm": 0.8653549775039087,
+      "learning_rate": 3.859419804845935e-06,
+      "loss": 0.5035,
+      "step": 3210
+    },
+    {
+      "epoch": 3.081613058089294,
+      "grad_norm": 1.219744652512554,
+      "learning_rate": 3.856158426120895e-06,
+      "loss": 0.47,
+      "step": 3211
+    },
+    {
+      "epoch": 3.082573211713874,
+      "grad_norm": 1.2564983811371668,
+      "learning_rate": 3.8528975608541095e-06,
+      "loss": 0.4151,
+      "step": 3212
+    },
+    {
+      "epoch": 3.0835333653384542,
+      "grad_norm": 0.7138516531906572,
+      "learning_rate": 3.849637210509346e-06,
+      "loss": 0.4206,
+      "step": 3213
+    },
+    {
+      "epoch": 3.084493518963034,
+      "grad_norm": 1.048381836978345,
+      "learning_rate": 3.846377376550144e-06,
+      "loss": 0.4641,
+      "step": 3214
+    },
+    {
+      "epoch": 3.085453672587614,
+      "grad_norm": 1.3334048711381585,
+      "learning_rate": 3.843118060439805e-06,
+      "loss": 0.3623,
+      "step": 3215
+    },
+    {
+      "epoch": 3.086413826212194,
+      "grad_norm": 1.138159930191426,
+      "learning_rate": 3.839859263641403e-06,
+      "loss": 0.4043,
+      "step": 3216
+    },
+    {
+      "epoch": 3.087373979836774,
+      "grad_norm": 1.0229895039801273,
+      "learning_rate": 3.836600987617776e-06,
+      "loss": 0.4506,
+      "step": 3217
+    },
+    {
+      "epoch": 3.0883341334613537,
+      "grad_norm": 0.8673136493237804,
+      "learning_rate": 3.833343233831529e-06,
+      "loss": 0.4152,
+      "step": 3218
+    },
+    {
+      "epoch": 3.089294287085934,
+      "grad_norm": 0.9357362030273287,
+      "learning_rate": 3.830086003745034e-06,
+      "loss": 0.5096,
+      "step": 3219
+    },
+    {
+      "epoch": 3.0902544407105137,
+      "grad_norm": 1.4772522514161999,
+      "learning_rate": 3.826829298820425e-06,
+      "loss": 0.4703,
+      "step": 3220
+    },
+    {
+      "epoch": 3.0912145943350935,
+      "grad_norm": 1.0072406751468546,
+      "learning_rate": 3.823573120519606e-06,
+      "loss": 0.4959,
+      "step": 3221
+    },
+    {
+      "epoch": 3.0921747479596737,
+      "grad_norm": 1.2202622857467345,
+      "learning_rate": 3.8203174703042345e-06,
+      "loss": 0.4395,
+      "step": 3222
+    },
+    {
+      "epoch": 3.0931349015842535,
+      "grad_norm": 0.861688638601738,
+      "learning_rate": 3.817062349635742e-06,
+      "loss": 0.4745,
+      "step": 3223
+    },
+    {
+      "epoch": 3.0940950552088333,
+      "grad_norm": 1.1104091449681204,
+      "learning_rate": 3.813807759975317e-06,
+      "loss": 0.4227,
+      "step": 3224
+    },
+    {
+      "epoch": 3.095055208833413,
+      "grad_norm": 1.1075278256066776,
+      "learning_rate": 3.8105537027839078e-06,
+      "loss": 0.4286,
+      "step": 3225
+    },
+    {
+      "epoch": 3.0960153624579934,
+      "grad_norm": 0.8914863475256624,
+      "learning_rate": 3.8073001795222274e-06,
+      "loss": 0.3922,
+      "step": 3226
+    },
+    {
+      "epoch": 3.096975516082573,
+      "grad_norm": 1.3212083607993594,
+      "learning_rate": 3.8040471916507467e-06,
+      "loss": 0.5374,
+      "step": 3227
+    },
+    {
+      "epoch": 3.097935669707153,
+      "grad_norm": 0.9732801603490799,
+      "learning_rate": 3.8007947406296975e-06,
+      "loss": 0.4555,
+      "step": 3228
+    },
+    {
+      "epoch": 3.098895823331733,
+      "grad_norm": 1.0735942347910503,
+      "learning_rate": 3.797542827919071e-06,
+      "loss": 0.3659,
+      "step": 3229
+    },
+    {
+      "epoch": 3.099855976956313,
+      "grad_norm": 1.3439091960514233,
+      "learning_rate": 3.7942914549786152e-06,
+      "loss": 0.4808,
+      "step": 3230
+    },
+    {
+      "epoch": 3.100816130580893,
+      "grad_norm": 1.450476722211154,
+      "learning_rate": 3.7910406232678363e-06,
+      "loss": 0.4025,
+      "step": 3231
+    },
+    {
+      "epoch": 3.101776284205473,
+      "grad_norm": 1.3777931479697445,
+      "learning_rate": 3.7877903342460002e-06,
+      "loss": 0.4184,
+      "step": 3232
+    },
+    {
+      "epoch": 3.102736437830053,
+      "grad_norm": 1.4940735049135367,
+      "learning_rate": 3.7845405893721237e-06,
+      "loss": 0.4585,
+      "step": 3233
+    },
+    {
+      "epoch": 3.1036965914546326,
+      "grad_norm": 0.9654084159876766,
+      "learning_rate": 3.781291390104986e-06,
+      "loss": 0.472,
+      "step": 3234
+    },
+    {
+      "epoch": 3.104656745079213,
+      "grad_norm": 1.6243116592159903,
+      "learning_rate": 3.7780427379031137e-06,
+      "loss": 0.4137,
+      "step": 3235
+    },
+    {
+      "epoch": 3.1056168987037926,
+      "grad_norm": 1.1887888248526208,
+      "learning_rate": 3.774794634224794e-06,
+      "loss": 0.6594,
+      "step": 3236
+    },
+    {
+      "epoch": 3.1065770523283724,
+      "grad_norm": 1.4731057093868527,
+      "learning_rate": 3.7715470805280695e-06,
+      "loss": 0.4132,
+      "step": 3237
+    },
+    {
+      "epoch": 3.1075372059529527,
+      "grad_norm": 1.7175492898097233,
+      "learning_rate": 3.7683000782707258e-06,
+      "loss": 0.5084,
+      "step": 3238
+    },
+    {
+      "epoch": 3.1084973595775325,
+      "grad_norm": 1.0061148986911161,
+      "learning_rate": 3.765053628910314e-06,
+      "loss": 0.4858,
+      "step": 3239
+    },
+    {
+      "epoch": 3.1094575132021123,
+      "grad_norm": 1.1488406903906454,
+      "learning_rate": 3.7618077339041244e-06,
+      "loss": 0.4184,
+      "step": 3240
+    },
+    {
+      "epoch": 3.110417666826692,
+      "grad_norm": 1.4276915640649592,
+      "learning_rate": 3.758562394709208e-06,
+      "loss": 0.4356,
+      "step": 3241
+    },
+    {
+      "epoch": 3.1113778204512723,
+      "grad_norm": 1.289728246681116,
+      "learning_rate": 3.7553176127823643e-06,
+      "loss": 0.5564,
+      "step": 3242
+    },
+    {
+      "epoch": 3.112337974075852,
+      "grad_norm": 1.0279051391307596,
+      "learning_rate": 3.7520733895801375e-06,
+      "loss": 0.5863,
+      "step": 3243
+    },
+    {
+      "epoch": 3.113298127700432,
+      "grad_norm": 1.0665447788363311,
+      "learning_rate": 3.748829726558827e-06,
+      "loss": 0.504,
+      "step": 3244
+    },
+    {
+      "epoch": 3.114258281325012,
+      "grad_norm": 0.9704432657322218,
+      "learning_rate": 3.745586625174479e-06,
+      "loss": 0.3969,
+      "step": 3245
+    },
+    {
+      "epoch": 3.115218434949592,
+      "grad_norm": 1.070127672460268,
+      "learning_rate": 3.7423440868828842e-06,
+      "loss": 0.5753,
+      "step": 3246
+    },
+    {
+      "epoch": 3.1161785885741717,
+      "grad_norm": 1.1184348251680445,
+      "learning_rate": 3.7391021131395866e-06,
+      "loss": 0.4954,
+      "step": 3247
+    },
+    {
+      "epoch": 3.117138742198752,
+      "grad_norm": 1.0837417056276069,
+      "learning_rate": 3.7358607053998717e-06,
+      "loss": 0.5256,
+      "step": 3248
+    },
+    {
+      "epoch": 3.1180988958233318,
+      "grad_norm": 1.2362034284400025,
+      "learning_rate": 3.732619865118772e-06,
+      "loss": 0.6156,
+      "step": 3249
+    },
+    {
+      "epoch": 3.1190590494479116,
+      "grad_norm": 0.9943377433158054,
+      "learning_rate": 3.7293795937510703e-06,
+      "loss": 0.5776,
+      "step": 3250
+    },
+    {
+      "epoch": 3.120019203072492,
+      "grad_norm": 1.0699939660758246,
+      "learning_rate": 3.7261398927512836e-06,
+      "loss": 0.5344,
+      "step": 3251
+    },
+    {
+      "epoch": 3.1209793566970716,
+      "grad_norm": 1.0025947534822273,
+      "learning_rate": 3.722900763573685e-06,
+      "loss": 0.5069,
+      "step": 3252
+    },
+    {
+      "epoch": 3.1219395103216514,
+      "grad_norm": 1.140528324525897,
+      "learning_rate": 3.71966220767228e-06,
+      "loss": 0.4209,
+      "step": 3253
+    },
+    {
+      "epoch": 3.122899663946231,
+      "grad_norm": 1.2978734786116022,
+      "learning_rate": 3.7164242265008244e-06,
+      "loss": 0.491,
+      "step": 3254
+    },
+    {
+      "epoch": 3.1238598175708114,
+      "grad_norm": 0.9999203946788728,
+      "learning_rate": 3.7131868215128135e-06,
+      "loss": 0.5186,
+      "step": 3255
+    },
+    {
+      "epoch": 3.1248199711953912,
+      "grad_norm": 1.057205022313916,
+      "learning_rate": 3.7099499941614823e-06,
+      "loss": 0.4692,
+      "step": 3256
+    },
+    {
+      "epoch": 3.125780124819971,
+      "grad_norm": 1.2980050492898259,
+      "learning_rate": 3.7067137458998092e-06,
+      "loss": 0.5775,
+      "step": 3257
+    },
+    {
+      "epoch": 3.1267402784445513,
+      "grad_norm": 1.1190426964775122,
+      "learning_rate": 3.7034780781805126e-06,
+      "loss": 0.4075,
+      "step": 3258
+    },
+    {
+      "epoch": 3.127700432069131,
+      "grad_norm": 1.152389270233681,
+      "learning_rate": 3.7002429924560458e-06,
+      "loss": 0.4699,
+      "step": 3259
+    },
+    {
+      "epoch": 3.128660585693711,
+      "grad_norm": 1.216668100461093,
+      "learning_rate": 3.6970084901786075e-06,
+      "loss": 0.4619,
+      "step": 3260
+    },
+    {
+      "epoch": 3.129620739318291,
+      "grad_norm": 1.0513476971501425,
+      "learning_rate": 3.6937745728001283e-06,
+      "loss": 0.4733,
+      "step": 3261
+    },
+    {
+      "epoch": 3.130580892942871,
+      "grad_norm": 0.7897814078083351,
+      "learning_rate": 3.6905412417722807e-06,
+      "loss": 0.4976,
+      "step": 3262
+    },
+    {
+      "epoch": 3.1315410465674507,
+      "grad_norm": 1.1588697994784707,
+      "learning_rate": 3.6873084985464734e-06,
+      "loss": 0.5196,
+      "step": 3263
+    },
+    {
+      "epoch": 3.132501200192031,
+      "grad_norm": 1.0634528742230984,
+      "learning_rate": 3.6840763445738485e-06,
+      "loss": 0.5228,
+      "step": 3264
+    },
+    {
+      "epoch": 3.1334613538166107,
+      "grad_norm": 1.0666700384191108,
+      "learning_rate": 3.6808447813052873e-06,
+      "loss": 0.5539,
+      "step": 3265
+    },
+    {
+      "epoch": 3.1344215074411905,
+      "grad_norm": 1.0515651543556919,
+      "learning_rate": 3.677613810191401e-06,
+      "loss": 0.4234,
+      "step": 3266
+    },
+    {
+      "epoch": 3.1353816610657703,
+      "grad_norm": 1.1951511818551965,
+      "learning_rate": 3.674383432682539e-06,
+      "loss": 0.4887,
+      "step": 3267
+    },
+    {
+      "epoch": 3.1363418146903506,
+      "grad_norm": 0.9813920067817109,
+      "learning_rate": 3.671153650228787e-06,
+      "loss": 0.4065,
+      "step": 3268
+    },
+    {
+      "epoch": 3.1373019683149304,
+      "grad_norm": 0.9676268553436997,
+      "learning_rate": 3.6679244642799544e-06,
+      "loss": 0.4306,
+      "step": 3269
+    },
+    {
+      "epoch": 3.13826212193951,
+      "grad_norm": 1.1881893731950244,
+      "learning_rate": 3.664695876285592e-06,
+      "loss": 0.5699,
+      "step": 3270
+    },
+    {
+      "epoch": 3.1392222755640904,
+      "grad_norm": 1.461591051048004,
+      "learning_rate": 3.6614678876949767e-06,
+      "loss": 0.4189,
+      "step": 3271
+    },
+    {
+      "epoch": 3.14018242918867,
+      "grad_norm": 1.0026465226135486,
+      "learning_rate": 3.658240499957118e-06,
+      "loss": 0.3928,
+      "step": 3272
+    },
+    {
+      "epoch": 3.14114258281325,
+      "grad_norm": 1.0357425119454975,
+      "learning_rate": 3.6550137145207577e-06,
+      "loss": 0.4824,
+      "step": 3273
+    },
+    {
+      "epoch": 3.1421027364378302,
+      "grad_norm": 1.2120623370860484,
+      "learning_rate": 3.651787532834362e-06,
+      "loss": 0.4207,
+      "step": 3274
+    },
+    {
+      "epoch": 3.14306289006241,
+      "grad_norm": 1.0440246953335104,
+      "learning_rate": 3.648561956346132e-06,
+      "loss": 0.5504,
+      "step": 3275
+    },
+    {
+      "epoch": 3.14402304368699,
+      "grad_norm": 1.0017968909914516,
+      "learning_rate": 3.645336986503995e-06,
+      "loss": 0.4758,
+      "step": 3276
+    },
+    {
+      "epoch": 3.14498319731157,
+      "grad_norm": 1.1240838343915827,
+      "learning_rate": 3.642112624755603e-06,
+      "loss": 0.3923,
+      "step": 3277
+    },
+    {
+      "epoch": 3.14594335093615,
+      "grad_norm": 1.107268652506849,
+      "learning_rate": 3.6388888725483394e-06,
+      "loss": 0.4472,
+      "step": 3278
+    },
+    {
+      "epoch": 3.1469035045607296,
+      "grad_norm": 1.2371282603532567,
+      "learning_rate": 3.6356657313293104e-06,
+      "loss": 0.4953,
+      "step": 3279
+    },
+    {
+      "epoch": 3.1478636581853094,
+      "grad_norm": 1.0116247415883994,
+      "learning_rate": 3.6324432025453494e-06,
+      "loss": 0.5046,
+      "step": 3280
+    },
+    {
+      "epoch": 3.1488238118098897,
+      "grad_norm": 1.0068164058837563,
+      "learning_rate": 3.62922128764302e-06,
+      "loss": 0.5694,
+      "step": 3281
+    },
+    {
+      "epoch": 3.1497839654344695,
+      "grad_norm": 1.4933537670353163,
+      "learning_rate": 3.6259999880685987e-06,
+      "loss": 0.4549,
+      "step": 3282
+    },
+    {
+      "epoch": 3.1507441190590493,
+      "grad_norm": 0.7830821285727417,
+      "learning_rate": 3.6227793052680983e-06,
+      "loss": 0.5753,
+      "step": 3283
+    },
+    {
+      "epoch": 3.1517042726836295,
+      "grad_norm": 0.9154062087447861,
+      "learning_rate": 3.6195592406872436e-06,
+      "loss": 0.5167,
+      "step": 3284
+    },
+    {
+      "epoch": 3.1526644263082093,
+      "grad_norm": 1.0552162978473834,
+      "learning_rate": 3.6163397957714895e-06,
+      "loss": 0.5033,
+      "step": 3285
+    },
+    {
+      "epoch": 3.153624579932789,
+      "grad_norm": 1.315451938632114,
+      "learning_rate": 3.6131209719660133e-06,
+      "loss": 0.5488,
+      "step": 3286
+    },
+    {
+      "epoch": 3.1545847335573693,
+      "grad_norm": 1.3608193150497787,
+      "learning_rate": 3.609902770715705e-06,
+      "loss": 0.5327,
+      "step": 3287
+    },
+    {
+      "epoch": 3.155544887181949,
+      "grad_norm": 1.2519325582184433,
+      "learning_rate": 3.6066851934651847e-06,
+      "loss": 0.4471,
+      "step": 3288
+    },
+    {
+      "epoch": 3.156505040806529,
+      "grad_norm": 1.1592147615083936,
+      "learning_rate": 3.6034682416587885e-06,
+      "loss": 0.4497,
+      "step": 3289
+    },
+    {
+      "epoch": 3.157465194431109,
+      "grad_norm": 0.9694913809238275,
+      "learning_rate": 3.600251916740569e-06,
+      "loss": 0.4233,
+      "step": 3290
+    },
+    {
+      "epoch": 3.158425348055689,
+      "grad_norm": 0.9701111295164058,
+      "learning_rate": 3.597036220154303e-06,
+      "loss": 0.4099,
+      "step": 3291
+    },
+    {
+      "epoch": 3.1593855016802688,
+      "grad_norm": 1.0199823864484145,
+      "learning_rate": 3.5938211533434797e-06,
+      "loss": 0.4941,
+      "step": 3292
+    },
+    {
+      "epoch": 3.160345655304849,
+      "grad_norm": 1.0669644487761487,
+      "learning_rate": 3.590606717751308e-06,
+      "loss": 0.4531,
+      "step": 3293
+    },
+    {
+      "epoch": 3.161305808929429,
+      "grad_norm": 0.771997878277345,
+      "learning_rate": 3.5873929148207187e-06,
+      "loss": 0.4119,
+      "step": 3294
+    },
+    {
+      "epoch": 3.1622659625540086,
+      "grad_norm": 1.0522984652052074,
+      "learning_rate": 3.5841797459943473e-06,
+      "loss": 0.426,
+      "step": 3295
+    },
+    {
+      "epoch": 3.1632261161785884,
+      "grad_norm": 1.0066860826212287,
+      "learning_rate": 3.5809672127145563e-06,
+      "loss": 0.4972,
+      "step": 3296
+    },
+    {
+      "epoch": 3.1641862698031686,
+      "grad_norm": 1.0884200144630383,
+      "learning_rate": 3.5777553164234114e-06,
+      "loss": 0.4472,
+      "step": 3297
+    },
+    {
+      "epoch": 3.1651464234277484,
+      "grad_norm": 1.5140542399282748,
+      "learning_rate": 3.5745440585627032e-06,
+      "loss": 0.4027,
+      "step": 3298
+    },
+    {
+      "epoch": 3.1661065770523282,
+      "grad_norm": 1.1461248235172055,
+      "learning_rate": 3.571333440573932e-06,
+      "loss": 0.3802,
+      "step": 3299
+    },
+    {
+      "epoch": 3.1670667306769085,
+      "grad_norm": 0.9912716650082614,
+      "learning_rate": 3.5681234638983054e-06,
+      "loss": 0.5616,
+      "step": 3300
+    },
+    {
+      "epoch": 3.1680268843014883,
+      "grad_norm": 1.186252350558142,
+      "learning_rate": 3.5649141299767504e-06,
+      "loss": 0.443,
+      "step": 3301
+    },
+    {
+      "epoch": 3.168987037926068,
+      "grad_norm": 0.730051744463476,
+      "learning_rate": 3.561705440249903e-06,
+      "loss": 0.5011,
+      "step": 3302
+    },
+    {
+      "epoch": 3.1699471915506483,
+      "grad_norm": 1.18465251140844,
+      "learning_rate": 3.558497396158108e-06,
+      "loss": 0.5261,
+      "step": 3303
+    },
+    {
+      "epoch": 3.170907345175228,
+      "grad_norm": 1.1108408003802177,
+      "learning_rate": 3.5552899991414248e-06,
+      "loss": 0.5155,
+      "step": 3304
+    },
+    {
+      "epoch": 3.171867498799808,
+      "grad_norm": 0.8408718415219331,
+      "learning_rate": 3.5520832506396156e-06,
+      "loss": 0.5019,
+      "step": 3305
+    },
+    {
+      "epoch": 3.172827652424388,
+      "grad_norm": 1.1925771987881413,
+      "learning_rate": 3.548877152092158e-06,
+      "loss": 0.4683,
+      "step": 3306
+    },
+    {
+      "epoch": 3.173787806048968,
+      "grad_norm": 1.2642793493238076,
+      "learning_rate": 3.5456717049382373e-06,
+      "loss": 0.5608,
+      "step": 3307
+    },
+    {
+      "epoch": 3.1747479596735477,
+      "grad_norm": 0.970020546344962,
+      "learning_rate": 3.5424669106167408e-06,
+      "loss": 0.4829,
+      "step": 3308
+    },
+    {
+      "epoch": 3.1757081132981275,
+      "grad_norm": 0.9906885100727404,
+      "learning_rate": 3.5392627705662686e-06,
+      "loss": 0.4891,
+      "step": 3309
+    },
+    {
+      "epoch": 3.1766682669227078,
+      "grad_norm": 1.0518499699182908,
+      "learning_rate": 3.5360592862251242e-06,
+      "loss": 0.4512,
+      "step": 3310
+    },
+    {
+      "epoch": 3.1776284205472876,
+      "grad_norm": 1.0474757515728943,
+      "learning_rate": 3.5328564590313155e-06,
+      "loss": 0.5234,
+      "step": 3311
+    },
+    {
+      "epoch": 3.1785885741718674,
+      "grad_norm": 1.2201900305719278,
+      "learning_rate": 3.529654290422563e-06,
+      "loss": 0.5018,
+      "step": 3312
+    },
+    {
+      "epoch": 3.1795487277964476,
+      "grad_norm": 1.6103658165194548,
+      "learning_rate": 3.526452781836279e-06,
+      "loss": 0.5324,
+      "step": 3313
+    },
+    {
+      "epoch": 3.1805088814210274,
+      "grad_norm": 1.2746683664480682,
+      "learning_rate": 3.5232519347095935e-06,
+      "loss": 0.464,
+      "step": 3314
+    },
+    {
+      "epoch": 3.181469035045607,
+      "grad_norm": 1.0960185852689768,
+      "learning_rate": 3.5200517504793254e-06,
+      "loss": 0.4887,
+      "step": 3315
+    },
+    {
+      "epoch": 3.1824291886701874,
+      "grad_norm": 0.9771584574866977,
+      "learning_rate": 3.516852230582008e-06,
+      "loss": 0.3827,
+      "step": 3316
+    },
+    {
+      "epoch": 3.183389342294767,
+      "grad_norm": 0.8402776598601022,
+      "learning_rate": 3.513653376453872e-06,
+      "loss": 0.533,
+      "step": 3317
+    },
+    {
+      "epoch": 3.184349495919347,
+      "grad_norm": 1.218000048801814,
+      "learning_rate": 3.510455189530845e-06,
+      "loss": 0.477,
+      "step": 3318
+    },
+    {
+      "epoch": 3.1853096495439273,
+      "grad_norm": 1.339773296808066,
+      "learning_rate": 3.507257671248561e-06,
+      "loss": 0.4487,
+      "step": 3319
+    },
+    {
+      "epoch": 3.186269803168507,
+      "grad_norm": 1.6925606341338868,
+      "learning_rate": 3.5040608230423533e-06,
+      "loss": 0.4381,
+      "step": 3320
+    },
+    {
+      "epoch": 3.187229956793087,
+      "grad_norm": 1.6901513748600399,
+      "learning_rate": 3.5008646463472506e-06,
+      "loss": 0.4492,
+      "step": 3321
+    },
+    {
+      "epoch": 3.1881901104176666,
+      "grad_norm": 1.4941852465466554,
+      "learning_rate": 3.497669142597984e-06,
+      "loss": 0.4977,
+      "step": 3322
+    },
+    {
+      "epoch": 3.189150264042247,
+      "grad_norm": 1.7854948637874957,
+      "learning_rate": 3.4944743132289793e-06,
+      "loss": 0.5432,
+      "step": 3323
+    },
+    {
+      "epoch": 3.1901104176668267,
+      "grad_norm": 1.4064662071931047,
+      "learning_rate": 3.4912801596743615e-06,
+      "loss": 0.5089,
+      "step": 3324
+    },
+    {
+      "epoch": 3.1910705712914065,
+      "grad_norm": 1.005166713978039,
+      "learning_rate": 3.4880866833679555e-06,
+      "loss": 0.556,
+      "step": 3325
+    },
+    {
+      "epoch": 3.1920307249159867,
+      "grad_norm": 1.162207357434463,
+      "learning_rate": 3.4848938857432727e-06,
+      "loss": 0.5408,
+      "step": 3326
+    },
+    {
+      "epoch": 3.1929908785405665,
+      "grad_norm": 1.3742301676493038,
+      "learning_rate": 3.4817017682335326e-06,
+      "loss": 0.3645,
+      "step": 3327
+    },
+    {
+      "epoch": 3.1939510321651463,
+      "grad_norm": 1.1408811361356186,
+      "learning_rate": 3.478510332271635e-06,
+      "loss": 0.4802,
+      "step": 3328
+    },
+    {
+      "epoch": 3.1949111857897265,
+      "grad_norm": 1.3935825313256787,
+      "learning_rate": 3.475319579290187e-06,
+      "loss": 0.4794,
+      "step": 3329
+    },
+    {
+      "epoch": 3.1958713394143063,
+      "grad_norm": 1.5822717757408087,
+      "learning_rate": 3.4721295107214835e-06,
+      "loss": 0.4593,
+      "step": 3330
+    },
+    {
+      "epoch": 3.196831493038886,
+      "grad_norm": 0.8524738384891366,
+      "learning_rate": 3.4689401279975087e-06,
+      "loss": 0.4832,
+      "step": 3331
+    },
+    {
+      "epoch": 3.1977916466634664,
+      "grad_norm": 1.0663359354385602,
+      "learning_rate": 3.4657514325499454e-06,
+      "loss": 0.5063,
+      "step": 3332
+    },
+    {
+      "epoch": 3.198751800288046,
+      "grad_norm": 0.9170238183896795,
+      "learning_rate": 3.4625634258101636e-06,
+      "loss": 0.4861,
+      "step": 3333
+    },
+    {
+      "epoch": 3.199711953912626,
+      "grad_norm": 1.001647895389685,
+      "learning_rate": 3.459376109209226e-06,
+      "loss": 0.5152,
+      "step": 3334
+    },
+    {
+      "epoch": 3.2006721075372058,
+      "grad_norm": 1.3001995421432293,
+      "learning_rate": 3.456189484177884e-06,
+      "loss": 0.4323,
+      "step": 3335
+    },
+    {
+      "epoch": 3.201632261161786,
+      "grad_norm": 1.012916159155316,
+      "learning_rate": 3.4530035521465803e-06,
+      "loss": 0.5205,
+      "step": 3336
+    },
+    {
+      "epoch": 3.202592414786366,
+      "grad_norm": 1.1970758320366057,
+      "learning_rate": 3.4498183145454445e-06,
+      "loss": 0.4571,
+      "step": 3337
+    },
+    {
+      "epoch": 3.2035525684109456,
+      "grad_norm": 1.180635387707229,
+      "learning_rate": 3.446633772804299e-06,
+      "loss": 0.4831,
+      "step": 3338
+    },
+    {
+      "epoch": 3.204512722035526,
+      "grad_norm": 0.9896897038481279,
+      "learning_rate": 3.443449928352646e-06,
+      "loss": 0.388,
+      "step": 3339
+    },
+    {
+      "epoch": 3.2054728756601056,
+      "grad_norm": 1.0014371263730026,
+      "learning_rate": 3.4402667826196845e-06,
+      "loss": 0.5363,
+      "step": 3340
+    },
+    {
+      "epoch": 3.2064330292846854,
+      "grad_norm": 1.1995139113871789,
+      "learning_rate": 3.4370843370342887e-06,
+      "loss": 0.5185,
+      "step": 3341
+    },
+    {
+      "epoch": 3.2073931829092657,
+      "grad_norm": 1.0838767809520797,
+      "learning_rate": 3.433902593025028e-06,
+      "loss": 0.4338,
+      "step": 3342
+    },
+    {
+      "epoch": 3.2083533365338455,
+      "grad_norm": 1.0049377683233403,
+      "learning_rate": 3.430721552020154e-06,
+      "loss": 0.5724,
+      "step": 3343
+    },
+    {
+      "epoch": 3.2093134901584253,
+      "grad_norm": 1.202614872907095,
+      "learning_rate": 3.4275412154475996e-06,
+      "loss": 0.4236,
+      "step": 3344
+    },
+    {
+      "epoch": 3.2102736437830055,
+      "grad_norm": 1.362368374088749,
+      "learning_rate": 3.4243615847349855e-06,
+      "loss": 0.4111,
+      "step": 3345
+    },
+    {
+      "epoch": 3.2112337974075853,
+      "grad_norm": 1.0263026967459457,
+      "learning_rate": 3.4211826613096135e-06,
+      "loss": 0.4428,
+      "step": 3346
+    },
+    {
+      "epoch": 3.212193951032165,
+      "grad_norm": 1.1244581590935658,
+      "learning_rate": 3.418004446598467e-06,
+      "loss": 0.5376,
+      "step": 3347
+    },
+    {
+      "epoch": 3.213154104656745,
+      "grad_norm": 0.6812781864592568,
+      "learning_rate": 3.4148269420282155e-06,
+      "loss": 0.5687,
+      "step": 3348
+    },
+    {
+      "epoch": 3.214114258281325,
+      "grad_norm": 1.4033588235401515,
+      "learning_rate": 3.4116501490252037e-06,
+      "loss": 0.5465,
+      "step": 3349
+    },
+    {
+      "epoch": 3.215074411905905,
+      "grad_norm": 0.8564337119299268,
+      "learning_rate": 3.4084740690154617e-06,
+      "loss": 0.5893,
+      "step": 3350
+    },
+    {
+      "epoch": 3.2160345655304847,
+      "grad_norm": 1.0117580984207832,
+      "learning_rate": 3.405298703424699e-06,
+      "loss": 0.4107,
+      "step": 3351
+    },
+    {
+      "epoch": 3.216994719155065,
+      "grad_norm": 1.1463898157504062,
+      "learning_rate": 3.4021240536783e-06,
+      "loss": 0.4721,
+      "step": 3352
+    },
+    {
+      "epoch": 3.2179548727796448,
+      "grad_norm": 1.181012503463394,
+      "learning_rate": 3.3989501212013344e-06,
+      "loss": 0.4971,
+      "step": 3353
+    },
+    {
+      "epoch": 3.2189150264042246,
+      "grad_norm": 1.3756866362828968,
+      "learning_rate": 3.3957769074185443e-06,
+      "loss": 0.4897,
+      "step": 3354
+    },
+    {
+      "epoch": 3.219875180028805,
+      "grad_norm": 1.2784531705669089,
+      "learning_rate": 3.392604413754351e-06,
+      "loss": 0.3867,
+      "step": 3355
+    },
+    {
+      "epoch": 3.2208353336533846,
+      "grad_norm": 0.917163876562131,
+      "learning_rate": 3.3894326416328575e-06,
+      "loss": 0.4702,
+      "step": 3356
+    },
+    {
+      "epoch": 3.2217954872779644,
+      "grad_norm": 1.0948257174927034,
+      "learning_rate": 3.3862615924778326e-06,
+      "loss": 0.4988,
+      "step": 3357
+    },
+    {
+      "epoch": 3.2227556409025446,
+      "grad_norm": 1.0310106773910908,
+      "learning_rate": 3.3830912677127313e-06,
+      "loss": 0.4795,
+      "step": 3358
+    },
+    {
+      "epoch": 3.2237157945271244,
+      "grad_norm": 1.3671893523526126,
+      "learning_rate": 3.3799216687606727e-06,
+      "loss": 0.4901,
+      "step": 3359
+    },
+    {
+      "epoch": 3.224675948151704,
+      "grad_norm": 0.9566398085583736,
+      "learning_rate": 3.3767527970444614e-06,
+      "loss": 0.5341,
+      "step": 3360
+    },
+    {
+      "epoch": 3.225636101776284,
+      "grad_norm": 1.1932589789270014,
+      "learning_rate": 3.3735846539865684e-06,
+      "loss": 0.4825,
+      "step": 3361
+    },
+    {
+      "epoch": 3.2265962554008643,
+      "grad_norm": 1.2833169196774847,
+      "learning_rate": 3.370417241009138e-06,
+      "loss": 0.485,
+      "step": 3362
+    },
+    {
+      "epoch": 3.227556409025444,
+      "grad_norm": 1.6946584550643682,
+      "learning_rate": 3.367250559533989e-06,
+      "loss": 0.337,
+      "step": 3363
+    },
+    {
+      "epoch": 3.228516562650024,
+      "grad_norm": 1.208036184721561,
+      "learning_rate": 3.3640846109826107e-06,
+      "loss": 0.5515,
+      "step": 3364
+    },
+    {
+      "epoch": 3.229476716274604,
+      "grad_norm": 1.5140749661296737,
+      "learning_rate": 3.3609193967761637e-06,
+      "loss": 0.5919,
+      "step": 3365
+    },
+    {
+      "epoch": 3.230436869899184,
+      "grad_norm": 1.0667318856557129,
+      "learning_rate": 3.3577549183354797e-06,
+      "loss": 0.5171,
+      "step": 3366
+    },
+    {
+      "epoch": 3.2313970235237637,
+      "grad_norm": 1.0646065738217612,
+      "learning_rate": 3.3545911770810565e-06,
+      "loss": 0.4703,
+      "step": 3367
+    },
+    {
+      "epoch": 3.232357177148344,
+      "grad_norm": 1.1498645585646654,
+      "learning_rate": 3.3514281744330644e-06,
+      "loss": 0.4772,
+      "step": 3368
+    },
+    {
+      "epoch": 3.2333173307729237,
+      "grad_norm": 1.2698844006251553,
+      "learning_rate": 3.348265911811346e-06,
+      "loss": 0.4839,
+      "step": 3369
+    },
+    {
+      "epoch": 3.2342774843975035,
+      "grad_norm": 1.1693955675875376,
+      "learning_rate": 3.345104390635401e-06,
+      "loss": 0.4786,
+      "step": 3370
+    },
+    {
+      "epoch": 3.2352376380220838,
+      "grad_norm": 1.1242271232796532,
+      "learning_rate": 3.3419436123244076e-06,
+      "loss": 0.5496,
+      "step": 3371
+    },
+    {
+      "epoch": 3.2361977916466635,
+      "grad_norm": 0.9245939122961161,
+      "learning_rate": 3.3387835782972e-06,
+      "loss": 0.4833,
+      "step": 3372
+    },
+    {
+      "epoch": 3.2371579452712433,
+      "grad_norm": 1.1901852269025686,
+      "learning_rate": 3.3356242899722877e-06,
+      "loss": 0.4112,
+      "step": 3373
+    },
+    {
+      "epoch": 3.238118098895823,
+      "grad_norm": 0.8720786666862682,
+      "learning_rate": 3.3324657487678423e-06,
+      "loss": 0.5028,
+      "step": 3374
+    },
+    {
+      "epoch": 3.2390782525204034,
+      "grad_norm": 1.101206910697223,
+      "learning_rate": 3.3293079561016957e-06,
+      "loss": 0.5186,
+      "step": 3375
+    },
+    {
+      "epoch": 3.240038406144983,
+      "grad_norm": 1.2504122943737073,
+      "learning_rate": 3.3261509133913484e-06,
+      "loss": 0.5274,
+      "step": 3376
+    },
+    {
+      "epoch": 3.240998559769563,
+      "grad_norm": 1.3041076369141953,
+      "learning_rate": 3.3229946220539657e-06,
+      "loss": 0.4301,
+      "step": 3377
+    },
+    {
+      "epoch": 3.241958713394143,
+      "grad_norm": 1.1048596462101312,
+      "learning_rate": 3.3198390835063685e-06,
+      "loss": 0.4669,
+      "step": 3378
+    },
+    {
+      "epoch": 3.242918867018723,
+      "grad_norm": 1.2118225930095048,
+      "learning_rate": 3.3166842991650493e-06,
+      "loss": 0.3906,
+      "step": 3379
+    },
+    {
+      "epoch": 3.243879020643303,
+      "grad_norm": 1.1081278091583657,
+      "learning_rate": 3.3135302704461526e-06,
+      "loss": 0.4835,
+      "step": 3380
+    },
+    {
+      "epoch": 3.244839174267883,
+      "grad_norm": 0.9183897912325273,
+      "learning_rate": 3.31037699876549e-06,
+      "loss": 0.4727,
+      "step": 3381
+    },
+    {
+      "epoch": 3.245799327892463,
+      "grad_norm": 1.2383478227332478,
+      "learning_rate": 3.307224485538534e-06,
+      "loss": 0.5473,
+      "step": 3382
+    },
+    {
+      "epoch": 3.2467594815170426,
+      "grad_norm": 0.9179539109617296,
+      "learning_rate": 3.3040727321804083e-06,
+      "loss": 0.4235,
+      "step": 3383
+    },
+    {
+      "epoch": 3.247719635141623,
+      "grad_norm": 1.0470684611668648,
+      "learning_rate": 3.300921740105908e-06,
+      "loss": 0.5502,
+      "step": 3384
+    },
+    {
+      "epoch": 3.2486797887662027,
+      "grad_norm": 1.1523109663948459,
+      "learning_rate": 3.297771510729474e-06,
+      "loss": 0.4698,
+      "step": 3385
+    },
+    {
+      "epoch": 3.2496399423907825,
+      "grad_norm": 0.7895437107468081,
+      "learning_rate": 3.2946220454652134e-06,
+      "loss": 0.4404,
+      "step": 3386
+    },
+    {
+      "epoch": 3.2506000960153623,
+      "grad_norm": 1.3301520489344474,
+      "learning_rate": 3.2914733457268876e-06,
+      "loss": 0.4792,
+      "step": 3387
+    },
+    {
+      "epoch": 3.2515602496399425,
+      "grad_norm": 1.1369329810424718,
+      "learning_rate": 3.288325412927914e-06,
+      "loss": 0.4875,
+      "step": 3388
+    },
+    {
+      "epoch": 3.2525204032645223,
+      "grad_norm": 1.0372480042346035,
+      "learning_rate": 3.285178248481366e-06,
+      "loss": 0.5181,
+      "step": 3389
+    },
+    {
+      "epoch": 3.253480556889102,
+      "grad_norm": 1.4600363522590416,
+      "learning_rate": 3.282031853799971e-06,
+      "loss": 0.5002,
+      "step": 3390
+    },
+    {
+      "epoch": 3.2544407105136823,
+      "grad_norm": 1.8264208969823579,
+      "learning_rate": 3.278886230296113e-06,
+      "loss": 0.4852,
+      "step": 3391
+    },
+    {
+      "epoch": 3.255400864138262,
+      "grad_norm": 1.2567708763509158,
+      "learning_rate": 3.2757413793818294e-06,
+      "loss": 0.362,
+      "step": 3392
+    },
+    {
+      "epoch": 3.256361017762842,
+      "grad_norm": 1.0573259497847434,
+      "learning_rate": 3.2725973024688073e-06,
+      "loss": 0.3959,
+      "step": 3393
+    },
+    {
+      "epoch": 3.257321171387422,
+      "grad_norm": 1.3673160118755916,
+      "learning_rate": 3.2694540009683926e-06,
+      "loss": 0.4424,
+      "step": 3394
+    },
+    {
+      "epoch": 3.258281325012002,
+      "grad_norm": 1.2415590684023268,
+      "learning_rate": 3.266311476291579e-06,
+      "loss": 0.4529,
+      "step": 3395
+    },
+    {
+      "epoch": 3.2592414786365818,
+      "grad_norm": 1.316246835233954,
+      "learning_rate": 3.2631697298490105e-06,
+      "loss": 0.5188,
+      "step": 3396
+    },
+    {
+      "epoch": 3.260201632261162,
+      "grad_norm": 1.8505419386055224,
+      "learning_rate": 3.2600287630509854e-06,
+      "loss": 0.4197,
+      "step": 3397
+    },
+    {
+      "epoch": 3.261161785885742,
+      "grad_norm": 1.0858363045616548,
+      "learning_rate": 3.256888577307448e-06,
+      "loss": 0.4654,
+      "step": 3398
+    },
+    {
+      "epoch": 3.2621219395103216,
+      "grad_norm": 1.149974975061263,
+      "learning_rate": 3.2537491740279946e-06,
+      "loss": 0.4222,
+      "step": 3399
+    },
+    {
+      "epoch": 3.2630820931349014,
+      "grad_norm": 1.1122434305885254,
+      "learning_rate": 3.250610554621875e-06,
+      "loss": 0.481,
+      "step": 3400
+    },
+    {
+      "epoch": 3.2640422467594816,
+      "grad_norm": 1.0259082741978534,
+      "learning_rate": 3.2474727204979728e-06,
+      "loss": 0.5001,
+      "step": 3401
+    },
+    {
+      "epoch": 3.2650024003840614,
+      "grad_norm": 0.9838965282016903,
+      "learning_rate": 3.2443356730648363e-06,
+      "loss": 0.5344,
+      "step": 3402
+    },
+    {
+      "epoch": 3.265962554008641,
+      "grad_norm": 1.0617414570144292,
+      "learning_rate": 3.2411994137306463e-06,
+      "loss": 0.5246,
+      "step": 3403
+    },
+    {
+      "epoch": 3.2669227076332215,
+      "grad_norm": 1.1538887232016177,
+      "learning_rate": 3.23806394390324e-06,
+      "loss": 0.4985,
+      "step": 3404
+    },
+    {
+      "epoch": 3.2678828612578013,
+      "grad_norm": 0.9960359606115007,
+      "learning_rate": 3.234929264990096e-06,
+      "loss": 0.4315,
+      "step": 3405
+    },
+    {
+      "epoch": 3.268843014882381,
+      "grad_norm": 1.327478953197582,
+      "learning_rate": 3.231795378398337e-06,
+      "loss": 0.4122,
+      "step": 3406
+    },
+    {
+      "epoch": 3.2698031685069613,
+      "grad_norm": 1.0052665100987026,
+      "learning_rate": 3.2286622855347306e-06,
+      "loss": 0.4576,
+      "step": 3407
+    },
+    {
+      "epoch": 3.270763322131541,
+      "grad_norm": 0.9796713931822297,
+      "learning_rate": 3.2255299878056913e-06,
+      "loss": 0.427,
+      "step": 3408
+    },
+    {
+      "epoch": 3.271723475756121,
+      "grad_norm": 1.1915662604324562,
+      "learning_rate": 3.2223984866172715e-06,
+      "loss": 0.4556,
+      "step": 3409
+    },
+    {
+      "epoch": 3.272683629380701,
+      "grad_norm": 1.1870264380788835,
+      "learning_rate": 3.21926778337517e-06,
+      "loss": 0.5556,
+      "step": 3410
+    },
+    {
+      "epoch": 3.273643783005281,
+      "grad_norm": 1.5499519945152198,
+      "learning_rate": 3.2161378794847255e-06,
+      "loss": 0.5996,
+      "step": 3411
+    },
+    {
+      "epoch": 3.2746039366298607,
+      "grad_norm": 1.0989856023449642,
+      "learning_rate": 3.2130087763509165e-06,
+      "loss": 0.4208,
+      "step": 3412
+    },
+    {
+      "epoch": 3.2755640902544405,
+      "grad_norm": 1.1789745802119442,
+      "learning_rate": 3.2098804753783696e-06,
+      "loss": 0.4553,
+      "step": 3413
+    },
+    {
+      "epoch": 3.2765242438790207,
+      "grad_norm": 1.0881776174621691,
+      "learning_rate": 3.2067529779713384e-06,
+      "loss": 0.4282,
+      "step": 3414
+    },
+    {
+      "epoch": 3.2774843975036005,
+      "grad_norm": 0.9339671894649936,
+      "learning_rate": 3.2036262855337304e-06,
+      "loss": 0.5593,
+      "step": 3415
+    },
+    {
+      "epoch": 3.2784445511281803,
+      "grad_norm": 1.086597740555038,
+      "learning_rate": 3.200500399469077e-06,
+      "loss": 0.4424,
+      "step": 3416
+    },
+    {
+      "epoch": 3.2794047047527606,
+      "grad_norm": 1.0465027172551948,
+      "learning_rate": 3.1973753211805597e-06,
+      "loss": 0.4828,
+      "step": 3417
+    },
+    {
+      "epoch": 3.2803648583773404,
+      "grad_norm": 1.0325905508363895,
+      "learning_rate": 3.1942510520709928e-06,
+      "loss": 0.5094,
+      "step": 3418
+    },
+    {
+      "epoch": 3.28132501200192,
+      "grad_norm": 1.15513350758389,
+      "learning_rate": 3.1911275935428256e-06,
+      "loss": 0.4294,
+      "step": 3419
+    },
+    {
+      "epoch": 3.2822851656265004,
+      "grad_norm": 1.518021291501259,
+      "learning_rate": 3.1880049469981468e-06,
+      "loss": 0.4575,
+      "step": 3420
+    },
+    {
+      "epoch": 3.28324531925108,
+      "grad_norm": 1.4587076038083908,
+      "learning_rate": 3.1848831138386772e-06,
+      "loss": 0.5337,
+      "step": 3421
+    },
+    {
+      "epoch": 3.28420547287566,
+      "grad_norm": 0.8795338185889969,
+      "learning_rate": 3.1817620954657745e-06,
+      "loss": 0.4557,
+      "step": 3422
+    },
+    {
+      "epoch": 3.2851656265002402,
+      "grad_norm": 1.2431144860536827,
+      "learning_rate": 3.178641893280433e-06,
+      "loss": 0.4621,
+      "step": 3423
+    },
+    {
+      "epoch": 3.28612578012482,
+      "grad_norm": 1.5686525243364142,
+      "learning_rate": 3.1755225086832755e-06,
+      "loss": 0.4119,
+      "step": 3424
+    },
+    {
+      "epoch": 3.2870859337494,
+      "grad_norm": 0.7486346018277268,
+      "learning_rate": 3.17240394307456e-06,
+      "loss": 0.4562,
+      "step": 3425
+    },
+    {
+      "epoch": 3.2880460873739796,
+      "grad_norm": 1.2865116519965656,
+      "learning_rate": 3.169286197854181e-06,
+      "loss": 0.4699,
+      "step": 3426
+    },
+    {
+      "epoch": 3.28900624099856,
+      "grad_norm": 1.2580008937051055,
+      "learning_rate": 3.1661692744216564e-06,
+      "loss": 0.4202,
+      "step": 3427
+    },
+    {
+      "epoch": 3.2899663946231397,
+      "grad_norm": 1.3041173598167417,
+      "learning_rate": 3.163053174176145e-06,
+      "loss": 0.4516,
+      "step": 3428
+    },
+    {
+      "epoch": 3.2909265482477195,
+      "grad_norm": 1.484892422938102,
+      "learning_rate": 3.159937898516424e-06,
+      "loss": 0.4718,
+      "step": 3429
+    },
+    {
+      "epoch": 3.2918867018722997,
+      "grad_norm": 1.1966360423356948,
+      "learning_rate": 3.1568234488409126e-06,
+      "loss": 0.6373,
+      "step": 3430
+    },
+    {
+      "epoch": 3.2928468554968795,
+      "grad_norm": 1.0375136993167255,
+      "learning_rate": 3.153709826547654e-06,
+      "loss": 0.6153,
+      "step": 3431
+    },
+    {
+      "epoch": 3.2938070091214593,
+      "grad_norm": 1.3851694711256348,
+      "learning_rate": 3.1505970330343172e-06,
+      "loss": 0.5655,
+      "step": 3432
+    },
+    {
+      "epoch": 3.2947671627460395,
+      "grad_norm": 1.1507859670549394,
+      "learning_rate": 3.147485069698204e-06,
+      "loss": 0.3919,
+      "step": 3433
+    },
+    {
+      "epoch": 3.2957273163706193,
+      "grad_norm": 1.3178751563655322,
+      "learning_rate": 3.1443739379362396e-06,
+      "loss": 0.4346,
+      "step": 3434
+    },
+    {
+      "epoch": 3.296687469995199,
+      "grad_norm": 1.051420712945425,
+      "learning_rate": 3.1412636391449792e-06,
+      "loss": 0.5403,
+      "step": 3435
+    },
+    {
+      "epoch": 3.2976476236197794,
+      "grad_norm": 1.223455075404112,
+      "learning_rate": 3.1381541747206034e-06,
+      "loss": 0.4533,
+      "step": 3436
+    },
+    {
+      "epoch": 3.298607777244359,
+      "grad_norm": 0.7936865086870284,
+      "learning_rate": 3.1350455460589157e-06,
+      "loss": 0.3837,
+      "step": 3437
+    },
+    {
+      "epoch": 3.299567930868939,
+      "grad_norm": 1.0857151502458677,
+      "learning_rate": 3.131937754555347e-06,
+      "loss": 0.4601,
+      "step": 3438
+    },
+    {
+      "epoch": 3.3005280844935188,
+      "grad_norm": 0.9741083327453095,
+      "learning_rate": 3.128830801604953e-06,
+      "loss": 0.3819,
+      "step": 3439
+    },
+    {
+      "epoch": 3.301488238118099,
+      "grad_norm": 1.041916953774406,
+      "learning_rate": 3.125724688602409e-06,
+      "loss": 0.4653,
+      "step": 3440
+    },
+    {
+      "epoch": 3.302448391742679,
+      "grad_norm": 1.1966069536100719,
+      "learning_rate": 3.122619416942019e-06,
+      "loss": 0.3917,
+      "step": 3441
+    },
+    {
+      "epoch": 3.3034085453672586,
+      "grad_norm": 1.0193960101948056,
+      "learning_rate": 3.1195149880177027e-06,
+      "loss": 0.4986,
+      "step": 3442
+    },
+    {
+      "epoch": 3.304368698991839,
+      "grad_norm": 0.7881438157390115,
+      "learning_rate": 3.1164114032230063e-06,
+      "loss": 0.4653,
+      "step": 3443
+    },
+    {
+      "epoch": 3.3053288526164186,
+      "grad_norm": 0.8632803495885336,
+      "learning_rate": 3.1133086639511e-06,
+      "loss": 0.5546,
+      "step": 3444
+    },
+    {
+      "epoch": 3.3062890062409984,
+      "grad_norm": 1.1017526483484688,
+      "learning_rate": 3.1102067715947637e-06,
+      "loss": 0.626,
+      "step": 3445
+    },
+    {
+      "epoch": 3.3072491598655787,
+      "grad_norm": 1.2786833885885354,
+      "learning_rate": 3.10710572754641e-06,
+      "loss": 0.3558,
+      "step": 3446
+    },
+    {
+      "epoch": 3.3082093134901585,
+      "grad_norm": 0.9570696118206218,
+      "learning_rate": 3.104005533198058e-06,
+      "loss": 0.5335,
+      "step": 3447
+    },
+    {
+      "epoch": 3.3091694671147383,
+      "grad_norm": 1.3751514725149228,
+      "learning_rate": 3.100906189941356e-06,
+      "loss": 0.4759,
+      "step": 3448
+    },
+    {
+      "epoch": 3.3101296207393185,
+      "grad_norm": 1.2906503133724956,
+      "learning_rate": 3.0978076991675667e-06,
+      "loss": 0.461,
+      "step": 3449
+    },
+    {
+      "epoch": 3.3110897743638983,
+      "grad_norm": 1.4476935118879004,
+      "learning_rate": 3.0947100622675663e-06,
+      "loss": 0.5398,
+      "step": 3450
+    },
+    {
+      "epoch": 3.312049927988478,
+      "grad_norm": 0.9318079668790964,
+      "learning_rate": 3.091613280631853e-06,
+      "loss": 0.45,
+      "step": 3451
+    },
+    {
+      "epoch": 3.313010081613058,
+      "grad_norm": 1.1702058935334203,
+      "learning_rate": 3.08851735565054e-06,
+      "loss": 0.4344,
+      "step": 3452
+    },
+    {
+      "epoch": 3.313970235237638,
+      "grad_norm": 1.2461461360558348,
+      "learning_rate": 3.0854222887133513e-06,
+      "loss": 0.4286,
+      "step": 3453
+    },
+    {
+      "epoch": 3.314930388862218,
+      "grad_norm": 0.9763357955033047,
+      "learning_rate": 3.082328081209633e-06,
+      "loss": 0.5813,
+      "step": 3454
+    },
+    {
+      "epoch": 3.3158905424867977,
+      "grad_norm": 1.1330988220710863,
+      "learning_rate": 3.079234734528339e-06,
+      "loss": 0.4451,
+      "step": 3455
+    },
+    {
+      "epoch": 3.316850696111378,
+      "grad_norm": 1.1369137673997254,
+      "learning_rate": 3.07614225005804e-06,
+      "loss": 0.5123,
+      "step": 3456
+    },
+    {
+      "epoch": 3.3178108497359577,
+      "grad_norm": 0.9528213068630654,
+      "learning_rate": 3.0730506291869235e-06,
+      "loss": 0.4762,
+      "step": 3457
+    },
+    {
+      "epoch": 3.3187710033605375,
+      "grad_norm": 1.0022913952498307,
+      "learning_rate": 3.069959873302778e-06,
+      "loss": 0.4013,
+      "step": 3458
+    },
+    {
+      "epoch": 3.319731156985118,
+      "grad_norm": 1.1282189510727791,
+      "learning_rate": 3.0668699837930175e-06,
+      "loss": 0.5274,
+      "step": 3459
+    },
+    {
+      "epoch": 3.3206913106096976,
+      "grad_norm": 0.9659139619938799,
+      "learning_rate": 3.0637809620446547e-06,
+      "loss": 0.4892,
+      "step": 3460
+    },
+    {
+      "epoch": 3.3216514642342774,
+      "grad_norm": 0.8951935658245836,
+      "learning_rate": 3.0606928094443216e-06,
+      "loss": 0.4825,
+      "step": 3461
+    },
+    {
+      "epoch": 3.3226116178588576,
+      "grad_norm": 1.3330222605731297,
+      "learning_rate": 3.0576055273782574e-06,
+      "loss": 0.4465,
+      "step": 3462
+    },
+    {
+      "epoch": 3.3235717714834374,
+      "grad_norm": 1.0080412211279601,
+      "learning_rate": 3.0545191172323076e-06,
+      "loss": 0.4193,
+      "step": 3463
+    },
+    {
+      "epoch": 3.324531925108017,
+      "grad_norm": 1.0969758704723247,
+      "learning_rate": 3.0514335803919314e-06,
+      "loss": 0.4456,
+      "step": 3464
+    },
+    {
+      "epoch": 3.325492078732597,
+      "grad_norm": 1.3671975807837193,
+      "learning_rate": 3.048348918242191e-06,
+      "loss": 0.5039,
+      "step": 3465
+    },
+    {
+      "epoch": 3.3264522323571772,
+      "grad_norm": 1.1750093089391647,
+      "learning_rate": 3.045265132167759e-06,
+      "loss": 0.4663,
+      "step": 3466
+    },
+    {
+      "epoch": 3.327412385981757,
+      "grad_norm": 1.1575758311513453,
+      "learning_rate": 3.0421822235529154e-06,
+      "loss": 0.5039,
+      "step": 3467
+    },
+    {
+      "epoch": 3.328372539606337,
+      "grad_norm": 1.035873531954432,
+      "learning_rate": 3.0391001937815425e-06,
+      "loss": 0.4767,
+      "step": 3468
+    },
+    {
+      "epoch": 3.329332693230917,
+      "grad_norm": 1.4534583618470323,
+      "learning_rate": 3.036019044237132e-06,
+      "loss": 0.5261,
+      "step": 3469
+    },
+    {
+      "epoch": 3.330292846855497,
+      "grad_norm": 0.8341856090733533,
+      "learning_rate": 3.0329387763027794e-06,
+      "loss": 0.5003,
+      "step": 3470
+    },
+    {
+      "epoch": 3.3312530004800767,
+      "grad_norm": 1.0478543324000291,
+      "learning_rate": 3.029859391361183e-06,
+      "loss": 0.3969,
+      "step": 3471
+    },
+    {
+      "epoch": 3.332213154104657,
+      "grad_norm": 1.219496309750024,
+      "learning_rate": 3.026780890794647e-06,
+      "loss": 0.3882,
+      "step": 3472
+    },
+    {
+      "epoch": 3.3331733077292367,
+      "grad_norm": 0.9239846362761154,
+      "learning_rate": 3.0237032759850736e-06,
+      "loss": 0.5066,
+      "step": 3473
+    },
+    {
+      "epoch": 3.3341334613538165,
+      "grad_norm": 1.157510897432906,
+      "learning_rate": 3.020626548313976e-06,
+      "loss": 0.5152,
+      "step": 3474
+    },
+    {
+      "epoch": 3.3350936149783967,
+      "grad_norm": 1.2597328595495847,
+      "learning_rate": 3.017550709162464e-06,
+      "loss": 0.4714,
+      "step": 3475
+    },
+    {
+      "epoch": 3.3360537686029765,
+      "grad_norm": 0.9380508511247198,
+      "learning_rate": 3.014475759911246e-06,
+      "loss": 0.5032,
+      "step": 3476
+    },
+    {
+      "epoch": 3.3370139222275563,
+      "grad_norm": 1.0435923128503564,
+      "learning_rate": 3.0114017019406362e-06,
+      "loss": 0.4262,
+      "step": 3477
+    },
+    {
+      "epoch": 3.337974075852136,
+      "grad_norm": 1.541404588932945,
+      "learning_rate": 3.0083285366305447e-06,
+      "loss": 0.5179,
+      "step": 3478
+    },
+    {
+      "epoch": 3.3389342294767164,
+      "grad_norm": 1.3649451627519333,
+      "learning_rate": 3.0052562653604827e-06,
+      "loss": 0.4553,
+      "step": 3479
+    },
+    {
+      "epoch": 3.339894383101296,
+      "grad_norm": 1.2231930622706915,
+      "learning_rate": 3.002184889509562e-06,
+      "loss": 0.4557,
+      "step": 3480
+    },
+    {
+      "epoch": 3.340854536725876,
+      "grad_norm": 1.1887514312279568,
+      "learning_rate": 2.9991144104564883e-06,
+      "loss": 0.4144,
+      "step": 3481
+    },
+    {
+      "epoch": 3.341814690350456,
+      "grad_norm": 1.4304855176373867,
+      "learning_rate": 2.9960448295795667e-06,
+      "loss": 0.573,
+      "step": 3482
+    },
+    {
+      "epoch": 3.342774843975036,
+      "grad_norm": 1.1859827630645317,
+      "learning_rate": 2.9929761482567017e-06,
+      "loss": 0.5282,
+      "step": 3483
+    },
+    {
+      "epoch": 3.343734997599616,
+      "grad_norm": 1.1700470288724143,
+      "learning_rate": 2.9899083678653885e-06,
+      "loss": 0.4676,
+      "step": 3484
+    },
+    {
+      "epoch": 3.344695151224196,
+      "grad_norm": 1.3223092752673435,
+      "learning_rate": 2.9868414897827236e-06,
+      "loss": 0.3924,
+      "step": 3485
+    },
+    {
+      "epoch": 3.345655304848776,
+      "grad_norm": 0.993114079383793,
+      "learning_rate": 2.983775515385393e-06,
+      "loss": 0.5712,
+      "step": 3486
+    },
+    {
+      "epoch": 3.3466154584733556,
+      "grad_norm": 1.1411752022316926,
+      "learning_rate": 2.9807104460496807e-06,
+      "loss": 0.4888,
+      "step": 3487
+    },
+    {
+      "epoch": 3.347575612097936,
+      "grad_norm": 1.0269578683448715,
+      "learning_rate": 2.9776462831514664e-06,
+      "loss": 0.5026,
+      "step": 3488
+    },
+    {
+      "epoch": 3.3485357657225157,
+      "grad_norm": 1.0339014426150241,
+      "learning_rate": 2.9745830280662157e-06,
+      "loss": 0.4463,
+      "step": 3489
+    },
+    {
+      "epoch": 3.3494959193470955,
+      "grad_norm": 0.8775852931129492,
+      "learning_rate": 2.971520682168996e-06,
+      "loss": 0.4852,
+      "step": 3490
+    },
+    {
+      "epoch": 3.3504560729716752,
+      "grad_norm": 1.0932945047678582,
+      "learning_rate": 2.968459246834455e-06,
+      "loss": 0.4676,
+      "step": 3491
+    },
+    {
+      "epoch": 3.3514162265962555,
+      "grad_norm": 1.1250239389917458,
+      "learning_rate": 2.9653987234368443e-06,
+      "loss": 0.5186,
+      "step": 3492
+    },
+    {
+      "epoch": 3.3523763802208353,
+      "grad_norm": 0.9436454989420195,
+      "learning_rate": 2.9623391133499986e-06,
+      "loss": 0.4263,
+      "step": 3493
+    },
+    {
+      "epoch": 3.353336533845415,
+      "grad_norm": 1.0609472334448762,
+      "learning_rate": 2.959280417947342e-06,
+      "loss": 0.5188,
+      "step": 3494
+    },
+    {
+      "epoch": 3.3542966874699953,
+      "grad_norm": 1.202275563516378,
+      "learning_rate": 2.9562226386018935e-06,
+      "loss": 0.2793,
+      "step": 3495
+    },
+    {
+      "epoch": 3.355256841094575,
+      "grad_norm": 1.7083387156092242,
+      "learning_rate": 2.953165776686254e-06,
+      "loss": 0.5163,
+      "step": 3496
+    },
+    {
+      "epoch": 3.356216994719155,
+      "grad_norm": 1.1409952406855968,
+      "learning_rate": 2.9501098335726188e-06,
+      "loss": 0.5052,
+      "step": 3497
+    },
+    {
+      "epoch": 3.357177148343735,
+      "grad_norm": 1.186020755033467,
+      "learning_rate": 2.947054810632768e-06,
+      "loss": 0.4582,
+      "step": 3498
+    },
+    {
+      "epoch": 3.358137301968315,
+      "grad_norm": 1.4150439410912465,
+      "learning_rate": 2.9440007092380663e-06,
+      "loss": 0.4449,
+      "step": 3499
+    },
+    {
+      "epoch": 3.3590974555928947,
+      "grad_norm": 1.356711673122678,
+      "learning_rate": 2.9409475307594683e-06,
+      "loss": 0.491,
+      "step": 3500
+    },
+    {
+      "epoch": 3.360057609217475,
+      "grad_norm": 1.0576636480136845,
+      "learning_rate": 2.937895276567516e-06,
+      "loss": 0.5146,
+      "step": 3501
+    },
+    {
+      "epoch": 3.361017762842055,
+      "grad_norm": 0.9253348342532148,
+      "learning_rate": 2.934843948032329e-06,
+      "loss": 0.3877,
+      "step": 3502
+    },
+    {
+      "epoch": 3.3619779164666346,
+      "grad_norm": 1.0595590531486225,
+      "learning_rate": 2.931793546523622e-06,
+      "loss": 0.4221,
+      "step": 3503
+    },
+    {
+      "epoch": 3.3629380700912144,
+      "grad_norm": 1.5780005221790405,
+      "learning_rate": 2.9287440734106802e-06,
+      "loss": 0.3656,
+      "step": 3504
+    },
+    {
+      "epoch": 3.3638982237157946,
+      "grad_norm": 1.2094603353904014,
+      "learning_rate": 2.9256955300623847e-06,
+      "loss": 0.5271,
+      "step": 3505
+    },
+    {
+      "epoch": 3.3648583773403744,
+      "grad_norm": 1.1825058315142765,
+      "learning_rate": 2.9226479178471933e-06,
+      "loss": 0.3477,
+      "step": 3506
+    },
+    {
+      "epoch": 3.365818530964954,
+      "grad_norm": 1.3214146414963868,
+      "learning_rate": 2.919601238133145e-06,
+      "loss": 0.4159,
+      "step": 3507
+    },
+    {
+      "epoch": 3.3667786845895344,
+      "grad_norm": 1.1719105500897151,
+      "learning_rate": 2.916555492287865e-06,
+      "loss": 0.4692,
+      "step": 3508
+    },
+    {
+      "epoch": 3.3677388382141142,
+      "grad_norm": 1.8800779319989285,
+      "learning_rate": 2.913510681678552e-06,
+      "loss": 0.4837,
+      "step": 3509
+    },
+    {
+      "epoch": 3.368698991838694,
+      "grad_norm": 1.2531946213367224,
+      "learning_rate": 2.9104668076719876e-06,
+      "loss": 0.4717,
+      "step": 3510
+    },
+    {
+      "epoch": 3.3696591454632743,
+      "grad_norm": 1.1105983974295732,
+      "learning_rate": 2.9074238716345426e-06,
+      "loss": 0.3971,
+      "step": 3511
+    },
+    {
+      "epoch": 3.370619299087854,
+      "grad_norm": 0.8628587269532157,
+      "learning_rate": 2.9043818749321506e-06,
+      "loss": 0.4435,
+      "step": 3512
+    },
+    {
+      "epoch": 3.371579452712434,
+      "grad_norm": 1.2458667397479701,
+      "learning_rate": 2.9013408189303337e-06,
+      "loss": 0.4599,
+      "step": 3513
+    },
+    {
+      "epoch": 3.372539606337014,
+      "grad_norm": 1.2264588368349105,
+      "learning_rate": 2.8983007049941904e-06,
+      "loss": 0.5525,
+      "step": 3514
+    },
+    {
+      "epoch": 3.373499759961594,
+      "grad_norm": 1.4314881554786696,
+      "learning_rate": 2.8952615344883938e-06,
+      "loss": 0.4697,
+      "step": 3515
+    },
+    {
+      "epoch": 3.3744599135861737,
+      "grad_norm": 0.9575411876956582,
+      "learning_rate": 2.8922233087771988e-06,
+      "loss": 0.4158,
+      "step": 3516
+    },
+    {
+      "epoch": 3.3754200672107535,
+      "grad_norm": 1.2465922949189534,
+      "learning_rate": 2.889186029224428e-06,
+      "loss": 0.5241,
+      "step": 3517
+    },
+    {
+      "epoch": 3.3763802208353337,
+      "grad_norm": 1.0781670235659189,
+      "learning_rate": 2.8861496971934843e-06,
+      "loss": 0.4696,
+      "step": 3518
+    },
+    {
+      "epoch": 3.3773403744599135,
+      "grad_norm": 1.1216142983619832,
+      "learning_rate": 2.883114314047346e-06,
+      "loss": 0.5838,
+      "step": 3519
+    },
+    {
+      "epoch": 3.3783005280844933,
+      "grad_norm": 1.1020393649253364,
+      "learning_rate": 2.880079881148564e-06,
+      "loss": 0.4021,
+      "step": 3520
+    },
+    {
+      "epoch": 3.3792606817090736,
+      "grad_norm": 1.1408753452661533,
+      "learning_rate": 2.8770463998592645e-06,
+      "loss": 0.4518,
+      "step": 3521
+    },
+    {
+      "epoch": 3.3802208353336534,
+      "grad_norm": 1.4476366523641173,
+      "learning_rate": 2.8740138715411413e-06,
+      "loss": 0.535,
+      "step": 3522
+    },
+    {
+      "epoch": 3.381180988958233,
+      "grad_norm": 1.0947910518321444,
+      "learning_rate": 2.8709822975554636e-06,
+      "loss": 0.4504,
+      "step": 3523
+    },
+    {
+      "epoch": 3.3821411425828134,
+      "grad_norm": 1.083311343722618,
+      "learning_rate": 2.8679516792630783e-06,
+      "loss": 0.4013,
+      "step": 3524
+    },
+    {
+      "epoch": 3.383101296207393,
+      "grad_norm": 1.1621997871934076,
+      "learning_rate": 2.864922018024392e-06,
+      "loss": 0.4416,
+      "step": 3525
+    },
+    {
+      "epoch": 3.384061449831973,
+      "grad_norm": 0.8587905880392465,
+      "learning_rate": 2.8618933151993895e-06,
+      "loss": 0.4537,
+      "step": 3526
+    },
+    {
+      "epoch": 3.3850216034565532,
+      "grad_norm": 0.8986686761045902,
+      "learning_rate": 2.858865572147622e-06,
+      "loss": 0.5098,
+      "step": 3527
+    },
+    {
+      "epoch": 3.385981757081133,
+      "grad_norm": 1.5340809367773565,
+      "learning_rate": 2.8558387902282124e-06,
+      "loss": 0.5056,
+      "step": 3528
+    },
+    {
+      "epoch": 3.386941910705713,
+      "grad_norm": 0.9800974216612418,
+      "learning_rate": 2.852812970799852e-06,
+      "loss": 0.4697,
+      "step": 3529
+    },
+    {
+      "epoch": 3.3879020643302926,
+      "grad_norm": 1.3399154574672762,
+      "learning_rate": 2.849788115220794e-06,
+      "loss": 0.5324,
+      "step": 3530
+    },
+    {
+      "epoch": 3.388862217954873,
+      "grad_norm": 1.2917476442777003,
+      "learning_rate": 2.846764224848867e-06,
+      "loss": 0.4676,
+      "step": 3531
+    },
+    {
+      "epoch": 3.3898223715794527,
+      "grad_norm": 1.5463920595501937,
+      "learning_rate": 2.843741301041463e-06,
+      "loss": 0.4506,
+      "step": 3532
+    },
+    {
+      "epoch": 3.390782525204033,
+      "grad_norm": 1.0781539145409702,
+      "learning_rate": 2.8407193451555398e-06,
+      "loss": 0.4419,
+      "step": 3533
+    },
+    {
+      "epoch": 3.3917426788286127,
+      "grad_norm": 1.0027066731044258,
+      "learning_rate": 2.8376983585476226e-06,
+      "loss": 0.4063,
+      "step": 3534
+    },
+    {
+      "epoch": 3.3927028324531925,
+      "grad_norm": 0.8821407393548042,
+      "learning_rate": 2.834678342573797e-06,
+      "loss": 0.5158,
+      "step": 3535
+    },
+    {
+      "epoch": 3.3936629860777723,
+      "grad_norm": 1.1939673584672799,
+      "learning_rate": 2.8316592985897173e-06,
+      "loss": 0.463,
+      "step": 3536
+    },
+    {
+      "epoch": 3.3946231397023525,
+      "grad_norm": 1.1480599497638293,
+      "learning_rate": 2.8286412279506006e-06,
+      "loss": 0.445,
+      "step": 3537
+    },
+    {
+      "epoch": 3.3955832933269323,
+      "grad_norm": 0.9313188281494226,
+      "learning_rate": 2.8256241320112253e-06,
+      "loss": 0.5166,
+      "step": 3538
+    },
+    {
+      "epoch": 3.396543446951512,
+      "grad_norm": 1.006547046161028,
+      "learning_rate": 2.822608012125937e-06,
+      "loss": 0.5434,
+      "step": 3539
+    },
+    {
+      "epoch": 3.3975036005760924,
+      "grad_norm": 1.201997757291289,
+      "learning_rate": 2.8195928696486345e-06,
+      "loss": 0.4984,
+      "step": 3540
+    },
+    {
+      "epoch": 3.398463754200672,
+      "grad_norm": 0.9038008372857832,
+      "learning_rate": 2.8165787059327844e-06,
+      "loss": 0.5657,
+      "step": 3541
+    },
+    {
+      "epoch": 3.399423907825252,
+      "grad_norm": 1.1774393610346174,
+      "learning_rate": 2.8135655223314166e-06,
+      "loss": 0.3767,
+      "step": 3542
+    },
+    {
+      "epoch": 3.4003840614498317,
+      "grad_norm": 1.12599849774403,
+      "learning_rate": 2.810553320197113e-06,
+      "loss": 0.5037,
+      "step": 3543
+    },
+    {
+      "epoch": 3.401344215074412,
+      "grad_norm": 1.2062143158323828,
+      "learning_rate": 2.8075421008820197e-06,
+      "loss": 0.4462,
+      "step": 3544
+    },
+    {
+      "epoch": 3.4023043686989918,
+      "grad_norm": 1.2403997832602278,
+      "learning_rate": 2.804531865737842e-06,
+      "loss": 0.4407,
+      "step": 3545
+    },
+    {
+      "epoch": 3.403264522323572,
+      "grad_norm": 1.850016546542534,
+      "learning_rate": 2.8015226161158403e-06,
+      "loss": 0.4886,
+      "step": 3546
+    },
+    {
+      "epoch": 3.404224675948152,
+      "grad_norm": 1.1404929419636376,
+      "learning_rate": 2.798514353366838e-06,
+      "loss": 0.4044,
+      "step": 3547
+    },
+    {
+      "epoch": 3.4051848295727316,
+      "grad_norm": 1.3213709708007022,
+      "learning_rate": 2.7955070788412085e-06,
+      "loss": 0.5141,
+      "step": 3548
+    },
+    {
+      "epoch": 3.4061449831973114,
+      "grad_norm": 0.8699083415858886,
+      "learning_rate": 2.7925007938888867e-06,
+      "loss": 0.4399,
+      "step": 3549
+    },
+    {
+      "epoch": 3.4071051368218916,
+      "grad_norm": 1.3754971584958042,
+      "learning_rate": 2.789495499859361e-06,
+      "loss": 0.4446,
+      "step": 3550
+    },
+    {
+      "epoch": 3.4080652904464714,
+      "grad_norm": 1.5106588377195465,
+      "learning_rate": 2.786491198101677e-06,
+      "loss": 0.5024,
+      "step": 3551
+    },
+    {
+      "epoch": 3.4090254440710512,
+      "grad_norm": 1.5740256115254103,
+      "learning_rate": 2.783487889964435e-06,
+      "loss": 0.5005,
+      "step": 3552
+    },
+    {
+      "epoch": 3.4099855976956315,
+      "grad_norm": 1.4239322372170375,
+      "learning_rate": 2.7804855767957833e-06,
+      "loss": 0.4261,
+      "step": 3553
+    },
+    {
+      "epoch": 3.4109457513202113,
+      "grad_norm": 0.7441920668705134,
+      "learning_rate": 2.7774842599434294e-06,
+      "loss": 0.5355,
+      "step": 3554
+    },
+    {
+      "epoch": 3.411905904944791,
+      "grad_norm": 1.1964170644097079,
+      "learning_rate": 2.7744839407546374e-06,
+      "loss": 0.5028,
+      "step": 3555
+    },
+    {
+      "epoch": 3.412866058569371,
+      "grad_norm": 0.8917016245174008,
+      "learning_rate": 2.7714846205762134e-06,
+      "loss": 0.4768,
+      "step": 3556
+    },
+    {
+      "epoch": 3.413826212193951,
+      "grad_norm": 1.0888707464088432,
+      "learning_rate": 2.7684863007545227e-06,
+      "loss": 0.5231,
+      "step": 3557
+    },
+    {
+      "epoch": 3.414786365818531,
+      "grad_norm": 1.0868139743790344,
+      "learning_rate": 2.765488982635477e-06,
+      "loss": 0.5151,
+      "step": 3558
+    },
+    {
+      "epoch": 3.415746519443111,
+      "grad_norm": 0.9436745273446261,
+      "learning_rate": 2.7624926675645415e-06,
+      "loss": 0.4133,
+      "step": 3559
+    },
+    {
+      "epoch": 3.416706673067691,
+      "grad_norm": 1.0352478830995882,
+      "learning_rate": 2.759497356886731e-06,
+      "loss": 0.5861,
+      "step": 3560
+    },
+    {
+      "epoch": 3.4176668266922707,
+      "grad_norm": 1.0762675806754036,
+      "learning_rate": 2.7565030519466063e-06,
+      "loss": 0.4028,
+      "step": 3561
+    },
+    {
+      "epoch": 3.4186269803168505,
+      "grad_norm": 0.8724198562506894,
+      "learning_rate": 2.75350975408828e-06,
+      "loss": 0.4266,
+      "step": 3562
+    },
+    {
+      "epoch": 3.4195871339414308,
+      "grad_norm": 1.2616615183996187,
+      "learning_rate": 2.7505174646554112e-06,
+      "loss": 0.4518,
+      "step": 3563
+    },
+    {
+      "epoch": 3.4205472875660106,
+      "grad_norm": 1.324485618575122,
+      "learning_rate": 2.7475261849912073e-06,
+      "loss": 0.4463,
+      "step": 3564
+    },
+    {
+      "epoch": 3.4215074411905904,
+      "grad_norm": 1.0485211087613273,
+      "learning_rate": 2.744535916438423e-06,
+      "loss": 0.539,
+      "step": 3565
+    },
+    {
+      "epoch": 3.4224675948151706,
+      "grad_norm": 1.3669013683457083,
+      "learning_rate": 2.7415466603393538e-06,
+      "loss": 0.6324,
+      "step": 3566
+    },
+    {
+      "epoch": 3.4234277484397504,
+      "grad_norm": 0.9702366821080235,
+      "learning_rate": 2.738558418035846e-06,
+      "loss": 0.517,
+      "step": 3567
+    },
+    {
+      "epoch": 3.42438790206433,
+      "grad_norm": 1.0843548366284068,
+      "learning_rate": 2.7355711908692942e-06,
+      "loss": 0.3979,
+      "step": 3568
+    },
+    {
+      "epoch": 3.42534805568891,
+      "grad_norm": 1.0300795464968182,
+      "learning_rate": 2.732584980180628e-06,
+      "loss": 0.3735,
+      "step": 3569
+    },
+    {
+      "epoch": 3.4263082093134902,
+      "grad_norm": 1.1337504956207816,
+      "learning_rate": 2.7295997873103286e-06,
+      "loss": 0.3978,
+      "step": 3570
+    },
+    {
+      "epoch": 3.42726836293807,
+      "grad_norm": 1.0188041814822386,
+      "learning_rate": 2.726615613598411e-06,
+      "loss": 0.4734,
+      "step": 3571
+    },
+    {
+      "epoch": 3.4282285165626503,
+      "grad_norm": 1.8133667201452484,
+      "learning_rate": 2.7236324603844467e-06,
+      "loss": 0.4625,
+      "step": 3572
+    },
+    {
+      "epoch": 3.42918867018723,
+      "grad_norm": 2.04669729109423,
+      "learning_rate": 2.7206503290075388e-06,
+      "loss": 0.5092,
+      "step": 3573
+    },
+    {
+      "epoch": 3.43014882381181,
+      "grad_norm": 1.285179302856492,
+      "learning_rate": 2.7176692208063317e-06,
+      "loss": 0.4538,
+      "step": 3574
+    },
+    {
+      "epoch": 3.4311089774363897,
+      "grad_norm": 1.0888419094057504,
+      "learning_rate": 2.7146891371190144e-06,
+      "loss": 0.4493,
+      "step": 3575
+    },
+    {
+      "epoch": 3.43206913106097,
+      "grad_norm": 1.1711896679652405,
+      "learning_rate": 2.7117100792833144e-06,
+      "loss": 0.493,
+      "step": 3576
+    },
+    {
+      "epoch": 3.4330292846855497,
+      "grad_norm": 1.140479659389972,
+      "learning_rate": 2.7087320486364997e-06,
+      "loss": 0.3899,
+      "step": 3577
+    },
+    {
+      "epoch": 3.4339894383101295,
+      "grad_norm": 1.1330798335546821,
+      "learning_rate": 2.7057550465153782e-06,
+      "loss": 0.4961,
+      "step": 3578
+    },
+    {
+      "epoch": 3.4349495919347097,
+      "grad_norm": 1.4964408783536312,
+      "learning_rate": 2.7027790742562905e-06,
+      "loss": 0.511,
+      "step": 3579
+    },
+    {
+      "epoch": 3.4359097455592895,
+      "grad_norm": 1.2967733590915969,
+      "learning_rate": 2.6998041331951196e-06,
+      "loss": 0.4862,
+      "step": 3580
+    },
+    {
+      "epoch": 3.4368698991838693,
+      "grad_norm": 1.2047896517623808,
+      "learning_rate": 2.6968302246672863e-06,
+      "loss": 0.4532,
+      "step": 3581
+    },
+    {
+      "epoch": 3.437830052808449,
+      "grad_norm": 1.1207257158205917,
+      "learning_rate": 2.693857350007746e-06,
+      "loss": 0.455,
+      "step": 3582
+    },
+    {
+      "epoch": 3.4387902064330294,
+      "grad_norm": 1.217969904538037,
+      "learning_rate": 2.6908855105509912e-06,
+      "loss": 0.4445,
+      "step": 3583
+    },
+    {
+      "epoch": 3.439750360057609,
+      "grad_norm": 1.2049534130896251,
+      "learning_rate": 2.6879147076310457e-06,
+      "loss": 0.4064,
+      "step": 3584
+    },
+    {
+      "epoch": 3.4407105136821894,
+      "grad_norm": 1.3761988292840837,
+      "learning_rate": 2.6849449425814713e-06,
+      "loss": 0.4026,
+      "step": 3585
+    },
+    {
+      "epoch": 3.441670667306769,
+      "grad_norm": 0.9415572000589764,
+      "learning_rate": 2.6819762167353695e-06,
+      "loss": 0.383,
+      "step": 3586
+    },
+    {
+      "epoch": 3.442630820931349,
+      "grad_norm": 0.9470684110040503,
+      "learning_rate": 2.6790085314253645e-06,
+      "loss": 0.43,
+      "step": 3587
+    },
+    {
+      "epoch": 3.4435909745559288,
+      "grad_norm": 0.894013486586558,
+      "learning_rate": 2.6760418879836186e-06,
+      "loss": 0.4578,
+      "step": 3588
+    },
+    {
+      "epoch": 3.444551128180509,
+      "grad_norm": 0.8156437597110491,
+      "learning_rate": 2.673076287741828e-06,
+      "loss": 0.4178,
+      "step": 3589
+    },
+    {
+      "epoch": 3.445511281805089,
+      "grad_norm": 1.0825128113462217,
+      "learning_rate": 2.670111732031218e-06,
+      "loss": 0.4715,
+      "step": 3590
+    },
+    {
+      "epoch": 3.4464714354296686,
+      "grad_norm": 1.0144452624565126,
+      "learning_rate": 2.6671482221825484e-06,
+      "loss": 0.5427,
+      "step": 3591
+    },
+    {
+      "epoch": 3.447431589054249,
+      "grad_norm": 1.0665060436734815,
+      "learning_rate": 2.6641857595261035e-06,
+      "loss": 0.4447,
+      "step": 3592
+    },
+    {
+      "epoch": 3.4483917426788286,
+      "grad_norm": 0.9795414216047997,
+      "learning_rate": 2.661224345391703e-06,
+      "loss": 0.4444,
+      "step": 3593
+    },
+    {
+      "epoch": 3.4493518963034084,
+      "grad_norm": 0.8353413950195184,
+      "learning_rate": 2.658263981108693e-06,
+      "loss": 0.5798,
+      "step": 3594
+    },
+    {
+      "epoch": 3.4503120499279882,
+      "grad_norm": 0.8687455116406125,
+      "learning_rate": 2.65530466800595e-06,
+      "loss": 0.5348,
+      "step": 3595
+    },
+    {
+      "epoch": 3.4512722035525685,
+      "grad_norm": 1.148995894038912,
+      "learning_rate": 2.652346407411881e-06,
+      "loss": 0.3702,
+      "step": 3596
+    },
+    {
+      "epoch": 3.4522323571771483,
+      "grad_norm": 1.2989800722974758,
+      "learning_rate": 2.649389200654412e-06,
+      "loss": 0.4112,
+      "step": 3597
+    },
+    {
+      "epoch": 3.4531925108017285,
+      "grad_norm": 0.9463725991275451,
+      "learning_rate": 2.6464330490610026e-06,
+      "loss": 0.4448,
+      "step": 3598
+    },
+    {
+      "epoch": 3.4541526644263083,
+      "grad_norm": 1.318494074442644,
+      "learning_rate": 2.6434779539586432e-06,
+      "loss": 0.5126,
+      "step": 3599
+    },
+    {
+      "epoch": 3.455112818050888,
+      "grad_norm": 1.3358485591347307,
+      "learning_rate": 2.640523916673838e-06,
+      "loss": 0.4431,
+      "step": 3600
+    },
+    {
+      "epoch": 3.456072971675468,
+      "grad_norm": 1.4512534259842509,
+      "learning_rate": 2.6375709385326255e-06,
+      "loss": 0.4406,
+      "step": 3601
+    },
+    {
+      "epoch": 3.457033125300048,
+      "grad_norm": 1.1449355325344546,
+      "learning_rate": 2.6346190208605653e-06,
+      "loss": 0.4695,
+      "step": 3602
+    },
+    {
+      "epoch": 3.457993278924628,
+      "grad_norm": 0.9559669934907598,
+      "learning_rate": 2.6316681649827423e-06,
+      "loss": 0.4964,
+      "step": 3603
+    },
+    {
+      "epoch": 3.4589534325492077,
+      "grad_norm": 1.201264073736653,
+      "learning_rate": 2.6287183722237663e-06,
+      "loss": 0.3984,
+      "step": 3604
+    },
+    {
+      "epoch": 3.459913586173788,
+      "grad_norm": 1.096836226946159,
+      "learning_rate": 2.625769643907763e-06,
+      "loss": 0.5035,
+      "step": 3605
+    },
+    {
+      "epoch": 3.4608737397983678,
+      "grad_norm": 0.7278545545970976,
+      "learning_rate": 2.622821981358388e-06,
+      "loss": 0.5073,
+      "step": 3606
+    },
+    {
+      "epoch": 3.4618338934229476,
+      "grad_norm": 1.3266279163313695,
+      "learning_rate": 2.6198753858988147e-06,
+      "loss": 0.4638,
+      "step": 3607
+    },
+    {
+      "epoch": 3.4627940470475274,
+      "grad_norm": 0.8010194816184371,
+      "learning_rate": 2.616929858851739e-06,
+      "loss": 0.3906,
+      "step": 3608
+    },
+    {
+      "epoch": 3.4637542006721076,
+      "grad_norm": 1.160469371575079,
+      "learning_rate": 2.613985401539377e-06,
+      "loss": 0.4674,
+      "step": 3609
+    },
+    {
+      "epoch": 3.4647143542966874,
+      "grad_norm": 1.385240494381929,
+      "learning_rate": 2.6110420152834614e-06,
+      "loss": 0.6116,
+      "step": 3610
+    },
+    {
+      "epoch": 3.4656745079212676,
+      "grad_norm": 1.4254500913885508,
+      "learning_rate": 2.6080997014052474e-06,
+      "loss": 0.4868,
+      "step": 3611
+    },
+    {
+      "epoch": 3.4666346615458474,
+      "grad_norm": 1.0313280287612518,
+      "learning_rate": 2.6051584612255132e-06,
+      "loss": 0.4166,
+      "step": 3612
+    },
+    {
+      "epoch": 3.4675948151704272,
+      "grad_norm": 0.9944099859028148,
+      "learning_rate": 2.6022182960645436e-06,
+      "loss": 0.4879,
+      "step": 3613
+    },
+    {
+      "epoch": 3.468554968795007,
+      "grad_norm": 0.9995198484098955,
+      "learning_rate": 2.5992792072421526e-06,
+      "loss": 0.4761,
+      "step": 3614
+    },
+    {
+      "epoch": 3.4695151224195873,
+      "grad_norm": 1.216734458299888,
+      "learning_rate": 2.596341196077659e-06,
+      "loss": 0.5055,
+      "step": 3615
+    },
+    {
+      "epoch": 3.470475276044167,
+      "grad_norm": 1.246247200010548,
+      "learning_rate": 2.593404263889911e-06,
+      "loss": 0.5008,
+      "step": 3616
+    },
+    {
+      "epoch": 3.471435429668747,
+      "grad_norm": 0.9976808037461659,
+      "learning_rate": 2.5904684119972656e-06,
+      "loss": 0.5077,
+      "step": 3617
+    },
+    {
+      "epoch": 3.472395583293327,
+      "grad_norm": 0.8269283769717857,
+      "learning_rate": 2.587533641717592e-06,
+      "loss": 0.4821,
+      "step": 3618
+    },
+    {
+      "epoch": 3.473355736917907,
+      "grad_norm": 1.2400549875225575,
+      "learning_rate": 2.5845999543682787e-06,
+      "loss": 0.5868,
+      "step": 3619
+    },
+    {
+      "epoch": 3.4743158905424867,
+      "grad_norm": 0.9329814175183276,
+      "learning_rate": 2.581667351266226e-06,
+      "loss": 0.6003,
+      "step": 3620
+    },
+    {
+      "epoch": 3.475276044167067,
+      "grad_norm": 1.3338791789490658,
+      "learning_rate": 2.5787358337278493e-06,
+      "loss": 0.4371,
+      "step": 3621
+    },
+    {
+      "epoch": 3.4762361977916467,
+      "grad_norm": 1.1511633734554696,
+      "learning_rate": 2.575805403069077e-06,
+      "loss": 0.4362,
+      "step": 3622
+    },
+    {
+      "epoch": 3.4771963514162265,
+      "grad_norm": 1.1202526138333042,
+      "learning_rate": 2.5728760606053443e-06,
+      "loss": 0.4908,
+      "step": 3623
+    },
+    {
+      "epoch": 3.4781565050408068,
+      "grad_norm": 1.249918071885655,
+      "learning_rate": 2.5699478076516034e-06,
+      "loss": 0.4459,
+      "step": 3624
+    },
+    {
+      "epoch": 3.4791166586653866,
+      "grad_norm": 0.9026035841529765,
+      "learning_rate": 2.5670206455223156e-06,
+      "loss": 0.536,
+      "step": 3625
+    },
+    {
+      "epoch": 3.4800768122899663,
+      "grad_norm": 1.1902872385852585,
+      "learning_rate": 2.5640945755314527e-06,
+      "loss": 0.4952,
+      "step": 3626
+    },
+    {
+      "epoch": 3.481036965914546,
+      "grad_norm": 1.0066852241962887,
+      "learning_rate": 2.5611695989924988e-06,
+      "loss": 0.4986,
+      "step": 3627
+    },
+    {
+      "epoch": 3.4819971195391264,
+      "grad_norm": 0.8493448562993611,
+      "learning_rate": 2.5582457172184405e-06,
+      "loss": 0.4542,
+      "step": 3628
+    },
+    {
+      "epoch": 3.482957273163706,
+      "grad_norm": 0.9521082941092199,
+      "learning_rate": 2.5553229315217763e-06,
+      "loss": 0.3476,
+      "step": 3629
+    },
+    {
+      "epoch": 3.483917426788286,
+      "grad_norm": 1.140189369962596,
+      "learning_rate": 2.5524012432145203e-06,
+      "loss": 0.5449,
+      "step": 3630
+    },
+    {
+      "epoch": 3.484877580412866,
+      "grad_norm": 0.9634567981887172,
+      "learning_rate": 2.5494806536081806e-06,
+      "loss": 0.4575,
+      "step": 3631
+    },
+    {
+      "epoch": 3.485837734037446,
+      "grad_norm": 0.8307670440375307,
+      "learning_rate": 2.5465611640137813e-06,
+      "loss": 0.4691,
+      "step": 3632
+    },
+    {
+      "epoch": 3.486797887662026,
+      "grad_norm": 1.2732339891636295,
+      "learning_rate": 2.543642775741849e-06,
+      "loss": 0.4743,
+      "step": 3633
+    },
+    {
+      "epoch": 3.487758041286606,
+      "grad_norm": 1.397784544861311,
+      "learning_rate": 2.540725490102418e-06,
+      "loss": 0.5432,
+      "step": 3634
+    },
+    {
+      "epoch": 3.488718194911186,
+      "grad_norm": 0.9880987144603872,
+      "learning_rate": 2.5378093084050287e-06,
+      "loss": 0.5647,
+      "step": 3635
+    },
+    {
+      "epoch": 3.4896783485357656,
+      "grad_norm": 0.8861042098478457,
+      "learning_rate": 2.534894231958719e-06,
+      "loss": 0.5291,
+      "step": 3636
+    },
+    {
+      "epoch": 3.490638502160346,
+      "grad_norm": 0.9065219243847623,
+      "learning_rate": 2.531980262072038e-06,
+      "loss": 0.4777,
+      "step": 3637
+    },
+    {
+      "epoch": 3.4915986557849257,
+      "grad_norm": 1.1297295787094745,
+      "learning_rate": 2.5290674000530365e-06,
+      "loss": 0.5404,
+      "step": 3638
+    },
+    {
+      "epoch": 3.4925588094095055,
+      "grad_norm": 1.646709559764397,
+      "learning_rate": 2.526155647209265e-06,
+      "loss": 0.4869,
+      "step": 3639
+    },
+    {
+      "epoch": 3.4935189630340853,
+      "grad_norm": 0.7720437600850697,
+      "learning_rate": 2.523245004847783e-06,
+      "loss": 0.5778,
+      "step": 3640
+    },
+    {
+      "epoch": 3.4944791166586655,
+      "grad_norm": 0.9884900889150412,
+      "learning_rate": 2.520335474275141e-06,
+      "loss": 0.4142,
+      "step": 3641
+    },
+    {
+      "epoch": 3.4954392702832453,
+      "grad_norm": 1.124525370980486,
+      "learning_rate": 2.5174270567973957e-06,
+      "loss": 0.5587,
+      "step": 3642
+    },
+    {
+      "epoch": 3.496399423907825,
+      "grad_norm": 0.9719887183464196,
+      "learning_rate": 2.5145197537201116e-06,
+      "loss": 0.5514,
+      "step": 3643
+    },
+    {
+      "epoch": 3.4973595775324053,
+      "grad_norm": 1.2104387401344738,
+      "learning_rate": 2.51161356634834e-06,
+      "loss": 0.5069,
+      "step": 3644
+    },
+    {
+      "epoch": 3.498319731156985,
+      "grad_norm": 1.2346000448649872,
+      "learning_rate": 2.5087084959866403e-06,
+      "loss": 0.4351,
+      "step": 3645
+    },
+    {
+      "epoch": 3.499279884781565,
+      "grad_norm": 1.099044269053787,
+      "learning_rate": 2.5058045439390634e-06,
+      "loss": 0.491,
+      "step": 3646
+    },
+    {
+      "epoch": 3.5002400384061447,
+      "grad_norm": 1.4946661581040868,
+      "learning_rate": 2.502901711509167e-06,
+      "loss": 0.4149,
+      "step": 3647
+    },
+    {
+      "epoch": 3.501200192030725,
+      "grad_norm": 1.227057661512799,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.5229,
+      "step": 3648
+    },
+    {
+      "epoch": 3.5021603456553048,
+      "grad_norm": 1.2370509801878777,
+      "learning_rate": 2.4970994107141106e-06,
+      "loss": 0.4831,
+      "step": 3649
+    },
+    {
+      "epoch": 3.503120499279885,
+      "grad_norm": 1.2958315162178096,
+      "learning_rate": 2.4941999449535393e-06,
+      "loss": 0.3693,
+      "step": 3650
+    },
+    {
+      "epoch": 3.504080652904465,
+      "grad_norm": 1.0522251409633594,
+      "learning_rate": 2.491301604019828e-06,
+      "loss": 0.5309,
+      "step": 3651
+    },
+    {
+      "epoch": 3.5050408065290446,
+      "grad_norm": 1.2251693069715273,
+      "learning_rate": 2.4884043892140096e-06,
+      "loss": 0.4234,
+      "step": 3652
+    },
+    {
+      "epoch": 3.5060009601536244,
+      "grad_norm": 1.056679063466086,
+      "learning_rate": 2.4855083018366154e-06,
+      "loss": 0.4025,
+      "step": 3653
+    },
+    {
+      "epoch": 3.5069611137782046,
+      "grad_norm": 1.0109892189050866,
+      "learning_rate": 2.4826133431876644e-06,
+      "loss": 0.4746,
+      "step": 3654
+    },
+    {
+      "epoch": 3.5079212674027844,
+      "grad_norm": 0.8900546374984372,
+      "learning_rate": 2.479719514566672e-06,
+      "loss": 0.4223,
+      "step": 3655
+    },
+    {
+      "epoch": 3.5088814210273642,
+      "grad_norm": 1.1033470328591162,
+      "learning_rate": 2.4768268172726535e-06,
+      "loss": 0.472,
+      "step": 3656
+    },
+    {
+      "epoch": 3.5098415746519445,
+      "grad_norm": 1.270969298574359,
+      "learning_rate": 2.473935252604103e-06,
+      "loss": 0.5337,
+      "step": 3657
+    },
+    {
+      "epoch": 3.5108017282765243,
+      "grad_norm": 1.479006407731475,
+      "learning_rate": 2.4710448218590176e-06,
+      "loss": 0.5184,
+      "step": 3658
+    },
+    {
+      "epoch": 3.511761881901104,
+      "grad_norm": 1.1883063504227078,
+      "learning_rate": 2.468155526334875e-06,
+      "loss": 0.5456,
+      "step": 3659
+    },
+    {
+      "epoch": 3.512722035525684,
+      "grad_norm": 1.0164127178929956,
+      "learning_rate": 2.465267367328655e-06,
+      "loss": 0.4789,
+      "step": 3660
+    },
+    {
+      "epoch": 3.513682189150264,
+      "grad_norm": 1.1188398770666201,
+      "learning_rate": 2.46238034613682e-06,
+      "loss": 0.4018,
+      "step": 3661
+    },
+    {
+      "epoch": 3.514642342774844,
+      "grad_norm": 1.0569981650165516,
+      "learning_rate": 2.4594944640553207e-06,
+      "loss": 0.4344,
+      "step": 3662
+    },
+    {
+      "epoch": 3.515602496399424,
+      "grad_norm": 1.3233370416968462,
+      "learning_rate": 2.4566097223795997e-06,
+      "loss": 0.527,
+      "step": 3663
+    },
+    {
+      "epoch": 3.516562650024004,
+      "grad_norm": 1.1378472162358964,
+      "learning_rate": 2.453726122404587e-06,
+      "loss": 0.5237,
+      "step": 3664
+    },
+    {
+      "epoch": 3.5175228036485837,
+      "grad_norm": 1.0450132913587404,
+      "learning_rate": 2.4508436654247004e-06,
+      "loss": 0.4229,
+      "step": 3665
+    },
+    {
+      "epoch": 3.5184829572731635,
+      "grad_norm": 0.9468624461124091,
+      "learning_rate": 2.4479623527338447e-06,
+      "loss": 0.5138,
+      "step": 3666
+    },
+    {
+      "epoch": 3.5194431108977438,
+      "grad_norm": 1.1454995223310345,
+      "learning_rate": 2.4450821856254074e-06,
+      "loss": 0.4455,
+      "step": 3667
+    },
+    {
+      "epoch": 3.5204032645223235,
+      "grad_norm": 1.2594858516678467,
+      "learning_rate": 2.442203165392266e-06,
+      "loss": 0.3976,
+      "step": 3668
+    },
+    {
+      "epoch": 3.5213634181469033,
+      "grad_norm": 1.6642733147102062,
+      "learning_rate": 2.439325293326783e-06,
+      "loss": 0.3881,
+      "step": 3669
+    },
+    {
+      "epoch": 3.5223235717714836,
+      "grad_norm": 1.1145416415512275,
+      "learning_rate": 2.4364485707208037e-06,
+      "loss": 0.5226,
+      "step": 3670
+    },
+    {
+      "epoch": 3.5232837253960634,
+      "grad_norm": 1.3938203305124055,
+      "learning_rate": 2.433572998865659e-06,
+      "loss": 0.4091,
+      "step": 3671
+    },
+    {
+      "epoch": 3.524243879020643,
+      "grad_norm": 0.9522141730008076,
+      "learning_rate": 2.430698579052161e-06,
+      "loss": 0.5209,
+      "step": 3672
+    },
+    {
+      "epoch": 3.525204032645223,
+      "grad_norm": 1.0579883556910796,
+      "learning_rate": 2.4278253125706044e-06,
+      "loss": 0.5121,
+      "step": 3673
+    },
+    {
+      "epoch": 3.526164186269803,
+      "grad_norm": 1.0666069425645939,
+      "learning_rate": 2.424953200710773e-06,
+      "loss": 0.468,
+      "step": 3674
+    },
+    {
+      "epoch": 3.527124339894383,
+      "grad_norm": 1.124981702042489,
+      "learning_rate": 2.4220822447619223e-06,
+      "loss": 0.4579,
+      "step": 3675
+    },
+    {
+      "epoch": 3.5280844935189632,
+      "grad_norm": 1.1505524021429214,
+      "learning_rate": 2.419212446012796e-06,
+      "loss": 0.4537,
+      "step": 3676
+    },
+    {
+      "epoch": 3.529044647143543,
+      "grad_norm": 0.9687320319338322,
+      "learning_rate": 2.416343805751615e-06,
+      "loss": 0.4503,
+      "step": 3677
+    },
+    {
+      "epoch": 3.530004800768123,
+      "grad_norm": 1.3101841403266583,
+      "learning_rate": 2.4134763252660825e-06,
+      "loss": 0.386,
+      "step": 3678
+    },
+    {
+      "epoch": 3.5309649543927026,
+      "grad_norm": 1.2122305542938274,
+      "learning_rate": 2.4106100058433814e-06,
+      "loss": 0.4864,
+      "step": 3679
+    },
+    {
+      "epoch": 3.531925108017283,
+      "grad_norm": 1.4978240894195236,
+      "learning_rate": 2.4077448487701667e-06,
+      "loss": 0.5496,
+      "step": 3680
+    },
+    {
+      "epoch": 3.5328852616418627,
+      "grad_norm": 1.3428200095632699,
+      "learning_rate": 2.4048808553325804e-06,
+      "loss": 0.5052,
+      "step": 3681
+    },
+    {
+      "epoch": 3.5338454152664425,
+      "grad_norm": 1.039762695895337,
+      "learning_rate": 2.4020180268162374e-06,
+      "loss": 0.565,
+      "step": 3682
+    },
+    {
+      "epoch": 3.5348055688910227,
+      "grad_norm": 1.0547541209262872,
+      "learning_rate": 2.3991563645062317e-06,
+      "loss": 0.4556,
+      "step": 3683
+    },
+    {
+      "epoch": 3.5357657225156025,
+      "grad_norm": 1.1631940207140978,
+      "learning_rate": 2.3962958696871343e-06,
+      "loss": 0.4225,
+      "step": 3684
+    },
+    {
+      "epoch": 3.5367258761401823,
+      "grad_norm": 1.2427090021569067,
+      "learning_rate": 2.393436543642987e-06,
+      "loss": 0.3593,
+      "step": 3685
+    },
+    {
+      "epoch": 3.537686029764762,
+      "grad_norm": 1.2190538515419043,
+      "learning_rate": 2.3905783876573108e-06,
+      "loss": 0.508,
+      "step": 3686
+    },
+    {
+      "epoch": 3.5386461833893423,
+      "grad_norm": 1.1922326644684484,
+      "learning_rate": 2.387721403013107e-06,
+      "loss": 0.5042,
+      "step": 3687
+    },
+    {
+      "epoch": 3.539606337013922,
+      "grad_norm": 1.1304810096194196,
+      "learning_rate": 2.3848655909928392e-06,
+      "loss": 0.5237,
+      "step": 3688
+    },
+    {
+      "epoch": 3.5405664906385024,
+      "grad_norm": 1.019042176877277,
+      "learning_rate": 2.382010952878456e-06,
+      "loss": 0.4779,
+      "step": 3689
+    },
+    {
+      "epoch": 3.541526644263082,
+      "grad_norm": 1.039931316432492,
+      "learning_rate": 2.379157489951367e-06,
+      "loss": 0.5413,
+      "step": 3690
+    },
+    {
+      "epoch": 3.542486797887662,
+      "grad_norm": 1.1195582239210486,
+      "learning_rate": 2.376305203492468e-06,
+      "loss": 0.4932,
+      "step": 3691
+    },
+    {
+      "epoch": 3.5434469515122418,
+      "grad_norm": 1.00174933850564,
+      "learning_rate": 2.373454094782119e-06,
+      "loss": 0.4171,
+      "step": 3692
+    },
+    {
+      "epoch": 3.544407105136822,
+      "grad_norm": 1.0735340158142546,
+      "learning_rate": 2.370604165100148e-06,
+      "loss": 0.4101,
+      "step": 3693
+    },
+    {
+      "epoch": 3.545367258761402,
+      "grad_norm": 1.1214307020282934,
+      "learning_rate": 2.367755415725862e-06,
+      "loss": 0.4242,
+      "step": 3694
+    },
+    {
+      "epoch": 3.5463274123859816,
+      "grad_norm": 1.1645266080572896,
+      "learning_rate": 2.3649078479380316e-06,
+      "loss": 0.4577,
+      "step": 3695
+    },
+    {
+      "epoch": 3.547287566010562,
+      "grad_norm": 1.5721307352224687,
+      "learning_rate": 2.3620614630149013e-06,
+      "loss": 0.4848,
+      "step": 3696
+    },
+    {
+      "epoch": 3.5482477196351416,
+      "grad_norm": 1.3104680164652156,
+      "learning_rate": 2.359216262234184e-06,
+      "loss": 0.438,
+      "step": 3697
+    },
+    {
+      "epoch": 3.5492078732597214,
+      "grad_norm": 1.0779816651842788,
+      "learning_rate": 2.3563722468730567e-06,
+      "loss": 0.5184,
+      "step": 3698
+    },
+    {
+      "epoch": 3.550168026884301,
+      "grad_norm": 0.837558838380594,
+      "learning_rate": 2.3535294182081666e-06,
+      "loss": 0.4584,
+      "step": 3699
+    },
+    {
+      "epoch": 3.5511281805088815,
+      "grad_norm": 0.9892452703556603,
+      "learning_rate": 2.350687777515635e-06,
+      "loss": 0.5487,
+      "step": 3700
+    },
+    {
+      "epoch": 3.5520883341334613,
+      "grad_norm": 1.0156020785944047,
+      "learning_rate": 2.3478473260710387e-06,
+      "loss": 0.5219,
+      "step": 3701
+    },
+    {
+      "epoch": 3.5530484877580415,
+      "grad_norm": 1.0201654504271127,
+      "learning_rate": 2.34500806514943e-06,
+      "loss": 0.5055,
+      "step": 3702
+    },
+    {
+      "epoch": 3.5540086413826213,
+      "grad_norm": 0.7851140133339441,
+      "learning_rate": 2.3421699960253168e-06,
+      "loss": 0.4465,
+      "step": 3703
+    },
+    {
+      "epoch": 3.554968795007201,
+      "grad_norm": 0.9613923870294074,
+      "learning_rate": 2.3393331199726837e-06,
+      "loss": 0.5031,
+      "step": 3704
+    },
+    {
+      "epoch": 3.555928948631781,
+      "grad_norm": 1.0917314279342007,
+      "learning_rate": 2.3364974382649737e-06,
+      "loss": 0.4907,
+      "step": 3705
+    },
+    {
+      "epoch": 3.556889102256361,
+      "grad_norm": 1.1549333470707257,
+      "learning_rate": 2.3336629521750902e-06,
+      "loss": 0.4393,
+      "step": 3706
+    },
+    {
+      "epoch": 3.557849255880941,
+      "grad_norm": 1.1844655097667611,
+      "learning_rate": 2.3308296629754067e-06,
+      "loss": 0.398,
+      "step": 3707
+    },
+    {
+      "epoch": 3.558809409505521,
+      "grad_norm": 1.4718673944191167,
+      "learning_rate": 2.327997571937755e-06,
+      "loss": 0.5234,
+      "step": 3708
+    },
+    {
+      "epoch": 3.559769563130101,
+      "grad_norm": 0.7899354720749071,
+      "learning_rate": 2.3251666803334316e-06,
+      "loss": 0.3849,
+      "step": 3709
+    },
+    {
+      "epoch": 3.5607297167546808,
+      "grad_norm": 0.9801400054494465,
+      "learning_rate": 2.3223369894331947e-06,
+      "loss": 0.4092,
+      "step": 3710
+    },
+    {
+      "epoch": 3.5616898703792605,
+      "grad_norm": 1.2336551635147301,
+      "learning_rate": 2.3195085005072587e-06,
+      "loss": 0.4218,
+      "step": 3711
+    },
+    {
+      "epoch": 3.5626500240038403,
+      "grad_norm": 0.8657123624681815,
+      "learning_rate": 2.3166812148253045e-06,
+      "loss": 0.5174,
+      "step": 3712
+    },
+    {
+      "epoch": 3.5636101776284206,
+      "grad_norm": 1.1406124090395062,
+      "learning_rate": 2.313855133656469e-06,
+      "loss": 0.3652,
+      "step": 3713
+    },
+    {
+      "epoch": 3.5645703312530004,
+      "grad_norm": 1.0904647275948893,
+      "learning_rate": 2.3110302582693512e-06,
+      "loss": 0.4873,
+      "step": 3714
+    },
+    {
+      "epoch": 3.5655304848775806,
+      "grad_norm": 0.969364953364631,
+      "learning_rate": 2.3082065899320084e-06,
+      "loss": 0.4624,
+      "step": 3715
+    },
+    {
+      "epoch": 3.5664906385021604,
+      "grad_norm": 0.9765307121391327,
+      "learning_rate": 2.3053841299119524e-06,
+      "loss": 0.5346,
+      "step": 3716
+    },
+    {
+      "epoch": 3.56745079212674,
+      "grad_norm": 1.0690333161157577,
+      "learning_rate": 2.3025628794761544e-06,
+      "loss": 0.503,
+      "step": 3717
+    },
+    {
+      "epoch": 3.56841094575132,
+      "grad_norm": 0.862838816483104,
+      "learning_rate": 2.2997428398910497e-06,
+      "loss": 0.4878,
+      "step": 3718
+    },
+    {
+      "epoch": 3.5693710993759002,
+      "grad_norm": 1.2005659050029838,
+      "learning_rate": 2.2969240124225177e-06,
+      "loss": 0.4877,
+      "step": 3719
+    },
+    {
+      "epoch": 3.57033125300048,
+      "grad_norm": 0.919234353397464,
+      "learning_rate": 2.294106398335905e-06,
+      "loss": 0.3685,
+      "step": 3720
+    },
+    {
+      "epoch": 3.5712914066250603,
+      "grad_norm": 1.3125460441962709,
+      "learning_rate": 2.2912899988960015e-06,
+      "loss": 0.5297,
+      "step": 3721
+    },
+    {
+      "epoch": 3.57225156024964,
+      "grad_norm": 1.253709953333357,
+      "learning_rate": 2.288474815367065e-06,
+      "loss": 0.4826,
+      "step": 3722
+    },
+    {
+      "epoch": 3.57321171387422,
+      "grad_norm": 1.129404042598159,
+      "learning_rate": 2.2856608490128005e-06,
+      "loss": 0.4338,
+      "step": 3723
+    },
+    {
+      "epoch": 3.5741718674987997,
+      "grad_norm": 1.1505810636583158,
+      "learning_rate": 2.2828481010963647e-06,
+      "loss": 0.4836,
+      "step": 3724
+    },
+    {
+      "epoch": 3.5751320211233795,
+      "grad_norm": 1.246252816796723,
+      "learning_rate": 2.280036572880372e-06,
+      "loss": 0.5039,
+      "step": 3725
+    },
+    {
+      "epoch": 3.5760921747479597,
+      "grad_norm": 1.2784790157010364,
+      "learning_rate": 2.277226265626887e-06,
+      "loss": 0.4429,
+      "step": 3726
+    },
+    {
+      "epoch": 3.5770523283725395,
+      "grad_norm": 1.0481861440733469,
+      "learning_rate": 2.2744171805974257e-06,
+      "loss": 0.4873,
+      "step": 3727
+    },
+    {
+      "epoch": 3.5780124819971197,
+      "grad_norm": 1.2778175546190231,
+      "learning_rate": 2.27160931905296e-06,
+      "loss": 0.4806,
+      "step": 3728
+    },
+    {
+      "epoch": 3.5789726356216995,
+      "grad_norm": 1.098165453489468,
+      "learning_rate": 2.268802682253903e-06,
+      "loss": 0.4883,
+      "step": 3729
+    },
+    {
+      "epoch": 3.5799327892462793,
+      "grad_norm": 1.2114603753626112,
+      "learning_rate": 2.2659972714601264e-06,
+      "loss": 0.4976,
+      "step": 3730
+    },
+    {
+      "epoch": 3.580892942870859,
+      "grad_norm": 0.9533509049683406,
+      "learning_rate": 2.2631930879309525e-06,
+      "loss": 0.5182,
+      "step": 3731
+    },
+    {
+      "epoch": 3.5818530964954394,
+      "grad_norm": 1.4678657246619216,
+      "learning_rate": 2.2603901329251452e-06,
+      "loss": 0.4437,
+      "step": 3732
+    },
+    {
+      "epoch": 3.582813250120019,
+      "grad_norm": 1.1679398597192725,
+      "learning_rate": 2.2575884077009246e-06,
+      "loss": 0.5174,
+      "step": 3733
+    },
+    {
+      "epoch": 3.5837734037445994,
+      "grad_norm": 1.0503972593308595,
+      "learning_rate": 2.2547879135159495e-06,
+      "loss": 0.483,
+      "step": 3734
+    },
+    {
+      "epoch": 3.584733557369179,
+      "grad_norm": 0.9289524071299289,
+      "learning_rate": 2.2519886516273365e-06,
+      "loss": 0.4431,
+      "step": 3735
+    },
+    {
+      "epoch": 3.585693710993759,
+      "grad_norm": 1.209828943543199,
+      "learning_rate": 2.2491906232916457e-06,
+      "loss": 0.4678,
+      "step": 3736
+    },
+    {
+      "epoch": 3.586653864618339,
+      "grad_norm": 1.5376189296878735,
+      "learning_rate": 2.246393829764877e-06,
+      "loss": 0.4692,
+      "step": 3737
+    },
+    {
+      "epoch": 3.5876140182429186,
+      "grad_norm": 1.1288844498173878,
+      "learning_rate": 2.243598272302484e-06,
+      "loss": 0.4994,
+      "step": 3738
+    },
+    {
+      "epoch": 3.588574171867499,
+      "grad_norm": 1.3065944926194302,
+      "learning_rate": 2.240803952159362e-06,
+      "loss": 0.4572,
+      "step": 3739
+    },
+    {
+      "epoch": 3.5895343254920786,
+      "grad_norm": 1.265342252364368,
+      "learning_rate": 2.238010870589852e-06,
+      "loss": 0.4564,
+      "step": 3740
+    },
+    {
+      "epoch": 3.590494479116659,
+      "grad_norm": 0.9576001472932502,
+      "learning_rate": 2.23521902884774e-06,
+      "loss": 0.4557,
+      "step": 3741
+    },
+    {
+      "epoch": 3.5914546327412387,
+      "grad_norm": 1.0903177339210735,
+      "learning_rate": 2.23242842818625e-06,
+      "loss": 0.4634,
+      "step": 3742
+    },
+    {
+      "epoch": 3.5924147863658185,
+      "grad_norm": 1.4148767249512473,
+      "learning_rate": 2.2296390698580558e-06,
+      "loss": 0.4254,
+      "step": 3743
+    },
+    {
+      "epoch": 3.5933749399903983,
+      "grad_norm": 1.1502034934304621,
+      "learning_rate": 2.2268509551152684e-06,
+      "loss": 0.4472,
+      "step": 3744
+    },
+    {
+      "epoch": 3.5943350936149785,
+      "grad_norm": 1.0443676514435147,
+      "learning_rate": 2.2240640852094446e-06,
+      "loss": 0.5179,
+      "step": 3745
+    },
+    {
+      "epoch": 3.5952952472395583,
+      "grad_norm": 1.0334295753671674,
+      "learning_rate": 2.2212784613915816e-06,
+      "loss": 0.4259,
+      "step": 3746
+    },
+    {
+      "epoch": 3.5962554008641385,
+      "grad_norm": 1.335848923966106,
+      "learning_rate": 2.2184940849121123e-06,
+      "loss": 0.5216,
+      "step": 3747
+    },
+    {
+      "epoch": 3.5972155544887183,
+      "grad_norm": 0.9548860517089505,
+      "learning_rate": 2.215710957020914e-06,
+      "loss": 0.5389,
+      "step": 3748
+    },
+    {
+      "epoch": 3.598175708113298,
+      "grad_norm": 1.041411201420578,
+      "learning_rate": 2.2129290789673085e-06,
+      "loss": 0.5052,
+      "step": 3749
+    },
+    {
+      "epoch": 3.599135861737878,
+      "grad_norm": 1.1026477327746016,
+      "learning_rate": 2.2101484520000456e-06,
+      "loss": 0.4706,
+      "step": 3750
+    },
+    {
+      "epoch": 3.600096015362458,
+      "grad_norm": 1.0318213135724337,
+      "learning_rate": 2.2073690773673203e-06,
+      "loss": 0.4997,
+      "step": 3751
+    },
+    {
+      "epoch": 3.601056168987038,
+      "grad_norm": 0.8807303009517201,
+      "learning_rate": 2.2045909563167643e-06,
+      "loss": 0.4689,
+      "step": 3752
+    },
+    {
+      "epoch": 3.6020163226116177,
+      "grad_norm": 0.9186984880227923,
+      "learning_rate": 2.201814090095448e-06,
+      "loss": 0.3609,
+      "step": 3753
+    },
+    {
+      "epoch": 3.602976476236198,
+      "grad_norm": 0.8214188000362983,
+      "learning_rate": 2.1990384799498767e-06,
+      "loss": 0.5032,
+      "step": 3754
+    },
+    {
+      "epoch": 3.603936629860778,
+      "grad_norm": 1.182496994616656,
+      "learning_rate": 2.196264127125989e-06,
+      "loss": 0.3753,
+      "step": 3755
+    },
+    {
+      "epoch": 3.6048967834853576,
+      "grad_norm": 1.1228670731642691,
+      "learning_rate": 2.193491032869165e-06,
+      "loss": 0.4307,
+      "step": 3756
+    },
+    {
+      "epoch": 3.6058569371099374,
+      "grad_norm": 1.137019141890346,
+      "learning_rate": 2.190719198424216e-06,
+      "loss": 0.5454,
+      "step": 3757
+    },
+    {
+      "epoch": 3.6068170907345176,
+      "grad_norm": 1.0369007609946983,
+      "learning_rate": 2.1879486250353894e-06,
+      "loss": 0.5174,
+      "step": 3758
+    },
+    {
+      "epoch": 3.6077772443590974,
+      "grad_norm": 1.3671917447731123,
+      "learning_rate": 2.185179313946368e-06,
+      "loss": 0.4332,
+      "step": 3759
+    },
+    {
+      "epoch": 3.6087373979836777,
+      "grad_norm": 1.3303459448543085,
+      "learning_rate": 2.1824112664002612e-06,
+      "loss": 0.475,
+      "step": 3760
+    },
+    {
+      "epoch": 3.6096975516082574,
+      "grad_norm": 1.151641920586827,
+      "learning_rate": 2.1796444836396174e-06,
+      "loss": 0.4152,
+      "step": 3761
+    },
+    {
+      "epoch": 3.6106577052328372,
+      "grad_norm": 1.2268694170133585,
+      "learning_rate": 2.1768789669064205e-06,
+      "loss": 0.4331,
+      "step": 3762
+    },
+    {
+      "epoch": 3.611617858857417,
+      "grad_norm": 1.1768682100365933,
+      "learning_rate": 2.174114717442076e-06,
+      "loss": 0.508,
+      "step": 3763
+    },
+    {
+      "epoch": 3.6125780124819973,
+      "grad_norm": 1.3976781563151333,
+      "learning_rate": 2.1713517364874297e-06,
+      "loss": 0.5216,
+      "step": 3764
+    },
+    {
+      "epoch": 3.613538166106577,
+      "grad_norm": 1.131551678961866,
+      "learning_rate": 2.1685900252827485e-06,
+      "loss": 0.5117,
+      "step": 3765
+    },
+    {
+      "epoch": 3.614498319731157,
+      "grad_norm": 0.9706726329158502,
+      "learning_rate": 2.1658295850677408e-06,
+      "loss": 0.4293,
+      "step": 3766
+    },
+    {
+      "epoch": 3.615458473355737,
+      "grad_norm": 1.2978353247982275,
+      "learning_rate": 2.1630704170815386e-06,
+      "loss": 0.5624,
+      "step": 3767
+    },
+    {
+      "epoch": 3.616418626980317,
+      "grad_norm": 1.116083473577353,
+      "learning_rate": 2.160312522562699e-06,
+      "loss": 0.4809,
+      "step": 3768
+    },
+    {
+      "epoch": 3.6173787806048967,
+      "grad_norm": 1.3468108530134666,
+      "learning_rate": 2.1575559027492136e-06,
+      "loss": 0.4204,
+      "step": 3769
+    },
+    {
+      "epoch": 3.6183389342294765,
+      "grad_norm": 1.5594933534085889,
+      "learning_rate": 2.1548005588785e-06,
+      "loss": 0.45,
+      "step": 3770
+    },
+    {
+      "epoch": 3.6192990878540567,
+      "grad_norm": 0.8602918576594968,
+      "learning_rate": 2.1520464921874014e-06,
+      "loss": 0.4982,
+      "step": 3771
+    },
+    {
+      "epoch": 3.6202592414786365,
+      "grad_norm": 1.2003644889457366,
+      "learning_rate": 2.1492937039121914e-06,
+      "loss": 0.466,
+      "step": 3772
+    },
+    {
+      "epoch": 3.6212193951032168,
+      "grad_norm": 1.223810743019791,
+      "learning_rate": 2.1465421952885633e-06,
+      "loss": 0.4404,
+      "step": 3773
+    },
+    {
+      "epoch": 3.6221795487277966,
+      "grad_norm": 1.1282777272302784,
+      "learning_rate": 2.1437919675516407e-06,
+      "loss": 0.4875,
+      "step": 3774
+    },
+    {
+      "epoch": 3.6231397023523764,
+      "grad_norm": 1.4238496666103069,
+      "learning_rate": 2.1410430219359757e-06,
+      "loss": 0.4999,
+      "step": 3775
+    },
+    {
+      "epoch": 3.624099855976956,
+      "grad_norm": 1.1399952075571127,
+      "learning_rate": 2.1382953596755364e-06,
+      "loss": 0.516,
+      "step": 3776
+    },
+    {
+      "epoch": 3.6250600096015364,
+      "grad_norm": 1.1675105494430775,
+      "learning_rate": 2.1355489820037216e-06,
+      "loss": 0.484,
+      "step": 3777
+    },
+    {
+      "epoch": 3.626020163226116,
+      "grad_norm": 1.5866934246571585,
+      "learning_rate": 2.132803890153346e-06,
+      "loss": 0.4361,
+      "step": 3778
+    },
+    {
+      "epoch": 3.626980316850696,
+      "grad_norm": 1.2903098040377423,
+      "learning_rate": 2.1300600853566577e-06,
+      "loss": 0.4428,
+      "step": 3779
+    },
+    {
+      "epoch": 3.6279404704752762,
+      "grad_norm": 1.1407984683048416,
+      "learning_rate": 2.12731756884532e-06,
+      "loss": 0.4933,
+      "step": 3780
+    },
+    {
+      "epoch": 3.628900624099856,
+      "grad_norm": 1.027459534244755,
+      "learning_rate": 2.1245763418504173e-06,
+      "loss": 0.4886,
+      "step": 3781
+    },
+    {
+      "epoch": 3.629860777724436,
+      "grad_norm": 1.0942181619953522,
+      "learning_rate": 2.121836405602458e-06,
+      "loss": 0.51,
+      "step": 3782
+    },
+    {
+      "epoch": 3.6308209313490156,
+      "grad_norm": 0.7417101969205426,
+      "learning_rate": 2.1190977613313707e-06,
+      "loss": 0.5201,
+      "step": 3783
+    },
+    {
+      "epoch": 3.631781084973596,
+      "grad_norm": 1.0582103684843327,
+      "learning_rate": 2.1163604102665023e-06,
+      "loss": 0.6581,
+      "step": 3784
+    },
+    {
+      "epoch": 3.6327412385981757,
+      "grad_norm": 1.6004720545643263,
+      "learning_rate": 2.113624353636624e-06,
+      "loss": 0.4421,
+      "step": 3785
+    },
+    {
+      "epoch": 3.633701392222756,
+      "grad_norm": 1.2380137171182835,
+      "learning_rate": 2.110889592669918e-06,
+      "loss": 0.4854,
+      "step": 3786
+    },
+    {
+      "epoch": 3.6346615458473357,
+      "grad_norm": 1.0655190265487184,
+      "learning_rate": 2.1081561285939904e-06,
+      "loss": 0.447,
+      "step": 3787
+    },
+    {
+      "epoch": 3.6356216994719155,
+      "grad_norm": 1.0190173251848118,
+      "learning_rate": 2.1054239626358648e-06,
+      "loss": 0.5404,
+      "step": 3788
+    },
+    {
+      "epoch": 3.6365818530964953,
+      "grad_norm": 0.9016277543333088,
+      "learning_rate": 2.102693096021981e-06,
+      "loss": 0.4645,
+      "step": 3789
+    },
+    {
+      "epoch": 3.6375420067210755,
+      "grad_norm": 1.0905249306293494,
+      "learning_rate": 2.0999635299781976e-06,
+      "loss": 0.529,
+      "step": 3790
+    },
+    {
+      "epoch": 3.6385021603456553,
+      "grad_norm": 1.0432333904590128,
+      "learning_rate": 2.0972352657297845e-06,
+      "loss": 0.411,
+      "step": 3791
+    },
+    {
+      "epoch": 3.639462313970235,
+      "grad_norm": 1.0182766520315567,
+      "learning_rate": 2.0945083045014296e-06,
+      "loss": 0.5278,
+      "step": 3792
+    },
+    {
+      "epoch": 3.6404224675948154,
+      "grad_norm": 1.0280915175157022,
+      "learning_rate": 2.0917826475172427e-06,
+      "loss": 0.5295,
+      "step": 3793
+    },
+    {
+      "epoch": 3.641382621219395,
+      "grad_norm": 1.3129083329449196,
+      "learning_rate": 2.089058296000736e-06,
+      "loss": 0.4705,
+      "step": 3794
+    },
+    {
+      "epoch": 3.642342774843975,
+      "grad_norm": 1.4819794363441172,
+      "learning_rate": 2.0863352511748466e-06,
+      "loss": 0.5643,
+      "step": 3795
+    },
+    {
+      "epoch": 3.6433029284685547,
+      "grad_norm": 1.048438135848867,
+      "learning_rate": 2.083613514261913e-06,
+      "loss": 0.4837,
+      "step": 3796
+    },
+    {
+      "epoch": 3.644263082093135,
+      "grad_norm": 1.0991255701955989,
+      "learning_rate": 2.0808930864837007e-06,
+      "loss": 0.4736,
+      "step": 3797
+    },
+    {
+      "epoch": 3.645223235717715,
+      "grad_norm": 1.060065825731226,
+      "learning_rate": 2.07817396906138e-06,
+      "loss": 0.506,
+      "step": 3798
+    },
+    {
+      "epoch": 3.646183389342295,
+      "grad_norm": 1.0305012563717666,
+      "learning_rate": 2.0754561632155307e-06,
+      "loss": 0.426,
+      "step": 3799
+    },
+    {
+      "epoch": 3.647143542966875,
+      "grad_norm": 0.931002501810837,
+      "learning_rate": 2.072739670166148e-06,
+      "loss": 0.463,
+      "step": 3800
+    },
+    {
+      "epoch": 3.6481036965914546,
+      "grad_norm": 1.0176314728629507,
+      "learning_rate": 2.070024491132636e-06,
+      "loss": 0.5337,
+      "step": 3801
+    },
+    {
+      "epoch": 3.6490638502160344,
+      "grad_norm": 0.8311913086495,
+      "learning_rate": 2.067310627333811e-06,
+      "loss": 0.4432,
+      "step": 3802
+    },
+    {
+      "epoch": 3.6500240038406147,
+      "grad_norm": 1.014117237799301,
+      "learning_rate": 2.064598079987898e-06,
+      "loss": 0.4402,
+      "step": 3803
+    },
+    {
+      "epoch": 3.6509841574651944,
+      "grad_norm": 0.6819680264762884,
+      "learning_rate": 2.0618868503125273e-06,
+      "loss": 0.5504,
+      "step": 3804
+    },
+    {
+      "epoch": 3.6519443110897742,
+      "grad_norm": 1.1872475867322425,
+      "learning_rate": 2.059176939524741e-06,
+      "loss": 0.4089,
+      "step": 3805
+    },
+    {
+      "epoch": 3.6529044647143545,
+      "grad_norm": 1.1212236818671708,
+      "learning_rate": 2.056468348840995e-06,
+      "loss": 0.4413,
+      "step": 3806
+    },
+    {
+      "epoch": 3.6538646183389343,
+      "grad_norm": 1.0698093019264092,
+      "learning_rate": 2.0537610794771405e-06,
+      "loss": 0.4692,
+      "step": 3807
+    },
+    {
+      "epoch": 3.654824771963514,
+      "grad_norm": 0.8909055328097167,
+      "learning_rate": 2.0510551326484457e-06,
+      "loss": 0.418,
+      "step": 3808
+    },
+    {
+      "epoch": 3.655784925588094,
+      "grad_norm": 1.3020564706671507,
+      "learning_rate": 2.0483505095695755e-06,
+      "loss": 0.4975,
+      "step": 3809
+    },
+    {
+      "epoch": 3.656745079212674,
+      "grad_norm": 1.003057278453496,
+      "learning_rate": 2.045647211454613e-06,
+      "loss": 0.5627,
+      "step": 3810
+    },
+    {
+      "epoch": 3.657705232837254,
+      "grad_norm": 1.0211388788908284,
+      "learning_rate": 2.0429452395170372e-06,
+      "loss": 0.4647,
+      "step": 3811
+    },
+    {
+      "epoch": 3.658665386461834,
+      "grad_norm": 1.504892499000519,
+      "learning_rate": 2.0402445949697335e-06,
+      "loss": 0.4165,
+      "step": 3812
+    },
+    {
+      "epoch": 3.659625540086414,
+      "grad_norm": 1.0857535171455013,
+      "learning_rate": 2.037545279024993e-06,
+      "loss": 0.4049,
+      "step": 3813
+    },
+    {
+      "epoch": 3.6605856937109937,
+      "grad_norm": 1.1869996238194045,
+      "learning_rate": 2.0348472928945096e-06,
+      "loss": 0.3666,
+      "step": 3814
+    },
+    {
+      "epoch": 3.6615458473355735,
+      "grad_norm": 1.168459723293314,
+      "learning_rate": 2.0321506377893807e-06,
+      "loss": 0.4781,
+      "step": 3815
+    },
+    {
+      "epoch": 3.6625060009601538,
+      "grad_norm": 1.0862836544926067,
+      "learning_rate": 2.029455314920108e-06,
+      "loss": 0.4357,
+      "step": 3816
+    },
+    {
+      "epoch": 3.6634661545847336,
+      "grad_norm": 1.056344464364202,
+      "learning_rate": 2.026761325496589e-06,
+      "loss": 0.3488,
+      "step": 3817
+    },
+    {
+      "epoch": 3.6644263082093134,
+      "grad_norm": 1.6004071202908687,
+      "learning_rate": 2.024068670728127e-06,
+      "loss": 0.6168,
+      "step": 3818
+    },
+    {
+      "epoch": 3.6653864618338936,
+      "grad_norm": 2.1330535107668678,
+      "learning_rate": 2.021377351823432e-06,
+      "loss": 0.527,
+      "step": 3819
+    },
+    {
+      "epoch": 3.6663466154584734,
+      "grad_norm": 1.1470659153663634,
+      "learning_rate": 2.0186873699906017e-06,
+      "loss": 0.4922,
+      "step": 3820
+    },
+    {
+      "epoch": 3.667306769083053,
+      "grad_norm": 1.0598011721769491,
+      "learning_rate": 2.0159987264371445e-06,
+      "loss": 0.519,
+      "step": 3821
+    },
+    {
+      "epoch": 3.668266922707633,
+      "grad_norm": 0.9138033948010229,
+      "learning_rate": 2.0133114223699577e-06,
+      "loss": 0.4927,
+      "step": 3822
+    },
+    {
+      "epoch": 3.6692270763322132,
+      "grad_norm": 0.9235137477631645,
+      "learning_rate": 2.01062545899535e-06,
+      "loss": 0.566,
+      "step": 3823
+    },
+    {
+      "epoch": 3.670187229956793,
+      "grad_norm": 1.1116614261931967,
+      "learning_rate": 2.007940837519021e-06,
+      "loss": 0.4614,
+      "step": 3824
+    },
+    {
+      "epoch": 3.6711473835813733,
+      "grad_norm": 0.9962640331207523,
+      "learning_rate": 2.0052575591460636e-06,
+      "loss": 0.4634,
+      "step": 3825
+    },
+    {
+      "epoch": 3.672107537205953,
+      "grad_norm": 1.001008425099837,
+      "learning_rate": 2.002575625080977e-06,
+      "loss": 0.5174,
+      "step": 3826
+    },
+    {
+      "epoch": 3.673067690830533,
+      "grad_norm": 1.0279895057906892,
+      "learning_rate": 1.9998950365276514e-06,
+      "loss": 0.5231,
+      "step": 3827
+    },
+    {
+      "epoch": 3.6740278444551127,
+      "grad_norm": 1.2631330092758803,
+      "learning_rate": 1.997215794689375e-06,
+      "loss": 0.5177,
+      "step": 3828
+    },
+    {
+      "epoch": 3.674987998079693,
+      "grad_norm": 1.0452875649365183,
+      "learning_rate": 1.9945379007688314e-06,
+      "loss": 0.5364,
+      "step": 3829
+    },
+    {
+      "epoch": 3.6759481517042727,
+      "grad_norm": 1.1260108577949433,
+      "learning_rate": 1.991861355968096e-06,
+      "loss": 0.5284,
+      "step": 3830
+    },
+    {
+      "epoch": 3.6769083053288525,
+      "grad_norm": 0.7148469060758215,
+      "learning_rate": 1.989186161488643e-06,
+      "loss": 0.4806,
+      "step": 3831
+    },
+    {
+      "epoch": 3.6778684589534327,
+      "grad_norm": 1.063952880342718,
+      "learning_rate": 1.986512318531338e-06,
+      "loss": 0.6073,
+      "step": 3832
+    },
+    {
+      "epoch": 3.6788286125780125,
+      "grad_norm": 1.137127963521302,
+      "learning_rate": 1.9838398282964412e-06,
+      "loss": 0.4586,
+      "step": 3833
+    },
+    {
+      "epoch": 3.6797887662025923,
+      "grad_norm": 1.1803226032299732,
+      "learning_rate": 1.9811686919836054e-06,
+      "loss": 0.389,
+      "step": 3834
+    },
+    {
+      "epoch": 3.680748919827172,
+      "grad_norm": 0.9911285489470352,
+      "learning_rate": 1.9784989107918736e-06,
+      "loss": 0.5239,
+      "step": 3835
+    },
+    {
+      "epoch": 3.6817090734517524,
+      "grad_norm": 0.8276777288490931,
+      "learning_rate": 1.97583048591968e-06,
+      "loss": 0.4406,
+      "step": 3836
+    },
+    {
+      "epoch": 3.682669227076332,
+      "grad_norm": 0.947486413230033,
+      "learning_rate": 1.9731634185648584e-06,
+      "loss": 0.4647,
+      "step": 3837
+    },
+    {
+      "epoch": 3.6836293807009124,
+      "grad_norm": 1.1926616009097457,
+      "learning_rate": 1.970497709924622e-06,
+      "loss": 0.552,
+      "step": 3838
+    },
+    {
+      "epoch": 3.684589534325492,
+      "grad_norm": 1.3684420975143503,
+      "learning_rate": 1.9678333611955814e-06,
+      "loss": 0.3903,
+      "step": 3839
+    },
+    {
+      "epoch": 3.685549687950072,
+      "grad_norm": 1.023355141836186,
+      "learning_rate": 1.965170373573729e-06,
+      "loss": 0.5019,
+      "step": 3840
+    },
+    {
+      "epoch": 3.686509841574652,
+      "grad_norm": 1.6823560265523865,
+      "learning_rate": 1.9625087482544574e-06,
+      "loss": 0.5438,
+      "step": 3841
+    },
+    {
+      "epoch": 3.687469995199232,
+      "grad_norm": 1.1916669996347646,
+      "learning_rate": 1.959848486432542e-06,
+      "loss": 0.4666,
+      "step": 3842
+    },
+    {
+      "epoch": 3.688430148823812,
+      "grad_norm": 0.9046535451881969,
+      "learning_rate": 1.957189589302141e-06,
+      "loss": 0.5544,
+      "step": 3843
+    },
+    {
+      "epoch": 3.6893903024483916,
+      "grad_norm": 1.3042539268476607,
+      "learning_rate": 1.9545320580568077e-06,
+      "loss": 0.3806,
+      "step": 3844
+    },
+    {
+      "epoch": 3.690350456072972,
+      "grad_norm": 1.4284881962942269,
+      "learning_rate": 1.95187589388948e-06,
+      "loss": 0.4429,
+      "step": 3845
+    },
+    {
+      "epoch": 3.6913106096975516,
+      "grad_norm": 1.1642344349877685,
+      "learning_rate": 1.9492210979924805e-06,
+      "loss": 0.4918,
+      "step": 3846
+    },
+    {
+      "epoch": 3.6922707633221314,
+      "grad_norm": 0.799241305967219,
+      "learning_rate": 1.9465676715575203e-06,
+      "loss": 0.4752,
+      "step": 3847
+    },
+    {
+      "epoch": 3.6932309169467112,
+      "grad_norm": 1.0028058610874377,
+      "learning_rate": 1.9439156157756917e-06,
+      "loss": 0.5389,
+      "step": 3848
+    },
+    {
+      "epoch": 3.6941910705712915,
+      "grad_norm": 1.093868951632084,
+      "learning_rate": 1.9412649318374737e-06,
+      "loss": 0.5107,
+      "step": 3849
+    },
+    {
+      "epoch": 3.6951512241958713,
+      "grad_norm": 1.2293346312824376,
+      "learning_rate": 1.938615620932735e-06,
+      "loss": 0.4363,
+      "step": 3850
+    },
+    {
+      "epoch": 3.6961113778204515,
+      "grad_norm": 1.5130933913838918,
+      "learning_rate": 1.9359676842507186e-06,
+      "loss": 0.5077,
+      "step": 3851
+    },
+    {
+      "epoch": 3.6970715314450313,
+      "grad_norm": 1.1238119096822885,
+      "learning_rate": 1.933321122980057e-06,
+      "loss": 0.528,
+      "step": 3852
+    },
+    {
+      "epoch": 3.698031685069611,
+      "grad_norm": 1.2140115401556606,
+      "learning_rate": 1.9306759383087593e-06,
+      "loss": 0.4128,
+      "step": 3853
+    },
+    {
+      "epoch": 3.698991838694191,
+      "grad_norm": 1.8592799022004682,
+      "learning_rate": 1.9280321314242252e-06,
+      "loss": 0.3679,
+      "step": 3854
+    },
+    {
+      "epoch": 3.699951992318771,
+      "grad_norm": 1.2246118458849933,
+      "learning_rate": 1.925389703513232e-06,
+      "loss": 0.4778,
+      "step": 3855
+    },
+    {
+      "epoch": 3.700912145943351,
+      "grad_norm": 1.1750832993551101,
+      "learning_rate": 1.922748655761932e-06,
+      "loss": 0.5096,
+      "step": 3856
+    },
+    {
+      "epoch": 3.7018722995679307,
+      "grad_norm": 1.0519279990408164,
+      "learning_rate": 1.9201089893558677e-06,
+      "loss": 0.4687,
+      "step": 3857
+    },
+    {
+      "epoch": 3.702832453192511,
+      "grad_norm": 1.0062094032320303,
+      "learning_rate": 1.9174707054799558e-06,
+      "loss": 0.445,
+      "step": 3858
+    },
+    {
+      "epoch": 3.7037926068170908,
+      "grad_norm": 0.9081150690693638,
+      "learning_rate": 1.9148338053184933e-06,
+      "loss": 0.5094,
+      "step": 3859
+    },
+    {
+      "epoch": 3.7047527604416706,
+      "grad_norm": 0.9917559328970503,
+      "learning_rate": 1.9121982900551592e-06,
+      "loss": 0.535,
+      "step": 3860
+    },
+    {
+      "epoch": 3.7057129140662504,
+      "grad_norm": 1.5002787408890847,
+      "learning_rate": 1.909564160873004e-06,
+      "loss": 0.4836,
+      "step": 3861
+    },
+    {
+      "epoch": 3.7066730676908306,
+      "grad_norm": 1.1177091307108697,
+      "learning_rate": 1.9069314189544602e-06,
+      "loss": 0.4152,
+      "step": 3862
+    },
+    {
+      "epoch": 3.7076332213154104,
+      "grad_norm": 1.0614380295231434,
+      "learning_rate": 1.9043000654813433e-06,
+      "loss": 0.4936,
+      "step": 3863
+    },
+    {
+      "epoch": 3.7085933749399906,
+      "grad_norm": 1.1500221965488515,
+      "learning_rate": 1.9016701016348338e-06,
+      "loss": 0.4418,
+      "step": 3864
+    },
+    {
+      "epoch": 3.7095535285645704,
+      "grad_norm": 1.3080713530254886,
+      "learning_rate": 1.8990415285954976e-06,
+      "loss": 0.4278,
+      "step": 3865
+    },
+    {
+      "epoch": 3.7105136821891502,
+      "grad_norm": 1.000435917697426,
+      "learning_rate": 1.896414347543269e-06,
+      "loss": 0.5226,
+      "step": 3866
+    },
+    {
+      "epoch": 3.71147383581373,
+      "grad_norm": 1.0253677138621886,
+      "learning_rate": 1.8937885596574656e-06,
+      "loss": 0.5902,
+      "step": 3867
+    },
+    {
+      "epoch": 3.7124339894383103,
+      "grad_norm": 1.7138951431289862,
+      "learning_rate": 1.8911641661167762e-06,
+      "loss": 0.4834,
+      "step": 3868
+    },
+    {
+      "epoch": 3.71339414306289,
+      "grad_norm": 0.9001297497384394,
+      "learning_rate": 1.8885411680992594e-06,
+      "loss": 0.472,
+      "step": 3869
+    },
+    {
+      "epoch": 3.71435429668747,
+      "grad_norm": 1.2134955052246705,
+      "learning_rate": 1.885919566782352e-06,
+      "loss": 0.3747,
+      "step": 3870
+    },
+    {
+      "epoch": 3.71531445031205,
+      "grad_norm": 1.2486109383089363,
+      "learning_rate": 1.8832993633428643e-06,
+      "loss": 0.5055,
+      "step": 3871
+    },
+    {
+      "epoch": 3.71627460393663,
+      "grad_norm": 0.9578237580606398,
+      "learning_rate": 1.880680558956976e-06,
+      "loss": 0.4621,
+      "step": 3872
+    },
+    {
+      "epoch": 3.7172347575612097,
+      "grad_norm": 1.1294383846309326,
+      "learning_rate": 1.878063154800243e-06,
+      "loss": 0.4249,
+      "step": 3873
+    },
+    {
+      "epoch": 3.7181949111857895,
+      "grad_norm": 1.1396561743373135,
+      "learning_rate": 1.8754471520475863e-06,
+      "loss": 0.5239,
+      "step": 3874
+    },
+    {
+      "epoch": 3.7191550648103697,
+      "grad_norm": 1.4387812119997676,
+      "learning_rate": 1.8728325518733038e-06,
+      "loss": 0.4629,
+      "step": 3875
+    },
+    {
+      "epoch": 3.7201152184349495,
+      "grad_norm": 0.9601518279654647,
+      "learning_rate": 1.870219355451061e-06,
+      "loss": 0.4521,
+      "step": 3876
+    },
+    {
+      "epoch": 3.7210753720595298,
+      "grad_norm": 1.1369061406321572,
+      "learning_rate": 1.8676075639538943e-06,
+      "loss": 0.3902,
+      "step": 3877
+    },
+    {
+      "epoch": 3.7220355256841096,
+      "grad_norm": 1.264290690564499,
+      "learning_rate": 1.8649971785542097e-06,
+      "loss": 0.4139,
+      "step": 3878
+    },
+    {
+      "epoch": 3.7229956793086894,
+      "grad_norm": 0.9841744706554263,
+      "learning_rate": 1.8623882004237792e-06,
+      "loss": 0.4981,
+      "step": 3879
+    },
+    {
+      "epoch": 3.723955832933269,
+      "grad_norm": 1.2140239395794963,
+      "learning_rate": 1.8597806307337446e-06,
+      "loss": 0.4605,
+      "step": 3880
+    },
+    {
+      "epoch": 3.7249159865578494,
+      "grad_norm": 1.0122809883079258,
+      "learning_rate": 1.8571744706546213e-06,
+      "loss": 0.5375,
+      "step": 3881
+    },
+    {
+      "epoch": 3.725876140182429,
+      "grad_norm": 0.9299945433671742,
+      "learning_rate": 1.8545697213562813e-06,
+      "loss": 0.358,
+      "step": 3882
+    },
+    {
+      "epoch": 3.726836293807009,
+      "grad_norm": 1.188804231317955,
+      "learning_rate": 1.8519663840079721e-06,
+      "loss": 0.5174,
+      "step": 3883
+    },
+    {
+      "epoch": 3.727796447431589,
+      "grad_norm": 1.1092629350185208,
+      "learning_rate": 1.8493644597782996e-06,
+      "loss": 0.4704,
+      "step": 3884
+    },
+    {
+      "epoch": 3.728756601056169,
+      "grad_norm": 1.2262609049279667,
+      "learning_rate": 1.8467639498352442e-06,
+      "loss": 0.38,
+      "step": 3885
+    },
+    {
+      "epoch": 3.729716754680749,
+      "grad_norm": 0.9560647319306025,
+      "learning_rate": 1.8441648553461467e-06,
+      "loss": 0.5391,
+      "step": 3886
+    },
+    {
+      "epoch": 3.7306769083053286,
+      "grad_norm": 1.086947314689461,
+      "learning_rate": 1.8415671774777106e-06,
+      "loss": 0.4768,
+      "step": 3887
+    },
+    {
+      "epoch": 3.731637061929909,
+      "grad_norm": 1.0511073997157976,
+      "learning_rate": 1.8389709173960063e-06,
+      "loss": 0.4915,
+      "step": 3888
+    },
+    {
+      "epoch": 3.7325972155544886,
+      "grad_norm": 0.9677959792317167,
+      "learning_rate": 1.8363760762664678e-06,
+      "loss": 0.4843,
+      "step": 3889
+    },
+    {
+      "epoch": 3.733557369179069,
+      "grad_norm": 1.0713756992235126,
+      "learning_rate": 1.8337826552538911e-06,
+      "loss": 0.4454,
+      "step": 3890
+    },
+    {
+      "epoch": 3.7345175228036487,
+      "grad_norm": 0.9423399407475826,
+      "learning_rate": 1.8311906555224379e-06,
+      "loss": 0.5349,
+      "step": 3891
+    },
+    {
+      "epoch": 3.7354776764282285,
+      "grad_norm": 0.9525167268798441,
+      "learning_rate": 1.8286000782356245e-06,
+      "loss": 0.5551,
+      "step": 3892
+    },
+    {
+      "epoch": 3.7364378300528083,
+      "grad_norm": 0.904942346769752,
+      "learning_rate": 1.826010924556334e-06,
+      "loss": 0.5342,
+      "step": 3893
+    },
+    {
+      "epoch": 3.7373979836773885,
+      "grad_norm": 0.8432772139184227,
+      "learning_rate": 1.823423195646815e-06,
+      "loss": 0.4401,
+      "step": 3894
+    },
+    {
+      "epoch": 3.7383581373019683,
+      "grad_norm": 1.211176320508855,
+      "learning_rate": 1.820836892668666e-06,
+      "loss": 0.5373,
+      "step": 3895
+    },
+    {
+      "epoch": 3.739318290926548,
+      "grad_norm": 0.980186133683255,
+      "learning_rate": 1.8182520167828537e-06,
+      "loss": 0.467,
+      "step": 3896
+    },
+    {
+      "epoch": 3.7402784445511283,
+      "grad_norm": 1.1333613881232587,
+      "learning_rate": 1.8156685691496972e-06,
+      "loss": 0.4619,
+      "step": 3897
+    },
+    {
+      "epoch": 3.741238598175708,
+      "grad_norm": 1.079537447614277,
+      "learning_rate": 1.813086550928883e-06,
+      "loss": 0.4547,
+      "step": 3898
+    },
+    {
+      "epoch": 3.742198751800288,
+      "grad_norm": 1.6859175540977833,
+      "learning_rate": 1.8105059632794514e-06,
+      "loss": 0.5553,
+      "step": 3899
+    },
+    {
+      "epoch": 3.7431589054248677,
+      "grad_norm": 1.0163050132852784,
+      "learning_rate": 1.8079268073597978e-06,
+      "loss": 0.6202,
+      "step": 3900
+    },
+    {
+      "epoch": 3.744119059049448,
+      "grad_norm": 1.4034299593983564,
+      "learning_rate": 1.8053490843276788e-06,
+      "loss": 0.4142,
+      "step": 3901
+    },
+    {
+      "epoch": 3.7450792126740278,
+      "grad_norm": 0.8635569537928273,
+      "learning_rate": 1.802772795340207e-06,
+      "loss": 0.5466,
+      "step": 3902
+    },
+    {
+      "epoch": 3.746039366298608,
+      "grad_norm": 1.1086508319040311,
+      "learning_rate": 1.8001979415538507e-06,
+      "loss": 0.4346,
+      "step": 3903
+    },
+    {
+      "epoch": 3.746999519923188,
+      "grad_norm": 0.8020501928764563,
+      "learning_rate": 1.797624524124436e-06,
+      "loss": 0.5652,
+      "step": 3904
+    },
+    {
+      "epoch": 3.7479596735477676,
+      "grad_norm": 1.205370198851128,
+      "learning_rate": 1.7950525442071394e-06,
+      "loss": 0.4764,
+      "step": 3905
+    },
+    {
+      "epoch": 3.7489198271723474,
+      "grad_norm": 1.176167883572754,
+      "learning_rate": 1.792482002956495e-06,
+      "loss": 0.5206,
+      "step": 3906
+    },
+    {
+      "epoch": 3.7498799807969276,
+      "grad_norm": 1.5349338978715579,
+      "learning_rate": 1.7899129015263966e-06,
+      "loss": 0.4927,
+      "step": 3907
+    },
+    {
+      "epoch": 3.7508401344215074,
+      "grad_norm": 1.0906294491057715,
+      "learning_rate": 1.7873452410700809e-06,
+      "loss": 0.3882,
+      "step": 3908
+    },
+    {
+      "epoch": 3.7518002880460872,
+      "grad_norm": 0.9240064371691598,
+      "learning_rate": 1.7847790227401469e-06,
+      "loss": 0.5617,
+      "step": 3909
+    },
+    {
+      "epoch": 3.7527604416706675,
+      "grad_norm": 1.116016545720751,
+      "learning_rate": 1.7822142476885374e-06,
+      "loss": 0.4073,
+      "step": 3910
+    },
+    {
+      "epoch": 3.7537205952952473,
+      "grad_norm": 1.2140127137387613,
+      "learning_rate": 1.7796509170665571e-06,
+      "loss": 0.5582,
+      "step": 3911
+    },
+    {
+      "epoch": 3.754680748919827,
+      "grad_norm": 1.2331357315877847,
+      "learning_rate": 1.777089032024859e-06,
+      "loss": 0.4364,
+      "step": 3912
+    },
+    {
+      "epoch": 3.755640902544407,
+      "grad_norm": 1.4016512918660973,
+      "learning_rate": 1.774528593713441e-06,
+      "loss": 0.4629,
+      "step": 3913
+    },
+    {
+      "epoch": 3.756601056168987,
+      "grad_norm": 1.1567842910412773,
+      "learning_rate": 1.771969603281659e-06,
+      "loss": 0.5094,
+      "step": 3914
+    },
+    {
+      "epoch": 3.757561209793567,
+      "grad_norm": 0.9150352630096552,
+      "learning_rate": 1.7694120618782169e-06,
+      "loss": 0.4956,
+      "step": 3915
+    },
+    {
+      "epoch": 3.758521363418147,
+      "grad_norm": 1.2874225313145111,
+      "learning_rate": 1.7668559706511667e-06,
+      "loss": 0.4894,
+      "step": 3916
+    },
+    {
+      "epoch": 3.759481517042727,
+      "grad_norm": 0.8103318851271448,
+      "learning_rate": 1.7643013307479129e-06,
+      "loss": 0.4292,
+      "step": 3917
+    },
+    {
+      "epoch": 3.7604416706673067,
+      "grad_norm": 1.3081346179824862,
+      "learning_rate": 1.7617481433152022e-06,
+      "loss": 0.4678,
+      "step": 3918
+    },
+    {
+      "epoch": 3.7614018242918865,
+      "grad_norm": 1.600637321529164,
+      "learning_rate": 1.759196409499135e-06,
+      "loss": 0.4621,
+      "step": 3919
+    },
+    {
+      "epoch": 3.7623619779164668,
+      "grad_norm": 1.7248563076898713,
+      "learning_rate": 1.756646130445157e-06,
+      "loss": 0.4591,
+      "step": 3920
+    },
+    {
+      "epoch": 3.7633221315410466,
+      "grad_norm": 1.407207687240018,
+      "learning_rate": 1.7540973072980621e-06,
+      "loss": 0.5207,
+      "step": 3921
+    },
+    {
+      "epoch": 3.7642822851656264,
+      "grad_norm": 0.9463713629373397,
+      "learning_rate": 1.7515499412019903e-06,
+      "loss": 0.4003,
+      "step": 3922
+    },
+    {
+      "epoch": 3.7652424387902066,
+      "grad_norm": 1.1207872040845508,
+      "learning_rate": 1.7490040333004248e-06,
+      "loss": 0.4715,
+      "step": 3923
+    },
+    {
+      "epoch": 3.7662025924147864,
+      "grad_norm": 0.8120050174698782,
+      "learning_rate": 1.7464595847361954e-06,
+      "loss": 0.5019,
+      "step": 3924
+    },
+    {
+      "epoch": 3.767162746039366,
+      "grad_norm": 1.1402370705923843,
+      "learning_rate": 1.7439165966514831e-06,
+      "loss": 0.5053,
+      "step": 3925
+    },
+    {
+      "epoch": 3.768122899663946,
+      "grad_norm": 1.2563562511011135,
+      "learning_rate": 1.741375070187804e-06,
+      "loss": 0.4028,
+      "step": 3926
+    },
+    {
+      "epoch": 3.769083053288526,
+      "grad_norm": 2.0028719378457343,
+      "learning_rate": 1.7388350064860242e-06,
+      "loss": 0.4408,
+      "step": 3927
+    },
+    {
+      "epoch": 3.770043206913106,
+      "grad_norm": 1.5404120053122512,
+      "learning_rate": 1.7362964066863469e-06,
+      "loss": 0.4039,
+      "step": 3928
+    },
+    {
+      "epoch": 3.7710033605376863,
+      "grad_norm": 0.915029989906004,
+      "learning_rate": 1.7337592719283276e-06,
+      "loss": 0.4721,
+      "step": 3929
+    },
+    {
+      "epoch": 3.771963514162266,
+      "grad_norm": 1.1669503945125583,
+      "learning_rate": 1.7312236033508584e-06,
+      "loss": 0.4725,
+      "step": 3930
+    },
+    {
+      "epoch": 3.772923667786846,
+      "grad_norm": 1.0322905530871467,
+      "learning_rate": 1.7286894020921707e-06,
+      "loss": 0.3973,
+      "step": 3931
+    },
+    {
+      "epoch": 3.7738838214114256,
+      "grad_norm": 1.2102443686153808,
+      "learning_rate": 1.726156669289842e-06,
+      "loss": 0.4912,
+      "step": 3932
+    },
+    {
+      "epoch": 3.774843975036006,
+      "grad_norm": 0.9706334814101453,
+      "learning_rate": 1.7236254060807889e-06,
+      "loss": 0.4616,
+      "step": 3933
+    },
+    {
+      "epoch": 3.7758041286605857,
+      "grad_norm": 0.983761787230528,
+      "learning_rate": 1.7210956136012684e-06,
+      "loss": 0.5281,
+      "step": 3934
+    },
+    {
+      "epoch": 3.7767642822851655,
+      "grad_norm": 1.1635859888182039,
+      "learning_rate": 1.718567292986879e-06,
+      "loss": 0.6095,
+      "step": 3935
+    },
+    {
+      "epoch": 3.7777244359097457,
+      "grad_norm": 0.8723998414425806,
+      "learning_rate": 1.7160404453725543e-06,
+      "loss": 0.4562,
+      "step": 3936
+    },
+    {
+      "epoch": 3.7786845895343255,
+      "grad_norm": 1.1293150166627424,
+      "learning_rate": 1.7135150718925674e-06,
+      "loss": 0.5679,
+      "step": 3937
+    },
+    {
+      "epoch": 3.7796447431589053,
+      "grad_norm": 1.1096584802561167,
+      "learning_rate": 1.7109911736805385e-06,
+      "loss": 0.5861,
+      "step": 3938
+    },
+    {
+      "epoch": 3.780604896783485,
+      "grad_norm": 1.2762288013138434,
+      "learning_rate": 1.708468751869412e-06,
+      "loss": 0.5473,
+      "step": 3939
+    },
+    {
+      "epoch": 3.7815650504080653,
+      "grad_norm": 1.2029755140026448,
+      "learning_rate": 1.7059478075914799e-06,
+      "loss": 0.4715,
+      "step": 3940
+    },
+    {
+      "epoch": 3.782525204032645,
+      "grad_norm": 1.1678231158767183,
+      "learning_rate": 1.703428341978362e-06,
+      "loss": 0.4385,
+      "step": 3941
+    },
+    {
+      "epoch": 3.7834853576572254,
+      "grad_norm": 1.4223721118460644,
+      "learning_rate": 1.7009103561610252e-06,
+      "loss": 0.4775,
+      "step": 3942
+    },
+    {
+      "epoch": 3.784445511281805,
+      "grad_norm": 0.8085978878365441,
+      "learning_rate": 1.6983938512697655e-06,
+      "loss": 0.4579,
+      "step": 3943
+    },
+    {
+      "epoch": 3.785405664906385,
+      "grad_norm": 1.0128151559011735,
+      "learning_rate": 1.695878828434212e-06,
+      "loss": 0.5393,
+      "step": 3944
+    },
+    {
+      "epoch": 3.7863658185309648,
+      "grad_norm": 1.233514523118038,
+      "learning_rate": 1.6933652887833336e-06,
+      "loss": 0.4682,
+      "step": 3945
+    },
+    {
+      "epoch": 3.787325972155545,
+      "grad_norm": 0.7304627487638629,
+      "learning_rate": 1.6908532334454308e-06,
+      "loss": 0.4831,
+      "step": 3946
+    },
+    {
+      "epoch": 3.788286125780125,
+      "grad_norm": 1.16232494135482,
+      "learning_rate": 1.6883426635481392e-06,
+      "loss": 0.5213,
+      "step": 3947
+    },
+    {
+      "epoch": 3.7892462794047046,
+      "grad_norm": 0.7646794081219642,
+      "learning_rate": 1.6858335802184278e-06,
+      "loss": 0.4139,
+      "step": 3948
+    },
+    {
+      "epoch": 3.790206433029285,
+      "grad_norm": 1.3028968325156647,
+      "learning_rate": 1.683325984582594e-06,
+      "loss": 0.4575,
+      "step": 3949
+    },
+    {
+      "epoch": 3.7911665866538646,
+      "grad_norm": 1.4783428596196397,
+      "learning_rate": 1.680819877766272e-06,
+      "loss": 0.4673,
+      "step": 3950
+    },
+    {
+      "epoch": 3.7921267402784444,
+      "grad_norm": 1.498146910428719,
+      "learning_rate": 1.678315260894427e-06,
+      "loss": 0.5331,
+      "step": 3951
+    },
+    {
+      "epoch": 3.7930868939030242,
+      "grad_norm": 1.599070554279104,
+      "learning_rate": 1.6758121350913548e-06,
+      "loss": 0.4335,
+      "step": 3952
+    },
+    {
+      "epoch": 3.7940470475276045,
+      "grad_norm": 1.0546319831143243,
+      "learning_rate": 1.6733105014806834e-06,
+      "loss": 0.4151,
+      "step": 3953
+    },
+    {
+      "epoch": 3.7950072011521843,
+      "grad_norm": 1.0004624288599213,
+      "learning_rate": 1.6708103611853638e-06,
+      "loss": 0.3596,
+      "step": 3954
+    },
+    {
+      "epoch": 3.7959673547767645,
+      "grad_norm": 0.8841693193064396,
+      "learning_rate": 1.6683117153276878e-06,
+      "loss": 0.5316,
+      "step": 3955
+    },
+    {
+      "epoch": 3.7969275084013443,
+      "grad_norm": 1.1553617748917355,
+      "learning_rate": 1.6658145650292707e-06,
+      "loss": 0.5447,
+      "step": 3956
+    },
+    {
+      "epoch": 3.797887662025924,
+      "grad_norm": 1.0770525668556672,
+      "learning_rate": 1.6633189114110531e-06,
+      "loss": 0.4996,
+      "step": 3957
+    },
+    {
+      "epoch": 3.798847815650504,
+      "grad_norm": 1.2620691453511903,
+      "learning_rate": 1.6608247555933093e-06,
+      "loss": 0.4501,
+      "step": 3958
+    },
+    {
+      "epoch": 3.799807969275084,
+      "grad_norm": 1.4312504220509992,
+      "learning_rate": 1.6583320986956386e-06,
+      "loss": 0.4568,
+      "step": 3959
+    },
+    {
+      "epoch": 3.800768122899664,
+      "grad_norm": 1.3747786124916295,
+      "learning_rate": 1.6558409418369686e-06,
+      "loss": 0.6672,
+      "step": 3960
+    },
+    {
+      "epoch": 3.8017282765242437,
+      "grad_norm": 1.1357784839797793,
+      "learning_rate": 1.6533512861355543e-06,
+      "loss": 0.4539,
+      "step": 3961
+    },
+    {
+      "epoch": 3.802688430148824,
+      "grad_norm": 1.4784324229639423,
+      "learning_rate": 1.6508631327089725e-06,
+      "loss": 0.4767,
+      "step": 3962
+    },
+    {
+      "epoch": 3.8036485837734038,
+      "grad_norm": 1.749657398940357,
+      "learning_rate": 1.64837648267413e-06,
+      "loss": 0.528,
+      "step": 3963
+    },
+    {
+      "epoch": 3.8046087373979836,
+      "grad_norm": 1.4001479052399646,
+      "learning_rate": 1.6458913371472573e-06,
+      "loss": 0.4379,
+      "step": 3964
+    },
+    {
+      "epoch": 3.8055688910225633,
+      "grad_norm": 0.9724248120840239,
+      "learning_rate": 1.6434076972439106e-06,
+      "loss": 0.4493,
+      "step": 3965
+    },
+    {
+      "epoch": 3.8065290446471436,
+      "grad_norm": 1.2089777865100662,
+      "learning_rate": 1.6409255640789695e-06,
+      "loss": 0.3999,
+      "step": 3966
+    },
+    {
+      "epoch": 3.8074891982717234,
+      "grad_norm": 1.1890139752612174,
+      "learning_rate": 1.6384449387666351e-06,
+      "loss": 0.3957,
+      "step": 3967
+    },
+    {
+      "epoch": 3.8084493518963036,
+      "grad_norm": 1.0918089262696387,
+      "learning_rate": 1.6359658224204328e-06,
+      "loss": 0.5489,
+      "step": 3968
+    },
+    {
+      "epoch": 3.8094095055208834,
+      "grad_norm": 1.2735572997914448,
+      "learning_rate": 1.6334882161532161e-06,
+      "loss": 0.4587,
+      "step": 3969
+    },
+    {
+      "epoch": 3.810369659145463,
+      "grad_norm": 1.203926288656431,
+      "learning_rate": 1.6310121210771524e-06,
+      "loss": 0.4519,
+      "step": 3970
+    },
+    {
+      "epoch": 3.811329812770043,
+      "grad_norm": 1.2969788010262875,
+      "learning_rate": 1.628537538303736e-06,
+      "loss": 0.5325,
+      "step": 3971
+    },
+    {
+      "epoch": 3.8122899663946233,
+      "grad_norm": 1.1989039386120783,
+      "learning_rate": 1.6260644689437767e-06,
+      "loss": 0.4702,
+      "step": 3972
+    },
+    {
+      "epoch": 3.813250120019203,
+      "grad_norm": 1.0865728623294701,
+      "learning_rate": 1.6235929141074137e-06,
+      "loss": 0.419,
+      "step": 3973
+    },
+    {
+      "epoch": 3.814210273643783,
+      "grad_norm": 1.1972929334935527,
+      "learning_rate": 1.6211228749041007e-06,
+      "loss": 0.4221,
+      "step": 3974
+    },
+    {
+      "epoch": 3.815170427268363,
+      "grad_norm": 1.2058629895169104,
+      "learning_rate": 1.6186543524426101e-06,
+      "loss": 0.4712,
+      "step": 3975
+    },
+    {
+      "epoch": 3.816130580892943,
+      "grad_norm": 0.7150639392966994,
+      "learning_rate": 1.616187347831036e-06,
+      "loss": 0.3704,
+      "step": 3976
+    },
+    {
+      "epoch": 3.8170907345175227,
+      "grad_norm": 1.4134564371765639,
+      "learning_rate": 1.6137218621767902e-06,
+      "loss": 0.4629,
+      "step": 3977
+    },
+    {
+      "epoch": 3.8180508881421025,
+      "grad_norm": 1.0285428509236896,
+      "learning_rate": 1.6112578965866039e-06,
+      "loss": 0.4958,
+      "step": 3978
+    },
+    {
+      "epoch": 3.8190110417666827,
+      "grad_norm": 1.1106919237603665,
+      "learning_rate": 1.608795452166525e-06,
+      "loss": 0.5155,
+      "step": 3979
+    },
+    {
+      "epoch": 3.8199711953912625,
+      "grad_norm": 0.9984189443064169,
+      "learning_rate": 1.6063345300219162e-06,
+      "loss": 0.4263,
+      "step": 3980
+    },
+    {
+      "epoch": 3.8209313490158427,
+      "grad_norm": 1.0060149288925493,
+      "learning_rate": 1.6038751312574592e-06,
+      "loss": 0.5008,
+      "step": 3981
+    },
+    {
+      "epoch": 3.8218915026404225,
+      "grad_norm": 0.9101782682192767,
+      "learning_rate": 1.6014172569771564e-06,
+      "loss": 0.4947,
+      "step": 3982
+    },
+    {
+      "epoch": 3.8228516562650023,
+      "grad_norm": 1.1453386082332564,
+      "learning_rate": 1.598960908284316e-06,
+      "loss": 0.4514,
+      "step": 3983
+    },
+    {
+      "epoch": 3.823811809889582,
+      "grad_norm": 1.5280230530018748,
+      "learning_rate": 1.5965060862815713e-06,
+      "loss": 0.5467,
+      "step": 3984
+    },
+    {
+      "epoch": 3.8247719635141624,
+      "grad_norm": 1.5398681237936402,
+      "learning_rate": 1.5940527920708594e-06,
+      "loss": 0.4875,
+      "step": 3985
+    },
+    {
+      "epoch": 3.825732117138742,
+      "grad_norm": 1.3715231941699455,
+      "learning_rate": 1.5916010267534433e-06,
+      "loss": 0.5491,
+      "step": 3986
+    },
+    {
+      "epoch": 3.826692270763322,
+      "grad_norm": 1.132827396312227,
+      "learning_rate": 1.5891507914298936e-06,
+      "loss": 0.3903,
+      "step": 3987
+    },
+    {
+      "epoch": 3.827652424387902,
+      "grad_norm": 1.11042335309336,
+      "learning_rate": 1.5867020872000927e-06,
+      "loss": 0.4469,
+      "step": 3988
+    },
+    {
+      "epoch": 3.828612578012482,
+      "grad_norm": 1.2977517517389103,
+      "learning_rate": 1.5842549151632386e-06,
+      "loss": 0.4846,
+      "step": 3989
+    },
+    {
+      "epoch": 3.829572731637062,
+      "grad_norm": 1.195303421024155,
+      "learning_rate": 1.5818092764178405e-06,
+      "loss": 0.48,
+      "step": 3990
+    },
+    {
+      "epoch": 3.8305328852616416,
+      "grad_norm": 1.0073063884953934,
+      "learning_rate": 1.5793651720617203e-06,
+      "loss": 0.5019,
+      "step": 3991
+    },
+    {
+      "epoch": 3.831493038886222,
+      "grad_norm": 0.9365326092207242,
+      "learning_rate": 1.576922603192011e-06,
+      "loss": 0.5142,
+      "step": 3992
+    },
+    {
+      "epoch": 3.8324531925108016,
+      "grad_norm": 0.8995860933097729,
+      "learning_rate": 1.5744815709051532e-06,
+      "loss": 0.5226,
+      "step": 3993
+    },
+    {
+      "epoch": 3.833413346135382,
+      "grad_norm": 1.3349663616771277,
+      "learning_rate": 1.5720420762969012e-06,
+      "loss": 0.4431,
+      "step": 3994
+    },
+    {
+      "epoch": 3.8343734997599617,
+      "grad_norm": 1.5557802515426742,
+      "learning_rate": 1.5696041204623185e-06,
+      "loss": 0.483,
+      "step": 3995
+    },
+    {
+      "epoch": 3.8353336533845415,
+      "grad_norm": 1.0883915020823676,
+      "learning_rate": 1.5671677044957772e-06,
+      "loss": 0.4781,
+      "step": 3996
+    },
+    {
+      "epoch": 3.8362938070091213,
+      "grad_norm": 1.2148251581833054,
+      "learning_rate": 1.5647328294909598e-06,
+      "loss": 0.4284,
+      "step": 3997
+    },
+    {
+      "epoch": 3.8372539606337015,
+      "grad_norm": 1.245980954092881,
+      "learning_rate": 1.5622994965408527e-06,
+      "loss": 0.5778,
+      "step": 3998
+    },
+    {
+      "epoch": 3.8382141142582813,
+      "grad_norm": 1.2141459957545226,
+      "learning_rate": 1.5598677067377533e-06,
+      "loss": 0.4211,
+      "step": 3999
+    },
+    {
+      "epoch": 3.839174267882861,
+      "grad_norm": 1.2168386206877675,
+      "learning_rate": 1.55743746117327e-06,
+      "loss": 0.5021,
+      "step": 4000
+    },
+    {
+      "epoch": 3.8401344215074413,
+      "grad_norm": 0.9708078696586657,
+      "learning_rate": 1.5550087609383102e-06,
+      "loss": 0.5214,
+      "step": 4001
+    },
+    {
+      "epoch": 3.841094575132021,
+      "grad_norm": 1.0115662460484836,
+      "learning_rate": 1.5525816071230932e-06,
+      "loss": 0.4427,
+      "step": 4002
+    },
+    {
+      "epoch": 3.842054728756601,
+      "grad_norm": 1.1549305352553938,
+      "learning_rate": 1.5501560008171412e-06,
+      "loss": 0.5273,
+      "step": 4003
+    },
+    {
+      "epoch": 3.8430148823811807,
+      "grad_norm": 0.9841200403606181,
+      "learning_rate": 1.5477319431092835e-06,
+      "loss": 0.4897,
+      "step": 4004
+    },
+    {
+      "epoch": 3.843975036005761,
+      "grad_norm": 1.3534935188894743,
+      "learning_rate": 1.5453094350876563e-06,
+      "loss": 0.4788,
+      "step": 4005
+    },
+    {
+      "epoch": 3.8449351896303408,
+      "grad_norm": 0.9454490671210521,
+      "learning_rate": 1.5428884778396929e-06,
+      "loss": 0.4947,
+      "step": 4006
+    },
+    {
+      "epoch": 3.845895343254921,
+      "grad_norm": 0.982085650772257,
+      "learning_rate": 1.540469072452137e-06,
+      "loss": 0.5073,
+      "step": 4007
+    },
+    {
+      "epoch": 3.846855496879501,
+      "grad_norm": 1.1183833419766704,
+      "learning_rate": 1.5380512200110347e-06,
+      "loss": 0.4687,
+      "step": 4008
+    },
+    {
+      "epoch": 3.8478156505040806,
+      "grad_norm": 1.2972102500875846,
+      "learning_rate": 1.5356349216017336e-06,
+      "loss": 0.4551,
+      "step": 4009
+    },
+    {
+      "epoch": 3.8487758041286604,
+      "grad_norm": 1.7060004299228904,
+      "learning_rate": 1.5332201783088857e-06,
+      "loss": 0.4385,
+      "step": 4010
+    },
+    {
+      "epoch": 3.8497359577532406,
+      "grad_norm": 1.0363932977147206,
+      "learning_rate": 1.5308069912164403e-06,
+      "loss": 0.4592,
+      "step": 4011
+    },
+    {
+      "epoch": 3.8506961113778204,
+      "grad_norm": 0.9486440871438606,
+      "learning_rate": 1.5283953614076514e-06,
+      "loss": 0.5234,
+      "step": 4012
+    },
+    {
+      "epoch": 3.8516562650024007,
+      "grad_norm": 1.1287504733298028,
+      "learning_rate": 1.5259852899650784e-06,
+      "loss": 0.443,
+      "step": 4013
+    },
+    {
+      "epoch": 3.8526164186269805,
+      "grad_norm": 1.7121590389272714,
+      "learning_rate": 1.5235767779705717e-06,
+      "loss": 0.4277,
+      "step": 4014
+    },
+    {
+      "epoch": 3.8535765722515603,
+      "grad_norm": 1.1436615771183936,
+      "learning_rate": 1.5211698265052905e-06,
+      "loss": 0.5004,
+      "step": 4015
+    },
+    {
+      "epoch": 3.85453672587614,
+      "grad_norm": 0.9259865424761593,
+      "learning_rate": 1.518764436649684e-06,
+      "loss": 0.4672,
+      "step": 4016
+    },
+    {
+      "epoch": 3.85549687950072,
+      "grad_norm": 0.8042351206260997,
+      "learning_rate": 1.5163606094835115e-06,
+      "loss": 0.4571,
+      "step": 4017
+    },
+    {
+      "epoch": 3.8564570331253,
+      "grad_norm": 1.0242287750735588,
+      "learning_rate": 1.5139583460858249e-06,
+      "loss": 0.5616,
+      "step": 4018
+    },
+    {
+      "epoch": 3.85741718674988,
+      "grad_norm": 1.3296288625498407,
+      "learning_rate": 1.5115576475349715e-06,
+      "loss": 0.4741,
+      "step": 4019
+    },
+    {
+      "epoch": 3.85837734037446,
+      "grad_norm": 1.383565930716489,
+      "learning_rate": 1.5091585149086008e-06,
+      "loss": 0.4244,
+      "step": 4020
+    },
+    {
+      "epoch": 3.85933749399904,
+      "grad_norm": 1.118198473425921,
+      "learning_rate": 1.5067609492836572e-06,
+      "loss": 0.5082,
+      "step": 4021
+    },
+    {
+      "epoch": 3.8602976476236197,
+      "grad_norm": 1.3476105788891708,
+      "learning_rate": 1.5043649517363835e-06,
+      "loss": 0.4172,
+      "step": 4022
+    },
+    {
+      "epoch": 3.8612578012481995,
+      "grad_norm": 1.862910169367962,
+      "learning_rate": 1.5019705233423177e-06,
+      "loss": 0.5179,
+      "step": 4023
+    },
+    {
+      "epoch": 3.8622179548727797,
+      "grad_norm": 0.9927431150646678,
+      "learning_rate": 1.4995776651762906e-06,
+      "loss": 0.521,
+      "step": 4024
+    },
+    {
+      "epoch": 3.8631781084973595,
+      "grad_norm": 0.8802457449186407,
+      "learning_rate": 1.4971863783124308e-06,
+      "loss": 0.5404,
+      "step": 4025
+    },
+    {
+      "epoch": 3.86413826212194,
+      "grad_norm": 1.0088620252724212,
+      "learning_rate": 1.4947966638241662e-06,
+      "loss": 0.4715,
+      "step": 4026
+    },
+    {
+      "epoch": 3.8650984157465196,
+      "grad_norm": 1.1755045780960445,
+      "learning_rate": 1.4924085227842084e-06,
+      "loss": 0.4483,
+      "step": 4027
+    },
+    {
+      "epoch": 3.8660585693710994,
+      "grad_norm": 1.0702719556324969,
+      "learning_rate": 1.490021956264573e-06,
+      "loss": 0.4258,
+      "step": 4028
+    },
+    {
+      "epoch": 3.867018722995679,
+      "grad_norm": 0.8775672589847451,
+      "learning_rate": 1.487636965336558e-06,
+      "loss": 0.3576,
+      "step": 4029
+    },
+    {
+      "epoch": 3.867978876620259,
+      "grad_norm": 1.0924174648049145,
+      "learning_rate": 1.4852535510707662e-06,
+      "loss": 0.4449,
+      "step": 4030
+    },
+    {
+      "epoch": 3.868939030244839,
+      "grad_norm": 1.0990688153958843,
+      "learning_rate": 1.4828717145370863e-06,
+      "loss": 0.4891,
+      "step": 4031
+    },
+    {
+      "epoch": 3.869899183869419,
+      "grad_norm": 1.121158881487208,
+      "learning_rate": 1.4804914568046957e-06,
+      "loss": 0.4821,
+      "step": 4032
+    },
+    {
+      "epoch": 3.8708593374939992,
+      "grad_norm": 1.0268608560470882,
+      "learning_rate": 1.478112778942069e-06,
+      "loss": 0.4799,
+      "step": 4033
+    },
+    {
+      "epoch": 3.871819491118579,
+      "grad_norm": 1.4615768454751372,
+      "learning_rate": 1.4757356820169694e-06,
+      "loss": 0.4493,
+      "step": 4034
+    },
+    {
+      "epoch": 3.872779644743159,
+      "grad_norm": 1.7542196610056557,
+      "learning_rate": 1.4733601670964493e-06,
+      "loss": 0.4186,
+      "step": 4035
+    },
+    {
+      "epoch": 3.8737397983677386,
+      "grad_norm": 1.0868446926637694,
+      "learning_rate": 1.4709862352468541e-06,
+      "loss": 0.506,
+      "step": 4036
+    },
+    {
+      "epoch": 3.874699951992319,
+      "grad_norm": 1.1091506558753939,
+      "learning_rate": 1.4686138875338136e-06,
+      "loss": 0.475,
+      "step": 4037
+    },
+    {
+      "epoch": 3.8756601056168987,
+      "grad_norm": 1.4347493014909358,
+      "learning_rate": 1.4662431250222504e-06,
+      "loss": 0.4325,
+      "step": 4038
+    },
+    {
+      "epoch": 3.876620259241479,
+      "grad_norm": 0.8404888487332063,
+      "learning_rate": 1.4638739487763743e-06,
+      "loss": 0.3987,
+      "step": 4039
+    },
+    {
+      "epoch": 3.8775804128660587,
+      "grad_norm": 1.183921069828481,
+      "learning_rate": 1.4615063598596834e-06,
+      "loss": 0.5731,
+      "step": 4040
+    },
+    {
+      "epoch": 3.8785405664906385,
+      "grad_norm": 1.0639979381050415,
+      "learning_rate": 1.4591403593349652e-06,
+      "loss": 0.5338,
+      "step": 4041
+    },
+    {
+      "epoch": 3.8795007201152183,
+      "grad_norm": 1.1468457294566843,
+      "learning_rate": 1.4567759482642884e-06,
+      "loss": 0.4024,
+      "step": 4042
+    },
+    {
+      "epoch": 3.8804608737397985,
+      "grad_norm": 1.146199396866093,
+      "learning_rate": 1.4544131277090118e-06,
+      "loss": 0.4707,
+      "step": 4043
+    },
+    {
+      "epoch": 3.8814210273643783,
+      "grad_norm": 1.2054451240412933,
+      "learning_rate": 1.4520518987297854e-06,
+      "loss": 0.4779,
+      "step": 4044
+    },
+    {
+      "epoch": 3.882381180988958,
+      "grad_norm": 1.2773333772625617,
+      "learning_rate": 1.449692262386534e-06,
+      "loss": 0.4464,
+      "step": 4045
+    },
+    {
+      "epoch": 3.8833413346135384,
+      "grad_norm": 0.9227110359686093,
+      "learning_rate": 1.447334219738477e-06,
+      "loss": 0.5324,
+      "step": 4046
+    },
+    {
+      "epoch": 3.884301488238118,
+      "grad_norm": 1.3731221175775805,
+      "learning_rate": 1.4449777718441106e-06,
+      "loss": 0.5106,
+      "step": 4047
+    },
+    {
+      "epoch": 3.885261641862698,
+      "grad_norm": 1.2453865008405958,
+      "learning_rate": 1.4426229197612217e-06,
+      "loss": 0.5369,
+      "step": 4048
+    },
+    {
+      "epoch": 3.8862217954872778,
+      "grad_norm": 1.1758688469859149,
+      "learning_rate": 1.44026966454688e-06,
+      "loss": 0.5168,
+      "step": 4049
+    },
+    {
+      "epoch": 3.887181949111858,
+      "grad_norm": 1.1989467853498457,
+      "learning_rate": 1.4379180072574335e-06,
+      "loss": 0.5195,
+      "step": 4050
+    },
+    {
+      "epoch": 3.888142102736438,
+      "grad_norm": 1.1939247580457146,
+      "learning_rate": 1.4355679489485158e-06,
+      "loss": 0.4305,
+      "step": 4051
+    },
+    {
+      "epoch": 3.889102256361018,
+      "grad_norm": 0.6023502428972107,
+      "learning_rate": 1.433219490675045e-06,
+      "loss": 0.4727,
+      "step": 4052
+    },
+    {
+      "epoch": 3.890062409985598,
+      "grad_norm": 0.8724516734586564,
+      "learning_rate": 1.4308726334912176e-06,
+      "loss": 0.4893,
+      "step": 4053
+    },
+    {
+      "epoch": 3.8910225636101776,
+      "grad_norm": 1.2127240783439697,
+      "learning_rate": 1.4285273784505154e-06,
+      "loss": 0.4933,
+      "step": 4054
+    },
+    {
+      "epoch": 3.8919827172347574,
+      "grad_norm": 0.9695097636780937,
+      "learning_rate": 1.4261837266056938e-06,
+      "loss": 0.5214,
+      "step": 4055
+    },
+    {
+      "epoch": 3.8929428708593377,
+      "grad_norm": 1.160597549486381,
+      "learning_rate": 1.4238416790087945e-06,
+      "loss": 0.4482,
+      "step": 4056
+    },
+    {
+      "epoch": 3.8939030244839175,
+      "grad_norm": 0.8305336831110708,
+      "learning_rate": 1.4215012367111408e-06,
+      "loss": 0.4585,
+      "step": 4057
+    },
+    {
+      "epoch": 3.8948631781084972,
+      "grad_norm": 0.9272193865571398,
+      "learning_rate": 1.4191624007633286e-06,
+      "loss": 0.5662,
+      "step": 4058
+    },
+    {
+      "epoch": 3.8958233317330775,
+      "grad_norm": 1.0105506991936097,
+      "learning_rate": 1.41682517221524e-06,
+      "loss": 0.4465,
+      "step": 4059
+    },
+    {
+      "epoch": 3.8967834853576573,
+      "grad_norm": 1.1225112503216266,
+      "learning_rate": 1.4144895521160257e-06,
+      "loss": 0.5328,
+      "step": 4060
+    },
+    {
+      "epoch": 3.897743638982237,
+      "grad_norm": 1.279152969448898,
+      "learning_rate": 1.412155541514126e-06,
+      "loss": 0.5269,
+      "step": 4061
+    },
+    {
+      "epoch": 3.898703792606817,
+      "grad_norm": 1.8388135926587599,
+      "learning_rate": 1.4098231414572534e-06,
+      "loss": 0.4074,
+      "step": 4062
+    },
+    {
+      "epoch": 3.899663946231397,
+      "grad_norm": 1.6045605805300485,
+      "learning_rate": 1.4074923529923945e-06,
+      "loss": 0.566,
+      "step": 4063
+    },
+    {
+      "epoch": 3.900624099855977,
+      "grad_norm": 0.8785585405375341,
+      "learning_rate": 1.4051631771658159e-06,
+      "loss": 0.4144,
+      "step": 4064
+    },
+    {
+      "epoch": 3.901584253480557,
+      "grad_norm": 1.032110270713385,
+      "learning_rate": 1.402835615023061e-06,
+      "loss": 0.4382,
+      "step": 4065
+    },
+    {
+      "epoch": 3.902544407105137,
+      "grad_norm": 1.395893515938224,
+      "learning_rate": 1.400509667608947e-06,
+      "loss": 0.397,
+      "step": 4066
+    },
+    {
+      "epoch": 3.9035045607297167,
+      "grad_norm": 1.320399818859656,
+      "learning_rate": 1.398185335967568e-06,
+      "loss": 0.4619,
+      "step": 4067
+    },
+    {
+      "epoch": 3.9044647143542965,
+      "grad_norm": 1.1526727439294775,
+      "learning_rate": 1.3958626211422887e-06,
+      "loss": 0.3535,
+      "step": 4068
+    },
+    {
+      "epoch": 3.905424867978877,
+      "grad_norm": 1.4135671362045452,
+      "learning_rate": 1.3935415241757517e-06,
+      "loss": 0.5228,
+      "step": 4069
+    },
+    {
+      "epoch": 3.9063850216034566,
+      "grad_norm": 0.8189534748681332,
+      "learning_rate": 1.3912220461098763e-06,
+      "loss": 0.4531,
+      "step": 4070
+    },
+    {
+      "epoch": 3.9073451752280364,
+      "grad_norm": 1.1263794949554329,
+      "learning_rate": 1.3889041879858472e-06,
+      "loss": 0.4646,
+      "step": 4071
+    },
+    {
+      "epoch": 3.9083053288526166,
+      "grad_norm": 1.5344478431354853,
+      "learning_rate": 1.3865879508441294e-06,
+      "loss": 0.4896,
+      "step": 4072
+    },
+    {
+      "epoch": 3.9092654824771964,
+      "grad_norm": 1.392489823734356,
+      "learning_rate": 1.384273335724452e-06,
+      "loss": 0.5474,
+      "step": 4073
+    },
+    {
+      "epoch": 3.910225636101776,
+      "grad_norm": 1.170714564580904,
+      "learning_rate": 1.3819603436658258e-06,
+      "loss": 0.4303,
+      "step": 4074
+    },
+    {
+      "epoch": 3.911185789726356,
+      "grad_norm": 1.1368734226893547,
+      "learning_rate": 1.3796489757065284e-06,
+      "loss": 0.5351,
+      "step": 4075
+    },
+    {
+      "epoch": 3.9121459433509362,
+      "grad_norm": 1.0116935997285494,
+      "learning_rate": 1.3773392328841046e-06,
+      "loss": 0.5334,
+      "step": 4076
+    },
+    {
+      "epoch": 3.913106096975516,
+      "grad_norm": 1.2343182848929284,
+      "learning_rate": 1.3750311162353752e-06,
+      "loss": 0.4632,
+      "step": 4077
+    },
+    {
+      "epoch": 3.9140662506000963,
+      "grad_norm": 1.0308702206135525,
+      "learning_rate": 1.3727246267964288e-06,
+      "loss": 0.4387,
+      "step": 4078
+    },
+    {
+      "epoch": 3.915026404224676,
+      "grad_norm": 1.0340175430057703,
+      "learning_rate": 1.3704197656026242e-06,
+      "loss": 0.3849,
+      "step": 4079
+    },
+    {
+      "epoch": 3.915986557849256,
+      "grad_norm": 1.3134001215269133,
+      "learning_rate": 1.36811653368859e-06,
+      "loss": 0.5156,
+      "step": 4080
+    },
+    {
+      "epoch": 3.9169467114738357,
+      "grad_norm": 0.7877048705715364,
+      "learning_rate": 1.3658149320882196e-06,
+      "loss": 0.5337,
+      "step": 4081
+    },
+    {
+      "epoch": 3.917906865098416,
+      "grad_norm": 1.0372241944209186,
+      "learning_rate": 1.3635149618346793e-06,
+      "loss": 0.4775,
+      "step": 4082
+    },
+    {
+      "epoch": 3.9188670187229957,
+      "grad_norm": 1.4545514153391248,
+      "learning_rate": 1.3612166239604003e-06,
+      "loss": 0.4189,
+      "step": 4083
+    },
+    {
+      "epoch": 3.9198271723475755,
+      "grad_norm": 1.6400807256631689,
+      "learning_rate": 1.3589199194970825e-06,
+      "loss": 0.38,
+      "step": 4084
+    },
+    {
+      "epoch": 3.9207873259721557,
+      "grad_norm": 1.1161324377796293,
+      "learning_rate": 1.3566248494756934e-06,
+      "loss": 0.4868,
+      "step": 4085
+    },
+    {
+      "epoch": 3.9217474795967355,
+      "grad_norm": 0.8883548552685723,
+      "learning_rate": 1.3543314149264625e-06,
+      "loss": 0.5214,
+      "step": 4086
+    },
+    {
+      "epoch": 3.9227076332213153,
+      "grad_norm": 1.2708272295752705,
+      "learning_rate": 1.352039616878888e-06,
+      "loss": 0.3808,
+      "step": 4087
+    },
+    {
+      "epoch": 3.923667786845895,
+      "grad_norm": 0.8954602689209706,
+      "learning_rate": 1.3497494563617375e-06,
+      "loss": 0.488,
+      "step": 4088
+    },
+    {
+      "epoch": 3.9246279404704754,
+      "grad_norm": 1.4156612884362025,
+      "learning_rate": 1.347460934403036e-06,
+      "loss": 0.45,
+      "step": 4089
+    },
+    {
+      "epoch": 3.925588094095055,
+      "grad_norm": 1.18479486203872,
+      "learning_rate": 1.3451740520300787e-06,
+      "loss": 0.4398,
+      "step": 4090
+    },
+    {
+      "epoch": 3.9265482477196354,
+      "grad_norm": 1.0831389027518687,
+      "learning_rate": 1.3428888102694187e-06,
+      "loss": 0.4874,
+      "step": 4091
+    },
+    {
+      "epoch": 3.927508401344215,
+      "grad_norm": 0.9954881545826475,
+      "learning_rate": 1.3406052101468808e-06,
+      "loss": 0.4879,
+      "step": 4092
+    },
+    {
+      "epoch": 3.928468554968795,
+      "grad_norm": 0.9124216707052788,
+      "learning_rate": 1.338323252687549e-06,
+      "loss": 0.4502,
+      "step": 4093
+    },
+    {
+      "epoch": 3.929428708593375,
+      "grad_norm": 1.2429916730888522,
+      "learning_rate": 1.3360429389157664e-06,
+      "loss": 0.3754,
+      "step": 4094
+    },
+    {
+      "epoch": 3.930388862217955,
+      "grad_norm": 1.7210578064205935,
+      "learning_rate": 1.3337642698551428e-06,
+      "loss": 0.3892,
+      "step": 4095
+    },
+    {
+      "epoch": 3.931349015842535,
+      "grad_norm": 1.2540445183477316,
+      "learning_rate": 1.331487246528549e-06,
+      "loss": 0.4671,
+      "step": 4096
+    },
+    {
+      "epoch": 3.9323091694671146,
+      "grad_norm": 1.1752729101137955,
+      "learning_rate": 1.3292118699581153e-06,
+      "loss": 0.4323,
+      "step": 4097
+    },
+    {
+      "epoch": 3.933269323091695,
+      "grad_norm": 0.9954437339603511,
+      "learning_rate": 1.3269381411652366e-06,
+      "loss": 0.4047,
+      "step": 4098
+    },
+    {
+      "epoch": 3.9342294767162747,
+      "grad_norm": 0.9524777868909293,
+      "learning_rate": 1.324666061170562e-06,
+      "loss": 0.6096,
+      "step": 4099
+    },
+    {
+      "epoch": 3.9351896303408544,
+      "grad_norm": 1.0750406642574184,
+      "learning_rate": 1.3223956309940038e-06,
+      "loss": 0.4719,
+      "step": 4100
+    },
+    {
+      "epoch": 3.9361497839654342,
+      "grad_norm": 1.1347251012040827,
+      "learning_rate": 1.3201268516547382e-06,
+      "loss": 0.5203,
+      "step": 4101
+    },
+    {
+      "epoch": 3.9371099375900145,
+      "grad_norm": 1.0944097306950127,
+      "learning_rate": 1.317859724171192e-06,
+      "loss": 0.4772,
+      "step": 4102
+    },
+    {
+      "epoch": 3.9380700912145943,
+      "grad_norm": 1.3944724725377775,
+      "learning_rate": 1.3155942495610574e-06,
+      "loss": 0.4865,
+      "step": 4103
+    },
+    {
+      "epoch": 3.9390302448391745,
+      "grad_norm": 1.393872376217508,
+      "learning_rate": 1.3133304288412768e-06,
+      "loss": 0.4211,
+      "step": 4104
+    },
+    {
+      "epoch": 3.9399903984637543,
+      "grad_norm": 1.3494932929508532,
+      "learning_rate": 1.3110682630280603e-06,
+      "loss": 0.5309,
+      "step": 4105
+    },
+    {
+      "epoch": 3.940950552088334,
+      "grad_norm": 1.17868788523819,
+      "learning_rate": 1.3088077531368686e-06,
+      "loss": 0.4527,
+      "step": 4106
+    },
+    {
+      "epoch": 3.941910705712914,
+      "grad_norm": 1.2245565349535994,
+      "learning_rate": 1.3065489001824194e-06,
+      "loss": 0.5102,
+      "step": 4107
+    },
+    {
+      "epoch": 3.942870859337494,
+      "grad_norm": 0.9162536687842219,
+      "learning_rate": 1.3042917051786873e-06,
+      "loss": 0.5234,
+      "step": 4108
+    },
+    {
+      "epoch": 3.943831012962074,
+      "grad_norm": 1.2520030427754436,
+      "learning_rate": 1.3020361691389043e-06,
+      "loss": 0.3865,
+      "step": 4109
+    },
+    {
+      "epoch": 3.9447911665866537,
+      "grad_norm": 1.0770822645392424,
+      "learning_rate": 1.299782293075556e-06,
+      "loss": 0.4241,
+      "step": 4110
+    },
+    {
+      "epoch": 3.945751320211234,
+      "grad_norm": 0.9285368301633085,
+      "learning_rate": 1.297530078000384e-06,
+      "loss": 0.3585,
+      "step": 4111
+    },
+    {
+      "epoch": 3.9467114738358138,
+      "grad_norm": 1.0344233599670831,
+      "learning_rate": 1.295279524924382e-06,
+      "loss": 0.4348,
+      "step": 4112
+    },
+    {
+      "epoch": 3.9476716274603936,
+      "grad_norm": 0.9897457390323293,
+      "learning_rate": 1.2930306348577982e-06,
+      "loss": 0.4411,
+      "step": 4113
+    },
+    {
+      "epoch": 3.9486317810849734,
+      "grad_norm": 1.008796582147279,
+      "learning_rate": 1.2907834088101402e-06,
+      "loss": 0.468,
+      "step": 4114
+    },
+    {
+      "epoch": 3.9495919347095536,
+      "grad_norm": 1.0919656417194719,
+      "learning_rate": 1.2885378477901595e-06,
+      "loss": 0.4164,
+      "step": 4115
+    },
+    {
+      "epoch": 3.9505520883341334,
+      "grad_norm": 1.059155080492062,
+      "learning_rate": 1.2862939528058665e-06,
+      "loss": 0.4024,
+      "step": 4116
+    },
+    {
+      "epoch": 3.9515122419587136,
+      "grad_norm": 0.9145561364466925,
+      "learning_rate": 1.284051724864518e-06,
+      "loss": 0.4815,
+      "step": 4117
+    },
+    {
+      "epoch": 3.9524723955832934,
+      "grad_norm": 1.2025386746578184,
+      "learning_rate": 1.2818111649726306e-06,
+      "loss": 0.3387,
+      "step": 4118
+    },
+    {
+      "epoch": 3.9534325492078732,
+      "grad_norm": 1.3173067493707236,
+      "learning_rate": 1.2795722741359672e-06,
+      "loss": 0.3966,
+      "step": 4119
+    },
+    {
+      "epoch": 3.954392702832453,
+      "grad_norm": 0.820376797118698,
+      "learning_rate": 1.277335053359539e-06,
+      "loss": 0.5225,
+      "step": 4120
+    },
+    {
+      "epoch": 3.9553528564570333,
+      "grad_norm": 1.194067571446378,
+      "learning_rate": 1.2750995036476133e-06,
+      "loss": 0.4774,
+      "step": 4121
+    },
+    {
+      "epoch": 3.956313010081613,
+      "grad_norm": 0.8784531337871889,
+      "learning_rate": 1.2728656260037025e-06,
+      "loss": 0.5147,
+      "step": 4122
+    },
+    {
+      "epoch": 3.957273163706193,
+      "grad_norm": 1.1533340960198126,
+      "learning_rate": 1.2706334214305716e-06,
+      "loss": 0.506,
+      "step": 4123
+    },
+    {
+      "epoch": 3.958233317330773,
+      "grad_norm": 1.6351270381674599,
+      "learning_rate": 1.2684028909302348e-06,
+      "loss": 0.351,
+      "step": 4124
+    },
+    {
+      "epoch": 3.959193470955353,
+      "grad_norm": 0.731829491187216,
+      "learning_rate": 1.2661740355039499e-06,
+      "loss": 0.4465,
+      "step": 4125
+    },
+    {
+      "epoch": 3.9601536245799327,
+      "grad_norm": 1.0640778289239052,
+      "learning_rate": 1.2639468561522284e-06,
+      "loss": 0.474,
+      "step": 4126
+    },
+    {
+      "epoch": 3.9611137782045125,
+      "grad_norm": 2.016771247252845,
+      "learning_rate": 1.2617213538748268e-06,
+      "loss": 0.5078,
+      "step": 4127
+    },
+    {
+      "epoch": 3.9620739318290927,
+      "grad_norm": 1.1841314710691146,
+      "learning_rate": 1.2594975296707495e-06,
+      "loss": 0.5035,
+      "step": 4128
+    },
+    {
+      "epoch": 3.9630340854536725,
+      "grad_norm": 1.1036671376508644,
+      "learning_rate": 1.2572753845382484e-06,
+      "loss": 0.4209,
+      "step": 4129
+    },
+    {
+      "epoch": 3.9639942390782528,
+      "grad_norm": 1.0417370270306738,
+      "learning_rate": 1.255054919474819e-06,
+      "loss": 0.5141,
+      "step": 4130
+    },
+    {
+      "epoch": 3.9649543927028326,
+      "grad_norm": 1.1148626570006415,
+      "learning_rate": 1.252836135477203e-06,
+      "loss": 0.534,
+      "step": 4131
+    },
+    {
+      "epoch": 3.9659145463274124,
+      "grad_norm": 0.9617271472038451,
+      "learning_rate": 1.2506190335413943e-06,
+      "loss": 0.4632,
+      "step": 4132
+    },
+    {
+      "epoch": 3.966874699951992,
+      "grad_norm": 0.7577777482038071,
+      "learning_rate": 1.2484036146626205e-06,
+      "loss": 0.5134,
+      "step": 4133
+    },
+    {
+      "epoch": 3.9678348535765724,
+      "grad_norm": 1.1175602462362233,
+      "learning_rate": 1.2461898798353639e-06,
+      "loss": 0.4669,
+      "step": 4134
+    },
+    {
+      "epoch": 3.968795007201152,
+      "grad_norm": 0.9213626342830816,
+      "learning_rate": 1.2439778300533412e-06,
+      "loss": 0.4953,
+      "step": 4135
+    },
+    {
+      "epoch": 3.969755160825732,
+      "grad_norm": 1.1904391217338572,
+      "learning_rate": 1.241767466309522e-06,
+      "loss": 0.3834,
+      "step": 4136
+    },
+    {
+      "epoch": 3.9707153144503122,
+      "grad_norm": 1.0471316329462108,
+      "learning_rate": 1.2395587895961158e-06,
+      "loss": 0.4629,
+      "step": 4137
+    },
+    {
+      "epoch": 3.971675468074892,
+      "grad_norm": 0.9616781362666617,
+      "learning_rate": 1.2373518009045705e-06,
+      "loss": 0.5947,
+      "step": 4138
+    },
+    {
+      "epoch": 3.972635621699472,
+      "grad_norm": 1.4420661170373539,
+      "learning_rate": 1.2351465012255808e-06,
+      "loss": 0.4961,
+      "step": 4139
+    },
+    {
+      "epoch": 3.9735957753240516,
+      "grad_norm": 1.1630442478403853,
+      "learning_rate": 1.232942891549083e-06,
+      "loss": 0.4432,
+      "step": 4140
+    },
+    {
+      "epoch": 3.974555928948632,
+      "grad_norm": 1.15212310608021,
+      "learning_rate": 1.2307409728642539e-06,
+      "loss": 0.5117,
+      "step": 4141
+    },
+    {
+      "epoch": 3.9755160825732117,
+      "grad_norm": 1.265953078329887,
+      "learning_rate": 1.2285407461595122e-06,
+      "loss": 0.5225,
+      "step": 4142
+    },
+    {
+      "epoch": 3.976476236197792,
+      "grad_norm": 1.2350500785784575,
+      "learning_rate": 1.2263422124225132e-06,
+      "loss": 0.4487,
+      "step": 4143
+    },
+    {
+      "epoch": 3.9774363898223717,
+      "grad_norm": 1.1117089141723477,
+      "learning_rate": 1.2241453726401559e-06,
+      "loss": 0.4174,
+      "step": 4144
+    },
+    {
+      "epoch": 3.9783965434469515,
+      "grad_norm": 0.8972382030907151,
+      "learning_rate": 1.2219502277985817e-06,
+      "loss": 0.4392,
+      "step": 4145
+    },
+    {
+      "epoch": 3.9793566970715313,
+      "grad_norm": 1.1884767201653634,
+      "learning_rate": 1.2197567788831638e-06,
+      "loss": 0.482,
+      "step": 4146
+    },
+    {
+      "epoch": 3.9803168506961115,
+      "grad_norm": 1.0956299108243066,
+      "learning_rate": 1.2175650268785205e-06,
+      "loss": 0.4247,
+      "step": 4147
+    },
+    {
+      "epoch": 3.9812770043206913,
+      "grad_norm": 0.9047473318465206,
+      "learning_rate": 1.215374972768502e-06,
+      "loss": 0.3904,
+      "step": 4148
+    },
+    {
+      "epoch": 3.982237157945271,
+      "grad_norm": 1.4375665503159347,
+      "learning_rate": 1.213186617536204e-06,
+      "loss": 0.5551,
+      "step": 4149
+    },
+    {
+      "epoch": 3.9831973115698514,
+      "grad_norm": 1.050766956265714,
+      "learning_rate": 1.2109999621639561e-06,
+      "loss": 0.4392,
+      "step": 4150
+    },
+    {
+      "epoch": 3.984157465194431,
+      "grad_norm": 1.1837777696740561,
+      "learning_rate": 1.2088150076333222e-06,
+      "loss": 0.5144,
+      "step": 4151
+    },
+    {
+      "epoch": 3.985117618819011,
+      "grad_norm": 1.4306616871611946,
+      "learning_rate": 1.2066317549251055e-06,
+      "loss": 0.4162,
+      "step": 4152
+    },
+    {
+      "epoch": 3.9860777724435907,
+      "grad_norm": 1.7089999247541572,
+      "learning_rate": 1.2044502050193451e-06,
+      "loss": 0.4145,
+      "step": 4153
+    },
+    {
+      "epoch": 3.987037926068171,
+      "grad_norm": 1.2206210523280627,
+      "learning_rate": 1.2022703588953155e-06,
+      "loss": 0.4053,
+      "step": 4154
+    },
+    {
+      "epoch": 3.9879980796927508,
+      "grad_norm": 1.0254775921148231,
+      "learning_rate": 1.2000922175315278e-06,
+      "loss": 0.5259,
+      "step": 4155
+    },
+    {
+      "epoch": 3.988958233317331,
+      "grad_norm": 1.1665630930484832,
+      "learning_rate": 1.197915781905723e-06,
+      "loss": 0.4422,
+      "step": 4156
+    },
+    {
+      "epoch": 3.989918386941911,
+      "grad_norm": 1.1001199001856836,
+      "learning_rate": 1.1957410529948803e-06,
+      "loss": 0.5368,
+      "step": 4157
+    },
+    {
+      "epoch": 3.9908785405664906,
+      "grad_norm": 1.075018529315739,
+      "learning_rate": 1.1935680317752173e-06,
+      "loss": 0.4422,
+      "step": 4158
+    },
+    {
+      "epoch": 3.9918386941910704,
+      "grad_norm": 0.7857597868944044,
+      "learning_rate": 1.1913967192221742e-06,
+      "loss": 0.529,
+      "step": 4159
+    },
+    {
+      "epoch": 3.9927988478156506,
+      "grad_norm": 0.9411861184100021,
+      "learning_rate": 1.189227116310434e-06,
+      "loss": 0.5995,
+      "step": 4160
+    },
+    {
+      "epoch": 3.9937590014402304,
+      "grad_norm": 1.2649531390346653,
+      "learning_rate": 1.1870592240139033e-06,
+      "loss": 0.4276,
+      "step": 4161
+    },
+    {
+      "epoch": 3.9947191550648102,
+      "grad_norm": 1.0633526496414132,
+      "learning_rate": 1.18489304330573e-06,
+      "loss": 0.429,
+      "step": 4162
+    },
+    {
+      "epoch": 3.9956793086893905,
+      "grad_norm": 1.0737598014237413,
+      "learning_rate": 1.18272857515829e-06,
+      "loss": 0.3883,
+      "step": 4163
+    },
+    {
+      "epoch": 3.9966394623139703,
+      "grad_norm": 1.4406649557413433,
+      "learning_rate": 1.1805658205431868e-06,
+      "loss": 0.4952,
+      "step": 4164
+    },
+    {
+      "epoch": 3.99759961593855,
+      "grad_norm": 0.9857827185154254,
+      "learning_rate": 1.1784047804312598e-06,
+      "loss": 0.4536,
+      "step": 4165
+    },
+    {
+      "epoch": 3.99855976956313,
+      "grad_norm": 0.9085969015255203,
+      "learning_rate": 1.1762454557925756e-06,
+      "loss": 0.492,
+      "step": 4166
+    },
+    {
+      "epoch": 3.99951992318771,
+      "grad_norm": 1.2522481515344492,
+      "learning_rate": 1.174087847596433e-06,
+      "loss": 0.3796,
+      "step": 4167
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.3617478918866177,
+      "learning_rate": 1.1719319568113602e-06,
+      "loss": 0.4645,
+      "step": 4168
+    },
+    {
+      "epoch": 4.00096015362458,
+      "grad_norm": 1.0228089331127097,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.4895,
+      "step": 4169
+    },
+    {
+      "epoch": 4.00192030724916,
+      "grad_norm": 1.4808354983085061,
+      "learning_rate": 1.1676253313446706e-06,
+      "loss": 0.4577,
+      "step": 4170
+    },
+    {
+      "epoch": 4.002880460873739,
+      "grad_norm": 1.1169265009334828,
+      "learning_rate": 1.165474598596254e-06,
+      "loss": 0.3596,
+      "step": 4171
+    },
+    {
+      "epoch": 4.00384061449832,
+      "grad_norm": 0.9610152744839893,
+      "learning_rate": 1.1633255871253013e-06,
+      "loss": 0.4821,
+      "step": 4172
+    },
+    {
+      "epoch": 4.0048007681229,
+      "grad_norm": 1.100819168823591,
+      "learning_rate": 1.161178297896482e-06,
+      "loss": 0.3886,
+      "step": 4173
+    },
+    {
+      "epoch": 4.00576092174748,
+      "grad_norm": 1.560344140574424,
+      "learning_rate": 1.1590327318736887e-06,
+      "loss": 0.3792,
+      "step": 4174
+    },
+    {
+      "epoch": 4.0067210753720595,
+      "grad_norm": 1.1638371153671265,
+      "learning_rate": 1.1568888900200432e-06,
+      "loss": 0.3656,
+      "step": 4175
+    },
+    {
+      "epoch": 4.007681228996639,
+      "grad_norm": 0.9152074067317886,
+      "learning_rate": 1.1547467732978967e-06,
+      "loss": 0.5235,
+      "step": 4176
+    },
+    {
+      "epoch": 4.008641382621219,
+      "grad_norm": 1.0009183461893372,
+      "learning_rate": 1.152606382668819e-06,
+      "loss": 0.5299,
+      "step": 4177
+    },
+    {
+      "epoch": 4.0096015362458,
+      "grad_norm": 1.3534478469832056,
+      "learning_rate": 1.1504677190936115e-06,
+      "loss": 0.4403,
+      "step": 4178
+    },
+    {
+      "epoch": 4.0105616898703795,
+      "grad_norm": 1.1330371387665013,
+      "learning_rate": 1.148330783532292e-06,
+      "loss": 0.4855,
+      "step": 4179
+    },
+    {
+      "epoch": 4.011521843494959,
+      "grad_norm": 1.2623186318649344,
+      "learning_rate": 1.1461955769441141e-06,
+      "loss": 0.4975,
+      "step": 4180
+    },
+    {
+      "epoch": 4.012481997119539,
+      "grad_norm": 1.2575709816030505,
+      "learning_rate": 1.144062100287548e-06,
+      "loss": 0.459,
+      "step": 4181
+    },
+    {
+      "epoch": 4.013442150744119,
+      "grad_norm": 0.9310782530766164,
+      "learning_rate": 1.1419303545202854e-06,
+      "loss": 0.3379,
+      "step": 4182
+    },
+    {
+      "epoch": 4.014402304368699,
+      "grad_norm": 1.483922328937811,
+      "learning_rate": 1.1398003405992468e-06,
+      "loss": 0.3825,
+      "step": 4183
+    },
+    {
+      "epoch": 4.0153624579932785,
+      "grad_norm": 1.2249750724630295,
+      "learning_rate": 1.1376720594805719e-06,
+      "loss": 0.428,
+      "step": 4184
+    },
+    {
+      "epoch": 4.016322611617859,
+      "grad_norm": 1.2657921422968323,
+      "learning_rate": 1.1355455121196234e-06,
+      "loss": 0.3395,
+      "step": 4185
+    },
+    {
+      "epoch": 4.017282765242439,
+      "grad_norm": 1.2012124135035076,
+      "learning_rate": 1.1334206994709874e-06,
+      "loss": 0.4152,
+      "step": 4186
+    },
+    {
+      "epoch": 4.018242918867019,
+      "grad_norm": 1.0569575118521102,
+      "learning_rate": 1.1312976224884665e-06,
+      "loss": 0.4954,
+      "step": 4187
+    },
+    {
+      "epoch": 4.019203072491599,
+      "grad_norm": 1.1067383421150563,
+      "learning_rate": 1.1291762821250867e-06,
+      "loss": 0.3141,
+      "step": 4188
+    },
+    {
+      "epoch": 4.020163226116178,
+      "grad_norm": 0.9944493363007124,
+      "learning_rate": 1.1270566793331006e-06,
+      "loss": 0.4416,
+      "step": 4189
+    },
+    {
+      "epoch": 4.021123379740758,
+      "grad_norm": 1.082323844191331,
+      "learning_rate": 1.1249388150639702e-06,
+      "loss": 0.4461,
+      "step": 4190
+    },
+    {
+      "epoch": 4.022083533365339,
+      "grad_norm": 1.43474408959,
+      "learning_rate": 1.1228226902683847e-06,
+      "loss": 0.4098,
+      "step": 4191
+    },
+    {
+      "epoch": 4.023043686989919,
+      "grad_norm": 0.844896093678417,
+      "learning_rate": 1.1207083058962465e-06,
+      "loss": 0.4262,
+      "step": 4192
+    },
+    {
+      "epoch": 4.0240038406144985,
+      "grad_norm": 0.8683584109714455,
+      "learning_rate": 1.1185956628966832e-06,
+      "loss": 0.4131,
+      "step": 4193
+    },
+    {
+      "epoch": 4.024963994239078,
+      "grad_norm": 0.9444029751557702,
+      "learning_rate": 1.1164847622180392e-06,
+      "loss": 0.5157,
+      "step": 4194
+    },
+    {
+      "epoch": 4.025924147863658,
+      "grad_norm": 0.9856182532147272,
+      "learning_rate": 1.1143756048078707e-06,
+      "loss": 0.5141,
+      "step": 4195
+    },
+    {
+      "epoch": 4.026884301488238,
+      "grad_norm": 1.0951961014372649,
+      "learning_rate": 1.1122681916129586e-06,
+      "loss": 0.4607,
+      "step": 4196
+    },
+    {
+      "epoch": 4.027844455112818,
+      "grad_norm": 0.9031442935433569,
+      "learning_rate": 1.1101625235792978e-06,
+      "loss": 0.5384,
+      "step": 4197
+    },
+    {
+      "epoch": 4.028804608737398,
+      "grad_norm": 1.4404063615658798,
+      "learning_rate": 1.1080586016521e-06,
+      "loss": 0.3734,
+      "step": 4198
+    },
+    {
+      "epoch": 4.029764762361978,
+      "grad_norm": 1.0883888695253605,
+      "learning_rate": 1.1059564267757945e-06,
+      "loss": 0.4364,
+      "step": 4199
+    },
+    {
+      "epoch": 4.030724915986558,
+      "grad_norm": 1.4977311605432297,
+      "learning_rate": 1.1038559998940228e-06,
+      "loss": 0.4881,
+      "step": 4200
+    },
+    {
+      "epoch": 4.031685069611138,
+      "grad_norm": 0.8467897334018155,
+      "learning_rate": 1.1017573219496453e-06,
+      "loss": 0.5446,
+      "step": 4201
+    },
+    {
+      "epoch": 4.0326452232357175,
+      "grad_norm": 1.494093787775511,
+      "learning_rate": 1.0996603938847361e-06,
+      "loss": 0.4865,
+      "step": 4202
+    },
+    {
+      "epoch": 4.033605376860297,
+      "grad_norm": 1.1172721868839053,
+      "learning_rate": 1.0975652166405836e-06,
+      "loss": 0.4379,
+      "step": 4203
+    },
+    {
+      "epoch": 4.034565530484878,
+      "grad_norm": 1.3357464579781297,
+      "learning_rate": 1.0954717911576923e-06,
+      "loss": 0.3648,
+      "step": 4204
+    },
+    {
+      "epoch": 4.035525684109458,
+      "grad_norm": 1.0494931367800207,
+      "learning_rate": 1.0933801183757742e-06,
+      "loss": 0.4397,
+      "step": 4205
+    },
+    {
+      "epoch": 4.036485837734038,
+      "grad_norm": 1.093847679981737,
+      "learning_rate": 1.0912901992337626e-06,
+      "loss": 0.5484,
+      "step": 4206
+    },
+    {
+      "epoch": 4.037445991358617,
+      "grad_norm": 1.0098251961439262,
+      "learning_rate": 1.0892020346698e-06,
+      "loss": 0.2999,
+      "step": 4207
+    },
+    {
+      "epoch": 4.038406144983197,
+      "grad_norm": 0.8950132279843459,
+      "learning_rate": 1.087115625621239e-06,
+      "loss": 0.4694,
+      "step": 4208
+    },
+    {
+      "epoch": 4.039366298607777,
+      "grad_norm": 1.4797419378563925,
+      "learning_rate": 1.0850309730246472e-06,
+      "loss": 0.3842,
+      "step": 4209
+    },
+    {
+      "epoch": 4.040326452232357,
+      "grad_norm": 1.0555006824431175,
+      "learning_rate": 1.0829480778158025e-06,
+      "loss": 0.486,
+      "step": 4210
+    },
+    {
+      "epoch": 4.041286605856937,
+      "grad_norm": 1.29843019699927,
+      "learning_rate": 1.0808669409296951e-06,
+      "loss": 0.3899,
+      "step": 4211
+    },
+    {
+      "epoch": 4.042246759481517,
+      "grad_norm": 1.2177826969482468,
+      "learning_rate": 1.0787875633005256e-06,
+      "loss": 0.4925,
+      "step": 4212
+    },
+    {
+      "epoch": 4.043206913106097,
+      "grad_norm": 1.6704436579707684,
+      "learning_rate": 1.0767099458617025e-06,
+      "loss": 0.5365,
+      "step": 4213
+    },
+    {
+      "epoch": 4.044167066730677,
+      "grad_norm": 1.0136148725987781,
+      "learning_rate": 1.0746340895458475e-06,
+      "loss": 0.4631,
+      "step": 4214
+    },
+    {
+      "epoch": 4.045127220355257,
+      "grad_norm": 0.8765041498810147,
+      "learning_rate": 1.0725599952847894e-06,
+      "loss": 0.4409,
+      "step": 4215
+    },
+    {
+      "epoch": 4.046087373979836,
+      "grad_norm": 0.9427533038272904,
+      "learning_rate": 1.070487664009567e-06,
+      "loss": 0.4685,
+      "step": 4216
+    },
+    {
+      "epoch": 4.047047527604417,
+      "grad_norm": 0.9933754655194416,
+      "learning_rate": 1.0684170966504298e-06,
+      "loss": 0.5583,
+      "step": 4217
+    },
+    {
+      "epoch": 4.048007681228997,
+      "grad_norm": 1.158250101522896,
+      "learning_rate": 1.0663482941368303e-06,
+      "loss": 0.4586,
+      "step": 4218
+    },
+    {
+      "epoch": 4.048967834853577,
+      "grad_norm": 0.7532692687341203,
+      "learning_rate": 1.064281257397432e-06,
+      "loss": 0.4246,
+      "step": 4219
+    },
+    {
+      "epoch": 4.0499279884781565,
+      "grad_norm": 0.8551367076440082,
+      "learning_rate": 1.062215987360109e-06,
+      "loss": 0.4799,
+      "step": 4220
+    },
+    {
+      "epoch": 4.050888142102736,
+      "grad_norm": 1.2141855585896926,
+      "learning_rate": 1.0601524849519357e-06,
+      "loss": 0.4289,
+      "step": 4221
+    },
+    {
+      "epoch": 4.051848295727316,
+      "grad_norm": 1.0090877809711172,
+      "learning_rate": 1.0580907510991982e-06,
+      "loss": 0.4573,
+      "step": 4222
+    },
+    {
+      "epoch": 4.052808449351896,
+      "grad_norm": 1.0433680921569715,
+      "learning_rate": 1.0560307867273838e-06,
+      "loss": 0.445,
+      "step": 4223
+    },
+    {
+      "epoch": 4.053768602976477,
+      "grad_norm": 0.9502679652663969,
+      "learning_rate": 1.0539725927611915e-06,
+      "loss": 0.4869,
+      "step": 4224
+    },
+    {
+      "epoch": 4.054728756601056,
+      "grad_norm": 0.9674777898369518,
+      "learning_rate": 1.0519161701245234e-06,
+      "loss": 0.3871,
+      "step": 4225
+    },
+    {
+      "epoch": 4.055688910225636,
+      "grad_norm": 1.3676812062287418,
+      "learning_rate": 1.049861519740482e-06,
+      "loss": 0.4337,
+      "step": 4226
+    },
+    {
+      "epoch": 4.056649063850216,
+      "grad_norm": 1.7586505787627171,
+      "learning_rate": 1.04780864253138e-06,
+      "loss": 0.4283,
+      "step": 4227
+    },
+    {
+      "epoch": 4.057609217474796,
+      "grad_norm": 1.073864155625145,
+      "learning_rate": 1.0457575394187314e-06,
+      "loss": 0.4644,
+      "step": 4228
+    },
+    {
+      "epoch": 4.0585693710993755,
+      "grad_norm": 1.1607167995579928,
+      "learning_rate": 1.043708211323255e-06,
+      "loss": 0.362,
+      "step": 4229
+    },
+    {
+      "epoch": 4.059529524723956,
+      "grad_norm": 1.491228533906927,
+      "learning_rate": 1.0416606591648737e-06,
+      "loss": 0.3021,
+      "step": 4230
+    },
+    {
+      "epoch": 4.060489678348536,
+      "grad_norm": 1.0147464090050025,
+      "learning_rate": 1.039614883862709e-06,
+      "loss": 0.4277,
+      "step": 4231
+    },
+    {
+      "epoch": 4.061449831973116,
+      "grad_norm": 1.3055224272493493,
+      "learning_rate": 1.037570886335087e-06,
+      "loss": 0.3747,
+      "step": 4232
+    },
+    {
+      "epoch": 4.062409985597696,
+      "grad_norm": 0.8366133380077349,
+      "learning_rate": 1.0355286674995413e-06,
+      "loss": 0.4936,
+      "step": 4233
+    },
+    {
+      "epoch": 4.063370139222275,
+      "grad_norm": 0.9246694172987252,
+      "learning_rate": 1.0334882282727971e-06,
+      "loss": 0.4579,
+      "step": 4234
+    },
+    {
+      "epoch": 4.064330292846855,
+      "grad_norm": 1.4034925818936077,
+      "learning_rate": 1.0314495695707887e-06,
+      "loss": 0.4711,
+      "step": 4235
+    },
+    {
+      "epoch": 4.065290446471435,
+      "grad_norm": 1.0809037197410558,
+      "learning_rate": 1.029412692308645e-06,
+      "loss": 0.4059,
+      "step": 4236
+    },
+    {
+      "epoch": 4.066250600096016,
+      "grad_norm": 1.3469432718341132,
+      "learning_rate": 1.0273775974007017e-06,
+      "loss": 0.3732,
+      "step": 4237
+    },
+    {
+      "epoch": 4.0672107537205955,
+      "grad_norm": 0.8597826676320224,
+      "learning_rate": 1.0253442857604911e-06,
+      "loss": 0.4137,
+      "step": 4238
+    },
+    {
+      "epoch": 4.068170907345175,
+      "grad_norm": 1.524769301533565,
+      "learning_rate": 1.0233127583007424e-06,
+      "loss": 0.3453,
+      "step": 4239
+    },
+    {
+      "epoch": 4.069131060969755,
+      "grad_norm": 1.1619159943005153,
+      "learning_rate": 1.0212830159333886e-06,
+      "loss": 0.4099,
+      "step": 4240
+    },
+    {
+      "epoch": 4.070091214594335,
+      "grad_norm": 1.0543571257877578,
+      "learning_rate": 1.0192550595695584e-06,
+      "loss": 0.4851,
+      "step": 4241
+    },
+    {
+      "epoch": 4.071051368218915,
+      "grad_norm": 0.9272472552411923,
+      "learning_rate": 1.0172288901195805e-06,
+      "loss": 0.3227,
+      "step": 4242
+    },
+    {
+      "epoch": 4.072011521843495,
+      "grad_norm": 1.1430402130845945,
+      "learning_rate": 1.0152045084929813e-06,
+      "loss": 0.4766,
+      "step": 4243
+    },
+    {
+      "epoch": 4.072971675468075,
+      "grad_norm": 0.9113893300393455,
+      "learning_rate": 1.0131819155984818e-06,
+      "loss": 0.3991,
+      "step": 4244
+    },
+    {
+      "epoch": 4.073931829092655,
+      "grad_norm": 1.1436498993759299,
+      "learning_rate": 1.0111611123440041e-06,
+      "loss": 0.4557,
+      "step": 4245
+    },
+    {
+      "epoch": 4.074891982717235,
+      "grad_norm": 1.416402183830115,
+      "learning_rate": 1.0091420996366646e-06,
+      "loss": 0.4001,
+      "step": 4246
+    },
+    {
+      "epoch": 4.0758521363418145,
+      "grad_norm": 1.1313431713196778,
+      "learning_rate": 1.0071248783827764e-06,
+      "loss": 0.4447,
+      "step": 4247
+    },
+    {
+      "epoch": 4.076812289966394,
+      "grad_norm": 0.9859324415971613,
+      "learning_rate": 1.0051094494878505e-06,
+      "loss": 0.348,
+      "step": 4248
+    },
+    {
+      "epoch": 4.077772443590974,
+      "grad_norm": 0.917928864218637,
+      "learning_rate": 1.0030958138565883e-06,
+      "loss": 0.4204,
+      "step": 4249
+    },
+    {
+      "epoch": 4.078732597215555,
+      "grad_norm": 1.0581029548257181,
+      "learning_rate": 1.001083972392889e-06,
+      "loss": 0.4339,
+      "step": 4250
+    },
+    {
+      "epoch": 4.079692750840135,
+      "grad_norm": 1.178454031735879,
+      "learning_rate": 9.990739259998517e-07,
+      "loss": 0.4464,
+      "step": 4251
+    },
+    {
+      "epoch": 4.080652904464714,
+      "grad_norm": 0.7987610911287786,
+      "learning_rate": 9.970656755797598e-07,
+      "loss": 0.5038,
+      "step": 4252
+    },
+    {
+      "epoch": 4.081613058089294,
+      "grad_norm": 0.9193583831389514,
+      "learning_rate": 9.950592220340972e-07,
+      "loss": 0.4981,
+      "step": 4253
+    },
+    {
+      "epoch": 4.082573211713874,
+      "grad_norm": 0.8787063023269425,
+      "learning_rate": 9.930545662635393e-07,
+      "loss": 0.4153,
+      "step": 4254
+    },
+    {
+      "epoch": 4.083533365338454,
+      "grad_norm": 1.3368317566526455,
+      "learning_rate": 9.91051709167955e-07,
+      "loss": 0.3974,
+      "step": 4255
+    },
+    {
+      "epoch": 4.0844935189630345,
+      "grad_norm": 1.2647051227807544,
+      "learning_rate": 9.890506516464066e-07,
+      "loss": 0.425,
+      "step": 4256
+    },
+    {
+      "epoch": 4.085453672587614,
+      "grad_norm": 1.083904419924026,
+      "learning_rate": 9.870513945971444e-07,
+      "loss": 0.4483,
+      "step": 4257
+    },
+    {
+      "epoch": 4.086413826212194,
+      "grad_norm": 0.8488700025127257,
+      "learning_rate": 9.85053938917615e-07,
+      "loss": 0.4878,
+      "step": 4258
+    },
+    {
+      "epoch": 4.087373979836774,
+      "grad_norm": 0.8706164855245921,
+      "learning_rate": 9.83058285504455e-07,
+      "loss": 0.4989,
+      "step": 4259
+    },
+    {
+      "epoch": 4.088334133461354,
+      "grad_norm": 0.8575262207094165,
+      "learning_rate": 9.81064435253492e-07,
+      "loss": 0.4784,
+      "step": 4260
+    },
+    {
+      "epoch": 4.0892942870859335,
+      "grad_norm": 0.9887605308451878,
+      "learning_rate": 9.790723890597452e-07,
+      "loss": 0.4881,
+      "step": 4261
+    },
+    {
+      "epoch": 4.090254440710513,
+      "grad_norm": 1.1387912018631177,
+      "learning_rate": 9.770821478174207e-07,
+      "loss": 0.4661,
+      "step": 4262
+    },
+    {
+      "epoch": 4.091214594335094,
+      "grad_norm": 1.1655639448301596,
+      "learning_rate": 9.750937124199162e-07,
+      "loss": 0.4145,
+      "step": 4263
+    },
+    {
+      "epoch": 4.092174747959674,
+      "grad_norm": 1.0484353735066698,
+      "learning_rate": 9.731070837598227e-07,
+      "loss": 0.5036,
+      "step": 4264
+    },
+    {
+      "epoch": 4.0931349015842535,
+      "grad_norm": 1.4390833232501383,
+      "learning_rate": 9.71122262728913e-07,
+      "loss": 0.4491,
+      "step": 4265
+    },
+    {
+      "epoch": 4.094095055208833,
+      "grad_norm": 0.7749814621177304,
+      "learning_rate": 9.69139250218154e-07,
+      "loss": 0.4536,
+      "step": 4266
+    },
+    {
+      "epoch": 4.095055208833413,
+      "grad_norm": 1.4884168779464448,
+      "learning_rate": 9.67158047117696e-07,
+      "loss": 0.3593,
+      "step": 4267
+    },
+    {
+      "epoch": 4.096015362457993,
+      "grad_norm": 1.117155150474689,
+      "learning_rate": 9.651786543168834e-07,
+      "loss": 0.5245,
+      "step": 4268
+    },
+    {
+      "epoch": 4.096975516082574,
+      "grad_norm": 1.6872999701713718,
+      "learning_rate": 9.63201072704244e-07,
+      "loss": 0.4035,
+      "step": 4269
+    },
+    {
+      "epoch": 4.097935669707153,
+      "grad_norm": 0.954975072196789,
+      "learning_rate": 9.612253031674906e-07,
+      "loss": 0.3821,
+      "step": 4270
+    },
+    {
+      "epoch": 4.098895823331733,
+      "grad_norm": 0.951304544108571,
+      "learning_rate": 9.592513465935272e-07,
+      "loss": 0.5126,
+      "step": 4271
+    },
+    {
+      "epoch": 4.099855976956313,
+      "grad_norm": 0.921490814052639,
+      "learning_rate": 9.572792038684404e-07,
+      "loss": 0.4418,
+      "step": 4272
+    },
+    {
+      "epoch": 4.100816130580893,
+      "grad_norm": 1.7228555575088056,
+      "learning_rate": 9.553088758775048e-07,
+      "loss": 0.3493,
+      "step": 4273
+    },
+    {
+      "epoch": 4.101776284205473,
+      "grad_norm": 1.006455267047783,
+      "learning_rate": 9.533403635051813e-07,
+      "loss": 0.4744,
+      "step": 4274
+    },
+    {
+      "epoch": 4.102736437830052,
+      "grad_norm": 0.8733910351772793,
+      "learning_rate": 9.513736676351104e-07,
+      "loss": 0.4198,
+      "step": 4275
+    },
+    {
+      "epoch": 4.103696591454633,
+      "grad_norm": 1.4203884190388014,
+      "learning_rate": 9.494087891501213e-07,
+      "loss": 0.4427,
+      "step": 4276
+    },
+    {
+      "epoch": 4.104656745079213,
+      "grad_norm": 0.9757000809136535,
+      "learning_rate": 9.474457289322314e-07,
+      "loss": 0.5144,
+      "step": 4277
+    },
+    {
+      "epoch": 4.105616898703793,
+      "grad_norm": 1.1060393473010894,
+      "learning_rate": 9.454844878626329e-07,
+      "loss": 0.4511,
+      "step": 4278
+    },
+    {
+      "epoch": 4.1065770523283724,
+      "grad_norm": 1.3837663291815168,
+      "learning_rate": 9.435250668217088e-07,
+      "loss": 0.4659,
+      "step": 4279
+    },
+    {
+      "epoch": 4.107537205952952,
+      "grad_norm": 1.2887214163616825,
+      "learning_rate": 9.415674666890178e-07,
+      "loss": 0.4424,
+      "step": 4280
+    },
+    {
+      "epoch": 4.108497359577532,
+      "grad_norm": 1.1799751136383634,
+      "learning_rate": 9.396116883433104e-07,
+      "loss": 0.4207,
+      "step": 4281
+    },
+    {
+      "epoch": 4.109457513202113,
+      "grad_norm": 0.9030769312588787,
+      "learning_rate": 9.376577326625147e-07,
+      "loss": 0.5047,
+      "step": 4282
+    },
+    {
+      "epoch": 4.1104176668266925,
+      "grad_norm": 1.192159261759407,
+      "learning_rate": 9.35705600523737e-07,
+      "loss": 0.4577,
+      "step": 4283
+    },
+    {
+      "epoch": 4.111377820451272,
+      "grad_norm": 1.1049190893986787,
+      "learning_rate": 9.337552928032706e-07,
+      "loss": 0.4833,
+      "step": 4284
+    },
+    {
+      "epoch": 4.112337974075852,
+      "grad_norm": 0.8603276291123754,
+      "learning_rate": 9.318068103765871e-07,
+      "loss": 0.48,
+      "step": 4285
+    },
+    {
+      "epoch": 4.113298127700432,
+      "grad_norm": 0.95385485131521,
+      "learning_rate": 9.2986015411834e-07,
+      "loss": 0.4076,
+      "step": 4286
+    },
+    {
+      "epoch": 4.114258281325012,
+      "grad_norm": 1.0983527364540822,
+      "learning_rate": 9.279153249023637e-07,
+      "loss": 0.3698,
+      "step": 4287
+    },
+    {
+      "epoch": 4.1152184349495915,
+      "grad_norm": 0.7963206415542619,
+      "learning_rate": 9.259723236016688e-07,
+      "loss": 0.4433,
+      "step": 4288
+    },
+    {
+      "epoch": 4.116178588574172,
+      "grad_norm": 1.3193320038250071,
+      "learning_rate": 9.240311510884487e-07,
+      "loss": 0.3464,
+      "step": 4289
+    },
+    {
+      "epoch": 4.117138742198752,
+      "grad_norm": 1.058292953859479,
+      "learning_rate": 9.220918082340752e-07,
+      "loss": 0.4701,
+      "step": 4290
+    },
+    {
+      "epoch": 4.118098895823332,
+      "grad_norm": 0.8658520148648889,
+      "learning_rate": 9.201542959090987e-07,
+      "loss": 0.4166,
+      "step": 4291
+    },
+    {
+      "epoch": 4.119059049447912,
+      "grad_norm": 1.0574364621218113,
+      "learning_rate": 9.182186149832484e-07,
+      "loss": 0.5533,
+      "step": 4292
+    },
+    {
+      "epoch": 4.120019203072491,
+      "grad_norm": 0.94841809193125,
+      "learning_rate": 9.162847663254292e-07,
+      "loss": 0.4965,
+      "step": 4293
+    },
+    {
+      "epoch": 4.120979356697071,
+      "grad_norm": 1.2195490538076916,
+      "learning_rate": 9.143527508037243e-07,
+      "loss": 0.408,
+      "step": 4294
+    },
+    {
+      "epoch": 4.121939510321652,
+      "grad_norm": 0.9151466119180599,
+      "learning_rate": 9.124225692853995e-07,
+      "loss": 0.4738,
+      "step": 4295
+    },
+    {
+      "epoch": 4.122899663946232,
+      "grad_norm": 0.970946276289556,
+      "learning_rate": 9.104942226368879e-07,
+      "loss": 0.3946,
+      "step": 4296
+    },
+    {
+      "epoch": 4.123859817570811,
+      "grad_norm": 1.1728049596209218,
+      "learning_rate": 9.085677117238068e-07,
+      "loss": 0.3516,
+      "step": 4297
+    },
+    {
+      "epoch": 4.124819971195391,
+      "grad_norm": 0.9065424553789335,
+      "learning_rate": 9.066430374109419e-07,
+      "loss": 0.4066,
+      "step": 4298
+    },
+    {
+      "epoch": 4.125780124819971,
+      "grad_norm": 1.1154419366105508,
+      "learning_rate": 9.047202005622641e-07,
+      "loss": 0.4886,
+      "step": 4299
+    },
+    {
+      "epoch": 4.126740278444551,
+      "grad_norm": 0.908548924060711,
+      "learning_rate": 9.027992020409127e-07,
+      "loss": 0.5112,
+      "step": 4300
+    },
+    {
+      "epoch": 4.1277004320691315,
+      "grad_norm": 1.1014847745552856,
+      "learning_rate": 9.008800427092018e-07,
+      "loss": 0.4825,
+      "step": 4301
+    },
+    {
+      "epoch": 4.128660585693711,
+      "grad_norm": 1.0762917609699958,
+      "learning_rate": 8.989627234286225e-07,
+      "loss": 0.4954,
+      "step": 4302
+    },
+    {
+      "epoch": 4.129620739318291,
+      "grad_norm": 0.9915935095231558,
+      "learning_rate": 8.970472450598394e-07,
+      "loss": 0.4679,
+      "step": 4303
+    },
+    {
+      "epoch": 4.130580892942871,
+      "grad_norm": 1.2361143931951928,
+      "learning_rate": 8.951336084626894e-07,
+      "loss": 0.4544,
+      "step": 4304
+    },
+    {
+      "epoch": 4.131541046567451,
+      "grad_norm": 0.9749138566149839,
+      "learning_rate": 8.932218144961858e-07,
+      "loss": 0.407,
+      "step": 4305
+    },
+    {
+      "epoch": 4.1325012001920305,
+      "grad_norm": 1.1217454966773088,
+      "learning_rate": 8.913118640185086e-07,
+      "loss": 0.4004,
+      "step": 4306
+    },
+    {
+      "epoch": 4.13346135381661,
+      "grad_norm": 1.1038976678078036,
+      "learning_rate": 8.894037578870152e-07,
+      "loss": 0.3665,
+      "step": 4307
+    },
+    {
+      "epoch": 4.134421507441191,
+      "grad_norm": 0.9700248788900517,
+      "learning_rate": 8.874974969582379e-07,
+      "loss": 0.4415,
+      "step": 4308
+    },
+    {
+      "epoch": 4.135381661065771,
+      "grad_norm": 1.4278006870827293,
+      "learning_rate": 8.855930820878727e-07,
+      "loss": 0.4433,
+      "step": 4309
+    },
+    {
+      "epoch": 4.136341814690351,
+      "grad_norm": 1.1091730316071229,
+      "learning_rate": 8.836905141307933e-07,
+      "loss": 0.4636,
+      "step": 4310
+    },
+    {
+      "epoch": 4.13730196831493,
+      "grad_norm": 1.1142769292697934,
+      "learning_rate": 8.817897939410391e-07,
+      "loss": 0.4611,
+      "step": 4311
+    },
+    {
+      "epoch": 4.13826212193951,
+      "grad_norm": 0.9639399232985066,
+      "learning_rate": 8.79890922371826e-07,
+      "loss": 0.4674,
+      "step": 4312
+    },
+    {
+      "epoch": 4.13922227556409,
+      "grad_norm": 0.9840393238800248,
+      "learning_rate": 8.779939002755378e-07,
+      "loss": 0.4239,
+      "step": 4313
+    },
+    {
+      "epoch": 4.14018242918867,
+      "grad_norm": 1.3571010902250684,
+      "learning_rate": 8.760987285037248e-07,
+      "loss": 0.4589,
+      "step": 4314
+    },
+    {
+      "epoch": 4.14114258281325,
+      "grad_norm": 0.9979987934238568,
+      "learning_rate": 8.742054079071105e-07,
+      "loss": 0.4836,
+      "step": 4315
+    },
+    {
+      "epoch": 4.14210273643783,
+      "grad_norm": 1.104191561242044,
+      "learning_rate": 8.723139393355861e-07,
+      "loss": 0.3678,
+      "step": 4316
+    },
+    {
+      "epoch": 4.14306289006241,
+      "grad_norm": 0.7732049241607881,
+      "learning_rate": 8.704243236382115e-07,
+      "loss": 0.5113,
+      "step": 4317
+    },
+    {
+      "epoch": 4.14402304368699,
+      "grad_norm": 1.0667449949896786,
+      "learning_rate": 8.685365616632175e-07,
+      "loss": 0.4554,
+      "step": 4318
+    },
+    {
+      "epoch": 4.14498319731157,
+      "grad_norm": 1.1287351593004549,
+      "learning_rate": 8.666506542579961e-07,
+      "loss": 0.5003,
+      "step": 4319
+    },
+    {
+      "epoch": 4.145943350936149,
+      "grad_norm": 1.3527933770086522,
+      "learning_rate": 8.64766602269112e-07,
+      "loss": 0.3664,
+      "step": 4320
+    },
+    {
+      "epoch": 4.14690350456073,
+      "grad_norm": 0.9243479456992848,
+      "learning_rate": 8.628844065422992e-07,
+      "loss": 0.402,
+      "step": 4321
+    },
+    {
+      "epoch": 4.14786365818531,
+      "grad_norm": 1.0853048962497471,
+      "learning_rate": 8.610040679224518e-07,
+      "loss": 0.3226,
+      "step": 4322
+    },
+    {
+      "epoch": 4.14882381180989,
+      "grad_norm": 1.8494142670327685,
+      "learning_rate": 8.591255872536364e-07,
+      "loss": 0.4033,
+      "step": 4323
+    },
+    {
+      "epoch": 4.1497839654344695,
+      "grad_norm": 1.1547235391359174,
+      "learning_rate": 8.572489653790789e-07,
+      "loss": 0.4074,
+      "step": 4324
+    },
+    {
+      "epoch": 4.150744119059049,
+      "grad_norm": 0.8525740470029572,
+      "learning_rate": 8.553742031411783e-07,
+      "loss": 0.4835,
+      "step": 4325
+    },
+    {
+      "epoch": 4.151704272683629,
+      "grad_norm": 1.0738258050384755,
+      "learning_rate": 8.535013013814952e-07,
+      "loss": 0.401,
+      "step": 4326
+    },
+    {
+      "epoch": 4.15266442630821,
+      "grad_norm": 1.0421905543463854,
+      "learning_rate": 8.51630260940754e-07,
+      "loss": 0.4901,
+      "step": 4327
+    },
+    {
+      "epoch": 4.1536245799327896,
+      "grad_norm": 1.9839771489313465,
+      "learning_rate": 8.497610826588443e-07,
+      "loss": 0.5864,
+      "step": 4328
+    },
+    {
+      "epoch": 4.154584733557369,
+      "grad_norm": 0.8454376940347059,
+      "learning_rate": 8.478937673748211e-07,
+      "loss": 0.437,
+      "step": 4329
+    },
+    {
+      "epoch": 4.155544887181949,
+      "grad_norm": 0.9683212298990983,
+      "learning_rate": 8.460283159269023e-07,
+      "loss": 0.3862,
+      "step": 4330
+    },
+    {
+      "epoch": 4.156505040806529,
+      "grad_norm": 1.0770426460280165,
+      "learning_rate": 8.441647291524701e-07,
+      "loss": 0.4002,
+      "step": 4331
+    },
+    {
+      "epoch": 4.157465194431109,
+      "grad_norm": 0.8926379888513267,
+      "learning_rate": 8.423030078880667e-07,
+      "loss": 0.4265,
+      "step": 4332
+    },
+    {
+      "epoch": 4.1584253480556885,
+      "grad_norm": 0.8766685012161926,
+      "learning_rate": 8.404431529693996e-07,
+      "loss": 0.4383,
+      "step": 4333
+    },
+    {
+      "epoch": 4.159385501680269,
+      "grad_norm": 0.9426447039942522,
+      "learning_rate": 8.385851652313376e-07,
+      "loss": 0.3939,
+      "step": 4334
+    },
+    {
+      "epoch": 4.160345655304849,
+      "grad_norm": 1.2529311935400318,
+      "learning_rate": 8.367290455079125e-07,
+      "loss": 0.3699,
+      "step": 4335
+    },
+    {
+      "epoch": 4.161305808929429,
+      "grad_norm": 0.9787205617363444,
+      "learning_rate": 8.348747946323165e-07,
+      "loss": 0.3142,
+      "step": 4336
+    },
+    {
+      "epoch": 4.162265962554009,
+      "grad_norm": 1.3210305280646293,
+      "learning_rate": 8.330224134369014e-07,
+      "loss": 0.4638,
+      "step": 4337
+    },
+    {
+      "epoch": 4.163226116178588,
+      "grad_norm": 0.9801531970254048,
+      "learning_rate": 8.311719027531817e-07,
+      "loss": 0.4524,
+      "step": 4338
+    },
+    {
+      "epoch": 4.164186269803168,
+      "grad_norm": 1.031215780297556,
+      "learning_rate": 8.293232634118342e-07,
+      "loss": 0.3935,
+      "step": 4339
+    },
+    {
+      "epoch": 4.165146423427748,
+      "grad_norm": 1.1366654773442557,
+      "learning_rate": 8.274764962426912e-07,
+      "loss": 0.4191,
+      "step": 4340
+    },
+    {
+      "epoch": 4.166106577052329,
+      "grad_norm": 0.7775203896986472,
+      "learning_rate": 8.256316020747479e-07,
+      "loss": 0.4779,
+      "step": 4341
+    },
+    {
+      "epoch": 4.1670667306769085,
+      "grad_norm": 0.8980317212159564,
+      "learning_rate": 8.237885817361552e-07,
+      "loss": 0.4229,
+      "step": 4342
+    },
+    {
+      "epoch": 4.168026884301488,
+      "grad_norm": 0.9484593723923103,
+      "learning_rate": 8.219474360542273e-07,
+      "loss": 0.4587,
+      "step": 4343
+    },
+    {
+      "epoch": 4.168987037926068,
+      "grad_norm": 1.1054808758696812,
+      "learning_rate": 8.201081658554361e-07,
+      "loss": 0.3986,
+      "step": 4344
+    },
+    {
+      "epoch": 4.169947191550648,
+      "grad_norm": 1.527882440838069,
+      "learning_rate": 8.182707719654076e-07,
+      "loss": 0.347,
+      "step": 4345
+    },
+    {
+      "epoch": 4.170907345175228,
+      "grad_norm": 1.3505240312263833,
+      "learning_rate": 8.164352552089288e-07,
+      "loss": 0.4489,
+      "step": 4346
+    },
+    {
+      "epoch": 4.171867498799808,
+      "grad_norm": 1.6916111421535276,
+      "learning_rate": 8.146016164099446e-07,
+      "loss": 0.3571,
+      "step": 4347
+    },
+    {
+      "epoch": 4.172827652424388,
+      "grad_norm": 1.0646221367630506,
+      "learning_rate": 8.127698563915554e-07,
+      "loss": 0.5274,
+      "step": 4348
+    },
+    {
+      "epoch": 4.173787806048968,
+      "grad_norm": 1.2484292715766703,
+      "learning_rate": 8.109399759760194e-07,
+      "loss": 0.444,
+      "step": 4349
+    },
+    {
+      "epoch": 4.174747959673548,
+      "grad_norm": 1.1668223099744186,
+      "learning_rate": 8.091119759847493e-07,
+      "loss": 0.3771,
+      "step": 4350
+    },
+    {
+      "epoch": 4.1757081132981275,
+      "grad_norm": 0.9682240709956186,
+      "learning_rate": 8.07285857238313e-07,
+      "loss": 0.4213,
+      "step": 4351
+    },
+    {
+      "epoch": 4.176668266922707,
+      "grad_norm": 1.2349362734220277,
+      "learning_rate": 8.054616205564403e-07,
+      "loss": 0.4149,
+      "step": 4352
+    },
+    {
+      "epoch": 4.177628420547288,
+      "grad_norm": 1.4933877123150152,
+      "learning_rate": 8.036392667580084e-07,
+      "loss": 0.4818,
+      "step": 4353
+    },
+    {
+      "epoch": 4.178588574171868,
+      "grad_norm": 0.9192447638962328,
+      "learning_rate": 8.018187966610535e-07,
+      "loss": 0.4631,
+      "step": 4354
+    },
+    {
+      "epoch": 4.179548727796448,
+      "grad_norm": 0.980379117822449,
+      "learning_rate": 8.000002110827626e-07,
+      "loss": 0.4105,
+      "step": 4355
+    },
+    {
+      "epoch": 4.180508881421027,
+      "grad_norm": 1.319253235429286,
+      "learning_rate": 7.981835108394825e-07,
+      "loss": 0.4223,
+      "step": 4356
+    },
+    {
+      "epoch": 4.181469035045607,
+      "grad_norm": 0.9872882315648225,
+      "learning_rate": 7.963686967467111e-07,
+      "loss": 0.4224,
+      "step": 4357
+    },
+    {
+      "epoch": 4.182429188670187,
+      "grad_norm": 2.1465017034470275,
+      "learning_rate": 7.945557696190959e-07,
+      "loss": 0.3166,
+      "step": 4358
+    },
+    {
+      "epoch": 4.183389342294767,
+      "grad_norm": 0.8976886876840247,
+      "learning_rate": 7.927447302704422e-07,
+      "loss": 0.5547,
+      "step": 4359
+    },
+    {
+      "epoch": 4.1843494959193475,
+      "grad_norm": 1.2758054935655816,
+      "learning_rate": 7.909355795137058e-07,
+      "loss": 0.3662,
+      "step": 4360
+    },
+    {
+      "epoch": 4.185309649543927,
+      "grad_norm": 0.9514119430123684,
+      "learning_rate": 7.891283181609955e-07,
+      "loss": 0.3741,
+      "step": 4361
+    },
+    {
+      "epoch": 4.186269803168507,
+      "grad_norm": 0.9818243875674629,
+      "learning_rate": 7.873229470235722e-07,
+      "loss": 0.5299,
+      "step": 4362
+    },
+    {
+      "epoch": 4.187229956793087,
+      "grad_norm": 1.0348778856103753,
+      "learning_rate": 7.855194669118466e-07,
+      "loss": 0.3903,
+      "step": 4363
+    },
+    {
+      "epoch": 4.188190110417667,
+      "grad_norm": 0.9601524852459529,
+      "learning_rate": 7.837178786353811e-07,
+      "loss": 0.4132,
+      "step": 4364
+    },
+    {
+      "epoch": 4.189150264042246,
+      "grad_norm": 1.1093554204911817,
+      "learning_rate": 7.81918183002891e-07,
+      "loss": 0.4645,
+      "step": 4365
+    },
+    {
+      "epoch": 4.190110417666826,
+      "grad_norm": 1.2027994506192603,
+      "learning_rate": 7.801203808222396e-07,
+      "loss": 0.4558,
+      "step": 4366
+    },
+    {
+      "epoch": 4.191070571291407,
+      "grad_norm": 0.9758616955709369,
+      "learning_rate": 7.783244729004425e-07,
+      "loss": 0.3789,
+      "step": 4367
+    },
+    {
+      "epoch": 4.192030724915987,
+      "grad_norm": 1.067144813873389,
+      "learning_rate": 7.765304600436596e-07,
+      "loss": 0.4328,
+      "step": 4368
+    },
+    {
+      "epoch": 4.1929908785405665,
+      "grad_norm": 1.329729032766879,
+      "learning_rate": 7.747383430572081e-07,
+      "loss": 0.3728,
+      "step": 4369
+    },
+    {
+      "epoch": 4.193951032165146,
+      "grad_norm": 1.327894271246118,
+      "learning_rate": 7.729481227455498e-07,
+      "loss": 0.3563,
+      "step": 4370
+    },
+    {
+      "epoch": 4.194911185789726,
+      "grad_norm": 0.9759885002648725,
+      "learning_rate": 7.711597999122939e-07,
+      "loss": 0.4585,
+      "step": 4371
+    },
+    {
+      "epoch": 4.195871339414306,
+      "grad_norm": 1.0770740023102026,
+      "learning_rate": 7.693733753601995e-07,
+      "loss": 0.4242,
+      "step": 4372
+    },
+    {
+      "epoch": 4.196831493038887,
+      "grad_norm": 1.0916757540878639,
+      "learning_rate": 7.675888498911743e-07,
+      "loss": 0.3555,
+      "step": 4373
+    },
+    {
+      "epoch": 4.197791646663466,
+      "grad_norm": 0.9751146380302486,
+      "learning_rate": 7.658062243062725e-07,
+      "loss": 0.315,
+      "step": 4374
+    },
+    {
+      "epoch": 4.198751800288046,
+      "grad_norm": 0.7201333438208256,
+      "learning_rate": 7.640254994056967e-07,
+      "loss": 0.4189,
+      "step": 4375
+    },
+    {
+      "epoch": 4.199711953912626,
+      "grad_norm": 1.0943495629175872,
+      "learning_rate": 7.622466759887931e-07,
+      "loss": 0.3943,
+      "step": 4376
+    },
+    {
+      "epoch": 4.200672107537206,
+      "grad_norm": 1.097804637006829,
+      "learning_rate": 7.604697548540574e-07,
+      "loss": 0.4712,
+      "step": 4377
+    },
+    {
+      "epoch": 4.201632261161786,
+      "grad_norm": 1.134117033885051,
+      "learning_rate": 7.58694736799131e-07,
+      "loss": 0.4902,
+      "step": 4378
+    },
+    {
+      "epoch": 4.202592414786366,
+      "grad_norm": 1.2867175110808418,
+      "learning_rate": 7.569216226207999e-07,
+      "loss": 0.4477,
+      "step": 4379
+    },
+    {
+      "epoch": 4.203552568410946,
+      "grad_norm": 0.9635986344536933,
+      "learning_rate": 7.551504131149972e-07,
+      "loss": 0.4581,
+      "step": 4380
+    },
+    {
+      "epoch": 4.204512722035526,
+      "grad_norm": 1.2549734489048616,
+      "learning_rate": 7.533811090767984e-07,
+      "loss": 0.4083,
+      "step": 4381
+    },
+    {
+      "epoch": 4.205472875660106,
+      "grad_norm": 1.0346341836574662,
+      "learning_rate": 7.516137113004246e-07,
+      "loss": 0.4214,
+      "step": 4382
+    },
+    {
+      "epoch": 4.206433029284685,
+      "grad_norm": 1.1758488932610325,
+      "learning_rate": 7.498482205792457e-07,
+      "loss": 0.4676,
+      "step": 4383
+    },
+    {
+      "epoch": 4.207393182909265,
+      "grad_norm": 0.9125108956172018,
+      "learning_rate": 7.480846377057677e-07,
+      "loss": 0.4673,
+      "step": 4384
+    },
+    {
+      "epoch": 4.208353336533845,
+      "grad_norm": 1.4564810659659302,
+      "learning_rate": 7.463229634716462e-07,
+      "loss": 0.3907,
+      "step": 4385
+    },
+    {
+      "epoch": 4.209313490158426,
+      "grad_norm": 1.5349426567758955,
+      "learning_rate": 7.445631986676755e-07,
+      "loss": 0.4176,
+      "step": 4386
+    },
+    {
+      "epoch": 4.2102736437830055,
+      "grad_norm": 0.8116893514756358,
+      "learning_rate": 7.428053440837973e-07,
+      "loss": 0.3901,
+      "step": 4387
+    },
+    {
+      "epoch": 4.211233797407585,
+      "grad_norm": 1.0618448948486043,
+      "learning_rate": 7.410494005090946e-07,
+      "loss": 0.3491,
+      "step": 4388
+    },
+    {
+      "epoch": 4.212193951032165,
+      "grad_norm": 1.0709972050922476,
+      "learning_rate": 7.392953687317894e-07,
+      "loss": 0.4423,
+      "step": 4389
+    },
+    {
+      "epoch": 4.213154104656745,
+      "grad_norm": 1.375030001306386,
+      "learning_rate": 7.375432495392487e-07,
+      "loss": 0.3047,
+      "step": 4390
+    },
+    {
+      "epoch": 4.214114258281325,
+      "grad_norm": 0.952324512104103,
+      "learning_rate": 7.3579304371798e-07,
+      "loss": 0.463,
+      "step": 4391
+    },
+    {
+      "epoch": 4.215074411905905,
+      "grad_norm": 1.2558651253150959,
+      "learning_rate": 7.34044752053632e-07,
+      "loss": 0.3193,
+      "step": 4392
+    },
+    {
+      "epoch": 4.216034565530485,
+      "grad_norm": 0.8595265997554533,
+      "learning_rate": 7.322983753309959e-07,
+      "loss": 0.4808,
+      "step": 4393
+    },
+    {
+      "epoch": 4.216994719155065,
+      "grad_norm": 1.1045760086354,
+      "learning_rate": 7.305539143339985e-07,
+      "loss": 0.4097,
+      "step": 4394
+    },
+    {
+      "epoch": 4.217954872779645,
+      "grad_norm": 0.7915374479752565,
+      "learning_rate": 7.2881136984571e-07,
+      "loss": 0.4936,
+      "step": 4395
+    },
+    {
+      "epoch": 4.2189150264042246,
+      "grad_norm": 1.0908526339530273,
+      "learning_rate": 7.270707426483431e-07,
+      "loss": 0.6003,
+      "step": 4396
+    },
+    {
+      "epoch": 4.219875180028804,
+      "grad_norm": 0.9341987903260882,
+      "learning_rate": 7.253320335232439e-07,
+      "loss": 0.3379,
+      "step": 4397
+    },
+    {
+      "epoch": 4.220835333653384,
+      "grad_norm": 1.5674855072164864,
+      "learning_rate": 7.235952432509024e-07,
+      "loss": 0.4439,
+      "step": 4398
+    },
+    {
+      "epoch": 4.221795487277965,
+      "grad_norm": 0.9120035518167652,
+      "learning_rate": 7.218603726109413e-07,
+      "loss": 0.4495,
+      "step": 4399
+    },
+    {
+      "epoch": 4.222755640902545,
+      "grad_norm": 0.8944366659130141,
+      "learning_rate": 7.201274223821298e-07,
+      "loss": 0.4331,
+      "step": 4400
+    },
+    {
+      "epoch": 4.223715794527124,
+      "grad_norm": 0.9514219119648922,
+      "learning_rate": 7.183963933423699e-07,
+      "loss": 0.4905,
+      "step": 4401
+    },
+    {
+      "epoch": 4.224675948151704,
+      "grad_norm": 1.3943489187171438,
+      "learning_rate": 7.166672862687002e-07,
+      "loss": 0.3928,
+      "step": 4402
+    },
+    {
+      "epoch": 4.225636101776284,
+      "grad_norm": 1.214565473547559,
+      "learning_rate": 7.149401019372993e-07,
+      "loss": 0.3932,
+      "step": 4403
+    },
+    {
+      "epoch": 4.226596255400864,
+      "grad_norm": 1.5195610739463175,
+      "learning_rate": 7.132148411234818e-07,
+      "loss": 0.3896,
+      "step": 4404
+    },
+    {
+      "epoch": 4.2275564090254445,
+      "grad_norm": 1.0302111293927776,
+      "learning_rate": 7.11491504601699e-07,
+      "loss": 0.5536,
+      "step": 4405
+    },
+    {
+      "epoch": 4.228516562650024,
+      "grad_norm": 1.0380122600665802,
+      "learning_rate": 7.097700931455397e-07,
+      "loss": 0.4166,
+      "step": 4406
+    },
+    {
+      "epoch": 4.229476716274604,
+      "grad_norm": 1.0912850823669555,
+      "learning_rate": 7.080506075277243e-07,
+      "loss": 0.3937,
+      "step": 4407
+    },
+    {
+      "epoch": 4.230436869899184,
+      "grad_norm": 1.1944135216477008,
+      "learning_rate": 7.063330485201126e-07,
+      "loss": 0.488,
+      "step": 4408
+    },
+    {
+      "epoch": 4.231397023523764,
+      "grad_norm": 0.8459026612189624,
+      "learning_rate": 7.046174168936987e-07,
+      "loss": 0.3833,
+      "step": 4409
+    },
+    {
+      "epoch": 4.2323571771483435,
+      "grad_norm": 1.0128430552676349,
+      "learning_rate": 7.029037134186112e-07,
+      "loss": 0.5029,
+      "step": 4410
+    },
+    {
+      "epoch": 4.233317330772923,
+      "grad_norm": 1.3581601119845752,
+      "learning_rate": 7.011919388641148e-07,
+      "loss": 0.4926,
+      "step": 4411
+    },
+    {
+      "epoch": 4.234277484397504,
+      "grad_norm": 1.089810644774874,
+      "learning_rate": 6.994820939986025e-07,
+      "loss": 0.3769,
+      "step": 4412
+    },
+    {
+      "epoch": 4.235237638022084,
+      "grad_norm": 1.0701855565316365,
+      "learning_rate": 6.977741795896098e-07,
+      "loss": 0.5343,
+      "step": 4413
+    },
+    {
+      "epoch": 4.2361977916466635,
+      "grad_norm": 0.9992298903575676,
+      "learning_rate": 6.960681964038008e-07,
+      "loss": 0.4292,
+      "step": 4414
+    },
+    {
+      "epoch": 4.237157945271243,
+      "grad_norm": 1.1421889722876242,
+      "learning_rate": 6.943641452069705e-07,
+      "loss": 0.5045,
+      "step": 4415
+    },
+    {
+      "epoch": 4.238118098895823,
+      "grad_norm": 1.1337211290098332,
+      "learning_rate": 6.92662026764051e-07,
+      "loss": 0.4609,
+      "step": 4416
+    },
+    {
+      "epoch": 4.239078252520403,
+      "grad_norm": 1.3972395737905647,
+      "learning_rate": 6.909618418391045e-07,
+      "loss": 0.4463,
+      "step": 4417
+    },
+    {
+      "epoch": 4.240038406144984,
+      "grad_norm": 1.1080615797570958,
+      "learning_rate": 6.892635911953261e-07,
+      "loss": 0.5192,
+      "step": 4418
+    },
+    {
+      "epoch": 4.240998559769563,
+      "grad_norm": 1.3289900672497104,
+      "learning_rate": 6.875672755950441e-07,
+      "loss": 0.4912,
+      "step": 4419
+    },
+    {
+      "epoch": 4.241958713394143,
+      "grad_norm": 1.0115555082204035,
+      "learning_rate": 6.858728957997129e-07,
+      "loss": 0.4726,
+      "step": 4420
+    },
+    {
+      "epoch": 4.242918867018723,
+      "grad_norm": 1.3820913815000928,
+      "learning_rate": 6.841804525699236e-07,
+      "loss": 0.3195,
+      "step": 4421
+    },
+    {
+      "epoch": 4.243879020643303,
+      "grad_norm": 0.9404980902608489,
+      "learning_rate": 6.824899466653961e-07,
+      "loss": 0.4956,
+      "step": 4422
+    },
+    {
+      "epoch": 4.244839174267883,
+      "grad_norm": 1.0839776582644958,
+      "learning_rate": 6.808013788449791e-07,
+      "loss": 0.5658,
+      "step": 4423
+    },
+    {
+      "epoch": 4.245799327892462,
+      "grad_norm": 1.1769699753328842,
+      "learning_rate": 6.79114749866655e-07,
+      "loss": 0.451,
+      "step": 4424
+    },
+    {
+      "epoch": 4.246759481517043,
+      "grad_norm": 1.0114614360470604,
+      "learning_rate": 6.774300604875312e-07,
+      "loss": 0.5437,
+      "step": 4425
+    },
+    {
+      "epoch": 4.247719635141623,
+      "grad_norm": 1.1704788199467249,
+      "learning_rate": 6.757473114638458e-07,
+      "loss": 0.394,
+      "step": 4426
+    },
+    {
+      "epoch": 4.248679788766203,
+      "grad_norm": 1.015827273715958,
+      "learning_rate": 6.740665035509714e-07,
+      "loss": 0.4339,
+      "step": 4427
+    },
+    {
+      "epoch": 4.2496399423907825,
+      "grad_norm": 1.5050906563570037,
+      "learning_rate": 6.723876375034e-07,
+      "loss": 0.3998,
+      "step": 4428
+    },
+    {
+      "epoch": 4.250600096015362,
+      "grad_norm": 0.9893318347609339,
+      "learning_rate": 6.707107140747604e-07,
+      "loss": 0.477,
+      "step": 4429
+    },
+    {
+      "epoch": 4.251560249639942,
+      "grad_norm": 1.175298329064842,
+      "learning_rate": 6.690357340178011e-07,
+      "loss": 0.4447,
+      "step": 4430
+    },
+    {
+      "epoch": 4.252520403264523,
+      "grad_norm": 1.0576490355024406,
+      "learning_rate": 6.673626980844067e-07,
+      "loss": 0.4476,
+      "step": 4431
+    },
+    {
+      "epoch": 4.2534805568891025,
+      "grad_norm": 1.208711748055826,
+      "learning_rate": 6.656916070255842e-07,
+      "loss": 0.4743,
+      "step": 4432
+    },
+    {
+      "epoch": 4.254440710513682,
+      "grad_norm": 1.0782587346243588,
+      "learning_rate": 6.640224615914681e-07,
+      "loss": 0.4515,
+      "step": 4433
+    },
+    {
+      "epoch": 4.255400864138262,
+      "grad_norm": 1.0909404011006474,
+      "learning_rate": 6.62355262531319e-07,
+      "loss": 0.4656,
+      "step": 4434
+    },
+    {
+      "epoch": 4.256361017762842,
+      "grad_norm": 1.2066058971451301,
+      "learning_rate": 6.606900105935265e-07,
+      "loss": 0.4056,
+      "step": 4435
+    },
+    {
+      "epoch": 4.257321171387422,
+      "grad_norm": 1.365534030628649,
+      "learning_rate": 6.590267065256029e-07,
+      "loss": 0.3418,
+      "step": 4436
+    },
+    {
+      "epoch": 4.2582813250120015,
+      "grad_norm": 0.9844552332348545,
+      "learning_rate": 6.573653510741901e-07,
+      "loss": 0.4561,
+      "step": 4437
+    },
+    {
+      "epoch": 4.259241478636582,
+      "grad_norm": 1.0442577595603217,
+      "learning_rate": 6.557059449850494e-07,
+      "loss": 0.4649,
+      "step": 4438
+    },
+    {
+      "epoch": 4.260201632261162,
+      "grad_norm": 1.5771119655670944,
+      "learning_rate": 6.540484890030724e-07,
+      "loss": 0.3592,
+      "step": 4439
+    },
+    {
+      "epoch": 4.261161785885742,
+      "grad_norm": 1.1659416282352184,
+      "learning_rate": 6.523929838722726e-07,
+      "loss": 0.3645,
+      "step": 4440
+    },
+    {
+      "epoch": 4.262121939510322,
+      "grad_norm": 1.078049494468062,
+      "learning_rate": 6.507394303357894e-07,
+      "loss": 0.4677,
+      "step": 4441
+    },
+    {
+      "epoch": 4.263082093134901,
+      "grad_norm": 1.0508118717897696,
+      "learning_rate": 6.49087829135886e-07,
+      "loss": 0.4356,
+      "step": 4442
+    },
+    {
+      "epoch": 4.264042246759481,
+      "grad_norm": 1.4498334722820503,
+      "learning_rate": 6.474381810139446e-07,
+      "loss": 0.4595,
+      "step": 4443
+    },
+    {
+      "epoch": 4.265002400384062,
+      "grad_norm": 1.1094692463594185,
+      "learning_rate": 6.457904867104792e-07,
+      "loss": 0.4981,
+      "step": 4444
+    },
+    {
+      "epoch": 4.265962554008642,
+      "grad_norm": 0.9098243322565156,
+      "learning_rate": 6.441447469651213e-07,
+      "loss": 0.5511,
+      "step": 4445
+    },
+    {
+      "epoch": 4.2669227076332215,
+      "grad_norm": 0.9410277880142008,
+      "learning_rate": 6.42500962516624e-07,
+      "loss": 0.3342,
+      "step": 4446
+    },
+    {
+      "epoch": 4.267882861257801,
+      "grad_norm": 1.0298469402066288,
+      "learning_rate": 6.408591341028652e-07,
+      "loss": 0.3979,
+      "step": 4447
+    },
+    {
+      "epoch": 4.268843014882381,
+      "grad_norm": 1.035142320696121,
+      "learning_rate": 6.392192624608451e-07,
+      "loss": 0.5436,
+      "step": 4448
+    },
+    {
+      "epoch": 4.269803168506961,
+      "grad_norm": 1.0262178512812177,
+      "learning_rate": 6.375813483266835e-07,
+      "loss": 0.4625,
+      "step": 4449
+    },
+    {
+      "epoch": 4.270763322131541,
+      "grad_norm": 1.2681507987219254,
+      "learning_rate": 6.35945392435624e-07,
+      "loss": 0.3438,
+      "step": 4450
+    },
+    {
+      "epoch": 4.271723475756121,
+      "grad_norm": 1.0898167700551433,
+      "learning_rate": 6.343113955220276e-07,
+      "loss": 0.4123,
+      "step": 4451
+    },
+    {
+      "epoch": 4.272683629380701,
+      "grad_norm": 1.0630260987417366,
+      "learning_rate": 6.326793583193797e-07,
+      "loss": 0.4529,
+      "step": 4452
+    },
+    {
+      "epoch": 4.273643783005281,
+      "grad_norm": 1.0182487962488627,
+      "learning_rate": 6.310492815602831e-07,
+      "loss": 0.5655,
+      "step": 4453
+    },
+    {
+      "epoch": 4.274603936629861,
+      "grad_norm": 1.2327303090634718,
+      "learning_rate": 6.294211659764626e-07,
+      "loss": 0.3857,
+      "step": 4454
+    },
+    {
+      "epoch": 4.2755640902544405,
+      "grad_norm": 1.009455204999046,
+      "learning_rate": 6.277950122987631e-07,
+      "loss": 0.467,
+      "step": 4455
+    },
+    {
+      "epoch": 4.27652424387902,
+      "grad_norm": 1.3042464656689299,
+      "learning_rate": 6.261708212571438e-07,
+      "loss": 0.3489,
+      "step": 4456
+    },
+    {
+      "epoch": 4.277484397503601,
+      "grad_norm": 1.1822483571429399,
+      "learning_rate": 6.245485935806905e-07,
+      "loss": 0.3783,
+      "step": 4457
+    },
+    {
+      "epoch": 4.278444551128181,
+      "grad_norm": 0.9538121742017903,
+      "learning_rate": 6.229283299976041e-07,
+      "loss": 0.5196,
+      "step": 4458
+    },
+    {
+      "epoch": 4.279404704752761,
+      "grad_norm": 1.3927130292859722,
+      "learning_rate": 6.213100312352005e-07,
+      "loss": 0.4262,
+      "step": 4459
+    },
+    {
+      "epoch": 4.28036485837734,
+      "grad_norm": 1.3292202438041856,
+      "learning_rate": 6.196936980199181e-07,
+      "loss": 0.4451,
+      "step": 4460
+    },
+    {
+      "epoch": 4.28132501200192,
+      "grad_norm": 1.1446926937061679,
+      "learning_rate": 6.18079331077312e-07,
+      "loss": 0.4448,
+      "step": 4461
+    },
+    {
+      "epoch": 4.2822851656265,
+      "grad_norm": 0.774846484331113,
+      "learning_rate": 6.164669311320543e-07,
+      "loss": 0.4745,
+      "step": 4462
+    },
+    {
+      "epoch": 4.28324531925108,
+      "grad_norm": 1.4484757261162085,
+      "learning_rate": 6.148564989079347e-07,
+      "loss": 0.3511,
+      "step": 4463
+    },
+    {
+      "epoch": 4.2842054728756604,
+      "grad_norm": 0.9140691682864316,
+      "learning_rate": 6.13248035127857e-07,
+      "loss": 0.4921,
+      "step": 4464
+    },
+    {
+      "epoch": 4.28516562650024,
+      "grad_norm": 0.9016712253267545,
+      "learning_rate": 6.116415405138443e-07,
+      "loss": 0.4195,
+      "step": 4465
+    },
+    {
+      "epoch": 4.28612578012482,
+      "grad_norm": 1.035987517546343,
+      "learning_rate": 6.100370157870356e-07,
+      "loss": 0.4645,
+      "step": 4466
+    },
+    {
+      "epoch": 4.2870859337494,
+      "grad_norm": 1.1887084816991396,
+      "learning_rate": 6.08434461667684e-07,
+      "loss": 0.5482,
+      "step": 4467
+    },
+    {
+      "epoch": 4.28804608737398,
+      "grad_norm": 1.1201728937136024,
+      "learning_rate": 6.06833878875161e-07,
+      "loss": 0.405,
+      "step": 4468
+    },
+    {
+      "epoch": 4.289006240998559,
+      "grad_norm": 1.0066174407345565,
+      "learning_rate": 6.052352681279478e-07,
+      "loss": 0.4164,
+      "step": 4469
+    },
+    {
+      "epoch": 4.28996639462314,
+      "grad_norm": 1.0481544696809473,
+      "learning_rate": 6.036386301436448e-07,
+      "loss": 0.419,
+      "step": 4470
+    },
+    {
+      "epoch": 4.29092654824772,
+      "grad_norm": 1.0618489948336933,
+      "learning_rate": 6.020439656389676e-07,
+      "loss": 0.5291,
+      "step": 4471
+    },
+    {
+      "epoch": 4.2918867018723,
+      "grad_norm": 0.9405865659105404,
+      "learning_rate": 6.004512753297421e-07,
+      "loss": 0.4382,
+      "step": 4472
+    },
+    {
+      "epoch": 4.2928468554968795,
+      "grad_norm": 1.1299975350075617,
+      "learning_rate": 5.988605599309116e-07,
+      "loss": 0.4626,
+      "step": 4473
+    },
+    {
+      "epoch": 4.293807009121459,
+      "grad_norm": 1.0377816138637095,
+      "learning_rate": 5.97271820156528e-07,
+      "loss": 0.4099,
+      "step": 4474
+    },
+    {
+      "epoch": 4.294767162746039,
+      "grad_norm": 1.2302999158890702,
+      "learning_rate": 5.95685056719763e-07,
+      "loss": 0.3626,
+      "step": 4475
+    },
+    {
+      "epoch": 4.295727316370619,
+      "grad_norm": 0.9961375213603376,
+      "learning_rate": 5.941002703328974e-07,
+      "loss": 0.3926,
+      "step": 4476
+    },
+    {
+      "epoch": 4.2966874699952,
+      "grad_norm": 2.2479451107992428,
+      "learning_rate": 5.925174617073226e-07,
+      "loss": 0.434,
+      "step": 4477
+    },
+    {
+      "epoch": 4.297647623619779,
+      "grad_norm": 1.4175075649198223,
+      "learning_rate": 5.909366315535464e-07,
+      "loss": 0.4914,
+      "step": 4478
+    },
+    {
+      "epoch": 4.298607777244359,
+      "grad_norm": 0.8534080948985644,
+      "learning_rate": 5.893577805811856e-07,
+      "loss": 0.4018,
+      "step": 4479
+    },
+    {
+      "epoch": 4.299567930868939,
+      "grad_norm": 0.6793077800812086,
+      "learning_rate": 5.877809094989706e-07,
+      "loss": 0.5068,
+      "step": 4480
+    },
+    {
+      "epoch": 4.300528084493519,
+      "grad_norm": 1.1049764695956181,
+      "learning_rate": 5.862060190147417e-07,
+      "loss": 0.4824,
+      "step": 4481
+    },
+    {
+      "epoch": 4.3014882381180986,
+      "grad_norm": 0.9954385533117368,
+      "learning_rate": 5.84633109835449e-07,
+      "loss": 0.4763,
+      "step": 4482
+    },
+    {
+      "epoch": 4.302448391742679,
+      "grad_norm": 0.9487563599927328,
+      "learning_rate": 5.830621826671562e-07,
+      "loss": 0.4327,
+      "step": 4483
+    },
+    {
+      "epoch": 4.303408545367259,
+      "grad_norm": 1.5416099471982594,
+      "learning_rate": 5.814932382150346e-07,
+      "loss": 0.3276,
+      "step": 4484
+    },
+    {
+      "epoch": 4.304368698991839,
+      "grad_norm": 0.907646344471589,
+      "learning_rate": 5.799262771833675e-07,
+      "loss": 0.4563,
+      "step": 4485
+    },
+    {
+      "epoch": 4.305328852616419,
+      "grad_norm": 0.7292641239404976,
+      "learning_rate": 5.783613002755478e-07,
+      "loss": 0.4809,
+      "step": 4486
+    },
+    {
+      "epoch": 4.306289006240998,
+      "grad_norm": 1.4488767928829671,
+      "learning_rate": 5.767983081940731e-07,
+      "loss": 0.5004,
+      "step": 4487
+    },
+    {
+      "epoch": 4.307249159865578,
+      "grad_norm": 0.8757980682368414,
+      "learning_rate": 5.752373016405577e-07,
+      "loss": 0.4347,
+      "step": 4488
+    },
+    {
+      "epoch": 4.308209313490158,
+      "grad_norm": 1.1784110948522728,
+      "learning_rate": 5.736782813157204e-07,
+      "loss": 0.3357,
+      "step": 4489
+    },
+    {
+      "epoch": 4.309169467114739,
+      "grad_norm": 1.2082889439231683,
+      "learning_rate": 5.721212479193871e-07,
+      "loss": 0.47,
+      "step": 4490
+    },
+    {
+      "epoch": 4.3101296207393185,
+      "grad_norm": 0.9932145983115908,
+      "learning_rate": 5.705662021504943e-07,
+      "loss": 0.4699,
+      "step": 4491
+    },
+    {
+      "epoch": 4.311089774363898,
+      "grad_norm": 0.9426793040185881,
+      "learning_rate": 5.690131447070851e-07,
+      "loss": 0.4547,
+      "step": 4492
+    },
+    {
+      "epoch": 4.312049927988478,
+      "grad_norm": 0.7503710689496849,
+      "learning_rate": 5.674620762863109e-07,
+      "loss": 0.3817,
+      "step": 4493
+    },
+    {
+      "epoch": 4.313010081613058,
+      "grad_norm": 1.3958488505849693,
+      "learning_rate": 5.659129975844302e-07,
+      "loss": 0.3035,
+      "step": 4494
+    },
+    {
+      "epoch": 4.313970235237638,
+      "grad_norm": 1.2066119515708529,
+      "learning_rate": 5.643659092968062e-07,
+      "loss": 0.5303,
+      "step": 4495
+    },
+    {
+      "epoch": 4.314930388862218,
+      "grad_norm": 1.020308109195774,
+      "learning_rate": 5.628208121179107e-07,
+      "loss": 0.3165,
+      "step": 4496
+    },
+    {
+      "epoch": 4.315890542486798,
+      "grad_norm": 1.1745863963208425,
+      "learning_rate": 5.612777067413228e-07,
+      "loss": 0.4169,
+      "step": 4497
+    },
+    {
+      "epoch": 4.316850696111378,
+      "grad_norm": 1.1150202642573368,
+      "learning_rate": 5.597365938597243e-07,
+      "loss": 0.395,
+      "step": 4498
+    },
+    {
+      "epoch": 4.317810849735958,
+      "grad_norm": 1.1696750726671203,
+      "learning_rate": 5.581974741649066e-07,
+      "loss": 0.4384,
+      "step": 4499
+    },
+    {
+      "epoch": 4.3187710033605375,
+      "grad_norm": 1.3783733766796513,
+      "learning_rate": 5.566603483477607e-07,
+      "loss": 0.3204,
+      "step": 4500
+    },
+    {
+      "epoch": 4.319731156985117,
+      "grad_norm": 0.9876079093983445,
+      "learning_rate": 5.551252170982874e-07,
+      "loss": 0.4937,
+      "step": 4501
+    },
+    {
+      "epoch": 4.320691310609698,
+      "grad_norm": 0.9296670988101725,
+      "learning_rate": 5.535920811055928e-07,
+      "loss": 0.4967,
+      "step": 4502
+    },
+    {
+      "epoch": 4.321651464234278,
+      "grad_norm": 1.0871976192602064,
+      "learning_rate": 5.520609410578826e-07,
+      "loss": 0.5041,
+      "step": 4503
+    },
+    {
+      "epoch": 4.322611617858858,
+      "grad_norm": 0.9435435305415898,
+      "learning_rate": 5.505317976424712e-07,
+      "loss": 0.4403,
+      "step": 4504
+    },
+    {
+      "epoch": 4.323571771483437,
+      "grad_norm": 1.0117052877528614,
+      "learning_rate": 5.49004651545772e-07,
+      "loss": 0.4853,
+      "step": 4505
+    },
+    {
+      "epoch": 4.324531925108017,
+      "grad_norm": 1.2471165167903087,
+      "learning_rate": 5.474795034533064e-07,
+      "loss": 0.4208,
+      "step": 4506
+    },
+    {
+      "epoch": 4.325492078732597,
+      "grad_norm": 1.163381589307012,
+      "learning_rate": 5.459563540496987e-07,
+      "loss": 0.3361,
+      "step": 4507
+    },
+    {
+      "epoch": 4.326452232357177,
+      "grad_norm": 0.9948369298622194,
+      "learning_rate": 5.444352040186712e-07,
+      "loss": 0.4151,
+      "step": 4508
+    },
+    {
+      "epoch": 4.3274123859817575,
+      "grad_norm": 1.2134082766312317,
+      "learning_rate": 5.429160540430533e-07,
+      "loss": 0.4682,
+      "step": 4509
+    },
+    {
+      "epoch": 4.328372539606337,
+      "grad_norm": 1.4998570790890307,
+      "learning_rate": 5.413989048047752e-07,
+      "loss": 0.3206,
+      "step": 4510
+    },
+    {
+      "epoch": 4.329332693230917,
+      "grad_norm": 1.3959124111395018,
+      "learning_rate": 5.398837569848686e-07,
+      "loss": 0.322,
+      "step": 4511
+    },
+    {
+      "epoch": 4.330292846855497,
+      "grad_norm": 1.016386897809876,
+      "learning_rate": 5.38370611263469e-07,
+      "loss": 0.4268,
+      "step": 4512
+    },
+    {
+      "epoch": 4.331253000480077,
+      "grad_norm": 1.718250742759537,
+      "learning_rate": 5.368594683198087e-07,
+      "loss": 0.45,
+      "step": 4513
+    },
+    {
+      "epoch": 4.3322131541046565,
+      "grad_norm": 1.1866762837708889,
+      "learning_rate": 5.353503288322243e-07,
+      "loss": 0.3966,
+      "step": 4514
+    },
+    {
+      "epoch": 4.333173307729236,
+      "grad_norm": 0.9253850877437652,
+      "learning_rate": 5.338431934781535e-07,
+      "loss": 0.451,
+      "step": 4515
+    },
+    {
+      "epoch": 4.334133461353817,
+      "grad_norm": 1.2561321092435203,
+      "learning_rate": 5.323380629341324e-07,
+      "loss": 0.4467,
+      "step": 4516
+    },
+    {
+      "epoch": 4.335093614978397,
+      "grad_norm": 1.3632641866979487,
+      "learning_rate": 5.308349378757988e-07,
+      "loss": 0.4173,
+      "step": 4517
+    },
+    {
+      "epoch": 4.3360537686029765,
+      "grad_norm": 1.407036831917338,
+      "learning_rate": 5.293338189778874e-07,
+      "loss": 0.2751,
+      "step": 4518
+    },
+    {
+      "epoch": 4.337013922227556,
+      "grad_norm": 1.2780212410441019,
+      "learning_rate": 5.278347069142364e-07,
+      "loss": 0.4011,
+      "step": 4519
+    },
+    {
+      "epoch": 4.337974075852136,
+      "grad_norm": 1.1422680900779343,
+      "learning_rate": 5.263376023577821e-07,
+      "loss": 0.4181,
+      "step": 4520
+    },
+    {
+      "epoch": 4.338934229476716,
+      "grad_norm": 1.123026904967085,
+      "learning_rate": 5.248425059805557e-07,
+      "loss": 0.5086,
+      "step": 4521
+    },
+    {
+      "epoch": 4.339894383101297,
+      "grad_norm": 1.060183465329235,
+      "learning_rate": 5.23349418453692e-07,
+      "loss": 0.4731,
+      "step": 4522
+    },
+    {
+      "epoch": 4.340854536725876,
+      "grad_norm": 1.2510167414786553,
+      "learning_rate": 5.218583404474214e-07,
+      "loss": 0.4983,
+      "step": 4523
+    },
+    {
+      "epoch": 4.341814690350456,
+      "grad_norm": 1.2345514418247994,
+      "learning_rate": 5.203692726310738e-07,
+      "loss": 0.4174,
+      "step": 4524
+    },
+    {
+      "epoch": 4.342774843975036,
+      "grad_norm": 1.2788479147916623,
+      "learning_rate": 5.188822156730756e-07,
+      "loss": 0.4548,
+      "step": 4525
+    },
+    {
+      "epoch": 4.343734997599616,
+      "grad_norm": 0.8732961751624204,
+      "learning_rate": 5.173971702409491e-07,
+      "loss": 0.5299,
+      "step": 4526
+    },
+    {
+      "epoch": 4.344695151224196,
+      "grad_norm": 1.232851885938405,
+      "learning_rate": 5.159141370013165e-07,
+      "loss": 0.4358,
+      "step": 4527
+    },
+    {
+      "epoch": 4.345655304848776,
+      "grad_norm": 1.9851579327873001,
+      "learning_rate": 5.144331166198963e-07,
+      "loss": 0.3698,
+      "step": 4528
+    },
+    {
+      "epoch": 4.346615458473356,
+      "grad_norm": 1.1441699347447987,
+      "learning_rate": 5.129541097615009e-07,
+      "loss": 0.4816,
+      "step": 4529
+    },
+    {
+      "epoch": 4.347575612097936,
+      "grad_norm": 1.1372585439565508,
+      "learning_rate": 5.114771170900434e-07,
+      "loss": 0.3523,
+      "step": 4530
+    },
+    {
+      "epoch": 4.348535765722516,
+      "grad_norm": 1.1655901493154417,
+      "learning_rate": 5.100021392685267e-07,
+      "loss": 0.4205,
+      "step": 4531
+    },
+    {
+      "epoch": 4.3494959193470955,
+      "grad_norm": 1.5084867324248776,
+      "learning_rate": 5.08529176959055e-07,
+      "loss": 0.4414,
+      "step": 4532
+    },
+    {
+      "epoch": 4.350456072971675,
+      "grad_norm": 1.3280365995889196,
+      "learning_rate": 5.070582308228256e-07,
+      "loss": 0.3429,
+      "step": 4533
+    },
+    {
+      "epoch": 4.351416226596255,
+      "grad_norm": 1.074301942160981,
+      "learning_rate": 5.05589301520128e-07,
+      "loss": 0.4907,
+      "step": 4534
+    },
+    {
+      "epoch": 4.352376380220836,
+      "grad_norm": 1.2007583916027926,
+      "learning_rate": 5.041223897103509e-07,
+      "loss": 0.4296,
+      "step": 4535
+    },
+    {
+      "epoch": 4.3533365338454155,
+      "grad_norm": 0.8454902264937025,
+      "learning_rate": 5.026574960519747e-07,
+      "loss": 0.4284,
+      "step": 4536
+    },
+    {
+      "epoch": 4.354296687469995,
+      "grad_norm": 1.5634155969774313,
+      "learning_rate": 5.011946212025747e-07,
+      "loss": 0.4287,
+      "step": 4537
+    },
+    {
+      "epoch": 4.355256841094575,
+      "grad_norm": 1.2109126206771892,
+      "learning_rate": 4.997337658188207e-07,
+      "loss": 0.3732,
+      "step": 4538
+    },
+    {
+      "epoch": 4.356216994719155,
+      "grad_norm": 0.9904431828162874,
+      "learning_rate": 4.982749305564727e-07,
+      "loss": 0.4272,
+      "step": 4539
+    },
+    {
+      "epoch": 4.357177148343735,
+      "grad_norm": 0.9979274217998053,
+      "learning_rate": 4.968181160703877e-07,
+      "loss": 0.4326,
+      "step": 4540
+    },
+    {
+      "epoch": 4.3581373019683145,
+      "grad_norm": 1.0312499466279428,
+      "learning_rate": 4.953633230145144e-07,
+      "loss": 0.2995,
+      "step": 4541
+    },
+    {
+      "epoch": 4.359097455592895,
+      "grad_norm": 0.7224938097594399,
+      "learning_rate": 4.93910552041893e-07,
+      "loss": 0.372,
+      "step": 4542
+    },
+    {
+      "epoch": 4.360057609217475,
+      "grad_norm": 1.0709144019009873,
+      "learning_rate": 4.924598038046591e-07,
+      "loss": 0.397,
+      "step": 4543
+    },
+    {
+      "epoch": 4.361017762842055,
+      "grad_norm": 2.151572348900691,
+      "learning_rate": 4.910110789540346e-07,
+      "loss": 0.4301,
+      "step": 4544
+    },
+    {
+      "epoch": 4.361977916466635,
+      "grad_norm": 1.1224985143528168,
+      "learning_rate": 4.895643781403375e-07,
+      "loss": 0.4491,
+      "step": 4545
+    },
+    {
+      "epoch": 4.362938070091214,
+      "grad_norm": 1.2376849815327193,
+      "learning_rate": 4.881197020129797e-07,
+      "loss": 0.4396,
+      "step": 4546
+    },
+    {
+      "epoch": 4.363898223715794,
+      "grad_norm": 1.1241381462737696,
+      "learning_rate": 4.866770512204566e-07,
+      "loss": 0.4283,
+      "step": 4547
+    },
+    {
+      "epoch": 4.364858377340375,
+      "grad_norm": 1.0927577584920967,
+      "learning_rate": 4.852364264103621e-07,
+      "loss": 0.386,
+      "step": 4548
+    },
+    {
+      "epoch": 4.365818530964955,
+      "grad_norm": 0.9447907262807417,
+      "learning_rate": 4.837978282293731e-07,
+      "loss": 0.4059,
+      "step": 4549
+    },
+    {
+      "epoch": 4.366778684589534,
+      "grad_norm": 1.0765828788806624,
+      "learning_rate": 4.823612573232644e-07,
+      "loss": 0.4517,
+      "step": 4550
+    },
+    {
+      "epoch": 4.367738838214114,
+      "grad_norm": 1.1052584134893906,
+      "learning_rate": 4.809267143368979e-07,
+      "loss": 0.5089,
+      "step": 4551
+    },
+    {
+      "epoch": 4.368698991838694,
+      "grad_norm": 1.376233074781487,
+      "learning_rate": 4.794941999142222e-07,
+      "loss": 0.4837,
+      "step": 4552
+    },
+    {
+      "epoch": 4.369659145463274,
+      "grad_norm": 1.1806096422823795,
+      "learning_rate": 4.780637146982781e-07,
+      "loss": 0.3811,
+      "step": 4553
+    },
+    {
+      "epoch": 4.3706192990878545,
+      "grad_norm": 1.1241669745360072,
+      "learning_rate": 4.7663525933119625e-07,
+      "loss": 0.4013,
+      "step": 4554
+    },
+    {
+      "epoch": 4.371579452712434,
+      "grad_norm": 1.3671134807528549,
+      "learning_rate": 4.7520883445419387e-07,
+      "loss": 0.4703,
+      "step": 4555
+    },
+    {
+      "epoch": 4.372539606337014,
+      "grad_norm": 0.9831392593229334,
+      "learning_rate": 4.737844407075798e-07,
+      "loss": 0.4268,
+      "step": 4556
+    },
+    {
+      "epoch": 4.373499759961594,
+      "grad_norm": 1.2104135913434775,
+      "learning_rate": 4.723620787307465e-07,
+      "loss": 0.4959,
+      "step": 4557
+    },
+    {
+      "epoch": 4.374459913586174,
+      "grad_norm": 1.4830952674130176,
+      "learning_rate": 4.7094174916217817e-07,
+      "loss": 0.4121,
+      "step": 4558
+    },
+    {
+      "epoch": 4.3754200672107535,
+      "grad_norm": 1.2314396640248406,
+      "learning_rate": 4.69523452639446e-07,
+      "loss": 0.4283,
+      "step": 4559
+    },
+    {
+      "epoch": 4.376380220835333,
+      "grad_norm": 1.2594440126269677,
+      "learning_rate": 4.6810718979920734e-07,
+      "loss": 0.551,
+      "step": 4560
+    },
+    {
+      "epoch": 4.377340374459914,
+      "grad_norm": 1.093121773581581,
+      "learning_rate": 4.666929612772092e-07,
+      "loss": 0.4493,
+      "step": 4561
+    },
+    {
+      "epoch": 4.378300528084494,
+      "grad_norm": 1.137856047787528,
+      "learning_rate": 4.6528076770828045e-07,
+      "loss": 0.5164,
+      "step": 4562
+    },
+    {
+      "epoch": 4.379260681709074,
+      "grad_norm": 0.9180310069138364,
+      "learning_rate": 4.638706097263429e-07,
+      "loss": 0.3896,
+      "step": 4563
+    },
+    {
+      "epoch": 4.380220835333653,
+      "grad_norm": 1.0180404701188983,
+      "learning_rate": 4.6246248796440074e-07,
+      "loss": 0.382,
+      "step": 4564
+    },
+    {
+      "epoch": 4.381180988958233,
+      "grad_norm": 1.5199235971437013,
+      "learning_rate": 4.6105640305454337e-07,
+      "loss": 0.4492,
+      "step": 4565
+    },
+    {
+      "epoch": 4.382141142582813,
+      "grad_norm": 1.0314532935913057,
+      "learning_rate": 4.5965235562794754e-07,
+      "loss": 0.4477,
+      "step": 4566
+    },
+    {
+      "epoch": 4.383101296207393,
+      "grad_norm": 1.064903726380619,
+      "learning_rate": 4.582503463148763e-07,
+      "loss": 0.5554,
+      "step": 4567
+    },
+    {
+      "epoch": 4.384061449831973,
+      "grad_norm": 1.5702707118705925,
+      "learning_rate": 4.5685037574467574e-07,
+      "loss": 0.3958,
+      "step": 4568
+    },
+    {
+      "epoch": 4.385021603456553,
+      "grad_norm": 1.5379656344648542,
+      "learning_rate": 4.5545244454577863e-07,
+      "loss": 0.3598,
+      "step": 4569
+    },
+    {
+      "epoch": 4.385981757081133,
+      "grad_norm": 0.8654657383084474,
+      "learning_rate": 4.540565533457003e-07,
+      "loss": 0.4933,
+      "step": 4570
+    },
+    {
+      "epoch": 4.386941910705713,
+      "grad_norm": 1.7283784967091642,
+      "learning_rate": 4.5266270277104116e-07,
+      "loss": 0.3767,
+      "step": 4571
+    },
+    {
+      "epoch": 4.387902064330293,
+      "grad_norm": 1.2237399318454918,
+      "learning_rate": 4.512708934474863e-07,
+      "loss": 0.4854,
+      "step": 4572
+    },
+    {
+      "epoch": 4.388862217954872,
+      "grad_norm": 0.9246663696882009,
+      "learning_rate": 4.498811259998048e-07,
+      "loss": 0.5199,
+      "step": 4573
+    },
+    {
+      "epoch": 4.389822371579453,
+      "grad_norm": 0.9660510146192445,
+      "learning_rate": 4.4849340105184823e-07,
+      "loss": 0.5644,
+      "step": 4574
+    },
+    {
+      "epoch": 4.390782525204033,
+      "grad_norm": 0.9728257023773027,
+      "learning_rate": 4.4710771922654936e-07,
+      "loss": 0.4811,
+      "step": 4575
+    },
+    {
+      "epoch": 4.391742678828613,
+      "grad_norm": 0.9199310882092127,
+      "learning_rate": 4.4572408114592845e-07,
+      "loss": 0.4305,
+      "step": 4576
+    },
+    {
+      "epoch": 4.3927028324531925,
+      "grad_norm": 0.9779943800547182,
+      "learning_rate": 4.4434248743108587e-07,
+      "loss": 0.4469,
+      "step": 4577
+    },
+    {
+      "epoch": 4.393662986077772,
+      "grad_norm": 1.1789354200193212,
+      "learning_rate": 4.4296293870220276e-07,
+      "loss": 0.462,
+      "step": 4578
+    },
+    {
+      "epoch": 4.394623139702352,
+      "grad_norm": 0.9951559345660517,
+      "learning_rate": 4.415854355785443e-07,
+      "loss": 0.399,
+      "step": 4579
+    },
+    {
+      "epoch": 4.395583293326933,
+      "grad_norm": 0.9347679612202262,
+      "learning_rate": 4.4020997867845695e-07,
+      "loss": 0.4685,
+      "step": 4580
+    },
+    {
+      "epoch": 4.396543446951513,
+      "grad_norm": 1.40442396082108,
+      "learning_rate": 4.388365686193685e-07,
+      "loss": 0.4198,
+      "step": 4581
+    },
+    {
+      "epoch": 4.397503600576092,
+      "grad_norm": 1.1920837031436664,
+      "learning_rate": 4.3746520601778974e-07,
+      "loss": 0.452,
+      "step": 4582
+    },
+    {
+      "epoch": 4.398463754200672,
+      "grad_norm": 0.8465765408725517,
+      "learning_rate": 4.3609589148930766e-07,
+      "loss": 0.3935,
+      "step": 4583
+    },
+    {
+      "epoch": 4.399423907825252,
+      "grad_norm": 1.0887085174734115,
+      "learning_rate": 4.3472862564859496e-07,
+      "loss": 0.5193,
+      "step": 4584
+    },
+    {
+      "epoch": 4.400384061449832,
+      "grad_norm": 0.9376187730327918,
+      "learning_rate": 4.3336340910940186e-07,
+      "loss": 0.3983,
+      "step": 4585
+    },
+    {
+      "epoch": 4.4013442150744115,
+      "grad_norm": 1.0778029589212026,
+      "learning_rate": 4.320002424845604e-07,
+      "loss": 0.407,
+      "step": 4586
+    },
+    {
+      "epoch": 4.402304368698992,
+      "grad_norm": 1.2944301989546776,
+      "learning_rate": 4.3063912638598106e-07,
+      "loss": 0.3974,
+      "step": 4587
+    },
+    {
+      "epoch": 4.403264522323572,
+      "grad_norm": 1.031807254545532,
+      "learning_rate": 4.292800614246534e-07,
+      "loss": 0.4133,
+      "step": 4588
+    },
+    {
+      "epoch": 4.404224675948152,
+      "grad_norm": 0.8019802099081091,
+      "learning_rate": 4.2792304821064833e-07,
+      "loss": 0.448,
+      "step": 4589
+    },
+    {
+      "epoch": 4.405184829572732,
+      "grad_norm": 1.170700206301413,
+      "learning_rate": 4.265680873531136e-07,
+      "loss": 0.3703,
+      "step": 4590
+    },
+    {
+      "epoch": 4.406144983197311,
+      "grad_norm": 1.059936086357254,
+      "learning_rate": 4.2521517946027715e-07,
+      "loss": 0.4626,
+      "step": 4591
+    },
+    {
+      "epoch": 4.407105136821891,
+      "grad_norm": 1.0063863096673182,
+      "learning_rate": 4.2386432513944587e-07,
+      "loss": 0.5266,
+      "step": 4592
+    },
+    {
+      "epoch": 4.408065290446471,
+      "grad_norm": 0.9924005972054886,
+      "learning_rate": 4.225155249970014e-07,
+      "loss": 0.4086,
+      "step": 4593
+    },
+    {
+      "epoch": 4.409025444071052,
+      "grad_norm": 1.0385329428389618,
+      "learning_rate": 4.2116877963840786e-07,
+      "loss": 0.4435,
+      "step": 4594
+    },
+    {
+      "epoch": 4.4099855976956315,
+      "grad_norm": 1.037936602005678,
+      "learning_rate": 4.198240896682049e-07,
+      "loss": 0.4398,
+      "step": 4595
+    },
+    {
+      "epoch": 4.410945751320211,
+      "grad_norm": 1.6847855823495415,
+      "learning_rate": 4.18481455690008e-07,
+      "loss": 0.3883,
+      "step": 4596
+    },
+    {
+      "epoch": 4.411905904944791,
+      "grad_norm": 1.2269550769225495,
+      "learning_rate": 4.1714087830651226e-07,
+      "loss": 0.3814,
+      "step": 4597
+    },
+    {
+      "epoch": 4.412866058569371,
+      "grad_norm": 1.4977485360990013,
+      "learning_rate": 4.1580235811948753e-07,
+      "loss": 0.452,
+      "step": 4598
+    },
+    {
+      "epoch": 4.413826212193951,
+      "grad_norm": 1.1336189339111895,
+      "learning_rate": 4.14465895729782e-07,
+      "loss": 0.3192,
+      "step": 4599
+    },
+    {
+      "epoch": 4.414786365818531,
+      "grad_norm": 1.0160900366957863,
+      "learning_rate": 4.131314917373208e-07,
+      "loss": 0.4128,
+      "step": 4600
+    },
+    {
+      "epoch": 4.415746519443111,
+      "grad_norm": 0.9084506981875303,
+      "learning_rate": 4.1179914674110035e-07,
+      "loss": 0.5587,
+      "step": 4601
+    },
+    {
+      "epoch": 4.416706673067691,
+      "grad_norm": 1.0202889906469483,
+      "learning_rate": 4.1046886133919825e-07,
+      "loss": 0.3186,
+      "step": 4602
+    },
+    {
+      "epoch": 4.417666826692271,
+      "grad_norm": 1.184635343110861,
+      "learning_rate": 4.0914063612876473e-07,
+      "loss": 0.2861,
+      "step": 4603
+    },
+    {
+      "epoch": 4.4186269803168505,
+      "grad_norm": 1.0105819774998916,
+      "learning_rate": 4.0781447170602617e-07,
+      "loss": 0.4878,
+      "step": 4604
+    },
+    {
+      "epoch": 4.41958713394143,
+      "grad_norm": 0.916800302405071,
+      "learning_rate": 4.064903686662841e-07,
+      "loss": 0.4673,
+      "step": 4605
+    },
+    {
+      "epoch": 4.420547287566011,
+      "grad_norm": 1.241109346428586,
+      "learning_rate": 4.0516832760391256e-07,
+      "loss": 0.4678,
+      "step": 4606
+    },
+    {
+      "epoch": 4.421507441190591,
+      "grad_norm": 1.07599400529569,
+      "learning_rate": 4.0384834911236294e-07,
+      "loss": 0.3828,
+      "step": 4607
+    },
+    {
+      "epoch": 4.422467594815171,
+      "grad_norm": 1.1164020728168926,
+      "learning_rate": 4.025304337841607e-07,
+      "loss": 0.4583,
+      "step": 4608
+    },
+    {
+      "epoch": 4.42342774843975,
+      "grad_norm": 1.198347211801293,
+      "learning_rate": 4.012145822109015e-07,
+      "loss": 0.35,
+      "step": 4609
+    },
+    {
+      "epoch": 4.42438790206433,
+      "grad_norm": 0.9597224092712009,
+      "learning_rate": 3.9990079498325786e-07,
+      "loss": 0.4111,
+      "step": 4610
+    },
+    {
+      "epoch": 4.42534805568891,
+      "grad_norm": 1.0985588341273547,
+      "learning_rate": 3.985890726909747e-07,
+      "loss": 0.5562,
+      "step": 4611
+    },
+    {
+      "epoch": 4.42630820931349,
+      "grad_norm": 1.1466011011349508,
+      "learning_rate": 3.97279415922871e-07,
+      "loss": 0.4417,
+      "step": 4612
+    },
+    {
+      "epoch": 4.4272683629380705,
+      "grad_norm": 1.2435141211321377,
+      "learning_rate": 3.9597182526683763e-07,
+      "loss": 0.3718,
+      "step": 4613
+    },
+    {
+      "epoch": 4.42822851656265,
+      "grad_norm": 0.9479447933179591,
+      "learning_rate": 3.946663013098373e-07,
+      "loss": 0.3907,
+      "step": 4614
+    },
+    {
+      "epoch": 4.42918867018723,
+      "grad_norm": 1.2414848856718508,
+      "learning_rate": 3.9336284463790566e-07,
+      "loss": 0.3018,
+      "step": 4615
+    },
+    {
+      "epoch": 4.43014882381181,
+      "grad_norm": 1.2030546961756063,
+      "learning_rate": 3.9206145583615076e-07,
+      "loss": 0.3973,
+      "step": 4616
+    },
+    {
+      "epoch": 4.43110897743639,
+      "grad_norm": 1.081223155803227,
+      "learning_rate": 3.90762135488752e-07,
+      "loss": 0.5062,
+      "step": 4617
+    },
+    {
+      "epoch": 4.4320691310609694,
+      "grad_norm": 1.0381226103351264,
+      "learning_rate": 3.894648841789617e-07,
+      "loss": 0.5091,
+      "step": 4618
+    },
+    {
+      "epoch": 4.433029284685549,
+      "grad_norm": 2.100859562932593,
+      "learning_rate": 3.881697024890996e-07,
+      "loss": 0.4231,
+      "step": 4619
+    },
+    {
+      "epoch": 4.43398943831013,
+      "grad_norm": 1.0535313496527898,
+      "learning_rate": 3.8687659100056063e-07,
+      "loss": 0.537,
+      "step": 4620
+    },
+    {
+      "epoch": 4.43494959193471,
+      "grad_norm": 0.9168189758471965,
+      "learning_rate": 3.8558555029380937e-07,
+      "loss": 0.4552,
+      "step": 4621
+    },
+    {
+      "epoch": 4.4359097455592895,
+      "grad_norm": 0.8307751269320371,
+      "learning_rate": 3.8429658094837896e-07,
+      "loss": 0.3949,
+      "step": 4622
+    },
+    {
+      "epoch": 4.436869899183869,
+      "grad_norm": 1.0567023738650025,
+      "learning_rate": 3.830096835428737e-07,
+      "loss": 0.4792,
+      "step": 4623
+    },
+    {
+      "epoch": 4.437830052808449,
+      "grad_norm": 1.0118897659015862,
+      "learning_rate": 3.8172485865496876e-07,
+      "loss": 0.4811,
+      "step": 4624
+    },
+    {
+      "epoch": 4.438790206433029,
+      "grad_norm": 1.0944763064896719,
+      "learning_rate": 3.8044210686140784e-07,
+      "loss": 0.3227,
+      "step": 4625
+    },
+    {
+      "epoch": 4.43975036005761,
+      "grad_norm": 0.8597578977294977,
+      "learning_rate": 3.791614287380063e-07,
+      "loss": 0.4486,
+      "step": 4626
+    },
+    {
+      "epoch": 4.440710513682189,
+      "grad_norm": 1.3280114692241722,
+      "learning_rate": 3.778828248596439e-07,
+      "loss": 0.3981,
+      "step": 4627
+    },
+    {
+      "epoch": 4.441670667306769,
+      "grad_norm": 0.9824963134424053,
+      "learning_rate": 3.766062958002742e-07,
+      "loss": 0.4785,
+      "step": 4628
+    },
+    {
+      "epoch": 4.442630820931349,
+      "grad_norm": 1.0269967677537446,
+      "learning_rate": 3.7533184213291665e-07,
+      "loss": 0.5201,
+      "step": 4629
+    },
+    {
+      "epoch": 4.443590974555929,
+      "grad_norm": 1.2035620958243223,
+      "learning_rate": 3.7405946442966023e-07,
+      "loss": 0.4753,
+      "step": 4630
+    },
+    {
+      "epoch": 4.444551128180509,
+      "grad_norm": 0.9811362135399778,
+      "learning_rate": 3.727891632616626e-07,
+      "loss": 0.4886,
+      "step": 4631
+    },
+    {
+      "epoch": 4.445511281805089,
+      "grad_norm": 1.1164479040761177,
+      "learning_rate": 3.7152093919914644e-07,
+      "loss": 0.3473,
+      "step": 4632
+    },
+    {
+      "epoch": 4.446471435429669,
+      "grad_norm": 1.303352400684884,
+      "learning_rate": 3.702547928114053e-07,
+      "loss": 0.468,
+      "step": 4633
+    },
+    {
+      "epoch": 4.447431589054249,
+      "grad_norm": 0.8755270873828765,
+      "learning_rate": 3.689907246667984e-07,
+      "loss": 0.3892,
+      "step": 4634
+    },
+    {
+      "epoch": 4.448391742678829,
+      "grad_norm": 0.9283241839457869,
+      "learning_rate": 3.677287353327519e-07,
+      "loss": 0.4845,
+      "step": 4635
+    },
+    {
+      "epoch": 4.449351896303408,
+      "grad_norm": 1.049123002332921,
+      "learning_rate": 3.6646882537576144e-07,
+      "loss": 0.3652,
+      "step": 4636
+    },
+    {
+      "epoch": 4.450312049927988,
+      "grad_norm": 0.9071035520101988,
+      "learning_rate": 3.652109953613836e-07,
+      "loss": 0.401,
+      "step": 4637
+    },
+    {
+      "epoch": 4.451272203552568,
+      "grad_norm": 0.9180496356241947,
+      "learning_rate": 3.639552458542478e-07,
+      "loss": 0.4459,
+      "step": 4638
+    },
+    {
+      "epoch": 4.452232357177149,
+      "grad_norm": 1.1325922770177368,
+      "learning_rate": 3.627015774180459e-07,
+      "loss": 0.4246,
+      "step": 4639
+    },
+    {
+      "epoch": 4.4531925108017285,
+      "grad_norm": 1.1073477887363432,
+      "learning_rate": 3.614499906155355e-07,
+      "loss": 0.4852,
+      "step": 4640
+    },
+    {
+      "epoch": 4.454152664426308,
+      "grad_norm": 0.9385430408233402,
+      "learning_rate": 3.602004860085406e-07,
+      "loss": 0.4651,
+      "step": 4641
+    },
+    {
+      "epoch": 4.455112818050888,
+      "grad_norm": 1.0857351597359235,
+      "learning_rate": 3.589530641579514e-07,
+      "loss": 0.4507,
+      "step": 4642
+    },
+    {
+      "epoch": 4.456072971675468,
+      "grad_norm": 0.8128311834316073,
+      "learning_rate": 3.577077256237216e-07,
+      "loss": 0.4441,
+      "step": 4643
+    },
+    {
+      "epoch": 4.457033125300048,
+      "grad_norm": 1.0808085416470794,
+      "learning_rate": 3.564644709648707e-07,
+      "loss": 0.4415,
+      "step": 4644
+    },
+    {
+      "epoch": 4.4579932789246275,
+      "grad_norm": 1.1883011786961952,
+      "learning_rate": 3.5522330073948177e-07,
+      "loss": 0.3925,
+      "step": 4645
+    },
+    {
+      "epoch": 4.458953432549208,
+      "grad_norm": 1.0792365759954712,
+      "learning_rate": 3.53984215504703e-07,
+      "loss": 0.5069,
+      "step": 4646
+    },
+    {
+      "epoch": 4.459913586173788,
+      "grad_norm": 1.7224546507341325,
+      "learning_rate": 3.5274721581674664e-07,
+      "loss": 0.4679,
+      "step": 4647
+    },
+    {
+      "epoch": 4.460873739798368,
+      "grad_norm": 1.0270505359466444,
+      "learning_rate": 3.5151230223088915e-07,
+      "loss": 0.3891,
+      "step": 4648
+    },
+    {
+      "epoch": 4.461833893422948,
+      "grad_norm": 1.1936722087015554,
+      "learning_rate": 3.5027947530147045e-07,
+      "loss": 0.4193,
+      "step": 4649
+    },
+    {
+      "epoch": 4.462794047047527,
+      "grad_norm": 1.2134318744599935,
+      "learning_rate": 3.4904873558189055e-07,
+      "loss": 0.3959,
+      "step": 4650
+    },
+    {
+      "epoch": 4.463754200672107,
+      "grad_norm": 1.3284386265098087,
+      "learning_rate": 3.4782008362461817e-07,
+      "loss": 0.3829,
+      "step": 4651
+    },
+    {
+      "epoch": 4.464714354296688,
+      "grad_norm": 1.0788159580593186,
+      "learning_rate": 3.4659351998118204e-07,
+      "loss": 0.3785,
+      "step": 4652
+    },
+    {
+      "epoch": 4.465674507921268,
+      "grad_norm": 1.34088185550518,
+      "learning_rate": 3.4536904520217184e-07,
+      "loss": 0.3601,
+      "step": 4653
+    },
+    {
+      "epoch": 4.466634661545847,
+      "grad_norm": 0.9369895407189601,
+      "learning_rate": 3.441466598372423e-07,
+      "loss": 0.4827,
+      "step": 4654
+    },
+    {
+      "epoch": 4.467594815170427,
+      "grad_norm": 1.3212086612814424,
+      "learning_rate": 3.429263644351083e-07,
+      "loss": 0.3879,
+      "step": 4655
+    },
+    {
+      "epoch": 4.468554968795007,
+      "grad_norm": 1.5011069475026733,
+      "learning_rate": 3.417081595435484e-07,
+      "loss": 0.4307,
+      "step": 4656
+    },
+    {
+      "epoch": 4.469515122419587,
+      "grad_norm": 1.0083799795685993,
+      "learning_rate": 3.4049204570940185e-07,
+      "loss": 0.4076,
+      "step": 4657
+    },
+    {
+      "epoch": 4.4704752760441675,
+      "grad_norm": 1.241800592198575,
+      "learning_rate": 3.392780234785681e-07,
+      "loss": 0.4908,
+      "step": 4658
+    },
+    {
+      "epoch": 4.471435429668747,
+      "grad_norm": 0.9537669496623953,
+      "learning_rate": 3.3806609339600894e-07,
+      "loss": 0.4485,
+      "step": 4659
+    },
+    {
+      "epoch": 4.472395583293327,
+      "grad_norm": 1.2162317603348596,
+      "learning_rate": 3.36856256005747e-07,
+      "loss": 0.3165,
+      "step": 4660
+    },
+    {
+      "epoch": 4.473355736917907,
+      "grad_norm": 1.027184108405054,
+      "learning_rate": 3.356485118508657e-07,
+      "loss": 0.507,
+      "step": 4661
+    },
+    {
+      "epoch": 4.474315890542487,
+      "grad_norm": 0.935849857099973,
+      "learning_rate": 3.3444286147350955e-07,
+      "loss": 0.5405,
+      "step": 4662
+    },
+    {
+      "epoch": 4.4752760441670665,
+      "grad_norm": 1.3835524173254776,
+      "learning_rate": 3.332393054148786e-07,
+      "loss": 0.4296,
+      "step": 4663
+    },
+    {
+      "epoch": 4.476236197791646,
+      "grad_norm": 0.8944764702806797,
+      "learning_rate": 3.3203784421524056e-07,
+      "loss": 0.3917,
+      "step": 4664
+    },
+    {
+      "epoch": 4.477196351416227,
+      "grad_norm": 0.9643049038902026,
+      "learning_rate": 3.3083847841391514e-07,
+      "loss": 0.4942,
+      "step": 4665
+    },
+    {
+      "epoch": 4.478156505040807,
+      "grad_norm": 1.1445453507300092,
+      "learning_rate": 3.296412085492867e-07,
+      "loss": 0.3323,
+      "step": 4666
+    },
+    {
+      "epoch": 4.4791166586653866,
+      "grad_norm": 0.9241544707478011,
+      "learning_rate": 3.2844603515879647e-07,
+      "loss": 0.4916,
+      "step": 4667
+    },
+    {
+      "epoch": 4.480076812289966,
+      "grad_norm": 0.9935863549035892,
+      "learning_rate": 3.272529587789447e-07,
+      "loss": 0.5704,
+      "step": 4668
+    },
+    {
+      "epoch": 4.481036965914546,
+      "grad_norm": 0.8865031791214463,
+      "learning_rate": 3.2606197994529076e-07,
+      "loss": 0.4223,
+      "step": 4669
+    },
+    {
+      "epoch": 4.481997119539126,
+      "grad_norm": 1.3736724837829202,
+      "learning_rate": 3.2487309919245313e-07,
+      "loss": 0.2838,
+      "step": 4670
+    },
+    {
+      "epoch": 4.482957273163706,
+      "grad_norm": 1.3502526512112645,
+      "learning_rate": 3.236863170541066e-07,
+      "loss": 0.4459,
+      "step": 4671
+    },
+    {
+      "epoch": 4.483917426788286,
+      "grad_norm": 1.0425731557530016,
+      "learning_rate": 3.22501634062985e-07,
+      "loss": 0.372,
+      "step": 4672
+    },
+    {
+      "epoch": 4.484877580412866,
+      "grad_norm": 1.0981722570117114,
+      "learning_rate": 3.213190507508801e-07,
+      "loss": 0.4103,
+      "step": 4673
+    },
+    {
+      "epoch": 4.485837734037446,
+      "grad_norm": 0.9064090652531295,
+      "learning_rate": 3.2013856764864126e-07,
+      "loss": 0.423,
+      "step": 4674
+    },
+    {
+      "epoch": 4.486797887662026,
+      "grad_norm": 1.3331524416886484,
+      "learning_rate": 3.1896018528617456e-07,
+      "loss": 0.4736,
+      "step": 4675
+    },
+    {
+      "epoch": 4.487758041286606,
+      "grad_norm": 1.3950619923604852,
+      "learning_rate": 3.17783904192443e-07,
+      "loss": 0.4697,
+      "step": 4676
+    },
+    {
+      "epoch": 4.488718194911185,
+      "grad_norm": 1.356830743593381,
+      "learning_rate": 3.166097248954669e-07,
+      "loss": 0.3645,
+      "step": 4677
+    },
+    {
+      "epoch": 4.489678348535766,
+      "grad_norm": 0.9845654489031277,
+      "learning_rate": 3.15437647922322e-07,
+      "loss": 0.4774,
+      "step": 4678
+    },
+    {
+      "epoch": 4.490638502160346,
+      "grad_norm": 1.502083952953334,
+      "learning_rate": 3.1426767379914234e-07,
+      "loss": 0.3804,
+      "step": 4679
+    },
+    {
+      "epoch": 4.491598655784926,
+      "grad_norm": 0.933278485317963,
+      "learning_rate": 3.1309980305111674e-07,
+      "loss": 0.4455,
+      "step": 4680
+    },
+    {
+      "epoch": 4.4925588094095055,
+      "grad_norm": 0.9574803410164137,
+      "learning_rate": 3.119340362024875e-07,
+      "loss": 0.3957,
+      "step": 4681
+    },
+    {
+      "epoch": 4.493518963034085,
+      "grad_norm": 1.1913148882079458,
+      "learning_rate": 3.1077037377655827e-07,
+      "loss": 0.4653,
+      "step": 4682
+    },
+    {
+      "epoch": 4.494479116658665,
+      "grad_norm": 0.9520078309381338,
+      "learning_rate": 3.096088162956834e-07,
+      "loss": 0.4846,
+      "step": 4683
+    },
+    {
+      "epoch": 4.495439270283246,
+      "grad_norm": 1.1173728527292974,
+      "learning_rate": 3.084493642812725e-07,
+      "loss": 0.3963,
+      "step": 4684
+    },
+    {
+      "epoch": 4.4963994239078255,
+      "grad_norm": 1.051503282047858,
+      "learning_rate": 3.072920182537925e-07,
+      "loss": 0.3211,
+      "step": 4685
+    },
+    {
+      "epoch": 4.497359577532405,
+      "grad_norm": 1.0697288583757516,
+      "learning_rate": 3.0613677873276293e-07,
+      "loss": 0.4574,
+      "step": 4686
+    },
+    {
+      "epoch": 4.498319731156985,
+      "grad_norm": 0.9153817193755083,
+      "learning_rate": 3.049836462367589e-07,
+      "loss": 0.508,
+      "step": 4687
+    },
+    {
+      "epoch": 4.499279884781565,
+      "grad_norm": 1.0656737460462218,
+      "learning_rate": 3.038326212834103e-07,
+      "loss": 0.4427,
+      "step": 4688
+    },
+    {
+      "epoch": 4.500240038406145,
+      "grad_norm": 1.5554230690225583,
+      "learning_rate": 3.0268370438939767e-07,
+      "loss": 0.4265,
+      "step": 4689
+    },
+    {
+      "epoch": 4.5012001920307245,
+      "grad_norm": 0.9826281861393134,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.3888,
+      "step": 4690
+    },
+    {
+      "epoch": 4.502160345655305,
+      "grad_norm": 0.7502096687581634,
+      "learning_rate": 3.003921968413831e-07,
+      "loss": 0.4591,
+      "step": 4691
+    },
+    {
+      "epoch": 4.503120499279885,
+      "grad_norm": 1.3632152610718085,
+      "learning_rate": 2.9924960721601393e-07,
+      "loss": 0.3862,
+      "step": 4692
+    },
+    {
+      "epoch": 4.504080652904465,
+      "grad_norm": 1.283898241038968,
+      "learning_rate": 2.98109127707249e-07,
+      "loss": 0.3637,
+      "step": 4693
+    },
+    {
+      "epoch": 4.505040806529045,
+      "grad_norm": 1.1141174564795169,
+      "learning_rate": 2.9697075882703486e-07,
+      "loss": 0.4876,
+      "step": 4694
+    },
+    {
+      "epoch": 4.506000960153624,
+      "grad_norm": 1.4629851634358693,
+      "learning_rate": 2.95834501086375e-07,
+      "loss": 0.4729,
+      "step": 4695
+    },
+    {
+      "epoch": 4.506961113778204,
+      "grad_norm": 1.1765119852314543,
+      "learning_rate": 2.9470035499532425e-07,
+      "loss": 0.5148,
+      "step": 4696
+    },
+    {
+      "epoch": 4.507921267402784,
+      "grad_norm": 1.2243181961508938,
+      "learning_rate": 2.9356832106298584e-07,
+      "loss": 0.3447,
+      "step": 4697
+    },
+    {
+      "epoch": 4.508881421027365,
+      "grad_norm": 1.170493090676355,
+      "learning_rate": 2.9243839979751943e-07,
+      "loss": 0.4483,
+      "step": 4698
+    },
+    {
+      "epoch": 4.5098415746519445,
+      "grad_norm": 1.0775753225959048,
+      "learning_rate": 2.913105917061343e-07,
+      "loss": 0.4165,
+      "step": 4699
+    },
+    {
+      "epoch": 4.510801728276524,
+      "grad_norm": 0.9510803244993737,
+      "learning_rate": 2.9018489729509103e-07,
+      "loss": 0.4929,
+      "step": 4700
+    },
+    {
+      "epoch": 4.511761881901104,
+      "grad_norm": 0.8319880229367609,
+      "learning_rate": 2.8906131706970375e-07,
+      "loss": 0.5482,
+      "step": 4701
+    },
+    {
+      "epoch": 4.512722035525684,
+      "grad_norm": 0.8943703652143729,
+      "learning_rate": 2.879398515343329e-07,
+      "loss": 0.4644,
+      "step": 4702
+    },
+    {
+      "epoch": 4.5136821891502645,
+      "grad_norm": 1.5975482621957606,
+      "learning_rate": 2.8682050119239357e-07,
+      "loss": 0.4139,
+      "step": 4703
+    },
+    {
+      "epoch": 4.514642342774844,
+      "grad_norm": 1.3617928405004596,
+      "learning_rate": 2.857032665463499e-07,
+      "loss": 0.4734,
+      "step": 4704
+    },
+    {
+      "epoch": 4.515602496399424,
+      "grad_norm": 1.5159592096243613,
+      "learning_rate": 2.8458814809771694e-07,
+      "loss": 0.4307,
+      "step": 4705
+    },
+    {
+      "epoch": 4.516562650024004,
+      "grad_norm": 0.7069287483087185,
+      "learning_rate": 2.834751463470603e-07,
+      "loss": 0.4155,
+      "step": 4706
+    },
+    {
+      "epoch": 4.517522803648584,
+      "grad_norm": 1.1466902052701584,
+      "learning_rate": 2.823642617939926e-07,
+      "loss": 0.3466,
+      "step": 4707
+    },
+    {
+      "epoch": 4.5184829572731635,
+      "grad_norm": 1.14303972503422,
+      "learning_rate": 2.8125549493717984e-07,
+      "loss": 0.4133,
+      "step": 4708
+    },
+    {
+      "epoch": 4.519443110897743,
+      "grad_norm": 0.9522121915821085,
+      "learning_rate": 2.8014884627433404e-07,
+      "loss": 0.5797,
+      "step": 4709
+    },
+    {
+      "epoch": 4.520403264522324,
+      "grad_norm": 1.5584119740143225,
+      "learning_rate": 2.7904431630221937e-07,
+      "loss": 0.3904,
+      "step": 4710
+    },
+    {
+      "epoch": 4.521363418146904,
+      "grad_norm": 0.9210416961243966,
+      "learning_rate": 2.7794190551664604e-07,
+      "loss": 0.3868,
+      "step": 4711
+    },
+    {
+      "epoch": 4.522323571771484,
+      "grad_norm": 1.3442494699352647,
+      "learning_rate": 2.7684161441247584e-07,
+      "loss": 0.4929,
+      "step": 4712
+    },
+    {
+      "epoch": 4.523283725396063,
+      "grad_norm": 0.8237913680261635,
+      "learning_rate": 2.75743443483617e-07,
+      "loss": 0.3893,
+      "step": 4713
+    },
+    {
+      "epoch": 4.524243879020643,
+      "grad_norm": 0.9804219338583083,
+      "learning_rate": 2.746473932230281e-07,
+      "loss": 0.4672,
+      "step": 4714
+    },
+    {
+      "epoch": 4.525204032645223,
+      "grad_norm": 0.9146319451455552,
+      "learning_rate": 2.7355346412271223e-07,
+      "loss": 0.4603,
+      "step": 4715
+    },
+    {
+      "epoch": 4.526164186269803,
+      "grad_norm": 1.180443914481596,
+      "learning_rate": 2.7246165667372314e-07,
+      "loss": 0.4453,
+      "step": 4716
+    },
+    {
+      "epoch": 4.5271243398943835,
+      "grad_norm": 1.0442443022965664,
+      "learning_rate": 2.713719713661617e-07,
+      "loss": 0.3597,
+      "step": 4717
+    },
+    {
+      "epoch": 4.528084493518963,
+      "grad_norm": 0.9389450189934472,
+      "learning_rate": 2.702844086891765e-07,
+      "loss": 0.473,
+      "step": 4718
+    },
+    {
+      "epoch": 4.529044647143543,
+      "grad_norm": 1.1429502823703277,
+      "learning_rate": 2.6919896913096264e-07,
+      "loss": 0.4946,
+      "step": 4719
+    },
+    {
+      "epoch": 4.530004800768123,
+      "grad_norm": 1.285427504509259,
+      "learning_rate": 2.6811565317876154e-07,
+      "loss": 0.3764,
+      "step": 4720
+    },
+    {
+      "epoch": 4.530964954392703,
+      "grad_norm": 1.0897933396641855,
+      "learning_rate": 2.67034461318863e-07,
+      "loss": 0.4039,
+      "step": 4721
+    },
+    {
+      "epoch": 4.531925108017282,
+      "grad_norm": 1.0615716704044484,
+      "learning_rate": 2.659553940366016e-07,
+      "loss": 0.4663,
+      "step": 4722
+    },
+    {
+      "epoch": 4.532885261641862,
+      "grad_norm": 1.4840624933597613,
+      "learning_rate": 2.6487845181635994e-07,
+      "loss": 0.4214,
+      "step": 4723
+    },
+    {
+      "epoch": 4.533845415266443,
+      "grad_norm": 1.33164983121705,
+      "learning_rate": 2.638036351415668e-07,
+      "loss": 0.4708,
+      "step": 4724
+    },
+    {
+      "epoch": 4.534805568891023,
+      "grad_norm": 1.1187586323612524,
+      "learning_rate": 2.627309444946929e-07,
+      "loss": 0.4748,
+      "step": 4725
+    },
+    {
+      "epoch": 4.5357657225156025,
+      "grad_norm": 1.31466900113043,
+      "learning_rate": 2.6166038035725984e-07,
+      "loss": 0.3578,
+      "step": 4726
+    },
+    {
+      "epoch": 4.536725876140182,
+      "grad_norm": 1.4360734524150511,
+      "learning_rate": 2.605919432098331e-07,
+      "loss": 0.3739,
+      "step": 4727
+    },
+    {
+      "epoch": 4.537686029764762,
+      "grad_norm": 1.1550705092419071,
+      "learning_rate": 2.5952563353202064e-07,
+      "loss": 0.483,
+      "step": 4728
+    },
+    {
+      "epoch": 4.538646183389343,
+      "grad_norm": 0.8372637074357283,
+      "learning_rate": 2.584614518024786e-07,
+      "loss": 0.4724,
+      "step": 4729
+    },
+    {
+      "epoch": 4.539606337013923,
+      "grad_norm": 1.6567639736734052,
+      "learning_rate": 2.573993984989065e-07,
+      "loss": 0.3116,
+      "step": 4730
+    },
+    {
+      "epoch": 4.540566490638502,
+      "grad_norm": 1.4289361488618597,
+      "learning_rate": 2.5633947409804907e-07,
+      "loss": 0.3932,
+      "step": 4731
+    },
+    {
+      "epoch": 4.541526644263082,
+      "grad_norm": 1.2129696504945664,
+      "learning_rate": 2.552816790756951e-07,
+      "loss": 0.441,
+      "step": 4732
+    },
+    {
+      "epoch": 4.542486797887662,
+      "grad_norm": 1.2536367999467943,
+      "learning_rate": 2.542260139066771e-07,
+      "loss": 0.4802,
+      "step": 4733
+    },
+    {
+      "epoch": 4.543446951512242,
+      "grad_norm": 1.3447927634295767,
+      "learning_rate": 2.5317247906487153e-07,
+      "loss": 0.3472,
+      "step": 4734
+    },
+    {
+      "epoch": 4.544407105136822,
+      "grad_norm": 1.0619723245560864,
+      "learning_rate": 2.521210750231995e-07,
+      "loss": 0.4028,
+      "step": 4735
+    },
+    {
+      "epoch": 4.545367258761402,
+      "grad_norm": 0.7734107090656136,
+      "learning_rate": 2.510718022536246e-07,
+      "loss": 0.384,
+      "step": 4736
+    },
+    {
+      "epoch": 4.546327412385982,
+      "grad_norm": 1.1036974134222308,
+      "learning_rate": 2.5002466122715505e-07,
+      "loss": 0.4598,
+      "step": 4737
+    },
+    {
+      "epoch": 4.547287566010562,
+      "grad_norm": 1.2462577188622184,
+      "learning_rate": 2.489796524138394e-07,
+      "loss": 0.4144,
+      "step": 4738
+    },
+    {
+      "epoch": 4.548247719635142,
+      "grad_norm": 0.971740232758064,
+      "learning_rate": 2.479367762827739e-07,
+      "loss": 0.4539,
+      "step": 4739
+    },
+    {
+      "epoch": 4.549207873259721,
+      "grad_norm": 1.2436283674341448,
+      "learning_rate": 2.46896033302092e-07,
+      "loss": 0.4322,
+      "step": 4740
+    },
+    {
+      "epoch": 4.550168026884301,
+      "grad_norm": 0.9956978727370558,
+      "learning_rate": 2.4585742393897274e-07,
+      "loss": 0.4363,
+      "step": 4741
+    },
+    {
+      "epoch": 4.551128180508881,
+      "grad_norm": 1.1418167817812876,
+      "learning_rate": 2.44820948659637e-07,
+      "loss": 0.5482,
+      "step": 4742
+    },
+    {
+      "epoch": 4.552088334133462,
+      "grad_norm": 0.9323243316709064,
+      "learning_rate": 2.4378660792934715e-07,
+      "loss": 0.5052,
+      "step": 4743
+    },
+    {
+      "epoch": 4.5530484877580415,
+      "grad_norm": 1.0959351490461227,
+      "learning_rate": 2.4275440221240786e-07,
+      "loss": 0.3751,
+      "step": 4744
+    },
+    {
+      "epoch": 4.554008641382621,
+      "grad_norm": 1.0397603007711043,
+      "learning_rate": 2.4172433197216625e-07,
+      "loss": 0.4415,
+      "step": 4745
+    },
+    {
+      "epoch": 4.554968795007201,
+      "grad_norm": 1.1910560483330195,
+      "learning_rate": 2.4069639767100805e-07,
+      "loss": 0.4517,
+      "step": 4746
+    },
+    {
+      "epoch": 4.555928948631781,
+      "grad_norm": 0.9953672139473719,
+      "learning_rate": 2.39670599770363e-07,
+      "loss": 0.4191,
+      "step": 4747
+    },
+    {
+      "epoch": 4.556889102256361,
+      "grad_norm": 1.0480721588379325,
+      "learning_rate": 2.3864693873070054e-07,
+      "loss": 0.4109,
+      "step": 4748
+    },
+    {
+      "epoch": 4.5578492558809405,
+      "grad_norm": 1.2714793075569117,
+      "learning_rate": 2.376254150115309e-07,
+      "loss": 0.3871,
+      "step": 4749
+    },
+    {
+      "epoch": 4.558809409505521,
+      "grad_norm": 1.1473014772326129,
+      "learning_rate": 2.3660602907140672e-07,
+      "loss": 0.4092,
+      "step": 4750
+    },
+    {
+      "epoch": 4.559769563130101,
+      "grad_norm": 0.9050063687795113,
+      "learning_rate": 2.3558878136791697e-07,
+      "loss": 0.4069,
+      "step": 4751
+    },
+    {
+      "epoch": 4.560729716754681,
+      "grad_norm": 1.1588239489815373,
+      "learning_rate": 2.3457367235769468e-07,
+      "loss": 0.4068,
+      "step": 4752
+    },
+    {
+      "epoch": 4.5616898703792605,
+      "grad_norm": 1.4439982559960651,
+      "learning_rate": 2.335607024964115e-07,
+      "loss": 0.3089,
+      "step": 4753
+    },
+    {
+      "epoch": 4.56265002400384,
+      "grad_norm": 0.8533362222895106,
+      "learning_rate": 2.3254987223877868e-07,
+      "loss": 0.3971,
+      "step": 4754
+    },
+    {
+      "epoch": 4.563610177628421,
+      "grad_norm": 0.8538714424911488,
+      "learning_rate": 2.315411820385477e-07,
+      "loss": 0.4697,
+      "step": 4755
+    },
+    {
+      "epoch": 4.564570331253001,
+      "grad_norm": 0.8850294273129156,
+      "learning_rate": 2.3053463234850637e-07,
+      "loss": 0.4832,
+      "step": 4756
+    },
+    {
+      "epoch": 4.565530484877581,
+      "grad_norm": 1.6876140915223485,
+      "learning_rate": 2.295302236204866e-07,
+      "loss": 0.4635,
+      "step": 4757
+    },
+    {
+      "epoch": 4.56649063850216,
+      "grad_norm": 1.7123006347826433,
+      "learning_rate": 2.2852795630535661e-07,
+      "loss": 0.423,
+      "step": 4758
+    },
+    {
+      "epoch": 4.56745079212674,
+      "grad_norm": 1.2815384269336088,
+      "learning_rate": 2.2752783085302265e-07,
+      "loss": 0.4572,
+      "step": 4759
+    },
+    {
+      "epoch": 4.56841094575132,
+      "grad_norm": 1.2177641284463616,
+      "learning_rate": 2.2652984771243002e-07,
+      "loss": 0.3956,
+      "step": 4760
+    },
+    {
+      "epoch": 4.5693710993759,
+      "grad_norm": 1.0489753068784278,
+      "learning_rate": 2.2553400733156317e-07,
+      "loss": 0.3929,
+      "step": 4761
+    },
+    {
+      "epoch": 4.5703312530004805,
+      "grad_norm": 0.8411894507928341,
+      "learning_rate": 2.2454031015744448e-07,
+      "loss": 0.4813,
+      "step": 4762
+    },
+    {
+      "epoch": 4.57129140662506,
+      "grad_norm": 1.072517352365485,
+      "learning_rate": 2.2354875663613384e-07,
+      "loss": 0.3862,
+      "step": 4763
+    },
+    {
+      "epoch": 4.57225156024964,
+      "grad_norm": 1.0333875763883018,
+      "learning_rate": 2.2255934721272908e-07,
+      "loss": 0.4784,
+      "step": 4764
+    },
+    {
+      "epoch": 4.57321171387422,
+      "grad_norm": 0.7960202222219466,
+      "learning_rate": 2.2157208233136495e-07,
+      "loss": 0.3792,
+      "step": 4765
+    },
+    {
+      "epoch": 4.5741718674988,
+      "grad_norm": 1.1073194282711976,
+      "learning_rate": 2.2058696243521417e-07,
+      "loss": 0.5568,
+      "step": 4766
+    },
+    {
+      "epoch": 4.5751320211233795,
+      "grad_norm": 1.3864135854882582,
+      "learning_rate": 2.196039879664874e-07,
+      "loss": 0.4284,
+      "step": 4767
+    },
+    {
+      "epoch": 4.576092174747959,
+      "grad_norm": 1.3626765182526506,
+      "learning_rate": 2.186231593664312e-07,
+      "loss": 0.5581,
+      "step": 4768
+    },
+    {
+      "epoch": 4.57705232837254,
+      "grad_norm": 1.0512069642900155,
+      "learning_rate": 2.1764447707532722e-07,
+      "loss": 0.4252,
+      "step": 4769
+    },
+    {
+      "epoch": 4.57801248199712,
+      "grad_norm": 0.9769423191254444,
+      "learning_rate": 2.1666794153249792e-07,
+      "loss": 0.4284,
+      "step": 4770
+    },
+    {
+      "epoch": 4.5789726356216995,
+      "grad_norm": 0.8878212987644085,
+      "learning_rate": 2.1569355317629937e-07,
+      "loss": 0.3714,
+      "step": 4771
+    },
+    {
+      "epoch": 4.579932789246279,
+      "grad_norm": 1.2888709486320158,
+      "learning_rate": 2.1472131244412275e-07,
+      "loss": 0.3304,
+      "step": 4772
+    },
+    {
+      "epoch": 4.580892942870859,
+      "grad_norm": 0.8345795881164167,
+      "learning_rate": 2.1375121977239676e-07,
+      "loss": 0.4859,
+      "step": 4773
+    },
+    {
+      "epoch": 4.581853096495439,
+      "grad_norm": 1.20415974038489,
+      "learning_rate": 2.127832755965864e-07,
+      "loss": 0.4474,
+      "step": 4774
+    },
+    {
+      "epoch": 4.582813250120019,
+      "grad_norm": 0.9241071701133369,
+      "learning_rate": 2.118174803511913e-07,
+      "loss": 0.4495,
+      "step": 4775
+    },
+    {
+      "epoch": 4.583773403744599,
+      "grad_norm": 1.3261983159135524,
+      "learning_rate": 2.108538344697464e-07,
+      "loss": 0.4409,
+      "step": 4776
+    },
+    {
+      "epoch": 4.584733557369179,
+      "grad_norm": 1.0490455310698863,
+      "learning_rate": 2.0989233838482225e-07,
+      "loss": 0.386,
+      "step": 4777
+    },
+    {
+      "epoch": 4.585693710993759,
+      "grad_norm": 0.9553392408932545,
+      "learning_rate": 2.0893299252802367e-07,
+      "loss": 0.5054,
+      "step": 4778
+    },
+    {
+      "epoch": 4.586653864618339,
+      "grad_norm": 1.0530642504120253,
+      "learning_rate": 2.0797579732999063e-07,
+      "loss": 0.3779,
+      "step": 4779
+    },
+    {
+      "epoch": 4.587614018242919,
+      "grad_norm": 1.898932999851529,
+      "learning_rate": 2.0702075322039884e-07,
+      "loss": 0.375,
+      "step": 4780
+    },
+    {
+      "epoch": 4.588574171867499,
+      "grad_norm": 1.9203080044725702,
+      "learning_rate": 2.0606786062795702e-07,
+      "loss": 0.3602,
+      "step": 4781
+    },
+    {
+      "epoch": 4.589534325492079,
+      "grad_norm": 0.9058597459243907,
+      "learning_rate": 2.051171199804075e-07,
+      "loss": 0.4355,
+      "step": 4782
+    },
+    {
+      "epoch": 4.590494479116659,
+      "grad_norm": 1.2168556019858547,
+      "learning_rate": 2.0416853170452945e-07,
+      "loss": 0.4622,
+      "step": 4783
+    },
+    {
+      "epoch": 4.591454632741239,
+      "grad_norm": 1.4650957281350552,
+      "learning_rate": 2.032220962261322e-07,
+      "loss": 0.4145,
+      "step": 4784
+    },
+    {
+      "epoch": 4.5924147863658185,
+      "grad_norm": 1.30389333164178,
+      "learning_rate": 2.022778139700615e-07,
+      "loss": 0.4011,
+      "step": 4785
+    },
+    {
+      "epoch": 4.593374939990398,
+      "grad_norm": 1.9758087557470718,
+      "learning_rate": 2.0133568536019498e-07,
+      "loss": 0.4424,
+      "step": 4786
+    },
+    {
+      "epoch": 4.594335093614978,
+      "grad_norm": 1.0121911250520825,
+      "learning_rate": 2.0039571081944431e-07,
+      "loss": 0.4405,
+      "step": 4787
+    },
+    {
+      "epoch": 4.595295247239559,
+      "grad_norm": 1.1254257779385113,
+      "learning_rate": 1.994578907697542e-07,
+      "loss": 0.3679,
+      "step": 4788
+    },
+    {
+      "epoch": 4.5962554008641385,
+      "grad_norm": 1.157821565415064,
+      "learning_rate": 1.985222256321029e-07,
+      "loss": 0.4925,
+      "step": 4789
+    },
+    {
+      "epoch": 4.597215554488718,
+      "grad_norm": 1.1266668040253127,
+      "learning_rate": 1.9758871582649887e-07,
+      "loss": 0.4615,
+      "step": 4790
+    },
+    {
+      "epoch": 4.598175708113298,
+      "grad_norm": 1.1725380655090194,
+      "learning_rate": 1.9665736177198524e-07,
+      "loss": 0.4579,
+      "step": 4791
+    },
+    {
+      "epoch": 4.599135861737878,
+      "grad_norm": 1.0331884953100048,
+      "learning_rate": 1.9572816388663706e-07,
+      "loss": 0.3595,
+      "step": 4792
+    },
+    {
+      "epoch": 4.600096015362458,
+      "grad_norm": 0.8329083570637528,
+      "learning_rate": 1.9480112258756123e-07,
+      "loss": 0.3751,
+      "step": 4793
+    },
+    {
+      "epoch": 4.6010561689870375,
+      "grad_norm": 1.1219679308003967,
+      "learning_rate": 1.9387623829089764e-07,
+      "loss": 0.3678,
+      "step": 4794
+    },
+    {
+      "epoch": 4.602016322611618,
+      "grad_norm": 1.1426578543725208,
+      "learning_rate": 1.929535114118153e-07,
+      "loss": 0.4643,
+      "step": 4795
+    },
+    {
+      "epoch": 4.602976476236198,
+      "grad_norm": 1.839951916892278,
+      "learning_rate": 1.920329423645173e-07,
+      "loss": 0.3925,
+      "step": 4796
+    },
+    {
+      "epoch": 4.603936629860778,
+      "grad_norm": 1.3150107292337425,
+      "learning_rate": 1.9111453156223704e-07,
+      "loss": 0.3704,
+      "step": 4797
+    },
+    {
+      "epoch": 4.604896783485358,
+      "grad_norm": 1.709667630194013,
+      "learning_rate": 1.901982794172391e-07,
+      "loss": 0.3657,
+      "step": 4798
+    },
+    {
+      "epoch": 4.605856937109937,
+      "grad_norm": 1.23638543550137,
+      "learning_rate": 1.8928418634082068e-07,
+      "loss": 0.4846,
+      "step": 4799
+    },
+    {
+      "epoch": 4.606817090734517,
+      "grad_norm": 0.6429848154002672,
+      "learning_rate": 1.8837225274330572e-07,
+      "loss": 0.3755,
+      "step": 4800
+    },
+    {
+      "epoch": 4.607777244359097,
+      "grad_norm": 1.178232230479067,
+      "learning_rate": 1.87462479034054e-07,
+      "loss": 0.5208,
+      "step": 4801
+    },
+    {
+      "epoch": 4.608737397983678,
+      "grad_norm": 1.0899899828529767,
+      "learning_rate": 1.8655486562145275e-07,
+      "loss": 0.476,
+      "step": 4802
+    },
+    {
+      "epoch": 4.6096975516082574,
+      "grad_norm": 1.2574117619853196,
+      "learning_rate": 1.8564941291291883e-07,
+      "loss": 0.4262,
+      "step": 4803
+    },
+    {
+      "epoch": 4.610657705232837,
+      "grad_norm": 0.9061955696481128,
+      "learning_rate": 1.8474612131490098e-07,
+      "loss": 0.4249,
+      "step": 4804
+    },
+    {
+      "epoch": 4.611617858857417,
+      "grad_norm": 1.041083942578629,
+      "learning_rate": 1.8384499123287713e-07,
+      "loss": 0.4676,
+      "step": 4805
+    },
+    {
+      "epoch": 4.612578012481997,
+      "grad_norm": 1.2693179431520776,
+      "learning_rate": 1.8294602307135478e-07,
+      "loss": 0.3909,
+      "step": 4806
+    },
+    {
+      "epoch": 4.6135381661065775,
+      "grad_norm": 1.0226813030762867,
+      "learning_rate": 1.820492172338717e-07,
+      "loss": 0.4818,
+      "step": 4807
+    },
+    {
+      "epoch": 4.614498319731157,
+      "grad_norm": 1.870649909267636,
+      "learning_rate": 1.8115457412299363e-07,
+      "loss": 0.381,
+      "step": 4808
+    },
+    {
+      "epoch": 4.615458473355737,
+      "grad_norm": 0.8616460848106126,
+      "learning_rate": 1.802620941403166e-07,
+      "loss": 0.3836,
+      "step": 4809
+    },
+    {
+      "epoch": 4.616418626980317,
+      "grad_norm": 1.2083732396771987,
+      "learning_rate": 1.7937177768646508e-07,
+      "loss": 0.4336,
+      "step": 4810
+    },
+    {
+      "epoch": 4.617378780604897,
+      "grad_norm": 0.938800339545672,
+      "learning_rate": 1.7848362516109274e-07,
+      "loss": 0.4703,
+      "step": 4811
+    },
+    {
+      "epoch": 4.6183389342294765,
+      "grad_norm": 1.313447666563615,
+      "learning_rate": 1.775976369628829e-07,
+      "loss": 0.4097,
+      "step": 4812
+    },
+    {
+      "epoch": 4.619299087854056,
+      "grad_norm": 0.9220032408795396,
+      "learning_rate": 1.7671381348954408e-07,
+      "loss": 0.3955,
+      "step": 4813
+    },
+    {
+      "epoch": 4.620259241478637,
+      "grad_norm": 1.5912875171764935,
+      "learning_rate": 1.7583215513781725e-07,
+      "loss": 0.3773,
+      "step": 4814
+    },
+    {
+      "epoch": 4.621219395103217,
+      "grad_norm": 0.9212555151022918,
+      "learning_rate": 1.749526623034681e-07,
+      "loss": 0.474,
+      "step": 4815
+    },
+    {
+      "epoch": 4.622179548727797,
+      "grad_norm": 1.315833523722032,
+      "learning_rate": 1.7407533538129185e-07,
+      "loss": 0.3756,
+      "step": 4816
+    },
+    {
+      "epoch": 4.623139702352376,
+      "grad_norm": 1.2615289578029076,
+      "learning_rate": 1.732001747651113e-07,
+      "loss": 0.4456,
+      "step": 4817
+    },
+    {
+      "epoch": 4.624099855976956,
+      "grad_norm": 1.4158536370413721,
+      "learning_rate": 1.7232718084777666e-07,
+      "loss": 0.3716,
+      "step": 4818
+    },
+    {
+      "epoch": 4.625060009601536,
+      "grad_norm": 1.445241579427424,
+      "learning_rate": 1.7145635402116556e-07,
+      "loss": 0.4631,
+      "step": 4819
+    },
+    {
+      "epoch": 4.626020163226116,
+      "grad_norm": 0.9557225557144389,
+      "learning_rate": 1.7058769467618364e-07,
+      "loss": 0.4896,
+      "step": 4820
+    },
+    {
+      "epoch": 4.626980316850696,
+      "grad_norm": 1.3519843082438203,
+      "learning_rate": 1.697212032027612e-07,
+      "loss": 0.2894,
+      "step": 4821
+    },
+    {
+      "epoch": 4.627940470475276,
+      "grad_norm": 0.9224256638249925,
+      "learning_rate": 1.688568799898582e-07,
+      "loss": 0.3217,
+      "step": 4822
+    },
+    {
+      "epoch": 4.628900624099856,
+      "grad_norm": 0.9216992108440527,
+      "learning_rate": 1.6799472542545935e-07,
+      "loss": 0.3999,
+      "step": 4823
+    },
+    {
+      "epoch": 4.629860777724436,
+      "grad_norm": 0.9231941365283417,
+      "learning_rate": 1.671347398965767e-07,
+      "loss": 0.3665,
+      "step": 4824
+    },
+    {
+      "epoch": 4.630820931349016,
+      "grad_norm": 1.3808693258581342,
+      "learning_rate": 1.6627692378924975e-07,
+      "loss": 0.4512,
+      "step": 4825
+    },
+    {
+      "epoch": 4.631781084973595,
+      "grad_norm": 1.3485510873758093,
+      "learning_rate": 1.6542127748854108e-07,
+      "loss": 0.3721,
+      "step": 4826
+    },
+    {
+      "epoch": 4.632741238598175,
+      "grad_norm": 0.9918658002473777,
+      "learning_rate": 1.6456780137854346e-07,
+      "loss": 0.5436,
+      "step": 4827
+    },
+    {
+      "epoch": 4.633701392222756,
+      "grad_norm": 1.0361195140022446,
+      "learning_rate": 1.6371649584237093e-07,
+      "loss": 0.4174,
+      "step": 4828
+    },
+    {
+      "epoch": 4.634661545847336,
+      "grad_norm": 1.182004901649788,
+      "learning_rate": 1.6286736126216674e-07,
+      "loss": 0.3823,
+      "step": 4829
+    },
+    {
+      "epoch": 4.6356216994719155,
+      "grad_norm": 0.8701419208967126,
+      "learning_rate": 1.6202039801909763e-07,
+      "loss": 0.424,
+      "step": 4830
+    },
+    {
+      "epoch": 4.636581853096495,
+      "grad_norm": 1.1504272685963877,
+      "learning_rate": 1.6117560649335674e-07,
+      "loss": 0.4092,
+      "step": 4831
+    },
+    {
+      "epoch": 4.637542006721075,
+      "grad_norm": 0.9936816498598194,
+      "learning_rate": 1.6033298706416234e-07,
+      "loss": 0.6072,
+      "step": 4832
+    },
+    {
+      "epoch": 4.638502160345656,
+      "grad_norm": 0.91588079466234,
+      "learning_rate": 1.594925401097569e-07,
+      "loss": 0.416,
+      "step": 4833
+    },
+    {
+      "epoch": 4.639462313970236,
+      "grad_norm": 1.161219323651602,
+      "learning_rate": 1.5865426600740807e-07,
+      "loss": 0.462,
+      "step": 4834
+    },
+    {
+      "epoch": 4.640422467594815,
+      "grad_norm": 1.5677838114702125,
+      "learning_rate": 1.5781816513340764e-07,
+      "loss": 0.3762,
+      "step": 4835
+    },
+    {
+      "epoch": 4.641382621219395,
+      "grad_norm": 0.9077167818557987,
+      "learning_rate": 1.5698423786307314e-07,
+      "loss": 0.4921,
+      "step": 4836
+    },
+    {
+      "epoch": 4.642342774843975,
+      "grad_norm": 1.2432151317729871,
+      "learning_rate": 1.5615248457074516e-07,
+      "loss": 0.4517,
+      "step": 4837
+    },
+    {
+      "epoch": 4.643302928468555,
+      "grad_norm": 1.1282115531345462,
+      "learning_rate": 1.5532290562978948e-07,
+      "loss": 0.3731,
+      "step": 4838
+    },
+    {
+      "epoch": 4.6442630820931345,
+      "grad_norm": 1.0135208505894553,
+      "learning_rate": 1.5449550141259427e-07,
+      "loss": 0.4225,
+      "step": 4839
+    },
+    {
+      "epoch": 4.645223235717715,
+      "grad_norm": 1.2155129043556023,
+      "learning_rate": 1.5367027229057307e-07,
+      "loss": 0.5577,
+      "step": 4840
+    },
+    {
+      "epoch": 4.646183389342295,
+      "grad_norm": 1.2834642587348466,
+      "learning_rate": 1.5284721863416175e-07,
+      "loss": 0.503,
+      "step": 4841
+    },
+    {
+      "epoch": 4.647143542966875,
+      "grad_norm": 0.875103063905967,
+      "learning_rate": 1.5202634081282142e-07,
+      "loss": 0.5263,
+      "step": 4842
+    },
+    {
+      "epoch": 4.648103696591455,
+      "grad_norm": 0.8083863279715064,
+      "learning_rate": 1.512076391950351e-07,
+      "loss": 0.371,
+      "step": 4843
+    },
+    {
+      "epoch": 4.649063850216034,
+      "grad_norm": 0.9753449309385723,
+      "learning_rate": 1.5039111414830721e-07,
+      "loss": 0.4616,
+      "step": 4844
+    },
+    {
+      "epoch": 4.650024003840614,
+      "grad_norm": 1.3312206754583908,
+      "learning_rate": 1.4957676603917005e-07,
+      "loss": 0.3697,
+      "step": 4845
+    },
+    {
+      "epoch": 4.650984157465194,
+      "grad_norm": 0.8037315397522173,
+      "learning_rate": 1.4876459523317455e-07,
+      "loss": 0.4577,
+      "step": 4846
+    },
+    {
+      "epoch": 4.651944311089775,
+      "grad_norm": 1.1583517652034765,
+      "learning_rate": 1.479546020948952e-07,
+      "loss": 0.337,
+      "step": 4847
+    },
+    {
+      "epoch": 4.6529044647143545,
+      "grad_norm": 1.1034917280152576,
+      "learning_rate": 1.4714678698792894e-07,
+      "loss": 0.4576,
+      "step": 4848
+    },
+    {
+      "epoch": 4.653864618338934,
+      "grad_norm": 0.8404487653635586,
+      "learning_rate": 1.4634115027489627e-07,
+      "loss": 0.5042,
+      "step": 4849
+    },
+    {
+      "epoch": 4.654824771963514,
+      "grad_norm": 1.18548657659862,
+      "learning_rate": 1.4553769231743842e-07,
+      "loss": 0.3966,
+      "step": 4850
+    },
+    {
+      "epoch": 4.655784925588094,
+      "grad_norm": 1.055970806542243,
+      "learning_rate": 1.447364134762197e-07,
+      "loss": 0.4327,
+      "step": 4851
+    },
+    {
+      "epoch": 4.656745079212674,
+      "grad_norm": 0.9167219764458515,
+      "learning_rate": 1.439373141109246e-07,
+      "loss": 0.4236,
+      "step": 4852
+    },
+    {
+      "epoch": 4.6577052328372535,
+      "grad_norm": 1.4952176776863857,
+      "learning_rate": 1.4314039458026063e-07,
+      "loss": 0.3892,
+      "step": 4853
+    },
+    {
+      "epoch": 4.658665386461834,
+      "grad_norm": 1.4585786530087195,
+      "learning_rate": 1.4234565524195664e-07,
+      "loss": 0.4627,
+      "step": 4854
+    },
+    {
+      "epoch": 4.659625540086414,
+      "grad_norm": 1.3997699800947045,
+      "learning_rate": 1.4155309645276283e-07,
+      "loss": 0.4162,
+      "step": 4855
+    },
+    {
+      "epoch": 4.660585693710994,
+      "grad_norm": 0.9842333753736631,
+      "learning_rate": 1.407627185684507e-07,
+      "loss": 0.4124,
+      "step": 4856
+    },
+    {
+      "epoch": 4.6615458473355735,
+      "grad_norm": 0.863316922583818,
+      "learning_rate": 1.3997452194381034e-07,
+      "loss": 0.5363,
+      "step": 4857
+    },
+    {
+      "epoch": 4.662506000960153,
+      "grad_norm": 0.9762782180460298,
+      "learning_rate": 1.3918850693265763e-07,
+      "loss": 0.5241,
+      "step": 4858
+    },
+    {
+      "epoch": 4.663466154584734,
+      "grad_norm": 1.6979556539963534,
+      "learning_rate": 1.3840467388782418e-07,
+      "loss": 0.3888,
+      "step": 4859
+    },
+    {
+      "epoch": 4.664426308209314,
+      "grad_norm": 1.0081865531573144,
+      "learning_rate": 1.3762302316116527e-07,
+      "loss": 0.4033,
+      "step": 4860
+    },
+    {
+      "epoch": 4.665386461833894,
+      "grad_norm": 1.091767746541218,
+      "learning_rate": 1.3684355510355518e-07,
+      "loss": 0.3666,
+      "step": 4861
+    },
+    {
+      "epoch": 4.666346615458473,
+      "grad_norm": 0.8008406660157712,
+      "learning_rate": 1.3606627006488905e-07,
+      "loss": 0.5543,
+      "step": 4862
+    },
+    {
+      "epoch": 4.667306769083053,
+      "grad_norm": 1.3614175946932625,
+      "learning_rate": 1.3529116839408162e-07,
+      "loss": 0.373,
+      "step": 4863
+    },
+    {
+      "epoch": 4.668266922707633,
+      "grad_norm": 1.201104244046454,
+      "learning_rate": 1.3451825043906852e-07,
+      "loss": 0.4642,
+      "step": 4864
+    },
+    {
+      "epoch": 4.669227076332213,
+      "grad_norm": 1.0786220466963068,
+      "learning_rate": 1.337475165468033e-07,
+      "loss": 0.4308,
+      "step": 4865
+    },
+    {
+      "epoch": 4.6701872299567935,
+      "grad_norm": 1.1344337058104614,
+      "learning_rate": 1.3297896706326031e-07,
+      "loss": 0.4036,
+      "step": 4866
+    },
+    {
+      "epoch": 4.671147383581373,
+      "grad_norm": 1.2101706576117013,
+      "learning_rate": 1.3221260233343357e-07,
+      "loss": 0.46,
+      "step": 4867
+    },
+    {
+      "epoch": 4.672107537205953,
+      "grad_norm": 1.6304586067035027,
+      "learning_rate": 1.3144842270133573e-07,
+      "loss": 0.3969,
+      "step": 4868
+    },
+    {
+      "epoch": 4.673067690830533,
+      "grad_norm": 1.0671618880075804,
+      "learning_rate": 1.3068642850999957e-07,
+      "loss": 0.4138,
+      "step": 4869
+    },
+    {
+      "epoch": 4.674027844455113,
+      "grad_norm": 1.0223444659112357,
+      "learning_rate": 1.2992662010147428e-07,
+      "loss": 0.4489,
+      "step": 4870
+    },
+    {
+      "epoch": 4.6749879980796925,
+      "grad_norm": 1.2642173548965616,
+      "learning_rate": 1.2916899781683257e-07,
+      "loss": 0.3875,
+      "step": 4871
+    },
+    {
+      "epoch": 4.675948151704272,
+      "grad_norm": 1.1412850621951813,
+      "learning_rate": 1.2841356199616072e-07,
+      "loss": 0.4745,
+      "step": 4872
+    },
+    {
+      "epoch": 4.676908305328853,
+      "grad_norm": 1.3415299598720207,
+      "learning_rate": 1.276603129785664e-07,
+      "loss": 0.4558,
+      "step": 4873
+    },
+    {
+      "epoch": 4.677868458953433,
+      "grad_norm": 0.8547068664536178,
+      "learning_rate": 1.2690925110217523e-07,
+      "loss": 0.4189,
+      "step": 4874
+    },
+    {
+      "epoch": 4.6788286125780125,
+      "grad_norm": 0.970876090683359,
+      "learning_rate": 1.2616037670413084e-07,
+      "loss": 0.4404,
+      "step": 4875
+    },
+    {
+      "epoch": 4.679788766202592,
+      "grad_norm": 1.3348221156706659,
+      "learning_rate": 1.254136901205949e-07,
+      "loss": 0.3674,
+      "step": 4876
+    },
+    {
+      "epoch": 4.680748919827172,
+      "grad_norm": 1.0607047447901858,
+      "learning_rate": 1.2466919168674763e-07,
+      "loss": 0.4402,
+      "step": 4877
+    },
+    {
+      "epoch": 4.681709073451752,
+      "grad_norm": 1.3850927877470836,
+      "learning_rate": 1.2392688173678503e-07,
+      "loss": 0.377,
+      "step": 4878
+    },
+    {
+      "epoch": 4.682669227076333,
+      "grad_norm": 1.146457920670871,
+      "learning_rate": 1.2318676060392333e-07,
+      "loss": 0.4765,
+      "step": 4879
+    },
+    {
+      "epoch": 4.683629380700912,
+      "grad_norm": 1.3187711202669654,
+      "learning_rate": 1.2244882862039398e-07,
+      "loss": 0.3573,
+      "step": 4880
+    },
+    {
+      "epoch": 4.684589534325492,
+      "grad_norm": 1.0748003644300437,
+      "learning_rate": 1.217130861174476e-07,
+      "loss": 0.5518,
+      "step": 4881
+    },
+    {
+      "epoch": 4.685549687950072,
+      "grad_norm": 0.9285725553402598,
+      "learning_rate": 1.209795334253516e-07,
+      "loss": 0.478,
+      "step": 4882
+    },
+    {
+      "epoch": 4.686509841574652,
+      "grad_norm": 1.3821853568376987,
+      "learning_rate": 1.2024817087338826e-07,
+      "loss": 0.3444,
+      "step": 4883
+    },
+    {
+      "epoch": 4.687469995199232,
+      "grad_norm": 1.1346742359195487,
+      "learning_rate": 1.1951899878985983e-07,
+      "loss": 0.3314,
+      "step": 4884
+    },
+    {
+      "epoch": 4.688430148823812,
+      "grad_norm": 1.1054396444704837,
+      "learning_rate": 1.1879201750208291e-07,
+      "loss": 0.4511,
+      "step": 4885
+    },
+    {
+      "epoch": 4.689390302448392,
+      "grad_norm": 0.8909019999832792,
+      "learning_rate": 1.1806722733639253e-07,
+      "loss": 0.5304,
+      "step": 4886
+    },
+    {
+      "epoch": 4.690350456072972,
+      "grad_norm": 1.1991086945341605,
+      "learning_rate": 1.17344628618139e-07,
+      "loss": 0.4055,
+      "step": 4887
+    },
+    {
+      "epoch": 4.691310609697552,
+      "grad_norm": 0.888618511391377,
+      "learning_rate": 1.1662422167168841e-07,
+      "loss": 0.4966,
+      "step": 4888
+    },
+    {
+      "epoch": 4.692270763322131,
+      "grad_norm": 1.03938945352285,
+      "learning_rate": 1.159060068204243e-07,
+      "loss": 0.4609,
+      "step": 4889
+    },
+    {
+      "epoch": 4.693230916946711,
+      "grad_norm": 1.5886976054096131,
+      "learning_rate": 1.1518998438674656e-07,
+      "loss": 0.3648,
+      "step": 4890
+    },
+    {
+      "epoch": 4.694191070571291,
+      "grad_norm": 1.0111931849689184,
+      "learning_rate": 1.1447615469206863e-07,
+      "loss": 0.5052,
+      "step": 4891
+    },
+    {
+      "epoch": 4.695151224195872,
+      "grad_norm": 0.9451034878617085,
+      "learning_rate": 1.1376451805682143e-07,
+      "loss": 0.4494,
+      "step": 4892
+    },
+    {
+      "epoch": 4.6961113778204515,
+      "grad_norm": 1.0684130717439,
+      "learning_rate": 1.1305507480045108e-07,
+      "loss": 0.4941,
+      "step": 4893
+    },
+    {
+      "epoch": 4.697071531445031,
+      "grad_norm": 1.0331287204273547,
+      "learning_rate": 1.1234782524141896e-07,
+      "loss": 0.5403,
+      "step": 4894
+    },
+    {
+      "epoch": 4.698031685069611,
+      "grad_norm": 0.9787382958557056,
+      "learning_rate": 1.1164276969720278e-07,
+      "loss": 0.4327,
+      "step": 4895
+    },
+    {
+      "epoch": 4.698991838694191,
+      "grad_norm": 0.7971038099269354,
+      "learning_rate": 1.109399084842927e-07,
+      "loss": 0.4174,
+      "step": 4896
+    },
+    {
+      "epoch": 4.699951992318771,
+      "grad_norm": 1.0189260596115977,
+      "learning_rate": 1.1023924191819691e-07,
+      "loss": 0.4767,
+      "step": 4897
+    },
+    {
+      "epoch": 4.7009121459433505,
+      "grad_norm": 0.8119306713496375,
+      "learning_rate": 1.0954077031343658e-07,
+      "loss": 0.4478,
+      "step": 4898
+    },
+    {
+      "epoch": 4.701872299567931,
+      "grad_norm": 0.9838075867737415,
+      "learning_rate": 1.0884449398354868e-07,
+      "loss": 0.4201,
+      "step": 4899
+    },
+    {
+      "epoch": 4.702832453192511,
+      "grad_norm": 1.606158635230087,
+      "learning_rate": 1.0815041324108378e-07,
+      "loss": 0.4645,
+      "step": 4900
+    },
+    {
+      "epoch": 4.703792606817091,
+      "grad_norm": 1.2063985629173526,
+      "learning_rate": 1.0745852839760706e-07,
+      "loss": 0.3579,
+      "step": 4901
+    },
+    {
+      "epoch": 4.704752760441671,
+      "grad_norm": 0.8553193293134254,
+      "learning_rate": 1.0676883976369956e-07,
+      "loss": 0.4549,
+      "step": 4902
+    },
+    {
+      "epoch": 4.70571291406625,
+      "grad_norm": 1.002553399336827,
+      "learning_rate": 1.0608134764895306e-07,
+      "loss": 0.4566,
+      "step": 4903
+    },
+    {
+      "epoch": 4.70667306769083,
+      "grad_norm": 1.1243789394374941,
+      "learning_rate": 1.053960523619768e-07,
+      "loss": 0.3638,
+      "step": 4904
+    },
+    {
+      "epoch": 4.707633221315411,
+      "grad_norm": 1.2881083832451525,
+      "learning_rate": 1.0471295421039251e-07,
+      "loss": 0.4376,
+      "step": 4905
+    },
+    {
+      "epoch": 4.708593374939991,
+      "grad_norm": 0.9839190906016867,
+      "learning_rate": 1.040320535008349e-07,
+      "loss": 0.4521,
+      "step": 4906
+    },
+    {
+      "epoch": 4.70955352856457,
+      "grad_norm": 1.5004863605052488,
+      "learning_rate": 1.0335335053895391e-07,
+      "loss": 0.4402,
+      "step": 4907
+    },
+    {
+      "epoch": 4.71051368218915,
+      "grad_norm": 1.006124507240295,
+      "learning_rate": 1.0267684562941194e-07,
+      "loss": 0.3408,
+      "step": 4908
+    },
+    {
+      "epoch": 4.71147383581373,
+      "grad_norm": 1.1130247517133098,
+      "learning_rate": 1.0200253907588387e-07,
+      "loss": 0.4047,
+      "step": 4909
+    },
+    {
+      "epoch": 4.71243398943831,
+      "grad_norm": 1.2325609428027149,
+      "learning_rate": 1.0133043118105923e-07,
+      "loss": 0.4371,
+      "step": 4910
+    },
+    {
+      "epoch": 4.7133941430628905,
+      "grad_norm": 1.4056592768793215,
+      "learning_rate": 1.0066052224664058e-07,
+      "loss": 0.4301,
+      "step": 4911
+    },
+    {
+      "epoch": 4.71435429668747,
+      "grad_norm": 1.3159180364916698,
+      "learning_rate": 9.999281257334181e-08,
+      "loss": 0.4366,
+      "step": 4912
+    },
+    {
+      "epoch": 4.71531445031205,
+      "grad_norm": 1.4518304434272327,
+      "learning_rate": 9.932730246089262e-08,
+      "loss": 0.4812,
+      "step": 4913
+    },
+    {
+      "epoch": 4.71627460393663,
+      "grad_norm": 1.6672745203186197,
+      "learning_rate": 9.86639922080307e-08,
+      "loss": 0.4855,
+      "step": 4914
+    },
+    {
+      "epoch": 4.71723475756121,
+      "grad_norm": 0.9039540229330392,
+      "learning_rate": 9.800288211251119e-08,
+      "loss": 0.462,
+      "step": 4915
+    },
+    {
+      "epoch": 4.7181949111857895,
+      "grad_norm": 1.1369334787339889,
+      "learning_rate": 9.734397247109784e-08,
+      "loss": 0.4026,
+      "step": 4916
+    },
+    {
+      "epoch": 4.719155064810369,
+      "grad_norm": 0.8818340451726859,
+      "learning_rate": 9.668726357956903e-08,
+      "loss": 0.4413,
+      "step": 4917
+    },
+    {
+      "epoch": 4.72011521843495,
+      "grad_norm": 1.4896273099509099,
+      "learning_rate": 9.603275573271397e-08,
+      "loss": 0.4953,
+      "step": 4918
+    },
+    {
+      "epoch": 4.72107537205953,
+      "grad_norm": 1.0628416430946594,
+      "learning_rate": 9.538044922433432e-08,
+      "loss": 0.4605,
+      "step": 4919
+    },
+    {
+      "epoch": 4.72203552568411,
+      "grad_norm": 1.0993876630712076,
+      "learning_rate": 9.47303443472436e-08,
+      "loss": 0.3838,
+      "step": 4920
+    },
+    {
+      "epoch": 4.722995679308689,
+      "grad_norm": 1.168472568639651,
+      "learning_rate": 9.408244139326728e-08,
+      "loss": 0.4395,
+      "step": 4921
+    },
+    {
+      "epoch": 4.723955832933269,
+      "grad_norm": 1.1976832403588442,
+      "learning_rate": 9.34367406532416e-08,
+      "loss": 0.3265,
+      "step": 4922
+    },
+    {
+      "epoch": 4.724915986557849,
+      "grad_norm": 0.9147656733632449,
+      "learning_rate": 9.279324241701415e-08,
+      "loss": 0.5055,
+      "step": 4923
+    },
+    {
+      "epoch": 4.725876140182429,
+      "grad_norm": 1.1868320539821027,
+      "learning_rate": 9.215194697344554e-08,
+      "loss": 0.4473,
+      "step": 4924
+    },
+    {
+      "epoch": 4.726836293807009,
+      "grad_norm": 1.8420799450510248,
+      "learning_rate": 9.151285461040604e-08,
+      "loss": 0.3488,
+      "step": 4925
+    },
+    {
+      "epoch": 4.727796447431589,
+      "grad_norm": 1.2304430967818216,
+      "learning_rate": 9.08759656147773e-08,
+      "loss": 0.3462,
+      "step": 4926
+    },
+    {
+      "epoch": 4.728756601056169,
+      "grad_norm": 0.9520488195701418,
+      "learning_rate": 9.024128027245171e-08,
+      "loss": 0.5115,
+      "step": 4927
+    },
+    {
+      "epoch": 4.729716754680749,
+      "grad_norm": 0.7815379300936841,
+      "learning_rate": 8.960879886833307e-08,
+      "loss": 0.4298,
+      "step": 4928
+    },
+    {
+      "epoch": 4.730676908305329,
+      "grad_norm": 1.0394315489433514,
+      "learning_rate": 8.897852168633537e-08,
+      "loss": 0.6247,
+      "step": 4929
+    },
+    {
+      "epoch": 4.731637061929908,
+      "grad_norm": 1.1020238704377705,
+      "learning_rate": 8.835044900938339e-08,
+      "loss": 0.3987,
+      "step": 4930
+    },
+    {
+      "epoch": 4.732597215554489,
+      "grad_norm": 1.2263463836720976,
+      "learning_rate": 8.772458111941273e-08,
+      "loss": 0.4371,
+      "step": 4931
+    },
+    {
+      "epoch": 4.733557369179069,
+      "grad_norm": 1.3328171761452732,
+      "learning_rate": 8.7100918297367e-08,
+      "loss": 0.3241,
+      "step": 4932
+    },
+    {
+      "epoch": 4.734517522803649,
+      "grad_norm": 1.2551802685050564,
+      "learning_rate": 8.647946082320446e-08,
+      "loss": 0.5018,
+      "step": 4933
+    },
+    {
+      "epoch": 4.7354776764282285,
+      "grad_norm": 0.9142369627181196,
+      "learning_rate": 8.586020897588864e-08,
+      "loss": 0.3876,
+      "step": 4934
+    },
+    {
+      "epoch": 4.736437830052808,
+      "grad_norm": 1.1762289843339626,
+      "learning_rate": 8.524316303339609e-08,
+      "loss": 0.4208,
+      "step": 4935
+    },
+    {
+      "epoch": 4.737397983677388,
+      "grad_norm": 1.0550815448101691,
+      "learning_rate": 8.462832327271131e-08,
+      "loss": 0.4866,
+      "step": 4936
+    },
+    {
+      "epoch": 4.738358137301969,
+      "grad_norm": 1.058668993144802,
+      "learning_rate": 8.401568996983022e-08,
+      "loss": 0.4792,
+      "step": 4937
+    },
+    {
+      "epoch": 4.7393182909265485,
+      "grad_norm": 1.2542659226887585,
+      "learning_rate": 8.340526339975674e-08,
+      "loss": 0.3754,
+      "step": 4938
+    },
+    {
+      "epoch": 4.740278444551128,
+      "grad_norm": 0.9819182622733525,
+      "learning_rate": 8.279704383650555e-08,
+      "loss": 0.4479,
+      "step": 4939
+    },
+    {
+      "epoch": 4.741238598175708,
+      "grad_norm": 1.3028202373874764,
+      "learning_rate": 8.219103155309938e-08,
+      "loss": 0.3555,
+      "step": 4940
+    },
+    {
+      "epoch": 4.742198751800288,
+      "grad_norm": 0.8660422289707396,
+      "learning_rate": 8.158722682157005e-08,
+      "loss": 0.4085,
+      "step": 4941
+    },
+    {
+      "epoch": 4.743158905424868,
+      "grad_norm": 1.3875041639611745,
+      "learning_rate": 8.098562991296022e-08,
+      "loss": 0.4473,
+      "step": 4942
+    },
+    {
+      "epoch": 4.7441190590494475,
+      "grad_norm": 0.9151113514826871,
+      "learning_rate": 8.03862410973194e-08,
+      "loss": 0.484,
+      "step": 4943
+    },
+    {
+      "epoch": 4.745079212674028,
+      "grad_norm": 0.7596448931217988,
+      "learning_rate": 7.978906064370739e-08,
+      "loss": 0.4297,
+      "step": 4944
+    },
+    {
+      "epoch": 4.746039366298608,
+      "grad_norm": 1.2258444792057024,
+      "learning_rate": 7.919408882019142e-08,
+      "loss": 0.4487,
+      "step": 4945
+    },
+    {
+      "epoch": 4.746999519923188,
+      "grad_norm": 0.793912596275803,
+      "learning_rate": 7.86013258938495e-08,
+      "loss": 0.4838,
+      "step": 4946
+    },
+    {
+      "epoch": 4.747959673547768,
+      "grad_norm": 1.4583786481432341,
+      "learning_rate": 7.801077213076491e-08,
+      "loss": 0.4148,
+      "step": 4947
+    },
+    {
+      "epoch": 4.748919827172347,
+      "grad_norm": 1.113940011281248,
+      "learning_rate": 7.742242779603115e-08,
+      "loss": 0.4067,
+      "step": 4948
+    },
+    {
+      "epoch": 4.749879980796927,
+      "grad_norm": 1.4092380218749927,
+      "learning_rate": 7.683629315375086e-08,
+      "loss": 0.4295,
+      "step": 4949
+    },
+    {
+      "epoch": 4.750840134421507,
+      "grad_norm": 0.9293678014901214,
+      "learning_rate": 7.625236846703243e-08,
+      "loss": 0.43,
+      "step": 4950
+    },
+    {
+      "epoch": 4.751800288046088,
+      "grad_norm": 1.6534193883886423,
+      "learning_rate": 7.5670653997994e-08,
+      "loss": 0.3444,
+      "step": 4951
+    },
+    {
+      "epoch": 4.7527604416706675,
+      "grad_norm": 0.8622578441674419,
+      "learning_rate": 7.509115000776112e-08,
+      "loss": 0.4508,
+      "step": 4952
+    },
+    {
+      "epoch": 4.753720595295247,
+      "grad_norm": 1.4874994288456962,
+      "learning_rate": 7.451385675646627e-08,
+      "loss": 0.4145,
+      "step": 4953
+    },
+    {
+      "epoch": 4.754680748919827,
+      "grad_norm": 1.171764515099023,
+      "learning_rate": 7.393877450325105e-08,
+      "loss": 0.3923,
+      "step": 4954
+    },
+    {
+      "epoch": 4.755640902544407,
+      "grad_norm": 1.2786274251313396,
+      "learning_rate": 7.336590350626282e-08,
+      "loss": 0.4632,
+      "step": 4955
+    },
+    {
+      "epoch": 4.756601056168987,
+      "grad_norm": 1.1696736479222372,
+      "learning_rate": 7.279524402265758e-08,
+      "loss": 0.3443,
+      "step": 4956
+    },
+    {
+      "epoch": 4.757561209793567,
+      "grad_norm": 0.9854702719230389,
+      "learning_rate": 7.222679630859875e-08,
+      "loss": 0.4123,
+      "step": 4957
+    },
+    {
+      "epoch": 4.758521363418147,
+      "grad_norm": 0.7923822962355402,
+      "learning_rate": 7.166056061925609e-08,
+      "loss": 0.5323,
+      "step": 4958
+    },
+    {
+      "epoch": 4.759481517042727,
+      "grad_norm": 0.7356812366740414,
+      "learning_rate": 7.109653720880682e-08,
+      "loss": 0.3986,
+      "step": 4959
+    },
+    {
+      "epoch": 4.760441670667307,
+      "grad_norm": 1.1674277613161905,
+      "learning_rate": 7.053472633043457e-08,
+      "loss": 0.437,
+      "step": 4960
+    },
+    {
+      "epoch": 4.7614018242918865,
+      "grad_norm": 1.3706100197867934,
+      "learning_rate": 6.99751282363309e-08,
+      "loss": 0.3233,
+      "step": 4961
+    },
+    {
+      "epoch": 4.762361977916466,
+      "grad_norm": 1.1604441130532235,
+      "learning_rate": 6.941774317769268e-08,
+      "loss": 0.4747,
+      "step": 4962
+    },
+    {
+      "epoch": 4.763322131541047,
+      "grad_norm": 1.3250108388330313,
+      "learning_rate": 6.886257140472475e-08,
+      "loss": 0.5501,
+      "step": 4963
+    },
+    {
+      "epoch": 4.764282285165627,
+      "grad_norm": 1.0212396586151098,
+      "learning_rate": 6.830961316663775e-08,
+      "loss": 0.4548,
+      "step": 4964
+    },
+    {
+      "epoch": 4.765242438790207,
+      "grad_norm": 1.1639294654175862,
+      "learning_rate": 6.775886871164872e-08,
+      "loss": 0.3857,
+      "step": 4965
+    },
+    {
+      "epoch": 4.766202592414786,
+      "grad_norm": 0.8051939909618316,
+      "learning_rate": 6.721033828698098e-08,
+      "loss": 0.3928,
+      "step": 4966
+    },
+    {
+      "epoch": 4.767162746039366,
+      "grad_norm": 1.459218130958606,
+      "learning_rate": 6.666402213886313e-08,
+      "loss": 0.3408,
+      "step": 4967
+    },
+    {
+      "epoch": 4.768122899663946,
+      "grad_norm": 1.3492964232221831,
+      "learning_rate": 6.61199205125318e-08,
+      "loss": 0.3791,
+      "step": 4968
+    },
+    {
+      "epoch": 4.769083053288526,
+      "grad_norm": 0.8929234512957442,
+      "learning_rate": 6.557803365222714e-08,
+      "loss": 0.4892,
+      "step": 4969
+    },
+    {
+      "epoch": 4.7700432069131065,
+      "grad_norm": 0.8843561755935987,
+      "learning_rate": 6.503836180119849e-08,
+      "loss": 0.4556,
+      "step": 4970
+    },
+    {
+      "epoch": 4.771003360537686,
+      "grad_norm": 0.9240795086726811,
+      "learning_rate": 6.45009052016965e-08,
+      "loss": 0.4414,
+      "step": 4971
+    },
+    {
+      "epoch": 4.771963514162266,
+      "grad_norm": 0.6979346576133246,
+      "learning_rate": 6.396566409498039e-08,
+      "loss": 0.4472,
+      "step": 4972
+    },
+    {
+      "epoch": 4.772923667786846,
+      "grad_norm": 1.0614362995003612,
+      "learning_rate": 6.343263872131467e-08,
+      "loss": 0.3583,
+      "step": 4973
+    },
+    {
+      "epoch": 4.773883821411426,
+      "grad_norm": 0.9563491793609955,
+      "learning_rate": 6.290182931996846e-08,
+      "loss": 0.466,
+      "step": 4974
+    },
+    {
+      "epoch": 4.774843975036005,
+      "grad_norm": 0.8391276852505679,
+      "learning_rate": 6.237323612921675e-08,
+      "loss": 0.4898,
+      "step": 4975
+    },
+    {
+      "epoch": 4.775804128660585,
+      "grad_norm": 0.9597090155992283,
+      "learning_rate": 6.184685938633861e-08,
+      "loss": 0.398,
+      "step": 4976
+    },
+    {
+      "epoch": 4.776764282285166,
+      "grad_norm": 0.971137486645396,
+      "learning_rate": 6.132269932761947e-08,
+      "loss": 0.4431,
+      "step": 4977
+    },
+    {
+      "epoch": 4.777724435909746,
+      "grad_norm": 1.267122668625548,
+      "learning_rate": 6.080075618834891e-08,
+      "loss": 0.5065,
+      "step": 4978
+    },
+    {
+      "epoch": 4.7786845895343255,
+      "grad_norm": 0.9507539850267844,
+      "learning_rate": 6.028103020282228e-08,
+      "loss": 0.4919,
+      "step": 4979
+    },
+    {
+      "epoch": 4.779644743158905,
+      "grad_norm": 1.0701923975702317,
+      "learning_rate": 5.976352160433796e-08,
+      "loss": 0.4885,
+      "step": 4980
+    },
+    {
+      "epoch": 4.780604896783485,
+      "grad_norm": 1.1551069810306696,
+      "learning_rate": 5.9248230625200686e-08,
+      "loss": 0.5171,
+      "step": 4981
+    },
+    {
+      "epoch": 4.781565050408066,
+      "grad_norm": 1.412241643869421,
+      "learning_rate": 5.873515749671932e-08,
+      "loss": 0.3582,
+      "step": 4982
+    },
+    {
+      "epoch": 4.782525204032646,
+      "grad_norm": 0.8081650775943622,
+      "learning_rate": 5.822430244920685e-08,
+      "loss": 0.4219,
+      "step": 4983
+    },
+    {
+      "epoch": 4.783485357657225,
+      "grad_norm": 1.1116610362427617,
+      "learning_rate": 5.771566571197984e-08,
+      "loss": 0.4564,
+      "step": 4984
+    },
+    {
+      "epoch": 4.784445511281805,
+      "grad_norm": 1.0911799731899194,
+      "learning_rate": 5.72092475133601e-08,
+      "loss": 0.3598,
+      "step": 4985
+    },
+    {
+      "epoch": 4.785405664906385,
+      "grad_norm": 1.2795568684332475,
+      "learning_rate": 5.670504808067412e-08,
+      "loss": 0.4882,
+      "step": 4986
+    },
+    {
+      "epoch": 4.786365818530965,
+      "grad_norm": 1.2157900267444886,
+      "learning_rate": 5.620306764025085e-08,
+      "loss": 0.378,
+      "step": 4987
+    },
+    {
+      "epoch": 4.787325972155545,
+      "grad_norm": 1.1693247557414115,
+      "learning_rate": 5.57033064174245e-08,
+      "loss": 0.5115,
+      "step": 4988
+    },
+    {
+      "epoch": 4.788286125780125,
+      "grad_norm": 1.3610464976675363,
+      "learning_rate": 5.5205764636532266e-08,
+      "loss": 0.441,
+      "step": 4989
+    },
+    {
+      "epoch": 4.789246279404705,
+      "grad_norm": 1.1057349733443953,
+      "learning_rate": 5.471044252091551e-08,
+      "loss": 0.4532,
+      "step": 4990
+    },
+    {
+      "epoch": 4.790206433029285,
+      "grad_norm": 1.380890432768807,
+      "learning_rate": 5.421734029291858e-08,
+      "loss": 0.4394,
+      "step": 4991
+    },
+    {
+      "epoch": 4.791166586653865,
+      "grad_norm": 1.6010994351481194,
+      "learning_rate": 5.372645817389055e-08,
+      "loss": 0.3936,
+      "step": 4992
+    },
+    {
+      "epoch": 4.792126740278444,
+      "grad_norm": 1.0701981759846952,
+      "learning_rate": 5.323779638418236e-08,
+      "loss": 0.4799,
+      "step": 4993
+    },
+    {
+      "epoch": 4.793086893903024,
+      "grad_norm": 1.455181412551221,
+      "learning_rate": 5.275135514314911e-08,
+      "loss": 0.555,
+      "step": 4994
+    },
+    {
+      "epoch": 4.794047047527604,
+      "grad_norm": 1.27645281125826,
+      "learning_rate": 5.226713466915001e-08,
+      "loss": 0.3923,
+      "step": 4995
+    },
+    {
+      "epoch": 4.795007201152185,
+      "grad_norm": 1.4609538739748678,
+      "learning_rate": 5.1785135179545643e-08,
+      "loss": 0.355,
+      "step": 4996
+    },
+    {
+      "epoch": 4.7959673547767645,
+      "grad_norm": 1.0796528565548074,
+      "learning_rate": 5.130535689070071e-08,
+      "loss": 0.5151,
+      "step": 4997
+    },
+    {
+      "epoch": 4.796927508401344,
+      "grad_norm": 1.0709646652401668,
+      "learning_rate": 5.0827800017982376e-08,
+      "loss": 0.4795,
+      "step": 4998
+    },
+    {
+      "epoch": 4.797887662025924,
+      "grad_norm": 2.156807557723183,
+      "learning_rate": 5.035246477576083e-08,
+      "loss": 0.3752,
+      "step": 4999
+    },
+    {
+      "epoch": 4.798847815650504,
+      "grad_norm": 0.9900267881433429,
+      "learning_rate": 4.9879351377408734e-08,
+      "loss": 0.4411,
+      "step": 5000
+    },
+    {
+      "epoch": 4.799807969275084,
+      "grad_norm": 1.3097281496901851,
+      "learning_rate": 4.940846003530231e-08,
+      "loss": 0.3953,
+      "step": 5001
+    },
+    {
+      "epoch": 4.8007681228996635,
+      "grad_norm": 1.5642871992519123,
+      "learning_rate": 4.893979096081969e-08,
+      "loss": 0.4368,
+      "step": 5002
+    },
+    {
+      "epoch": 4.801728276524244,
+      "grad_norm": 1.0188868463034577,
+      "learning_rate": 4.847334436434037e-08,
+      "loss": 0.4169,
+      "step": 5003
+    },
+    {
+      "epoch": 4.802688430148824,
+      "grad_norm": 1.0283832936470796,
+      "learning_rate": 4.800912045524797e-08,
+      "loss": 0.5007,
+      "step": 5004
+    },
+    {
+      "epoch": 4.803648583773404,
+      "grad_norm": 1.5029285350762907,
+      "learning_rate": 4.7547119441927466e-08,
+      "loss": 0.5229,
+      "step": 5005
+    },
+    {
+      "epoch": 4.8046087373979836,
+      "grad_norm": 1.0538901984147298,
+      "learning_rate": 4.7087341531766864e-08,
+      "loss": 0.3801,
+      "step": 5006
+    },
+    {
+      "epoch": 4.805568891022563,
+      "grad_norm": 1.324369478183412,
+      "learning_rate": 4.662978693115439e-08,
+      "loss": 0.4654,
+      "step": 5007
+    },
+    {
+      "epoch": 4.806529044647144,
+      "grad_norm": 1.333077486173929,
+      "learning_rate": 4.617445584548186e-08,
+      "loss": 0.4119,
+      "step": 5008
+    },
+    {
+      "epoch": 4.807489198271724,
+      "grad_norm": 1.0225880236638505,
+      "learning_rate": 4.5721348479143e-08,
+      "loss": 0.3842,
+      "step": 5009
+    },
+    {
+      "epoch": 4.808449351896304,
+      "grad_norm": 1.4060556577339138,
+      "learning_rate": 4.527046503553234e-08,
+      "loss": 0.4361,
+      "step": 5010
+    },
+    {
+      "epoch": 4.809409505520883,
+      "grad_norm": 1.1957036171980913,
+      "learning_rate": 4.4821805717046864e-08,
+      "loss": 0.501,
+      "step": 5011
+    },
+    {
+      "epoch": 4.810369659145463,
+      "grad_norm": 0.9787724410074183,
+      "learning_rate": 4.437537072508491e-08,
+      "loss": 0.3446,
+      "step": 5012
+    },
+    {
+      "epoch": 4.811329812770043,
+      "grad_norm": 0.996299467897215,
+      "learning_rate": 4.393116026004618e-08,
+      "loss": 0.4115,
+      "step": 5013
+    },
+    {
+      "epoch": 4.812289966394623,
+      "grad_norm": 1.0115549019320391,
+      "learning_rate": 4.3489174521332833e-08,
+      "loss": 0.3722,
+      "step": 5014
+    },
+    {
+      "epoch": 4.8132501200192035,
+      "grad_norm": 0.9126117651097011,
+      "learning_rate": 4.304941370734616e-08,
+      "loss": 0.5471,
+      "step": 5015
+    },
+    {
+      "epoch": 4.814210273643783,
+      "grad_norm": 0.7846085957859918,
+      "learning_rate": 4.261187801549105e-08,
+      "loss": 0.4737,
+      "step": 5016
+    },
+    {
+      "epoch": 4.815170427268363,
+      "grad_norm": 1.5608139318157948,
+      "learning_rate": 4.217656764217204e-08,
+      "loss": 0.3668,
+      "step": 5017
+    },
+    {
+      "epoch": 4.816130580892943,
+      "grad_norm": 1.099803488631457,
+      "learning_rate": 4.1743482782796166e-08,
+      "loss": 0.4644,
+      "step": 5018
+    },
+    {
+      "epoch": 4.817090734517523,
+      "grad_norm": 0.8158778432417666,
+      "learning_rate": 4.1312623631770134e-08,
+      "loss": 0.4825,
+      "step": 5019
+    },
+    {
+      "epoch": 4.8180508881421025,
+      "grad_norm": 0.9106789335231578,
+      "learning_rate": 4.08839903825009e-08,
+      "loss": 0.4384,
+      "step": 5020
+    },
+    {
+      "epoch": 4.819011041766682,
+      "grad_norm": 1.0926509474000448,
+      "learning_rate": 4.045758322739957e-08,
+      "loss": 0.4491,
+      "step": 5021
+    },
+    {
+      "epoch": 4.819971195391263,
+      "grad_norm": 1.5911219887566954,
+      "learning_rate": 4.003340235787412e-08,
+      "loss": 0.4435,
+      "step": 5022
+    },
+    {
+      "epoch": 4.820931349015843,
+      "grad_norm": 1.1856930480347105,
+      "learning_rate": 3.961144796433502e-08,
+      "loss": 0.4514,
+      "step": 5023
+    },
+    {
+      "epoch": 4.8218915026404225,
+      "grad_norm": 1.1726696751660066,
+      "learning_rate": 3.919172023619355e-08,
+      "loss": 0.4326,
+      "step": 5024
+    },
+    {
+      "epoch": 4.822851656265002,
+      "grad_norm": 1.734404149191527,
+      "learning_rate": 3.8774219361860655e-08,
+      "loss": 0.3781,
+      "step": 5025
+    },
+    {
+      "epoch": 4.823811809889582,
+      "grad_norm": 0.9805679935983692,
+      "learning_rate": 3.835894552874808e-08,
+      "loss": 0.4952,
+      "step": 5026
+    },
+    {
+      "epoch": 4.824771963514162,
+      "grad_norm": 0.7514170971102834,
+      "learning_rate": 3.794589892326839e-08,
+      "loss": 0.3744,
+      "step": 5027
+    },
+    {
+      "epoch": 4.825732117138742,
+      "grad_norm": 1.445732296687345,
+      "learning_rate": 3.75350797308327e-08,
+      "loss": 0.3805,
+      "step": 5028
+    },
+    {
+      "epoch": 4.826692270763322,
+      "grad_norm": 1.0839632046033663,
+      "learning_rate": 3.712648813585407e-08,
+      "loss": 0.4165,
+      "step": 5029
+    },
+    {
+      "epoch": 4.827652424387902,
+      "grad_norm": 0.9701534195938427,
+      "learning_rate": 3.672012432174465e-08,
+      "loss": 0.3739,
+      "step": 5030
+    },
+    {
+      "epoch": 4.828612578012482,
+      "grad_norm": 0.8296847653170805,
+      "learning_rate": 3.631598847091688e-08,
+      "loss": 0.4379,
+      "step": 5031
+    },
+    {
+      "epoch": 4.829572731637062,
+      "grad_norm": 1.0767524351925464,
+      "learning_rate": 3.591408076478342e-08,
+      "loss": 0.357,
+      "step": 5032
+    },
+    {
+      "epoch": 4.830532885261642,
+      "grad_norm": 1.1096053501471805,
+      "learning_rate": 3.5514401383755505e-08,
+      "loss": 0.4687,
+      "step": 5033
+    },
+    {
+      "epoch": 4.831493038886222,
+      "grad_norm": 1.4866482743973553,
+      "learning_rate": 3.511695050724573e-08,
+      "loss": 0.4369,
+      "step": 5034
+    },
+    {
+      "epoch": 4.832453192510802,
+      "grad_norm": 0.8958708445370497,
+      "learning_rate": 3.4721728313664715e-08,
+      "loss": 0.4407,
+      "step": 5035
+    },
+    {
+      "epoch": 4.833413346135382,
+      "grad_norm": 0.9294298654858368,
+      "learning_rate": 3.432873498042444e-08,
+      "loss": 0.5196,
+      "step": 5036
+    },
+    {
+      "epoch": 4.834373499759962,
+      "grad_norm": 0.9348920356925896,
+      "learning_rate": 3.3937970683934893e-08,
+      "loss": 0.4301,
+      "step": 5037
+    },
+    {
+      "epoch": 4.8353336533845415,
+      "grad_norm": 1.2536392714464282,
+      "learning_rate": 3.354943559960577e-08,
+      "loss": 0.4297,
+      "step": 5038
+    },
+    {
+      "epoch": 4.836293807009121,
+      "grad_norm": 0.8043557599899609,
+      "learning_rate": 3.3163129901847e-08,
+      "loss": 0.4234,
+      "step": 5039
+    },
+    {
+      "epoch": 4.837253960633701,
+      "grad_norm": 1.4866701324310767,
+      "learning_rate": 3.277905376406654e-08,
+      "loss": 0.4822,
+      "step": 5040
+    },
+    {
+      "epoch": 4.838214114258282,
+      "grad_norm": 1.0122077616772496,
+      "learning_rate": 3.2397207358672024e-08,
+      "loss": 0.4176,
+      "step": 5041
+    },
+    {
+      "epoch": 4.8391742678828615,
+      "grad_norm": 1.2445365938766917,
+      "learning_rate": 3.201759085707079e-08,
+      "loss": 0.4098,
+      "step": 5042
+    },
+    {
+      "epoch": 4.840134421507441,
+      "grad_norm": 0.9607029819505497,
+      "learning_rate": 3.164020442966764e-08,
+      "loss": 0.449,
+      "step": 5043
+    },
+    {
+      "epoch": 4.841094575132021,
+      "grad_norm": 1.0185484393961937,
+      "learning_rate": 3.126504824586818e-08,
+      "loss": 0.3862,
+      "step": 5044
+    },
+    {
+      "epoch": 4.842054728756601,
+      "grad_norm": 0.8549733541522867,
+      "learning_rate": 3.0892122474076025e-08,
+      "loss": 0.4063,
+      "step": 5045
+    },
+    {
+      "epoch": 4.843014882381181,
+      "grad_norm": 1.0335421384351278,
+      "learning_rate": 3.0521427281693386e-08,
+      "loss": 0.3933,
+      "step": 5046
+    },
+    {
+      "epoch": 4.8439750360057605,
+      "grad_norm": 1.2391368799898677,
+      "learning_rate": 3.015296283512048e-08,
+      "loss": 0.3339,
+      "step": 5047
+    },
+    {
+      "epoch": 4.844935189630341,
+      "grad_norm": 1.2509800578794419,
+      "learning_rate": 2.9786729299758343e-08,
+      "loss": 0.4457,
+      "step": 5048
+    },
+    {
+      "epoch": 4.845895343254921,
+      "grad_norm": 0.9671288407106493,
+      "learning_rate": 2.942272684000491e-08,
+      "loss": 0.54,
+      "step": 5049
+    },
+    {
+      "epoch": 4.846855496879501,
+      "grad_norm": 1.1520318135178587,
+      "learning_rate": 2.9060955619256702e-08,
+      "loss": 0.372,
+      "step": 5050
+    },
+    {
+      "epoch": 4.847815650504081,
+      "grad_norm": 0.9319144448116563,
+      "learning_rate": 2.8701415799908815e-08,
+      "loss": 0.4941,
+      "step": 5051
+    },
+    {
+      "epoch": 4.84877580412866,
+      "grad_norm": 1.070882565350098,
+      "learning_rate": 2.8344107543356035e-08,
+      "loss": 0.565,
+      "step": 5052
+    },
+    {
+      "epoch": 4.84973595775324,
+      "grad_norm": 1.164358503633835,
+      "learning_rate": 2.7989031009988398e-08,
+      "loss": 0.4455,
+      "step": 5053
+    },
+    {
+      "epoch": 4.85069611137782,
+      "grad_norm": 0.9576045972050365,
+      "learning_rate": 2.763618635919729e-08,
+      "loss": 0.4493,
+      "step": 5054
+    },
+    {
+      "epoch": 4.851656265002401,
+      "grad_norm": 0.9859052075934536,
+      "learning_rate": 2.7285573749370465e-08,
+      "loss": 0.4165,
+      "step": 5055
+    },
+    {
+      "epoch": 4.8526164186269805,
+      "grad_norm": 0.9759591952474944,
+      "learning_rate": 2.693719333789424e-08,
+      "loss": 0.3959,
+      "step": 5056
+    },
+    {
+      "epoch": 4.85357657225156,
+      "grad_norm": 1.6916228334762529,
+      "learning_rate": 2.6591045281152418e-08,
+      "loss": 0.4815,
+      "step": 5057
+    },
+    {
+      "epoch": 4.85453672587614,
+      "grad_norm": 1.1637623283854825,
+      "learning_rate": 2.6247129734528474e-08,
+      "loss": 0.4088,
+      "step": 5058
+    },
+    {
+      "epoch": 4.85549687950072,
+      "grad_norm": 0.9643874496992656,
+      "learning_rate": 2.5905446852400594e-08,
+      "loss": 0.5052,
+      "step": 5059
+    },
+    {
+      "epoch": 4.8564570331253005,
+      "grad_norm": 1.4048550230246748,
+      "learning_rate": 2.5565996788148306e-08,
+      "loss": 0.427,
+      "step": 5060
+    },
+    {
+      "epoch": 4.85741718674988,
+      "grad_norm": 0.9317953985958425,
+      "learning_rate": 2.5228779694145832e-08,
+      "loss": 0.3779,
+      "step": 5061
+    },
+    {
+      "epoch": 4.85837734037446,
+      "grad_norm": 0.8975106174930008,
+      "learning_rate": 2.4893795721766534e-08,
+      "loss": 0.4195,
+      "step": 5062
+    },
+    {
+      "epoch": 4.85933749399904,
+      "grad_norm": 0.7927814892674755,
+      "learning_rate": 2.4561045021382346e-08,
+      "loss": 0.485,
+      "step": 5063
+    },
+    {
+      "epoch": 4.86029764762362,
+      "grad_norm": 0.8720161667507453,
+      "learning_rate": 2.4230527742359344e-08,
+      "loss": 0.4973,
+      "step": 5064
+    },
+    {
+      "epoch": 4.8612578012481995,
+      "grad_norm": 1.2585093263141438,
+      "learning_rate": 2.3902244033065512e-08,
+      "loss": 0.4572,
+      "step": 5065
+    },
+    {
+      "epoch": 4.862217954872779,
+      "grad_norm": 1.6788789791069185,
+      "learning_rate": 2.3576194040861866e-08,
+      "loss": 0.4782,
+      "step": 5066
+    },
+    {
+      "epoch": 4.86317810849736,
+      "grad_norm": 1.414839687090021,
+      "learning_rate": 2.3252377912110214e-08,
+      "loss": 0.4326,
+      "step": 5067
+    },
+    {
+      "epoch": 4.86413826212194,
+      "grad_norm": 0.977068520548313,
+      "learning_rate": 2.2930795792167615e-08,
+      "loss": 0.3233,
+      "step": 5068
+    },
+    {
+      "epoch": 4.86509841574652,
+      "grad_norm": 0.868506054639685,
+      "learning_rate": 2.2611447825389154e-08,
+      "loss": 0.4162,
+      "step": 5069
+    },
+    {
+      "epoch": 4.866058569371099,
+      "grad_norm": 0.7750212166806832,
+      "learning_rate": 2.2294334155125718e-08,
+      "loss": 0.4173,
+      "step": 5070
+    },
+    {
+      "epoch": 4.867018722995679,
+      "grad_norm": 0.9064544188794017,
+      "learning_rate": 2.1979454923727327e-08,
+      "loss": 0.368,
+      "step": 5071
+    },
+    {
+      "epoch": 4.867978876620259,
+      "grad_norm": 0.992579494216837,
+      "learning_rate": 2.1666810272539806e-08,
+      "loss": 0.3901,
+      "step": 5072
+    },
+    {
+      "epoch": 4.868939030244839,
+      "grad_norm": 1.2247131772624031,
+      "learning_rate": 2.1356400341905337e-08,
+      "loss": 0.4961,
+      "step": 5073
+    },
+    {
+      "epoch": 4.8698991838694194,
+      "grad_norm": 1.2712259325875028,
+      "learning_rate": 2.1048225271164125e-08,
+      "loss": 0.4253,
+      "step": 5074
+    },
+    {
+      "epoch": 4.870859337493999,
+      "grad_norm": 1.1591275819276787,
+      "learning_rate": 2.0742285198652734e-08,
+      "loss": 0.4923,
+      "step": 5075
+    },
+    {
+      "epoch": 4.871819491118579,
+      "grad_norm": 0.8658085315099535,
+      "learning_rate": 2.0438580261704087e-08,
+      "loss": 0.4324,
+      "step": 5076
+    },
+    {
+      "epoch": 4.872779644743159,
+      "grad_norm": 1.039298014186302,
+      "learning_rate": 2.0137110596648025e-08,
+      "loss": 0.4165,
+      "step": 5077
+    },
+    {
+      "epoch": 4.873739798367739,
+      "grad_norm": 1.44535186697573,
+      "learning_rate": 1.9837876338811292e-08,
+      "loss": 0.3649,
+      "step": 5078
+    },
+    {
+      "epoch": 4.874699951992318,
+      "grad_norm": 1.245908551410283,
+      "learning_rate": 1.9540877622516995e-08,
+      "loss": 0.4106,
+      "step": 5079
+    },
+    {
+      "epoch": 4.875660105616898,
+      "grad_norm": 1.0781166639004742,
+      "learning_rate": 1.924611458108461e-08,
+      "loss": 0.3914,
+      "step": 5080
+    },
+    {
+      "epoch": 4.876620259241479,
+      "grad_norm": 1.4211230254085798,
+      "learning_rate": 1.8953587346829948e-08,
+      "loss": 0.383,
+      "step": 5081
+    },
+    {
+      "epoch": 4.877580412866059,
+      "grad_norm": 1.5517652180259598,
+      "learning_rate": 1.8663296051065762e-08,
+      "loss": 0.4071,
+      "step": 5082
+    },
+    {
+      "epoch": 4.8785405664906385,
+      "grad_norm": 1.2114114580023057,
+      "learning_rate": 1.8375240824100028e-08,
+      "loss": 0.378,
+      "step": 5083
+    },
+    {
+      "epoch": 4.879500720115218,
+      "grad_norm": 1.0513556944457456,
+      "learning_rate": 1.8089421795238205e-08,
+      "loss": 0.3535,
+      "step": 5084
+    },
+    {
+      "epoch": 4.880460873739798,
+      "grad_norm": 1.4968238903736093,
+      "learning_rate": 1.7805839092781553e-08,
+      "loss": 0.465,
+      "step": 5085
+    },
+    {
+      "epoch": 4.881421027364379,
+      "grad_norm": 1.1405007768093178,
+      "learning_rate": 1.7524492844026576e-08,
+      "loss": 0.4803,
+      "step": 5086
+    },
+    {
+      "epoch": 4.882381180988959,
+      "grad_norm": 1.3866638610378546,
+      "learning_rate": 1.724538317526725e-08,
+      "loss": 0.3496,
+      "step": 5087
+    },
+    {
+      "epoch": 4.883341334613538,
+      "grad_norm": 1.0187832194723374,
+      "learning_rate": 1.69685102117928e-08,
+      "loss": 0.3701,
+      "step": 5088
+    },
+    {
+      "epoch": 4.884301488238118,
+      "grad_norm": 1.4084212742864863,
+      "learning_rate": 1.6693874077888805e-08,
+      "loss": 0.438,
+      "step": 5089
+    },
+    {
+      "epoch": 4.885261641862698,
+      "grad_norm": 1.0536901913621108,
+      "learning_rate": 1.6421474896835544e-08,
+      "loss": 0.4685,
+      "step": 5090
+    },
+    {
+      "epoch": 4.886221795487278,
+      "grad_norm": 1.5334535275759493,
+      "learning_rate": 1.615131279091131e-08,
+      "loss": 0.3611,
+      "step": 5091
+    },
+    {
+      "epoch": 4.8871819491118575,
+      "grad_norm": 1.2611632826678283,
+      "learning_rate": 1.588338788138799e-08,
+      "loss": 0.577,
+      "step": 5092
+    },
+    {
+      "epoch": 4.888142102736438,
+      "grad_norm": 0.9408967358501868,
+      "learning_rate": 1.5617700288534377e-08,
+      "loss": 0.4033,
+      "step": 5093
+    },
+    {
+      "epoch": 4.889102256361018,
+      "grad_norm": 0.9182097698873428,
+      "learning_rate": 1.5354250131615622e-08,
+      "loss": 0.5375,
+      "step": 5094
+    },
+    {
+      "epoch": 4.890062409985598,
+      "grad_norm": 1.240463356630294,
+      "learning_rate": 1.5093037528890463e-08,
+      "loss": 0.3847,
+      "step": 5095
+    },
+    {
+      "epoch": 4.891022563610178,
+      "grad_norm": 1.0300359990394057,
+      "learning_rate": 1.4834062597615107e-08,
+      "loss": 0.3564,
+      "step": 5096
+    },
+    {
+      "epoch": 4.891982717234757,
+      "grad_norm": 1.1013158862073584,
+      "learning_rate": 1.4577325454041003e-08,
+      "loss": 0.4245,
+      "step": 5097
+    },
+    {
+      "epoch": 4.892942870859337,
+      "grad_norm": 1.106545797520422,
+      "learning_rate": 1.4322826213414298e-08,
+      "loss": 0.5749,
+      "step": 5098
+    },
+    {
+      "epoch": 4.893903024483917,
+      "grad_norm": 0.960040175368356,
+      "learning_rate": 1.4070564989976387e-08,
+      "loss": 0.3848,
+      "step": 5099
+    },
+    {
+      "epoch": 4.894863178108498,
+      "grad_norm": 1.3938887097504613,
+      "learning_rate": 1.3820541896965579e-08,
+      "loss": 0.4164,
+      "step": 5100
+    },
+    {
+      "epoch": 4.8958233317330775,
+      "grad_norm": 1.6583964106689468,
+      "learning_rate": 1.3572757046614315e-08,
+      "loss": 0.388,
+      "step": 5101
+    },
+    {
+      "epoch": 4.896783485357657,
+      "grad_norm": 1.3875353619009965,
+      "learning_rate": 1.3327210550150293e-08,
+      "loss": 0.4067,
+      "step": 5102
+    },
+    {
+      "epoch": 4.897743638982237,
+      "grad_norm": 0.9864938493522539,
+      "learning_rate": 1.308390251779701e-08,
+      "loss": 0.3546,
+      "step": 5103
+    },
+    {
+      "epoch": 4.898703792606817,
+      "grad_norm": 1.4634855282024979,
+      "learning_rate": 1.2842833058772653e-08,
+      "loss": 0.4595,
+      "step": 5104
+    },
+    {
+      "epoch": 4.899663946231397,
+      "grad_norm": 1.3928005048454455,
+      "learning_rate": 1.260400228129066e-08,
+      "loss": 0.32,
+      "step": 5105
+    },
+    {
+      "epoch": 4.9006240998559765,
+      "grad_norm": 1.3315193818393363,
+      "learning_rate": 1.2367410292560277e-08,
+      "loss": 0.4512,
+      "step": 5106
+    },
+    {
+      "epoch": 4.901584253480557,
+      "grad_norm": 1.402112709594031,
+      "learning_rate": 1.2133057198784326e-08,
+      "loss": 0.4613,
+      "step": 5107
+    },
+    {
+      "epoch": 4.902544407105137,
+      "grad_norm": 0.9662704664685008,
+      "learning_rate": 1.1900943105160878e-08,
+      "loss": 0.5237,
+      "step": 5108
+    },
+    {
+      "epoch": 4.903504560729717,
+      "grad_norm": 0.9720094231775382,
+      "learning_rate": 1.1671068115884921e-08,
+      "loss": 0.3733,
+      "step": 5109
+    },
+    {
+      "epoch": 4.9044647143542965,
+      "grad_norm": 1.328747938924517,
+      "learning_rate": 1.1443432334144466e-08,
+      "loss": 0.3991,
+      "step": 5110
+    },
+    {
+      "epoch": 4.905424867978876,
+      "grad_norm": 0.9644265096027246,
+      "learning_rate": 1.1218035862122223e-08,
+      "loss": 0.5053,
+      "step": 5111
+    },
+    {
+      "epoch": 4.906385021603457,
+      "grad_norm": 1.322136554202964,
+      "learning_rate": 1.09948788009967e-08,
+      "loss": 0.3923,
+      "step": 5112
+    },
+    {
+      "epoch": 4.907345175228037,
+      "grad_norm": 1.428478812006693,
+      "learning_rate": 1.0773961250939991e-08,
+      "loss": 0.4183,
+      "step": 5113
+    },
+    {
+      "epoch": 4.908305328852617,
+      "grad_norm": 1.5965621060221347,
+      "learning_rate": 1.0555283311121101e-08,
+      "loss": 0.4922,
+      "step": 5114
+    },
+    {
+      "epoch": 4.909265482477196,
+      "grad_norm": 1.1610655346822698,
+      "learning_rate": 1.0338845079700954e-08,
+      "loss": 0.4809,
+      "step": 5115
+    },
+    {
+      "epoch": 4.910225636101776,
+      "grad_norm": 0.8931306602718819,
+      "learning_rate": 1.0124646653836279e-08,
+      "loss": 0.5179,
+      "step": 5116
+    },
+    {
+      "epoch": 4.911185789726356,
+      "grad_norm": 1.1999844882912658,
+      "learning_rate": 9.912688129679603e-09,
+      "loss": 0.34,
+      "step": 5117
+    },
+    {
+      "epoch": 4.912145943350936,
+      "grad_norm": 1.524911140052869,
+      "learning_rate": 9.702969602375378e-09,
+      "loss": 0.3102,
+      "step": 5118
+    },
+    {
+      "epoch": 4.9131060969755165,
+      "grad_norm": 1.1888212322198293,
+      "learning_rate": 9.495491166065519e-09,
+      "loss": 0.4304,
+      "step": 5119
+    },
+    {
+      "epoch": 4.914066250600096,
+      "grad_norm": 0.8370339215857046,
+      "learning_rate": 9.290252913883856e-09,
+      "loss": 0.5062,
+      "step": 5120
+    },
+    {
+      "epoch": 4.915026404224676,
+      "grad_norm": 0.8838455024416435,
+      "learning_rate": 9.087254937960032e-09,
+      "loss": 0.415,
+      "step": 5121
+    },
+    {
+      "epoch": 4.915986557849256,
+      "grad_norm": 1.4903560415545014,
+      "learning_rate": 8.886497329417265e-09,
+      "loss": 0.3926,
+      "step": 5122
+    },
+    {
+      "epoch": 4.916946711473836,
+      "grad_norm": 0.9053606944330268,
+      "learning_rate": 8.687980178374023e-09,
+      "loss": 0.415,
+      "step": 5123
+    },
+    {
+      "epoch": 4.9179068650984155,
+      "grad_norm": 0.9647380161249303,
+      "learning_rate": 8.491703573942356e-09,
+      "loss": 0.4446,
+      "step": 5124
+    },
+    {
+      "epoch": 4.918867018722995,
+      "grad_norm": 1.3693617755150729,
+      "learning_rate": 8.29766760422901e-09,
+      "loss": 0.4558,
+      "step": 5125
+    },
+    {
+      "epoch": 4.919827172347576,
+      "grad_norm": 1.2179187424465863,
+      "learning_rate": 8.105872356333755e-09,
+      "loss": 0.3484,
+      "step": 5126
+    },
+    {
+      "epoch": 4.920787325972156,
+      "grad_norm": 1.5332175038998923,
+      "learning_rate": 7.916317916352168e-09,
+      "loss": 0.4337,
+      "step": 5127
+    },
+    {
+      "epoch": 4.9217474795967355,
+      "grad_norm": 1.0003918034790997,
+      "learning_rate": 7.729004369373405e-09,
+      "loss": 0.4873,
+      "step": 5128
+    },
+    {
+      "epoch": 4.922707633221315,
+      "grad_norm": 0.9205647155475724,
+      "learning_rate": 7.543931799479653e-09,
+      "loss": 0.4052,
+      "step": 5129
+    },
+    {
+      "epoch": 4.923667786845895,
+      "grad_norm": 1.4386377949692875,
+      "learning_rate": 7.3611002897489015e-09,
+      "loss": 0.4085,
+      "step": 5130
+    },
+    {
+      "epoch": 4.924627940470475,
+      "grad_norm": 1.3599067463748153,
+      "learning_rate": 7.180509922251611e-09,
+      "loss": 0.4257,
+      "step": 5131
+    },
+    {
+      "epoch": 4.925588094095055,
+      "grad_norm": 1.139163702921157,
+      "learning_rate": 7.002160778053491e-09,
+      "loss": 0.4469,
+      "step": 5132
+    },
+    {
+      "epoch": 4.926548247719635,
+      "grad_norm": 1.3645010341460393,
+      "learning_rate": 6.826052937212724e-09,
+      "loss": 0.417,
+      "step": 5133
+    },
+    {
+      "epoch": 4.927508401344215,
+      "grad_norm": 1.3643477934842472,
+      "learning_rate": 6.652186478783296e-09,
+      "loss": 0.4737,
+      "step": 5134
+    },
+    {
+      "epoch": 4.928468554968795,
+      "grad_norm": 0.8448587410954269,
+      "learning_rate": 6.480561480811109e-09,
+      "loss": 0.4151,
+      "step": 5135
+    },
+    {
+      "epoch": 4.929428708593375,
+      "grad_norm": 1.0354494543652244,
+      "learning_rate": 6.311178020336761e-09,
+      "loss": 0.4958,
+      "step": 5136
+    },
+    {
+      "epoch": 4.930388862217955,
+      "grad_norm": 1.2147326495868704,
+      "learning_rate": 6.14403617339554e-09,
+      "loss": 0.5032,
+      "step": 5137
+    },
+    {
+      "epoch": 4.931349015842535,
+      "grad_norm": 0.9696535144120022,
+      "learning_rate": 5.979136015015208e-09,
+      "loss": 0.5898,
+      "step": 5138
+    },
+    {
+      "epoch": 4.932309169467115,
+      "grad_norm": 0.9746702442415063,
+      "learning_rate": 5.8164776192171094e-09,
+      "loss": 0.398,
+      "step": 5139
+    },
+    {
+      "epoch": 4.933269323091695,
+      "grad_norm": 1.3285842986567158,
+      "learning_rate": 5.656061059017837e-09,
+      "loss": 0.4384,
+      "step": 5140
+    },
+    {
+      "epoch": 4.934229476716275,
+      "grad_norm": 1.3033312799211494,
+      "learning_rate": 5.4978864064259005e-09,
+      "loss": 0.3526,
+      "step": 5141
+    },
+    {
+      "epoch": 4.9351896303408544,
+      "grad_norm": 0.9126781389699786,
+      "learning_rate": 5.3419537324445044e-09,
+      "loss": 0.454,
+      "step": 5142
+    },
+    {
+      "epoch": 4.936149783965434,
+      "grad_norm": 1.0040080868664603,
+      "learning_rate": 5.18826310706988e-09,
+      "loss": 0.4419,
+      "step": 5143
+    },
+    {
+      "epoch": 4.937109937590014,
+      "grad_norm": 0.9796725198047096,
+      "learning_rate": 5.0368145992929495e-09,
+      "loss": 0.4626,
+      "step": 5144
+    },
+    {
+      "epoch": 4.938070091214595,
+      "grad_norm": 1.3805326857551758,
+      "learning_rate": 4.887608277096001e-09,
+      "loss": 0.3561,
+      "step": 5145
+    },
+    {
+      "epoch": 4.9390302448391745,
+      "grad_norm": 1.0428810022184316,
+      "learning_rate": 4.740644207457678e-09,
+      "loss": 0.5836,
+      "step": 5146
+    },
+    {
+      "epoch": 4.939990398463754,
+      "grad_norm": 1.238303417596389,
+      "learning_rate": 4.5959224563474345e-09,
+      "loss": 0.4241,
+      "step": 5147
+    },
+    {
+      "epoch": 4.940950552088334,
+      "grad_norm": 1.235297425078616,
+      "learning_rate": 4.4534430887299694e-09,
+      "loss": 0.4415,
+      "step": 5148
+    },
+    {
+      "epoch": 4.941910705712914,
+      "grad_norm": 1.5714181493805923,
+      "learning_rate": 4.3132061685619005e-09,
+      "loss": 0.4448,
+      "step": 5149
+    },
+    {
+      "epoch": 4.942870859337494,
+      "grad_norm": 1.4210360472425325,
+      "learning_rate": 4.1752117587950945e-09,
+      "loss": 0.3297,
+      "step": 5150
+    },
+    {
+      "epoch": 4.9438310129620735,
+      "grad_norm": 0.906595718766946,
+      "learning_rate": 4.0394599213733346e-09,
+      "loss": 0.4804,
+      "step": 5151
+    },
+    {
+      "epoch": 4.944791166586654,
+      "grad_norm": 1.6793102175034818,
+      "learning_rate": 3.905950717233986e-09,
+      "loss": 0.3784,
+      "step": 5152
+    },
+    {
+      "epoch": 4.945751320211234,
+      "grad_norm": 1.0401983747434733,
+      "learning_rate": 3.774684206308554e-09,
+      "loss": 0.3924,
+      "step": 5153
+    },
+    {
+      "epoch": 4.946711473835814,
+      "grad_norm": 1.0420523529139116,
+      "learning_rate": 3.6456604475199055e-09,
+      "loss": 0.4338,
+      "step": 5154
+    },
+    {
+      "epoch": 4.947671627460394,
+      "grad_norm": 1.0285533862088443,
+      "learning_rate": 3.5188794987867093e-09,
+      "loss": 0.5015,
+      "step": 5155
+    },
+    {
+      "epoch": 4.948631781084973,
+      "grad_norm": 1.0652557416318498,
+      "learning_rate": 3.394341417018998e-09,
+      "loss": 0.417,
+      "step": 5156
+    },
+    {
+      "epoch": 4.949591934709553,
+      "grad_norm": 0.8211314402017954,
+      "learning_rate": 3.272046258120942e-09,
+      "loss": 0.4742,
+      "step": 5157
+    },
+    {
+      "epoch": 4.950552088334134,
+      "grad_norm": 0.9730651546065674,
+      "learning_rate": 3.151994076988629e-09,
+      "loss": 0.4185,
+      "step": 5158
+    },
+    {
+      "epoch": 4.951512241958714,
+      "grad_norm": 1.550655918576398,
+      "learning_rate": 3.034184927513395e-09,
+      "loss": 0.353,
+      "step": 5159
+    },
+    {
+      "epoch": 4.952472395583293,
+      "grad_norm": 1.0009479773329741,
+      "learning_rate": 2.9186188625779376e-09,
+      "loss": 0.3225,
+      "step": 5160
+    },
+    {
+      "epoch": 4.953432549207873,
+      "grad_norm": 1.1117566628566564,
+      "learning_rate": 2.805295934058538e-09,
+      "loss": 0.3916,
+      "step": 5161
+    },
+    {
+      "epoch": 4.954392702832453,
+      "grad_norm": 0.9949207272477235,
+      "learning_rate": 2.694216192824506e-09,
+      "loss": 0.4388,
+      "step": 5162
+    },
+    {
+      "epoch": 4.955352856457033,
+      "grad_norm": 0.8870075449913225,
+      "learning_rate": 2.5853796887387316e-09,
+      "loss": 0.4247,
+      "step": 5163
+    },
+    {
+      "epoch": 4.9563130100816135,
+      "grad_norm": 1.1345176979813556,
+      "learning_rate": 2.47878647065658e-09,
+      "loss": 0.3433,
+      "step": 5164
+    },
+    {
+      "epoch": 4.957273163706193,
+      "grad_norm": 1.0651711888218474,
+      "learning_rate": 2.374436586426443e-09,
+      "loss": 0.4421,
+      "step": 5165
+    },
+    {
+      "epoch": 4.958233317330773,
+      "grad_norm": 1.2129388831427408,
+      "learning_rate": 2.27233008289085e-09,
+      "loss": 0.4443,
+      "step": 5166
+    },
+    {
+      "epoch": 4.959193470955353,
+      "grad_norm": 1.3664212214230247,
+      "learning_rate": 2.1724670058825838e-09,
+      "loss": 0.4382,
+      "step": 5167
+    },
+    {
+      "epoch": 4.960153624579933,
+      "grad_norm": 1.04263073981288,
+      "learning_rate": 2.074847400230784e-09,
+      "loss": 0.398,
+      "step": 5168
+    },
+    {
+      "epoch": 4.9611137782045125,
+      "grad_norm": 1.1722104802613906,
+      "learning_rate": 1.9794713097548433e-09,
+      "loss": 0.5069,
+      "step": 5169
+    },
+    {
+      "epoch": 4.962073931829092,
+      "grad_norm": 0.8243771578752006,
+      "learning_rate": 1.886338777268293e-09,
+      "loss": 0.4242,
+      "step": 5170
+    },
+    {
+      "epoch": 4.963034085453673,
+      "grad_norm": 1.1138825072860248,
+      "learning_rate": 1.7954498445771352e-09,
+      "loss": 0.489,
+      "step": 5171
+    },
+    {
+      "epoch": 4.963994239078253,
+      "grad_norm": 1.1685493389282025,
+      "learning_rate": 1.7068045524815114e-09,
+      "loss": 0.4924,
+      "step": 5172
+    },
+    {
+      "epoch": 4.964954392702833,
+      "grad_norm": 0.8628981710697552,
+      "learning_rate": 1.6204029407718147e-09,
+      "loss": 0.4389,
+      "step": 5173
+    },
+    {
+      "epoch": 4.965914546327412,
+      "grad_norm": 0.9752931832801603,
+      "learning_rate": 1.5362450482336865e-09,
+      "loss": 0.5359,
+      "step": 5174
+    },
+    {
+      "epoch": 4.966874699951992,
+      "grad_norm": 0.9944015147203397,
+      "learning_rate": 1.4543309126446858e-09,
+      "loss": 0.4481,
+      "step": 5175
+    },
+    {
+      "epoch": 4.967834853576572,
+      "grad_norm": 1.6993954156604336,
+      "learning_rate": 1.374660570774844e-09,
+      "loss": 0.3301,
+      "step": 5176
+    },
+    {
+      "epoch": 4.968795007201152,
+      "grad_norm": 0.7550334465591608,
+      "learning_rate": 1.2972340583877751e-09,
+      "loss": 0.4681,
+      "step": 5177
+    },
+    {
+      "epoch": 4.969755160825732,
+      "grad_norm": 1.4877848652477177,
+      "learning_rate": 1.2220514102390114e-09,
+      "loss": 0.468,
+      "step": 5178
+    },
+    {
+      "epoch": 4.970715314450312,
+      "grad_norm": 1.2603219991752135,
+      "learning_rate": 1.149112660077667e-09,
+      "loss": 0.3782,
+      "step": 5179
+    },
+    {
+      "epoch": 4.971675468074892,
+      "grad_norm": 1.5932806416619052,
+      "learning_rate": 1.0784178406442192e-09,
+      "loss": 0.3616,
+      "step": 5180
+    },
+    {
+      "epoch": 4.972635621699472,
+      "grad_norm": 0.9456974079383582,
+      "learning_rate": 1.0099669836743931e-09,
+      "loss": 0.4605,
+      "step": 5181
+    },
+    {
+      "epoch": 4.973595775324052,
+      "grad_norm": 1.0605197578624472,
+      "learning_rate": 9.43760119893611e-10,
+      "loss": 0.3115,
+      "step": 5182
+    },
+    {
+      "epoch": 4.974555928948631,
+      "grad_norm": 0.9515697992755997,
+      "learning_rate": 8.797972790219877e-10,
+      "loss": 0.3714,
+      "step": 5183
+    },
+    {
+      "epoch": 4.975516082573212,
+      "grad_norm": 1.4573860961194787,
+      "learning_rate": 8.180784897715565e-10,
+      "loss": 0.457,
+      "step": 5184
+    },
+    {
+      "epoch": 4.976476236197792,
+      "grad_norm": 1.015072706107275,
+      "learning_rate": 7.586037798479329e-10,
+      "loss": 0.4341,
+      "step": 5185
+    },
+    {
+      "epoch": 4.977436389822372,
+      "grad_norm": 0.8891637134337853,
+      "learning_rate": 7.013731759475395e-10,
+      "loss": 0.4277,
+      "step": 5186
+    },
+    {
+      "epoch": 4.9783965434469515,
+      "grad_norm": 1.232368610611868,
+      "learning_rate": 6.463867037614923e-10,
+      "loss": 0.3846,
+      "step": 5187
+    },
+    {
+      "epoch": 4.979356697071531,
+      "grad_norm": 0.8693382528761674,
+      "learning_rate": 5.936443879728249e-10,
+      "loss": 0.4826,
+      "step": 5188
+    },
+    {
+      "epoch": 4.980316850696111,
+      "grad_norm": 0.8860195787172938,
+      "learning_rate": 5.431462522559327e-10,
+      "loss": 0.4861,
+      "step": 5189
+    },
+    {
+      "epoch": 4.981277004320692,
+      "grad_norm": 0.9964801950407475,
+      "learning_rate": 4.948923192793497e-10,
+      "loss": 0.5072,
+      "step": 5190
+    },
+    {
+      "epoch": 4.9822371579452716,
+      "grad_norm": 0.9280557523722671,
+      "learning_rate": 4.4888261070408223e-10,
+      "loss": 0.5056,
+      "step": 5191
+    },
+    {
+      "epoch": 4.983197311569851,
+      "grad_norm": 1.129209496152681,
+      "learning_rate": 4.0511714718305396e-10,
+      "loss": 0.53,
+      "step": 5192
+    },
+    {
+      "epoch": 4.984157465194431,
+      "grad_norm": 1.2597070459144213,
+      "learning_rate": 3.6359594836277156e-10,
+      "loss": 0.4663,
+      "step": 5193
+    },
+    {
+      "epoch": 4.985117618819011,
+      "grad_norm": 0.9751585238349448,
+      "learning_rate": 3.24319032880549e-10,
+      "loss": 0.382,
+      "step": 5194
+    },
+    {
+      "epoch": 4.986077772443591,
+      "grad_norm": 1.481484926785769,
+      "learning_rate": 2.872864183678381e-10,
+      "loss": 0.3733,
+      "step": 5195
+    },
+    {
+      "epoch": 4.9870379260681705,
+      "grad_norm": 1.4032654679426264,
+      "learning_rate": 2.5249812144856335e-10,
+      "loss": 0.4836,
+      "step": 5196
+    },
+    {
+      "epoch": 4.987998079692751,
+      "grad_norm": 0.9277705194905103,
+      "learning_rate": 2.1995415773856666e-10,
+      "loss": 0.5167,
+      "step": 5197
+    },
+    {
+      "epoch": 4.988958233317331,
+      "grad_norm": 1.2258570435859233,
+      "learning_rate": 1.8965454184671773e-10,
+      "loss": 0.387,
+      "step": 5198
+    },
+    {
+      "epoch": 4.989918386941911,
+      "grad_norm": 0.9924078353603917,
+      "learning_rate": 1.615992873732486e-10,
+      "loss": 0.4717,
+      "step": 5199
+    },
+    {
+      "epoch": 4.990878540566491,
+      "grad_norm": 0.8285429982259878,
+      "learning_rate": 1.3578840691308438e-10,
+      "loss": 0.424,
+      "step": 5200
+    },
+    {
+      "epoch": 4.99183869419107,
+      "grad_norm": 0.9218774212055307,
+      "learning_rate": 1.1222191205195743e-10,
+      "loss": 0.4128,
+      "step": 5201
+    },
+    {
+      "epoch": 4.99279884781565,
+      "grad_norm": 1.2712175336651776,
+      "learning_rate": 9.089981336807275e-11,
+      "loss": 0.3499,
+      "step": 5202
+    },
+    {
+      "epoch": 4.99375900144023,
+      "grad_norm": 0.9820496496723373,
+      "learning_rate": 7.182212043377323e-11,
+      "loss": 0.3324,
+      "step": 5203
+    },
+    {
+      "epoch": 4.994719155064811,
+      "grad_norm": 1.2192713071695829,
+      "learning_rate": 5.4988841812209095e-11,
+      "loss": 0.3896,
+      "step": 5204
+    },
+    {
+      "epoch": 4.9956793086893905,
+      "grad_norm": 1.1385866933018771,
+      "learning_rate": 4.039998505900311e-11,
+      "loss": 0.405,
+      "step": 5205
+    },
+    {
+      "epoch": 4.99663946231397,
+      "grad_norm": 1.0288482489490556,
+      "learning_rate": 2.8055556724471132e-11,
+      "loss": 0.5846,
+      "step": 5206
+    },
+    {
+      "epoch": 4.99759961593855,
+      "grad_norm": 1.2764357875772903,
+      "learning_rate": 1.7955562348626054e-11,
+      "loss": 0.5251,
+      "step": 5207
+    },
+    {
+      "epoch": 4.99855976956313,
+      "grad_norm": 1.2640086373758026,
+      "learning_rate": 1.0100006465618706e-11,
+      "loss": 0.3624,
+      "step": 5208
+    },
+    {
+      "epoch": 4.99951992318771,
+      "grad_norm": 1.049982002681759,
+      "learning_rate": 4.4888926020725254e-12,
+      "loss": 0.4065,
+      "step": 5209
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.049982002681759,
+      "learning_rate": 1.1222232765284447e-12,
+      "loss": 0.362,
+      "step": 5210
+    },
+    {
+      "epoch": 5.0,
+      "step": 5210,
+      "total_flos": 432224790970368.0,
+      "train_loss": 0.5731080526353759,
+      "train_runtime": 83358.552,
+      "train_samples_per_second": 0.5,
+      "train_steps_per_second": 0.063
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5210,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 432224790970368.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}