diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7252 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.999242079733212,
+  "eval_steps": 500,
+  "global_step": 1030,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004850689707442777,
+      "grad_norm": 6.339065858846749,
+      "learning_rate": 7.766990291262136e-07,
+      "loss": 1.0147,
+      "step": 1
+    },
+    {
+      "epoch": 0.009701379414885554,
+      "grad_norm": 6.370901874406726,
+      "learning_rate": 1.5533980582524272e-06,
+      "loss": 1.0217,
+      "step": 2
+    },
+    {
+      "epoch": 0.01455206912232833,
+      "grad_norm": 6.253076167725343,
+      "learning_rate": 2.330097087378641e-06,
+      "loss": 1.0059,
+      "step": 3
+    },
+    {
+      "epoch": 0.01940275882977111,
+      "grad_norm": 5.8627331903052005,
+      "learning_rate": 3.1067961165048544e-06,
+      "loss": 0.9987,
+      "step": 4
+    },
+    {
+      "epoch": 0.024253448537213885,
+      "grad_norm": 4.624724616975588,
+      "learning_rate": 3.883495145631068e-06,
+      "loss": 0.9654,
+      "step": 5
+    },
+    {
+      "epoch": 0.02910413824465666,
+      "grad_norm": 2.701973440331149,
+      "learning_rate": 4.660194174757282e-06,
+      "loss": 0.9221,
+      "step": 6
+    },
+    {
+      "epoch": 0.03395482795209944,
+      "grad_norm": 2.5821400220833683,
+      "learning_rate": 5.436893203883496e-06,
+      "loss": 0.9118,
+      "step": 7
+    },
+    {
+      "epoch": 0.03880551765954222,
+      "grad_norm": 3.1473551148693146,
+      "learning_rate": 6.213592233009709e-06,
+      "loss": 0.8818,
+      "step": 8
+    },
+    {
+      "epoch": 0.04365620736698499,
+      "grad_norm": 3.8230150062051638,
+      "learning_rate": 6.990291262135923e-06,
+      "loss": 0.8912,
+      "step": 9
+    },
+    {
+      "epoch": 0.04850689707442777,
+      "grad_norm": 3.4669236063777715,
+      "learning_rate": 7.766990291262136e-06,
+      "loss": 0.8779,
+      "step": 10
+    },
+    {
+      "epoch": 0.053357586781870546,
+      "grad_norm": 2.712055876575345,
+      "learning_rate": 8.54368932038835e-06,
+      "loss": 0.8328,
+      "step": 11
+    },
+    {
+      "epoch": 0.05820827648931332,
+      "grad_norm": 2.574263966000136,
+      "learning_rate": 9.320388349514565e-06,
+      "loss": 0.8149,
+      "step": 12
+    },
+    {
+      "epoch": 0.0630589661967561,
+      "grad_norm": 1.8016385817876701,
+      "learning_rate": 1.0097087378640778e-05,
+      "loss": 0.7941,
+      "step": 13
+    },
+    {
+      "epoch": 0.06790965590419888,
+      "grad_norm": 1.16110696712433,
+      "learning_rate": 1.0873786407766991e-05,
+      "loss": 0.7751,
+      "step": 14
+    },
+    {
+      "epoch": 0.07276034561164166,
+      "grad_norm": 1.5441894400920566,
+      "learning_rate": 1.1650485436893204e-05,
+      "loss": 0.7603,
+      "step": 15
+    },
+    {
+      "epoch": 0.07761103531908443,
+      "grad_norm": 1.3570174190036193,
+      "learning_rate": 1.2427184466019418e-05,
+      "loss": 0.7456,
+      "step": 16
+    },
+    {
+      "epoch": 0.08246172502652721,
+      "grad_norm": 1.0160879152766609,
+      "learning_rate": 1.3203883495145633e-05,
+      "loss": 0.74,
+      "step": 17
+    },
+    {
+      "epoch": 0.08731241473396999,
+      "grad_norm": 1.195923679791525,
+      "learning_rate": 1.3980582524271846e-05,
+      "loss": 0.7223,
+      "step": 18
+    },
+    {
+      "epoch": 0.09216310444141276,
+      "grad_norm": 1.0381307779091873,
+      "learning_rate": 1.475728155339806e-05,
+      "loss": 0.7149,
+      "step": 19
+    },
+    {
+      "epoch": 0.09701379414885554,
+      "grad_norm": 0.9191697728302082,
+      "learning_rate": 1.5533980582524273e-05,
+      "loss": 0.7032,
+      "step": 20
+    },
+    {
+      "epoch": 0.10186448385629832,
+      "grad_norm": 1.0389109685950821,
+      "learning_rate": 1.6310679611650486e-05,
+      "loss": 0.697,
+      "step": 21
+    },
+    {
+      "epoch": 0.10671517356374109,
+      "grad_norm": 0.8528569833940303,
+      "learning_rate": 1.70873786407767e-05,
+      "loss": 0.6913,
+      "step": 22
+    },
+    {
+      "epoch": 0.11156586327118387,
+      "grad_norm": 0.7397574673832126,
+      "learning_rate": 1.7864077669902916e-05,
+      "loss": 0.6844,
+      "step": 23
+    },
+    {
+      "epoch": 0.11641655297862664,
+      "grad_norm": 0.6762376097915315,
+      "learning_rate": 1.864077669902913e-05,
+      "loss": 0.6807,
+      "step": 24
+    },
+    {
+      "epoch": 0.12126724268606942,
+      "grad_norm": 0.6801312007046909,
+      "learning_rate": 1.9417475728155343e-05,
+      "loss": 0.6651,
+      "step": 25
+    },
+    {
+      "epoch": 0.1261179323935122,
+      "grad_norm": 0.5373129321939298,
+      "learning_rate": 2.0194174757281556e-05,
+      "loss": 0.6661,
+      "step": 26
+    },
+    {
+      "epoch": 0.13096862210095497,
+      "grad_norm": 0.6155691125010336,
+      "learning_rate": 2.097087378640777e-05,
+      "loss": 0.6655,
+      "step": 27
+    },
+    {
+      "epoch": 0.13581931180839776,
+      "grad_norm": 0.5373412410981904,
+      "learning_rate": 2.1747572815533982e-05,
+      "loss": 0.6541,
+      "step": 28
+    },
+    {
+      "epoch": 0.14067000151584053,
+      "grad_norm": 0.5722094683121568,
+      "learning_rate": 2.2524271844660196e-05,
+      "loss": 0.6534,
+      "step": 29
+    },
+    {
+      "epoch": 0.14552069122328332,
+      "grad_norm": 0.6369873796903149,
+      "learning_rate": 2.330097087378641e-05,
+      "loss": 0.6536,
+      "step": 30
+    },
+    {
+      "epoch": 0.15037138093072608,
+      "grad_norm": 0.5246684440675834,
+      "learning_rate": 2.4077669902912622e-05,
+      "loss": 0.6545,
+      "step": 31
+    },
+    {
+      "epoch": 0.15522207063816887,
+      "grad_norm": 0.504247506683658,
+      "learning_rate": 2.4854368932038836e-05,
+      "loss": 0.6314,
+      "step": 32
+    },
+    {
+      "epoch": 0.16007276034561163,
+      "grad_norm": 0.5566944063536889,
+      "learning_rate": 2.5631067961165052e-05,
+      "loss": 0.6373,
+      "step": 33
+    },
+    {
+      "epoch": 0.16492345005305442,
+      "grad_norm": 0.8570989167580252,
+      "learning_rate": 2.6407766990291266e-05,
+      "loss": 0.63,
+      "step": 34
+    },
+    {
+      "epoch": 0.16977413976049718,
+      "grad_norm": 1.8667995731915865,
+      "learning_rate": 2.718446601941748e-05,
+      "loss": 0.6456,
+      "step": 35
+    },
+    {
+      "epoch": 0.17462482946793997,
+      "grad_norm": 0.5164194413667431,
+      "learning_rate": 2.7961165048543692e-05,
+      "loss": 0.6308,
+      "step": 36
+    },
+    {
+      "epoch": 0.17947551917538274,
+      "grad_norm": 1.6098330943831782,
+      "learning_rate": 2.8737864077669905e-05,
+      "loss": 0.6398,
+      "step": 37
+    },
+    {
+      "epoch": 0.18432620888282553,
+      "grad_norm": 1.1492129920694993,
+      "learning_rate": 2.951456310679612e-05,
+      "loss": 0.6276,
+      "step": 38
+    },
+    {
+      "epoch": 0.18917689859026832,
+      "grad_norm": 0.7127614761088336,
+      "learning_rate": 3.0291262135922332e-05,
+      "loss": 0.6218,
+      "step": 39
+    },
+    {
+      "epoch": 0.19402758829771108,
+      "grad_norm": 1.0312378150228299,
+      "learning_rate": 3.1067961165048545e-05,
+      "loss": 0.6253,
+      "step": 40
+    },
+    {
+      "epoch": 0.19887827800515387,
+      "grad_norm": 1.7574387168351864,
+      "learning_rate": 3.184466019417476e-05,
+      "loss": 0.6281,
+      "step": 41
+    },
+    {
+      "epoch": 0.20372896771259663,
+      "grad_norm": 0.9416599187328968,
+      "learning_rate": 3.262135922330097e-05,
+      "loss": 0.6235,
+      "step": 42
+    },
+    {
+      "epoch": 0.20857965742003942,
+      "grad_norm": 1.8301787236679616,
+      "learning_rate": 3.339805825242719e-05,
+      "loss": 0.622,
+      "step": 43
+    },
+    {
+      "epoch": 0.21343034712748218,
+      "grad_norm": 0.9292078210446757,
+      "learning_rate": 3.41747572815534e-05,
+      "loss": 0.6107,
+      "step": 44
+    },
+    {
+      "epoch": 0.21828103683492497,
+      "grad_norm": 1.994214362456412,
+      "learning_rate": 3.4951456310679615e-05,
+      "loss": 0.6151,
+      "step": 45
+    },
+    {
+      "epoch": 0.22313172654236774,
+      "grad_norm": 1.3619344540131681,
+      "learning_rate": 3.572815533980583e-05,
+      "loss": 0.6089,
+      "step": 46
+    },
+    {
+      "epoch": 0.22798241624981053,
+      "grad_norm": 1.6323821476629805,
+      "learning_rate": 3.650485436893204e-05,
+      "loss": 0.6165,
+      "step": 47
+    },
+    {
+      "epoch": 0.2328331059572533,
+      "grad_norm": 1.1474300438640261,
+      "learning_rate": 3.728155339805826e-05,
+      "loss": 0.6104,
+      "step": 48
+    },
+    {
+      "epoch": 0.23768379566469608,
+      "grad_norm": 1.1936059623728144,
+      "learning_rate": 3.805825242718447e-05,
+      "loss": 0.6086,
+      "step": 49
+    },
+    {
+      "epoch": 0.24253448537213884,
+      "grad_norm": 1.4126137333521573,
+      "learning_rate": 3.8834951456310685e-05,
+      "loss": 0.6123,
+      "step": 50
+    },
+    {
+      "epoch": 0.24738517507958163,
+      "grad_norm": 0.7835607330331523,
+      "learning_rate": 3.9611650485436895e-05,
+      "loss": 0.5979,
+      "step": 51
+    },
+    {
+      "epoch": 0.2522358647870244,
+      "grad_norm": 1.3575257469759314,
+      "learning_rate": 4.038834951456311e-05,
+      "loss": 0.6134,
+      "step": 52
+    },
+    {
+      "epoch": 0.25708655449446716,
+      "grad_norm": 0.9580505107131282,
+      "learning_rate": 4.116504854368932e-05,
+      "loss": 0.5987,
+      "step": 53
+    },
+    {
+      "epoch": 0.26193724420190995,
+      "grad_norm": 1.1647956860260527,
+      "learning_rate": 4.194174757281554e-05,
+      "loss": 0.6134,
+      "step": 54
+    },
+    {
+      "epoch": 0.26678793390935274,
+      "grad_norm": 0.9133078407230598,
+      "learning_rate": 4.271844660194175e-05,
+      "loss": 0.5978,
+      "step": 55
+    },
+    {
+      "epoch": 0.2716386236167955,
+      "grad_norm": 1.3505836368180404,
+      "learning_rate": 4.3495145631067965e-05,
+      "loss": 0.5944,
+      "step": 56
+    },
+    {
+      "epoch": 0.2764893133242383,
+      "grad_norm": 1.1628512424723636,
+      "learning_rate": 4.4271844660194175e-05,
+      "loss": 0.6052,
+      "step": 57
+    },
+    {
+      "epoch": 0.28134000303168105,
+      "grad_norm": 1.6421048813051027,
+      "learning_rate": 4.504854368932039e-05,
+      "loss": 0.6008,
+      "step": 58
+    },
+    {
+      "epoch": 0.28619069273912384,
+      "grad_norm": 1.3019802198524983,
+      "learning_rate": 4.58252427184466e-05,
+      "loss": 0.594,
+      "step": 59
+    },
+    {
+      "epoch": 0.29104138244656663,
+      "grad_norm": 1.2774402772060065,
+      "learning_rate": 4.660194174757282e-05,
+      "loss": 0.5892,
+      "step": 60
+    },
+    {
+      "epoch": 0.2958920721540094,
+      "grad_norm": 1.4435670386305743,
+      "learning_rate": 4.737864077669903e-05,
+      "loss": 0.5931,
+      "step": 61
+    },
+    {
+      "epoch": 0.30074276186145216,
+      "grad_norm": 1.0284857540916943,
+      "learning_rate": 4.8155339805825245e-05,
+      "loss": 0.5932,
+      "step": 62
+    },
+    {
+      "epoch": 0.30559345156889495,
+      "grad_norm": 1.1698103896183938,
+      "learning_rate": 4.8932038834951454e-05,
+      "loss": 0.5948,
+      "step": 63
+    },
+    {
+      "epoch": 0.31044414127633774,
+      "grad_norm": 1.434142725222452,
+      "learning_rate": 4.970873786407767e-05,
+      "loss": 0.5897,
+      "step": 64
+    },
+    {
+      "epoch": 0.3152948309837805,
+      "grad_norm": 1.3482179068151203,
+      "learning_rate": 5.0485436893203895e-05,
+      "loss": 0.5824,
+      "step": 65
+    },
+    {
+      "epoch": 0.32014552069122326,
+      "grad_norm": 1.3069227496554443,
+      "learning_rate": 5.1262135922330105e-05,
+      "loss": 0.5802,
+      "step": 66
+    },
+    {
+      "epoch": 0.32499621039866605,
+      "grad_norm": 1.1366236056516827,
+      "learning_rate": 5.203883495145632e-05,
+      "loss": 0.5804,
+      "step": 67
+    },
+    {
+      "epoch": 0.32984690010610884,
+      "grad_norm": 1.6243350603336242,
+      "learning_rate": 5.281553398058253e-05,
+      "loss": 0.5812,
+      "step": 68
+    },
+    {
+      "epoch": 0.33469758981355163,
+      "grad_norm": 1.0822457196908746,
+      "learning_rate": 5.359223300970875e-05,
+      "loss": 0.5862,
+      "step": 69
+    },
+    {
+      "epoch": 0.33954827952099437,
+      "grad_norm": 1.6360957248140573,
+      "learning_rate": 5.436893203883496e-05,
+      "loss": 0.5912,
+      "step": 70
+    },
+    {
+      "epoch": 0.34439896922843716,
+      "grad_norm": 2.3477845490488813,
+      "learning_rate": 5.5145631067961174e-05,
+      "loss": 0.5906,
+      "step": 71
+    },
+    {
+      "epoch": 0.34924965893587995,
+      "grad_norm": 1.105543560593242,
+      "learning_rate": 5.5922330097087384e-05,
+      "loss": 0.5824,
+      "step": 72
+    },
+    {
+      "epoch": 0.35410034864332274,
+      "grad_norm": 4.137098681881185,
+      "learning_rate": 5.66990291262136e-05,
+      "loss": 0.6359,
+      "step": 73
+    },
+    {
+      "epoch": 0.3589510383507655,
+      "grad_norm": 3.8847979837997033,
+      "learning_rate": 5.747572815533981e-05,
+      "loss": 0.6486,
+      "step": 74
+    },
+    {
+      "epoch": 0.36380172805820826,
+      "grad_norm": 1.1747128429519862,
+      "learning_rate": 5.825242718446603e-05,
+      "loss": 0.595,
+      "step": 75
+    },
+    {
+      "epoch": 0.36865241776565105,
+      "grad_norm": 3.008245900701061,
+      "learning_rate": 5.902912621359224e-05,
+      "loss": 0.6387,
+      "step": 76
+    },
+    {
+      "epoch": 0.37350310747309384,
+      "grad_norm": 2.3713075132931554,
+      "learning_rate": 5.9805825242718454e-05,
+      "loss": 0.6344,
+      "step": 77
+    },
+    {
+      "epoch": 0.37835379718053663,
+      "grad_norm": 1.6213053074921984,
+      "learning_rate": 6.0582524271844664e-05,
+      "loss": 0.6048,
+      "step": 78
+    },
+    {
+      "epoch": 0.38320448688797937,
+      "grad_norm": 1.6024154837501339,
+      "learning_rate": 6.135922330097087e-05,
+      "loss": 0.6203,
+      "step": 79
+    },
+    {
+      "epoch": 0.38805517659542216,
+      "grad_norm": 1.1825030677591377,
+      "learning_rate": 6.213592233009709e-05,
+      "loss": 0.6052,
+      "step": 80
+    },
+    {
+      "epoch": 0.39290586630286495,
+      "grad_norm": 1.3964525731128163,
+      "learning_rate": 6.291262135922331e-05,
+      "loss": 0.6112,
+      "step": 81
+    },
+    {
+      "epoch": 0.39775655601030774,
+      "grad_norm": 1.1205074621871551,
+      "learning_rate": 6.368932038834952e-05,
+      "loss": 0.5977,
+      "step": 82
+    },
+    {
+      "epoch": 0.4026072457177505,
+      "grad_norm": 1.045620374565707,
+      "learning_rate": 6.446601941747573e-05,
+      "loss": 0.5914,
+      "step": 83
+    },
+    {
+      "epoch": 0.40745793542519326,
+      "grad_norm": 1.3974586249408472,
+      "learning_rate": 6.524271844660194e-05,
+      "loss": 0.5918,
+      "step": 84
+    },
+    {
+      "epoch": 0.41230862513263605,
+      "grad_norm": 1.0818483302602913,
+      "learning_rate": 6.601941747572816e-05,
+      "loss": 0.5948,
+      "step": 85
+    },
+    {
+      "epoch": 0.41715931484007884,
+      "grad_norm": 0.9808456957793906,
+      "learning_rate": 6.679611650485438e-05,
+      "loss": 0.5839,
+      "step": 86
+    },
+    {
+      "epoch": 0.4220100045475216,
+      "grad_norm": 1.2035779456517084,
+      "learning_rate": 6.757281553398058e-05,
+      "loss": 0.5833,
+      "step": 87
+    },
+    {
+      "epoch": 0.42686069425496437,
+      "grad_norm": 1.6887623926979713,
+      "learning_rate": 6.83495145631068e-05,
+      "loss": 0.5818,
+      "step": 88
+    },
+    {
+      "epoch": 0.43171138396240716,
+      "grad_norm": 0.8023218391013366,
+      "learning_rate": 6.912621359223301e-05,
+      "loss": 0.5863,
+      "step": 89
+    },
+    {
+      "epoch": 0.43656207366984995,
+      "grad_norm": 1.574900348178855,
+      "learning_rate": 6.990291262135923e-05,
+      "loss": 0.5821,
+      "step": 90
+    },
+    {
+      "epoch": 0.4414127633772927,
+      "grad_norm": 0.9288518542917786,
+      "learning_rate": 7.067961165048545e-05,
+      "loss": 0.5814,
+      "step": 91
+    },
+    {
+      "epoch": 0.4462634530847355,
+      "grad_norm": 1.6871845487045471,
+      "learning_rate": 7.145631067961166e-05,
+      "loss": 0.5819,
+      "step": 92
+    },
+    {
+      "epoch": 0.45111414279217826,
+      "grad_norm": 1.147490028185953,
+      "learning_rate": 7.223300970873787e-05,
+      "loss": 0.5752,
+      "step": 93
+    },
+    {
+      "epoch": 0.45596483249962105,
+      "grad_norm": 1.0734179177901382,
+      "learning_rate": 7.300970873786408e-05,
+      "loss": 0.5786,
+      "step": 94
+    },
+    {
+      "epoch": 0.46081552220706384,
+      "grad_norm": 1.2263367009960806,
+      "learning_rate": 7.37864077669903e-05,
+      "loss": 0.5789,
+      "step": 95
+    },
+    {
+      "epoch": 0.4656662119145066,
+      "grad_norm": 1.4570032389620742,
+      "learning_rate": 7.456310679611652e-05,
+      "loss": 0.5745,
+      "step": 96
+    },
+    {
+      "epoch": 0.47051690162194937,
+      "grad_norm": 1.3246870726440927,
+      "learning_rate": 7.533980582524272e-05,
+      "loss": 0.5775,
+      "step": 97
+    },
+    {
+      "epoch": 0.47536759132939216,
+      "grad_norm": 0.9415857506868542,
+      "learning_rate": 7.611650485436894e-05,
+      "loss": 0.5699,
+      "step": 98
+    },
+    {
+      "epoch": 0.48021828103683495,
+      "grad_norm": 1.2384384474151087,
+      "learning_rate": 7.689320388349515e-05,
+      "loss": 0.5733,
+      "step": 99
+    },
+    {
+      "epoch": 0.4850689707442777,
+      "grad_norm": 1.5627749991572353,
+      "learning_rate": 7.766990291262137e-05,
+      "loss": 0.5735,
+      "step": 100
+    },
+    {
+      "epoch": 0.4899196604517205,
+      "grad_norm": 1.0078484211944914,
+      "learning_rate": 7.844660194174757e-05,
+      "loss": 0.5733,
+      "step": 101
+    },
+    {
+      "epoch": 0.49477035015916326,
+      "grad_norm": 1.6421211712488573,
+      "learning_rate": 7.922330097087379e-05,
+      "loss": 0.576,
+      "step": 102
+    },
+    {
+      "epoch": 0.49962103986660605,
+      "grad_norm": 0.8416126904816602,
+      "learning_rate": 8e-05,
+      "loss": 0.5697,
+      "step": 103
+    },
+    {
+      "epoch": 0.5044717295740488,
+      "grad_norm": 1.7540458195736903,
+      "learning_rate": 7.999977029531286e-05,
+      "loss": 0.5799,
+      "step": 104
+    },
+    {
+      "epoch": 0.5093224192814916,
+      "grad_norm": 1.050447231844734,
+      "learning_rate": 7.999908118388965e-05,
+      "loss": 0.5756,
+      "step": 105
+    },
+    {
+      "epoch": 0.5141731089889343,
+      "grad_norm": 1.394330877960354,
+      "learning_rate": 7.999793267364497e-05,
+      "loss": 0.5713,
+      "step": 106
+    },
+    {
+      "epoch": 0.5190237986963772,
+      "grad_norm": 1.5873678777006228,
+      "learning_rate": 7.999632477776974e-05,
+      "loss": 0.5733,
+      "step": 107
+    },
+    {
+      "epoch": 0.5238744884038199,
+      "grad_norm": 0.8115732123781836,
+      "learning_rate": 7.9994257514731e-05,
+      "loss": 0.5661,
+      "step": 108
+    },
+    {
+      "epoch": 0.5287251781112627,
+      "grad_norm": 1.1857754936411384,
+      "learning_rate": 7.999173090827177e-05,
+      "loss": 0.5719,
+      "step": 109
+    },
+    {
+      "epoch": 0.5335758678187055,
+      "grad_norm": 0.8066157800150973,
+      "learning_rate": 7.998874498741072e-05,
+      "loss": 0.5695,
+      "step": 110
+    },
+    {
+      "epoch": 0.5384265575261482,
+      "grad_norm": 1.4739854945603235,
+      "learning_rate": 7.998529978644183e-05,
+      "loss": 0.5712,
+      "step": 111
+    },
+    {
+      "epoch": 0.543277247233591,
+      "grad_norm": 0.778875512416499,
+      "learning_rate": 7.998139534493407e-05,
+      "loss": 0.5609,
+      "step": 112
+    },
+    {
+      "epoch": 0.5481279369410338,
+      "grad_norm": 0.9919152170469479,
+      "learning_rate": 7.997703170773084e-05,
+      "loss": 0.5648,
+      "step": 113
+    },
+    {
+      "epoch": 0.5529786266484766,
+      "grad_norm": 1.3093959851041357,
+      "learning_rate": 7.997220892494955e-05,
+      "loss": 0.5757,
+      "step": 114
+    },
+    {
+      "epoch": 0.5578293163559194,
+      "grad_norm": 0.9066117120369992,
+      "learning_rate": 7.996692705198097e-05,
+      "loss": 0.566,
+      "step": 115
+    },
+    {
+      "epoch": 0.5626800060633621,
+      "grad_norm": 1.2498381901490132,
+      "learning_rate": 7.996118614948869e-05,
+      "loss": 0.5757,
+      "step": 116
+    },
+    {
+      "epoch": 0.567530695770805,
+      "grad_norm": 1.0385189418340641,
+      "learning_rate": 7.995498628340827e-05,
+      "loss": 0.5697,
+      "step": 117
+    },
+    {
+      "epoch": 0.5723813854782477,
+      "grad_norm": 1.3065564110858372,
+      "learning_rate": 7.994832752494667e-05,
+      "loss": 0.5672,
+      "step": 118
+    },
+    {
+      "epoch": 0.5772320751856904,
+      "grad_norm": 0.9524061899396458,
+      "learning_rate": 7.994120995058127e-05,
+      "loss": 0.5624,
+      "step": 119
+    },
+    {
+      "epoch": 0.5820827648931333,
+      "grad_norm": 1.018979775317657,
+      "learning_rate": 7.993363364205907e-05,
+      "loss": 0.554,
+      "step": 120
+    },
+    {
+      "epoch": 0.586933454600576,
+      "grad_norm": 0.9276898951496916,
+      "learning_rate": 7.992559868639576e-05,
+      "loss": 0.556,
+      "step": 121
+    },
+    {
+      "epoch": 0.5917841443080188,
+      "grad_norm": 1.1930439188526805,
+      "learning_rate": 7.99171051758747e-05,
+      "loss": 0.5526,
+      "step": 122
+    },
+    {
+      "epoch": 0.5966348340154616,
+      "grad_norm": 0.6826863389197481,
+      "learning_rate": 7.990815320804583e-05,
+      "loss": 0.5609,
+      "step": 123
+    },
+    {
+      "epoch": 0.6014855237229043,
+      "grad_norm": 0.8061980858751947,
+      "learning_rate": 7.98987428857246e-05,
+      "loss": 0.5586,
+      "step": 124
+    },
+    {
+      "epoch": 0.6063362134303472,
+      "grad_norm": 0.9239948149603757,
+      "learning_rate": 7.988887431699079e-05,
+      "loss": 0.5507,
+      "step": 125
+    },
+    {
+      "epoch": 0.6111869031377899,
+      "grad_norm": 1.0465957889843347,
+      "learning_rate": 7.987854761518719e-05,
+      "loss": 0.5568,
+      "step": 126
+    },
+    {
+      "epoch": 0.6160375928452326,
+      "grad_norm": 0.8308130357408615,
+      "learning_rate": 7.986776289891842e-05,
+      "loss": 0.5591,
+      "step": 127
+    },
+    {
+      "epoch": 0.6208882825526755,
+      "grad_norm": 1.0246744766037437,
+      "learning_rate": 7.985652029204946e-05,
+      "loss": 0.5563,
+      "step": 128
+    },
+    {
+      "epoch": 0.6257389722601182,
+      "grad_norm": 1.3238612418839921,
+      "learning_rate": 7.984481992370429e-05,
+      "loss": 0.5491,
+      "step": 129
+    },
+    {
+      "epoch": 0.630589661967561,
+      "grad_norm": 0.6948258976249133,
+      "learning_rate": 7.983266192826437e-05,
+      "loss": 0.5418,
+      "step": 130
+    },
+    {
+      "epoch": 0.6354403516750038,
+      "grad_norm": 0.7288790333090353,
+      "learning_rate": 7.982004644536716e-05,
+      "loss": 0.5441,
+      "step": 131
+    },
+    {
+      "epoch": 0.6402910413824465,
+      "grad_norm": 1.0943204811023435,
+      "learning_rate": 7.98069736199044e-05,
+      "loss": 0.5493,
+      "step": 132
+    },
+    {
+      "epoch": 0.6451417310898894,
+      "grad_norm": 0.8706827515570799,
+      "learning_rate": 7.979344360202055e-05,
+      "loss": 0.5465,
+      "step": 133
+    },
+    {
+      "epoch": 0.6499924207973321,
+      "grad_norm": 0.9127380693761118,
+      "learning_rate": 7.977945654711108e-05,
+      "loss": 0.5475,
+      "step": 134
+    },
+    {
+      "epoch": 0.654843110504775,
+      "grad_norm": 0.9510188119086359,
+      "learning_rate": 7.976501261582056e-05,
+      "loss": 0.543,
+      "step": 135
+    },
+    {
+      "epoch": 0.6596938002122177,
+      "grad_norm": 0.9350726421156861,
+      "learning_rate": 7.975011197404092e-05,
+      "loss": 0.5525,
+      "step": 136
+    },
+    {
+      "epoch": 0.6645444899196604,
+      "grad_norm": 1.272258773056705,
+      "learning_rate": 7.973475479290956e-05,
+      "loss": 0.5518,
+      "step": 137
+    },
+    {
+      "epoch": 0.6693951796271033,
+      "grad_norm": 0.9765265325518906,
+      "learning_rate": 7.971894124880727e-05,
+      "loss": 0.5417,
+      "step": 138
+    },
+    {
+      "epoch": 0.674245869334546,
+      "grad_norm": 1.1917501609756302,
+      "learning_rate": 7.970267152335632e-05,
+      "loss": 0.5464,
+      "step": 139
+    },
+    {
+      "epoch": 0.6790965590419887,
+      "grad_norm": 0.6719257463868904,
+      "learning_rate": 7.968594580341832e-05,
+      "loss": 0.544,
+      "step": 140
+    },
+    {
+      "epoch": 0.6839472487494316,
+      "grad_norm": 0.7137208028607956,
+      "learning_rate": 7.966876428109209e-05,
+      "loss": 0.5351,
+      "step": 141
+    },
+    {
+      "epoch": 0.6887979384568743,
+      "grad_norm": 0.7449688014358767,
+      "learning_rate": 7.965112715371144e-05,
+      "loss": 0.5397,
+      "step": 142
+    },
+    {
+      "epoch": 0.6936486281643172,
+      "grad_norm": 0.5377305872108858,
+      "learning_rate": 7.96330346238429e-05,
+      "loss": 0.5346,
+      "step": 143
+    },
+    {
+      "epoch": 0.6984993178717599,
+      "grad_norm": 0.5856757070627496,
+      "learning_rate": 7.961448689928341e-05,
+      "loss": 0.5395,
+      "step": 144
+    },
+    {
+      "epoch": 0.7033500075792026,
+      "grad_norm": 0.6000325566823206,
+      "learning_rate": 7.959548419305796e-05,
+      "loss": 0.5447,
+      "step": 145
+    },
+    {
+      "epoch": 0.7082006972866455,
+      "grad_norm": 0.7819963257560868,
+      "learning_rate": 7.957602672341707e-05,
+      "loss": 0.5364,
+      "step": 146
+    },
+    {
+      "epoch": 0.7130513869940882,
+      "grad_norm": 1.1223449508846108,
+      "learning_rate": 7.955611471383433e-05,
+      "loss": 0.5381,
+      "step": 147
+    },
+    {
+      "epoch": 0.717902076701531,
+      "grad_norm": 1.1418981667975974,
+      "learning_rate": 7.953574839300385e-05,
+      "loss": 0.5381,
+      "step": 148
+    },
+    {
+      "epoch": 0.7227527664089738,
+      "grad_norm": 0.758286759296052,
+      "learning_rate": 7.95149279948376e-05,
+      "loss": 0.5398,
+      "step": 149
+    },
+    {
+      "epoch": 0.7276034561164165,
+      "grad_norm": 0.7637204957772546,
+      "learning_rate": 7.949365375846271e-05,
+      "loss": 0.5386,
+      "step": 150
+    },
+    {
+      "epoch": 0.7324541458238594,
+      "grad_norm": 0.6982030938329856,
+      "learning_rate": 7.94719259282188e-05,
+      "loss": 0.5328,
+      "step": 151
+    },
+    {
+      "epoch": 0.7373048355313021,
+      "grad_norm": 0.7115887055025976,
+      "learning_rate": 7.944974475365506e-05,
+      "loss": 0.5406,
+      "step": 152
+    },
+    {
+      "epoch": 0.7421555252387448,
+      "grad_norm": 0.6914212445412167,
+      "learning_rate": 7.94271104895275e-05,
+      "loss": 0.5375,
+      "step": 153
+    },
+    {
+      "epoch": 0.7470062149461877,
+      "grad_norm": 0.6376946136665823,
+      "learning_rate": 7.940402339579596e-05,
+      "loss": 0.5322,
+      "step": 154
+    },
+    {
+      "epoch": 0.7518569046536304,
+      "grad_norm": 0.662468788270689,
+      "learning_rate": 7.93804837376211e-05,
+      "loss": 0.5312,
+      "step": 155
+    },
+    {
+      "epoch": 0.7567075943610733,
+      "grad_norm": 0.8116591959883654,
+      "learning_rate": 7.935649178536142e-05,
+      "loss": 0.5362,
+      "step": 156
+    },
+    {
+      "epoch": 0.761558284068516,
+      "grad_norm": 1.2251954995336705,
+      "learning_rate": 7.93320478145701e-05,
+      "loss": 0.5454,
+      "step": 157
+    },
+    {
+      "epoch": 0.7664089737759587,
+      "grad_norm": 1.1753293382340935,
+      "learning_rate": 7.93071521059919e-05,
+      "loss": 0.5369,
+      "step": 158
+    },
+    {
+      "epoch": 0.7712596634834016,
+      "grad_norm": 0.5797209510428332,
+      "learning_rate": 7.928180494555983e-05,
+      "loss": 0.5255,
+      "step": 159
+    },
+    {
+      "epoch": 0.7761103531908443,
+      "grad_norm": 0.9260629876609666,
+      "learning_rate": 7.925600662439201e-05,
+      "loss": 0.535,
+      "step": 160
+    },
+    {
+      "epoch": 0.780961042898287,
+      "grad_norm": 1.2828815170548864,
+      "learning_rate": 7.922975743878817e-05,
+      "loss": 0.5293,
+      "step": 161
+    },
+    {
+      "epoch": 0.7858117326057299,
+      "grad_norm": 0.5171701751512903,
+      "learning_rate": 7.92030576902264e-05,
+      "loss": 0.5298,
+      "step": 162
+    },
+    {
+      "epoch": 0.7906624223131726,
+      "grad_norm": 0.9531992162184508,
+      "learning_rate": 7.917590768535952e-05,
+      "loss": 0.5315,
+      "step": 163
+    },
+    {
+      "epoch": 0.7955131120206155,
+      "grad_norm": 1.1106333536493764,
+      "learning_rate": 7.914830773601173e-05,
+      "loss": 0.5279,
+      "step": 164
+    },
+    {
+      "epoch": 0.8003638017280582,
+      "grad_norm": 0.45675301699833715,
+      "learning_rate": 7.912025815917489e-05,
+      "loss": 0.5372,
+      "step": 165
+    },
+    {
+      "epoch": 0.805214491435501,
+      "grad_norm": 0.7189143844280501,
+      "learning_rate": 7.909175927700499e-05,
+      "loss": 0.535,
+      "step": 166
+    },
+    {
+      "epoch": 0.8100651811429438,
+      "grad_norm": 0.8686823215812186,
+      "learning_rate": 7.906281141681839e-05,
+      "loss": 0.5333,
+      "step": 167
+    },
+    {
+      "epoch": 0.8149158708503865,
+      "grad_norm": 0.588440050652225,
+      "learning_rate": 7.903341491108798e-05,
+      "loss": 0.5289,
+      "step": 168
+    },
+    {
+      "epoch": 0.8197665605578294,
+      "grad_norm": 0.563873615495661,
+      "learning_rate": 7.900357009743958e-05,
+      "loss": 0.5331,
+      "step": 169
+    },
+    {
+      "epoch": 0.8246172502652721,
+      "grad_norm": 0.545022127050129,
+      "learning_rate": 7.897327731864784e-05,
+      "loss": 0.5266,
+      "step": 170
+    },
+    {
+      "epoch": 0.8294679399727148,
+      "grad_norm": 0.5592040867673563,
+      "learning_rate": 7.894253692263244e-05,
+      "loss": 0.522,
+      "step": 171
+    },
+    {
+      "epoch": 0.8343186296801577,
+      "grad_norm": 0.6324827822327501,
+      "learning_rate": 7.891134926245402e-05,
+      "loss": 0.5297,
+      "step": 172
+    },
+    {
+      "epoch": 0.8391693193876004,
+      "grad_norm": 0.8244822385641454,
+      "learning_rate": 7.887971469631016e-05,
+      "loss": 0.5319,
+      "step": 173
+    },
+    {
+      "epoch": 0.8440200090950432,
+      "grad_norm": 1.2087031591715138,
+      "learning_rate": 7.884763358753129e-05,
+      "loss": 0.5408,
+      "step": 174
+    },
+    {
+      "epoch": 0.848870698802486,
+      "grad_norm": 0.7427926592130248,
+      "learning_rate": 7.881510630457643e-05,
+      "loss": 0.5326,
+      "step": 175
+    },
+    {
+      "epoch": 0.8537213885099287,
+      "grad_norm": 0.5074413186000275,
+      "learning_rate": 7.878213322102908e-05,
+      "loss": 0.5281,
+      "step": 176
+    },
+    {
+      "epoch": 0.8585720782173716,
+      "grad_norm": 0.7191422226961306,
+      "learning_rate": 7.874871471559282e-05,
+      "loss": 0.5269,
+      "step": 177
+    },
+    {
+      "epoch": 0.8634227679248143,
+      "grad_norm": 0.7351841845171684,
+      "learning_rate": 7.8714851172087e-05,
+      "loss": 0.5326,
+      "step": 178
+    },
+    {
+      "epoch": 0.868273457632257,
+      "grad_norm": 0.661668481852005,
+      "learning_rate": 7.868054297944237e-05,
+      "loss": 0.5312,
+      "step": 179
+    },
+    {
+      "epoch": 0.8731241473396999,
+      "grad_norm": 0.6257843185205204,
+      "learning_rate": 7.864579053169657e-05,
+      "loss": 0.5265,
+      "step": 180
+    },
+    {
+      "epoch": 0.8779748370471426,
+      "grad_norm": 0.6272508186215946,
+      "learning_rate": 7.86105942279896e-05,
+      "loss": 0.5242,
+      "step": 181
+    },
+    {
+      "epoch": 0.8828255267545854,
+      "grad_norm": 0.5859524441087289,
+      "learning_rate": 7.857495447255925e-05,
+      "loss": 0.5117,
+      "step": 182
+    },
+    {
+      "epoch": 0.8876762164620282,
+      "grad_norm": 0.5314342493506371,
+      "learning_rate": 7.853887167473646e-05,
+      "loss": 0.5275,
+      "step": 183
+    },
+    {
+      "epoch": 0.892526906169471,
+      "grad_norm": 0.5894396606474899,
+      "learning_rate": 7.850234624894064e-05,
+      "loss": 0.5236,
+      "step": 184
+    },
+    {
+      "epoch": 0.8973775958769138,
+      "grad_norm": 0.7789309413273731,
+      "learning_rate": 7.846537861467485e-05,
+      "loss": 0.5269,
+      "step": 185
+    },
+    {
+      "epoch": 0.9022282855843565,
+      "grad_norm": 0.9749399305462054,
+      "learning_rate": 7.842796919652104e-05,
+      "loss": 0.5177,
+      "step": 186
+    },
+    {
+      "epoch": 0.9070789752917993,
+      "grad_norm": 0.8944031183061603,
+      "learning_rate": 7.839011842413514e-05,
+      "loss": 0.5236,
+      "step": 187
+    },
+    {
+      "epoch": 0.9119296649992421,
+      "grad_norm": 0.6072136307213187,
+      "learning_rate": 7.835182673224212e-05,
+      "loss": 0.5237,
+      "step": 188
+    },
+    {
+      "epoch": 0.9167803547066848,
+      "grad_norm": 0.44815995922956803,
+      "learning_rate": 7.831309456063107e-05,
+      "loss": 0.5193,
+      "step": 189
+    },
+    {
+      "epoch": 0.9216310444141277,
+      "grad_norm": 0.5362363058315522,
+      "learning_rate": 7.827392235415005e-05,
+      "loss": 0.5242,
+      "step": 190
+    },
+    {
+      "epoch": 0.9264817341215704,
+      "grad_norm": 0.5675209533538658,
+      "learning_rate": 7.823431056270103e-05,
+      "loss": 0.5223,
+      "step": 191
+    },
+    {
+      "epoch": 0.9313324238290132,
+      "grad_norm": 0.5087214285921436,
+      "learning_rate": 7.81942596412347e-05,
+      "loss": 0.522,
+      "step": 192
+    },
+    {
+      "epoch": 0.936183113536456,
+      "grad_norm": 0.46687413642954234,
+      "learning_rate": 7.815377004974532e-05,
+      "loss": 0.509,
+      "step": 193
+    },
+    {
+      "epoch": 0.9410338032438987,
+      "grad_norm": 0.43127006261715695,
+      "learning_rate": 7.811284225326529e-05,
+      "loss": 0.522,
+      "step": 194
+    },
+    {
+      "epoch": 0.9458844929513415,
+      "grad_norm": 0.43502529088952246,
+      "learning_rate": 7.807147672185996e-05,
+      "loss": 0.5258,
+      "step": 195
+    },
+    {
+      "epoch": 0.9507351826587843,
+      "grad_norm": 0.5247431038277521,
+      "learning_rate": 7.802967393062219e-05,
+      "loss": 0.524,
+      "step": 196
+    },
+    {
+      "epoch": 0.955585872366227,
+      "grad_norm": 0.6873590048364063,
+      "learning_rate": 7.798743435966676e-05,
+      "loss": 0.5227,
+      "step": 197
+    },
+    {
+      "epoch": 0.9604365620736699,
+      "grad_norm": 0.8978763261543932,
+      "learning_rate": 7.794475849412512e-05,
+      "loss": 0.5143,
+      "step": 198
+    },
+    {
+      "epoch": 0.9652872517811126,
+      "grad_norm": 0.8491329782406589,
+      "learning_rate": 7.790164682413954e-05,
+      "loss": 0.5186,
+      "step": 199
+    },
+    {
+      "epoch": 0.9701379414885554,
+      "grad_norm": 0.5004070361840594,
+      "learning_rate": 7.785809984485765e-05,
+      "loss": 0.5185,
+      "step": 200
+    },
+    {
+      "epoch": 0.9749886311959982,
+      "grad_norm": 0.4620691718653308,
+      "learning_rate": 7.781411805642675e-05,
+      "loss": 0.5179,
+      "step": 201
+    },
+    {
+      "epoch": 0.979839320903441,
+      "grad_norm": 0.6078145800613444,
+      "learning_rate": 7.776970196398795e-05,
+      "loss": 0.5185,
+      "step": 202
+    },
+    {
+      "epoch": 0.9846900106108837,
+      "grad_norm": 0.6744449960922595,
+      "learning_rate": 7.77248520776705e-05,
+      "loss": 0.5223,
+      "step": 203
+    },
+    {
+      "epoch": 0.9895407003183265,
+      "grad_norm": 0.769858159356013,
+      "learning_rate": 7.767956891258585e-05,
+      "loss": 0.514,
+      "step": 204
+    },
+    {
+      "epoch": 0.9943913900257693,
+      "grad_norm": 0.9703823352556248,
+      "learning_rate": 7.763385298882177e-05,
+      "loss": 0.5227,
+      "step": 205
+    },
+    {
+      "epoch": 0.9992420797332121,
+      "grad_norm": 1.0790500827883502,
+      "learning_rate": 7.758770483143634e-05,
+      "loss": 0.5161,
+      "step": 206
+    },
+    {
+      "epoch": 1.0048506897074427,
+      "grad_norm": 0.979399858326228,
+      "learning_rate": 7.754112497045198e-05,
+      "loss": 0.5126,
+      "step": 207
+    },
+    {
+      "epoch": 1.0097013794148855,
+      "grad_norm": 0.8061955074549858,
+      "learning_rate": 7.749411394084931e-05,
+      "loss": 0.5093,
+      "step": 208
+    },
+    {
+      "epoch": 1.0145520691223284,
+      "grad_norm": 0.769561002371883,
+      "learning_rate": 7.744667228256102e-05,
+      "loss": 0.5129,
+      "step": 209
+    },
+    {
+      "epoch": 1.0194027588297712,
+      "grad_norm": 0.7563794216226459,
+      "learning_rate": 7.739880054046567e-05,
+      "loss": 0.504,
+      "step": 210
+    },
+    {
+      "epoch": 1.024253448537214,
+      "grad_norm": 0.4861095042439798,
+      "learning_rate": 7.735049926438143e-05,
+      "loss": 0.5008,
+      "step": 211
+    },
+    {
+      "epoch": 1.0291041382446566,
+      "grad_norm": 0.521763707621908,
+      "learning_rate": 7.730176900905978e-05,
+      "loss": 0.5039,
+      "step": 212
+    },
+    {
+      "epoch": 1.0339548279520994,
+      "grad_norm": 0.6847654097095645,
+      "learning_rate": 7.725261033417914e-05,
+      "loss": 0.4987,
+      "step": 213
+    },
+    {
+      "epoch": 1.0388055176595423,
+      "grad_norm": 0.45617035587869154,
+      "learning_rate": 7.720302380433838e-05,
+      "loss": 0.5082,
+      "step": 214
+    },
+    {
+      "epoch": 1.043656207366985,
+      "grad_norm": 0.5136841747695677,
+      "learning_rate": 7.715300998905045e-05,
+      "loss": 0.4903,
+      "step": 215
+    },
+    {
+      "epoch": 1.0485068970744278,
+      "grad_norm": 0.5201029165395914,
+      "learning_rate": 7.710256946273572e-05,
+      "loss": 0.5061,
+      "step": 216
+    },
+    {
+      "epoch": 1.0533575867818705,
+      "grad_norm": 0.5369451831716502,
+      "learning_rate": 7.705170280471546e-05,
+      "loss": 0.4923,
+      "step": 217
+    },
+    {
+      "epoch": 1.0582082764893133,
+      "grad_norm": 0.5292650633659572,
+      "learning_rate": 7.700041059920516e-05,
+      "loss": 0.4958,
+      "step": 218
+    },
+    {
+      "epoch": 1.0630589661967562,
+      "grad_norm": 0.38347803199441816,
+      "learning_rate": 7.694869343530781e-05,
+      "loss": 0.4949,
+      "step": 219
+    },
+    {
+      "epoch": 1.067909655904199,
+      "grad_norm": 0.42205008813365347,
+      "learning_rate": 7.689655190700719e-05,
+      "loss": 0.4958,
+      "step": 220
+    },
+    {
+      "epoch": 1.0727603456116417,
+      "grad_norm": 0.43326896920779806,
+      "learning_rate": 7.684398661316092e-05,
+      "loss": 0.5034,
+      "step": 221
+    },
+    {
+      "epoch": 1.0776110353190844,
+      "grad_norm": 0.3610486041319715,
+      "learning_rate": 7.679099815749377e-05,
+      "loss": 0.508,
+      "step": 222
+    },
+    {
+      "epoch": 1.0824617250265272,
+      "grad_norm": 0.4631910613487445,
+      "learning_rate": 7.673758714859052e-05,
+      "loss": 0.5012,
+      "step": 223
+    },
+    {
+      "epoch": 1.08731241473397,
+      "grad_norm": 0.5933537887890846,
+      "learning_rate": 7.668375419988918e-05,
+      "loss": 0.5003,
+      "step": 224
+    },
+    {
+      "epoch": 1.0921631044414128,
+      "grad_norm": 0.7536616463697116,
+      "learning_rate": 7.662949992967375e-05,
+      "loss": 0.4984,
+      "step": 225
+    },
+    {
+      "epoch": 1.0970137941488556,
+      "grad_norm": 0.8908783317227219,
+      "learning_rate": 7.657482496106725e-05,
+      "loss": 0.499,
+      "step": 226
+    },
+    {
+      "epoch": 1.1018644838562983,
+      "grad_norm": 0.8949365262495667,
+      "learning_rate": 7.651972992202449e-05,
+      "loss": 0.4964,
+      "step": 227
+    },
+    {
+      "epoch": 1.106715173563741,
+      "grad_norm": 0.7666080530207662,
+      "learning_rate": 7.646421544532492e-05,
+      "loss": 0.501,
+      "step": 228
+    },
+    {
+      "epoch": 1.1115658632711838,
+      "grad_norm": 0.5974548399149405,
+      "learning_rate": 7.640828216856532e-05,
+      "loss": 0.5019,
+      "step": 229
+    },
+    {
+      "epoch": 1.1164165529786267,
+      "grad_norm": 0.5208576739553936,
+      "learning_rate": 7.635193073415246e-05,
+      "loss": 0.4954,
+      "step": 230
+    },
+    {
+      "epoch": 1.1212672426860695,
+      "grad_norm": 0.4449218126901089,
+      "learning_rate": 7.62951617892958e-05,
+      "loss": 0.4966,
+      "step": 231
+    },
+    {
+      "epoch": 1.1261179323935122,
+      "grad_norm": 0.3413546355163314,
+      "learning_rate": 7.623797598599995e-05,
+      "loss": 0.4869,
+      "step": 232
+    },
+    {
+      "epoch": 1.130968622100955,
+      "grad_norm": 0.4021494906350838,
+      "learning_rate": 7.618037398105728e-05,
+      "loss": 0.4876,
+      "step": 233
+    },
+    {
+      "epoch": 1.1358193118083977,
+      "grad_norm": 0.5401093810240136,
+      "learning_rate": 7.612235643604031e-05,
+      "loss": 0.495,
+      "step": 234
+    },
+    {
+      "epoch": 1.1406700015158404,
+      "grad_norm": 0.6450843987922678,
+      "learning_rate": 7.606392401729415e-05,
+      "loss": 0.4953,
+      "step": 235
+    },
+    {
+      "epoch": 1.1455206912232834,
+      "grad_norm": 0.6413460171528268,
+      "learning_rate": 7.600507739592879e-05,
+      "loss": 0.4972,
+      "step": 236
+    },
+    {
+      "epoch": 1.150371380930726,
+      "grad_norm": 0.6276158556341717,
+      "learning_rate": 7.594581724781152e-05,
+      "loss": 0.4957,
+      "step": 237
+    },
+    {
+      "epoch": 1.1552220706381688,
+      "grad_norm": 0.7553018024510589,
+      "learning_rate": 7.588614425355898e-05,
+      "loss": 0.4955,
+      "step": 238
+    },
+    {
+      "epoch": 1.1600727603456116,
+      "grad_norm": 0.8968837770832118,
+      "learning_rate": 7.582605909852951e-05,
+      "loss": 0.4937,
+      "step": 239
+    },
+    {
+      "epoch": 1.1649234500530543,
+      "grad_norm": 0.7618870978199445,
+      "learning_rate": 7.576556247281522e-05,
+      "loss": 0.4969,
+      "step": 240
+    },
+    {
+      "epoch": 1.1697741397604973,
+      "grad_norm": 0.6531717186517569,
+      "learning_rate": 7.570465507123401e-05,
+      "loss": 0.5042,
+      "step": 241
+    },
+    {
+      "epoch": 1.17462482946794,
+      "grad_norm": 0.591515239054212,
+      "learning_rate": 7.564333759332167e-05,
+      "loss": 0.4906,
+      "step": 242
+    },
+    {
+      "epoch": 1.1794755191753827,
+      "grad_norm": 0.49797581923871925,
+      "learning_rate": 7.558161074332379e-05,
+      "loss": 0.4966,
+      "step": 243
+    },
+    {
+      "epoch": 1.1843262088828255,
+      "grad_norm": 0.5865910810734263,
+      "learning_rate": 7.551947523018774e-05,
+      "loss": 0.4997,
+      "step": 244
+    },
+    {
+      "epoch": 1.1891768985902682,
+      "grad_norm": 0.5003536808662635,
+      "learning_rate": 7.54569317675544e-05,
+      "loss": 0.4954,
+      "step": 245
+    },
+    {
+      "epoch": 1.1940275882977112,
+      "grad_norm": 0.4328856052217075,
+      "learning_rate": 7.539398107375015e-05,
+      "loss": 0.4979,
+      "step": 246
+    },
+    {
+      "epoch": 1.198878278005154,
+      "grad_norm": 0.5623661241326378,
+      "learning_rate": 7.533062387177843e-05,
+      "loss": 0.4982,
+      "step": 247
+    },
+    {
+      "epoch": 1.2037289677125966,
+      "grad_norm": 0.542587238982675,
+      "learning_rate": 7.526686088931156e-05,
+      "loss": 0.4991,
+      "step": 248
+    },
+    {
+      "epoch": 1.2085796574200394,
+      "grad_norm": 0.4949553127282243,
+      "learning_rate": 7.520269285868235e-05,
+      "loss": 0.4908,
+      "step": 249
+    },
+    {
+      "epoch": 1.213430347127482,
+      "grad_norm": 0.39917468694971237,
+      "learning_rate": 7.513812051687564e-05,
+      "loss": 0.4917,
+      "step": 250
+    },
+    {
+      "epoch": 1.218281036834925,
+      "grad_norm": 0.44831825611716425,
+      "learning_rate": 7.507314460551993e-05,
+      "loss": 0.4898,
+      "step": 251
+    },
+    {
+      "epoch": 1.2231317265423678,
+      "grad_norm": 0.4694728642442923,
+      "learning_rate": 7.500776587087878e-05,
+      "loss": 0.4929,
+      "step": 252
+    },
+    {
+      "epoch": 1.2279824162498105,
+      "grad_norm": 0.4548469944052975,
+      "learning_rate": 7.494198506384229e-05,
+      "loss": 0.4826,
+      "step": 253
+    },
+    {
+      "epoch": 1.2328331059572533,
+      "grad_norm": 0.6666754202357293,
+      "learning_rate": 7.487580293991844e-05,
+      "loss": 0.5021,
+      "step": 254
+    },
+    {
+      "epoch": 1.237683795664696,
+      "grad_norm": 0.6791995766586792,
+      "learning_rate": 7.480922025922443e-05,
+      "loss": 0.4974,
+      "step": 255
+    },
+    {
+      "epoch": 1.242534485372139,
+      "grad_norm": 0.48795487898431833,
+      "learning_rate": 7.474223778647796e-05,
+      "loss": 0.4934,
+      "step": 256
+    },
+    {
+      "epoch": 1.2473851750795817,
+      "grad_norm": 0.32007742081193336,
+      "learning_rate": 7.467485629098842e-05,
+      "loss": 0.4907,
+      "step": 257
+    },
+    {
+      "epoch": 1.2522358647870244,
+      "grad_norm": 0.2948779105297471,
+      "learning_rate": 7.460707654664807e-05,
+      "loss": 0.4974,
+      "step": 258
+    },
+    {
+      "epoch": 1.2570865544944672,
+      "grad_norm": 0.2954711132741566,
+      "learning_rate": 7.453889933192316e-05,
+      "loss": 0.4893,
+      "step": 259
+    },
+    {
+      "epoch": 1.26193724420191,
+      "grad_norm": 0.3234156565448516,
+      "learning_rate": 7.447032542984502e-05,
+      "loss": 0.4882,
+      "step": 260
+    },
+    {
+      "epoch": 1.2667879339093528,
+      "grad_norm": 0.37736907562809757,
+      "learning_rate": 7.440135562800093e-05,
+      "loss": 0.487,
+      "step": 261
+    },
+    {
+      "epoch": 1.2716386236167956,
+      "grad_norm": 0.4380964102542513,
+      "learning_rate": 7.433199071852526e-05,
+      "loss": 0.4965,
+      "step": 262
+    },
+    {
+      "epoch": 1.2764893133242383,
+      "grad_norm": 0.5377640268616285,
+      "learning_rate": 7.426223149809023e-05,
+      "loss": 0.4922,
+      "step": 263
+    },
+    {
+      "epoch": 1.281340003031681,
+      "grad_norm": 0.5447192325063156,
+      "learning_rate": 7.419207876789685e-05,
+      "loss": 0.4844,
+      "step": 264
+    },
+    {
+      "epoch": 1.2861906927391238,
+      "grad_norm": 0.5569387883590096,
+      "learning_rate": 7.412153333366567e-05,
+      "loss": 0.4887,
+      "step": 265
+    },
+    {
+      "epoch": 1.2910413824465667,
+      "grad_norm": 0.6127559785319273,
+      "learning_rate": 7.405059600562751e-05,
+      "loss": 0.4974,
+      "step": 266
+    },
+    {
+      "epoch": 1.2958920721540095,
+      "grad_norm": 0.5372861059475705,
+      "learning_rate": 7.397926759851425e-05,
+      "loss": 0.4946,
+      "step": 267
+    },
+    {
+      "epoch": 1.3007427618614522,
+      "grad_norm": 0.3715441206816088,
+      "learning_rate": 7.390754893154933e-05,
+      "loss": 0.4914,
+      "step": 268
+    },
+    {
+      "epoch": 1.305593451568895,
+      "grad_norm": 0.3468457624853246,
+      "learning_rate": 7.383544082843846e-05,
+      "loss": 0.4906,
+      "step": 269
+    },
+    {
+      "epoch": 1.3104441412763377,
+      "grad_norm": 0.37917435116867076,
+      "learning_rate": 7.376294411736009e-05,
+      "loss": 0.4877,
+      "step": 270
+    },
+    {
+      "epoch": 1.3152948309837806,
+      "grad_norm": 0.44133896758740837,
+      "learning_rate": 7.369005963095596e-05,
+      "loss": 0.4962,
+      "step": 271
+    },
+    {
+      "epoch": 1.3201455206912232,
+      "grad_norm": 0.5453886771483799,
+      "learning_rate": 7.361678820632145e-05,
+      "loss": 0.4918,
+      "step": 272
+    },
+    {
+      "epoch": 1.324996210398666,
+      "grad_norm": 0.6036321305780495,
+      "learning_rate": 7.354313068499607e-05,
+      "loss": 0.4892,
+      "step": 273
+    },
+    {
+      "epoch": 1.3298469001061088,
+      "grad_norm": 0.620838391242342,
+      "learning_rate": 7.346908791295369e-05,
+      "loss": 0.495,
+      "step": 274
+    },
+    {
+      "epoch": 1.3346975898135516,
+      "grad_norm": 0.5848284090531766,
+      "learning_rate": 7.339466074059292e-05,
+      "loss": 0.4862,
+      "step": 275
+    },
+    {
+      "epoch": 1.3395482795209943,
+      "grad_norm": 0.5915616847361488,
+      "learning_rate": 7.331985002272726e-05,
+      "loss": 0.4986,
+      "step": 276
+    },
+    {
+      "epoch": 1.344398969228437,
+      "grad_norm": 0.5871009859314463,
+      "learning_rate": 7.324465661857534e-05,
+      "loss": 0.4868,
+      "step": 277
+    },
+    {
+      "epoch": 1.34924965893588,
+      "grad_norm": 0.4325881770640798,
+      "learning_rate": 7.316908139175105e-05,
+      "loss": 0.4886,
+      "step": 278
+    },
+    {
+      "epoch": 1.3541003486433227,
+      "grad_norm": 0.342133338929163,
+      "learning_rate": 7.309312521025356e-05,
+      "loss": 0.4909,
+      "step": 279
+    },
+    {
+      "epoch": 1.3589510383507655,
+      "grad_norm": 0.4415991996544984,
+      "learning_rate": 7.301678894645742e-05,
+      "loss": 0.4915,
+      "step": 280
+    },
+    {
+      "epoch": 1.3638017280582082,
+      "grad_norm": 0.510599069072842,
+      "learning_rate": 7.294007347710251e-05,
+      "loss": 0.4935,
+      "step": 281
+    },
+    {
+      "epoch": 1.368652417765651,
+      "grad_norm": 0.5220646005978747,
+      "learning_rate": 7.286297968328397e-05,
+      "loss": 0.4834,
+      "step": 282
+    },
+    {
+      "epoch": 1.373503107473094,
+      "grad_norm": 0.4836929320122624,
+      "learning_rate": 7.27855084504421e-05,
+      "loss": 0.4953,
+      "step": 283
+    },
+    {
+      "epoch": 1.3783537971805366,
+      "grad_norm": 0.4678184328008887,
+      "learning_rate": 7.270766066835217e-05,
+      "loss": 0.4872,
+      "step": 284
+    },
+    {
+      "epoch": 1.3832044868879794,
+      "grad_norm": 0.4479819209137706,
+      "learning_rate": 7.262943723111419e-05,
+      "loss": 0.4916,
+      "step": 285
+    },
+    {
+      "epoch": 1.388055176595422,
+      "grad_norm": 0.44661460126650243,
+      "learning_rate": 7.255083903714266e-05,
+      "loss": 0.4866,
+      "step": 286
+    },
+    {
+      "epoch": 1.3929058663028648,
+      "grad_norm": 0.46700839086853646,
+      "learning_rate": 7.247186698915625e-05,
+      "loss": 0.4879,
+      "step": 287
+    },
+    {
+      "epoch": 1.3977565560103078,
+      "grad_norm": 0.5111356334340853,
+      "learning_rate": 7.239252199416749e-05,
+      "loss": 0.4812,
+      "step": 288
+    },
+    {
+      "epoch": 1.4026072457177505,
+      "grad_norm": 0.5198819509993116,
+      "learning_rate": 7.23128049634722e-05,
+      "loss": 0.4809,
+      "step": 289
+    },
+    {
+      "epoch": 1.4074579354251933,
+      "grad_norm": 0.5071119674882796,
+      "learning_rate": 7.223271681263916e-05,
+      "loss": 0.4839,
+      "step": 290
+    },
+    {
+      "epoch": 1.412308625132636,
+      "grad_norm": 0.4871736388487796,
+      "learning_rate": 7.215225846149957e-05,
+      "loss": 0.4899,
+      "step": 291
+    },
+    {
+      "epoch": 1.4171593148400787,
+      "grad_norm": 0.4644477177967194,
+      "learning_rate": 7.207143083413643e-05,
+      "loss": 0.4865,
+      "step": 292
+    },
+    {
+      "epoch": 1.4220100045475217,
+      "grad_norm": 0.5305010693129898,
+      "learning_rate": 7.1990234858874e-05,
+      "loss": 0.4876,
+      "step": 293
+    },
+    {
+      "epoch": 1.4268606942549644,
+      "grad_norm": 0.5381812012245146,
+      "learning_rate": 7.190867146826707e-05,
+      "loss": 0.4936,
+      "step": 294
+    },
+    {
+      "epoch": 1.4317113839624072,
+      "grad_norm": 0.42926452605809334,
+      "learning_rate": 7.182674159909031e-05,
+      "loss": 0.4845,
+      "step": 295
+    },
+    {
+      "epoch": 1.43656207366985,
+      "grad_norm": 0.42959757228135126,
+      "learning_rate": 7.174444619232745e-05,
+      "loss": 0.4952,
+      "step": 296
+    },
+    {
+      "epoch": 1.4414127633772926,
+      "grad_norm": 0.4994771109955089,
+      "learning_rate": 7.166178619316056e-05,
+      "loss": 0.4912,
+      "step": 297
+    },
+    {
+      "epoch": 1.4462634530847356,
+      "grad_norm": 0.4676921077663137,
+      "learning_rate": 7.157876255095906e-05,
+      "loss": 0.4875,
+      "step": 298
+    },
+    {
+      "epoch": 1.4511141427921783,
+      "grad_norm": 0.430209471088095,
+      "learning_rate": 7.149537621926895e-05,
+      "loss": 0.4862,
+      "step": 299
+    },
+    {
+      "epoch": 1.455964832499621,
+      "grad_norm": 0.5267214852499816,
+      "learning_rate": 7.14116281558018e-05,
+      "loss": 0.4879,
+      "step": 300
+    },
+    {
+      "epoch": 1.4608155222070638,
+      "grad_norm": 0.5169308603441447,
+      "learning_rate": 7.132751932242376e-05,
+      "loss": 0.4984,
+      "step": 301
+    },
+    {
+      "epoch": 1.4656662119145065,
+      "grad_norm": 0.3435076728807633,
+      "learning_rate": 7.124305068514444e-05,
+      "loss": 0.487,
+      "step": 302
+    },
+    {
+      "epoch": 1.4705169016219495,
+      "grad_norm": 0.2662958717194974,
+      "learning_rate": 7.1158223214106e-05,
+      "loss": 0.4878,
+      "step": 303
+    },
+    {
+      "epoch": 1.4753675913293922,
+      "grad_norm": 0.31770648802942325,
+      "learning_rate": 7.107303788357177e-05,
+      "loss": 0.4819,
+      "step": 304
+    },
+    {
+      "epoch": 1.480218281036835,
+      "grad_norm": 0.29599579651368213,
+      "learning_rate": 7.098749567191527e-05,
+      "loss": 0.4852,
+      "step": 305
+    },
+    {
+      "epoch": 1.4850689707442777,
+      "grad_norm": 0.3093179833213525,
+      "learning_rate": 7.090159756160886e-05,
+      "loss": 0.4877,
+      "step": 306
+    },
+    {
+      "epoch": 1.4899196604517204,
+      "grad_norm": 0.38971280696993216,
+      "learning_rate": 7.081534453921242e-05,
+      "loss": 0.4852,
+      "step": 307
+    },
+    {
+      "epoch": 1.4947703501591634,
+      "grad_norm": 0.3825128078756503,
+      "learning_rate": 7.072873759536217e-05,
+      "loss": 0.4913,
+      "step": 308
+    },
+    {
+      "epoch": 1.499621039866606,
+      "grad_norm": 0.2886624346515238,
+      "learning_rate": 7.064177772475912e-05,
+      "loss": 0.4798,
+      "step": 309
+    },
+    {
+      "epoch": 1.5044717295740488,
+      "grad_norm": 0.3712678777863381,
+      "learning_rate": 7.05544659261578e-05,
+      "loss": 0.4867,
+      "step": 310
+    },
+    {
+      "epoch": 1.5093224192814916,
+      "grad_norm": 0.42331161544955054,
+      "learning_rate": 7.046680320235466e-05,
+      "loss": 0.4871,
+      "step": 311
+    },
+    {
+      "epoch": 1.5141731089889343,
+      "grad_norm": 0.4018226592697959,
+      "learning_rate": 7.037879056017663e-05,
+      "loss": 0.4842,
+      "step": 312
+    },
+    {
+      "epoch": 1.5190237986963773,
+      "grad_norm": 0.3959692369182176,
+      "learning_rate": 7.029042901046952e-05,
+      "loss": 0.4802,
+      "step": 313
+    },
+    {
+      "epoch": 1.5238744884038198,
+      "grad_norm": 0.4052319536586186,
+      "learning_rate": 7.020171956808645e-05,
+      "loss": 0.4859,
+      "step": 314
+    },
+    {
+      "epoch": 1.5287251781112627,
+      "grad_norm": 0.40673400305291324,
+      "learning_rate": 7.011266325187615e-05,
+      "loss": 0.496,
+      "step": 315
+    },
+    {
+      "epoch": 1.5335758678187055,
+      "grad_norm": 0.44258718402573904,
+      "learning_rate": 7.002326108467129e-05,
+      "loss": 0.4864,
+      "step": 316
+    },
+    {
+      "epoch": 1.5384265575261482,
+      "grad_norm": 0.4022422201093314,
+      "learning_rate": 6.993351409327672e-05,
+      "loss": 0.4763,
+      "step": 317
+    },
+    {
+      "epoch": 1.5432772472335912,
+      "grad_norm": 0.39107692009497685,
+      "learning_rate": 6.984342330845764e-05,
+      "loss": 0.4952,
+      "step": 318
+    },
+    {
+      "epoch": 1.5481279369410337,
+      "grad_norm": 0.36401667233363455,
+      "learning_rate": 6.975298976492785e-05,
+      "loss": 0.4952,
+      "step": 319
+    },
+    {
+      "epoch": 1.5529786266484766,
+      "grad_norm": 0.3964209144895704,
+      "learning_rate": 6.966221450133779e-05,
+      "loss": 0.4901,
+      "step": 320
+    },
+    {
+      "epoch": 1.5578293163559194,
+      "grad_norm": 0.41002635948711413,
+      "learning_rate": 6.957109856026261e-05,
+      "loss": 0.4917,
+      "step": 321
+    },
+    {
+      "epoch": 1.562680006063362,
+      "grad_norm": 0.33043042887382146,
+      "learning_rate": 6.94796429881903e-05,
+      "loss": 0.4771,
+      "step": 322
+    },
+    {
+      "epoch": 1.567530695770805,
+      "grad_norm": 0.3258167342621945,
+      "learning_rate": 6.938784883550948e-05,
+      "loss": 0.4889,
+      "step": 323
+    },
+    {
+      "epoch": 1.5723813854782476,
+      "grad_norm": 0.33864970777234993,
+      "learning_rate": 6.929571715649755e-05,
+      "loss": 0.4866,
+      "step": 324
+    },
+    {
+      "epoch": 1.5772320751856905,
+      "grad_norm": 0.3799335436470155,
+      "learning_rate": 6.920324900930842e-05,
+      "loss": 0.4907,
+      "step": 325
+    },
+    {
+      "epoch": 1.5820827648931333,
+      "grad_norm": 0.38649667346458383,
+      "learning_rate": 6.911044545596042e-05,
+      "loss": 0.4854,
+      "step": 326
+    },
+    {
+      "epoch": 1.586933454600576,
+      "grad_norm": 0.3332457174533195,
+      "learning_rate": 6.901730756232411e-05,
+      "loss": 0.4895,
+      "step": 327
+    },
+    {
+      "epoch": 1.591784144308019,
+      "grad_norm": 0.3688581464993942,
+      "learning_rate": 6.892383639811005e-05,
+      "loss": 0.4958,
+      "step": 328
+    },
+    {
+      "epoch": 1.5966348340154615,
+      "grad_norm": 0.43127701691860393,
+      "learning_rate": 6.883003303685644e-05,
+      "loss": 0.4844,
+      "step": 329
+    },
+    {
+      "epoch": 1.6014855237229044,
+      "grad_norm": 0.5098788849460419,
+      "learning_rate": 6.87358985559169e-05,
+      "loss": 0.489,
+      "step": 330
+    },
+    {
+      "epoch": 1.6063362134303472,
+      "grad_norm": 0.531466303384909,
+      "learning_rate": 6.864143403644797e-05,
+      "loss": 0.4945,
+      "step": 331
+    },
+    {
+      "epoch": 1.61118690313779,
+      "grad_norm": 0.4575057116450561,
+      "learning_rate": 6.85466405633968e-05,
+      "loss": 0.4855,
+      "step": 332
+    },
+    {
+      "epoch": 1.6160375928452326,
+      "grad_norm": 0.42418027914564915,
+      "learning_rate": 6.845151922548865e-05,
+      "loss": 0.4783,
+      "step": 333
+    },
+    {
+      "epoch": 1.6208882825526754,
+      "grad_norm": 0.38431703300530295,
+      "learning_rate": 6.835607111521439e-05,
+      "loss": 0.4796,
+      "step": 334
+    },
+    {
+      "epoch": 1.6257389722601183,
+      "grad_norm": 0.325044588754403,
+      "learning_rate": 6.826029732881793e-05,
+      "loss": 0.4928,
+      "step": 335
+    },
+    {
+      "epoch": 1.630589661967561,
+      "grad_norm": 0.2723847494193817,
+      "learning_rate": 6.816419896628363e-05,
+      "loss": 0.4851,
+      "step": 336
+    },
+    {
+      "epoch": 1.6354403516750038,
+      "grad_norm": 0.25325170865464947,
+      "learning_rate": 6.806777713132374e-05,
+      "loss": 0.4826,
+      "step": 337
+    },
+    {
+      "epoch": 1.6402910413824465,
+      "grad_norm": 0.2793387163645126,
+      "learning_rate": 6.79710329313656e-05,
+      "loss": 0.4873,
+      "step": 338
+    },
+    {
+      "epoch": 1.6451417310898893,
+      "grad_norm": 0.38835730960557174,
+      "learning_rate": 6.787396747753903e-05,
+      "loss": 0.4744,
+      "step": 339
+    },
+    {
+      "epoch": 1.6499924207973322,
+      "grad_norm": 0.5896470389457479,
+      "learning_rate": 6.777658188466354e-05,
+      "loss": 0.4765,
+      "step": 340
+    },
+    {
+      "epoch": 1.654843110504775,
+      "grad_norm": 0.7717199781637745,
+      "learning_rate": 6.767887727123544e-05,
+      "loss": 0.4931,
+      "step": 341
+    },
+    {
+      "epoch": 1.6596938002122177,
+      "grad_norm": 1.0491183054565791,
+      "learning_rate": 6.758085475941516e-05,
+      "loss": 0.4875,
+      "step": 342
+    },
+    {
+      "epoch": 1.6645444899196604,
+      "grad_norm": 1.0963514331569248,
+      "learning_rate": 6.748251547501418e-05,
+      "loss": 0.4783,
+      "step": 343
+    },
+    {
+      "epoch": 1.6693951796271032,
+      "grad_norm": 0.6008540810802777,
+      "learning_rate": 6.738386054748226e-05,
+      "loss": 0.4836,
+      "step": 344
+    },
+    {
+      "epoch": 1.674245869334546,
+      "grad_norm": 0.4448291346545642,
+      "learning_rate": 6.728489110989434e-05,
+      "loss": 0.4883,
+      "step": 345
+    },
+    {
+      "epoch": 1.6790965590419886,
+      "grad_norm": 0.4942617693141126,
+      "learning_rate": 6.718560829893762e-05,
+      "loss": 0.4799,
+      "step": 346
+    },
+    {
+      "epoch": 1.6839472487494316,
+      "grad_norm": 0.4996733158915889,
+      "learning_rate": 6.708601325489844e-05,
+      "loss": 0.4872,
+      "step": 347
+    },
+    {
+      "epoch": 1.6887979384568743,
+      "grad_norm": 0.3715020675799419,
+      "learning_rate": 6.698610712164924e-05,
+      "loss": 0.4864,
+      "step": 348
+    },
+    {
+      "epoch": 1.693648628164317,
+      "grad_norm": 0.27977348496462506,
+      "learning_rate": 6.688589104663536e-05,
+      "loss": 0.4731,
+      "step": 349
+    },
+    {
+      "epoch": 1.69849931787176,
+      "grad_norm": 0.3329757061053683,
+      "learning_rate": 6.67853661808619e-05,
+      "loss": 0.4771,
+      "step": 350
+    },
+    {
+      "epoch": 1.7033500075792025,
+      "grad_norm": 0.35849519050184514,
+      "learning_rate": 6.668453367888052e-05,
+      "loss": 0.4867,
+      "step": 351
+    },
+    {
+      "epoch": 1.7082006972866455,
+      "grad_norm": 0.3093064728603477,
+      "learning_rate": 6.658339469877613e-05,
+      "loss": 0.478,
+      "step": 352
+    },
+    {
+      "epoch": 1.7130513869940882,
+      "grad_norm": 0.2983932071893944,
+      "learning_rate": 6.64819504021536e-05,
+      "loss": 0.4814,
+      "step": 353
+    },
+    {
+      "epoch": 1.717902076701531,
+      "grad_norm": 0.32895472335219694,
+      "learning_rate": 6.638020195412448e-05,
+      "loss": 0.4771,
+      "step": 354
+    },
+    {
+      "epoch": 1.722752766408974,
+      "grad_norm": 0.4277599821762847,
+      "learning_rate": 6.627815052329354e-05,
+      "loss": 0.4925,
+      "step": 355
+    },
+    {
+      "epoch": 1.7276034561164164,
+      "grad_norm": 0.4690652488456737,
+      "learning_rate": 6.617579728174535e-05,
+      "loss": 0.4854,
+      "step": 356
+    },
+    {
+      "epoch": 1.7324541458238594,
+      "grad_norm": 0.43487992599318925,
+      "learning_rate": 6.60731434050309e-05,
+      "loss": 0.4777,
+      "step": 357
+    },
+    {
+      "epoch": 1.737304835531302,
+      "grad_norm": 0.4398837483674929,
+      "learning_rate": 6.597019007215401e-05,
+      "loss": 0.4783,
+      "step": 358
+    },
+    {
+      "epoch": 1.7421555252387448,
+      "grad_norm": 0.43050412759021583,
+      "learning_rate": 6.586693846555788e-05,
+      "loss": 0.4743,
+      "step": 359
+    },
+    {
+      "epoch": 1.7470062149461878,
+      "grad_norm": 0.40202174280112624,
+      "learning_rate": 6.576338977111134e-05,
+      "loss": 0.48,
+      "step": 360
+    },
+    {
+      "epoch": 1.7518569046536303,
+      "grad_norm": 0.36259944082265505,
+      "learning_rate": 6.565954517809543e-05,
+      "loss": 0.4747,
+      "step": 361
+    },
+    {
+      "epoch": 1.7567075943610733,
+      "grad_norm": 0.28300782787905476,
+      "learning_rate": 6.555540587918968e-05,
+      "loss": 0.4778,
+      "step": 362
+    },
+    {
+      "epoch": 1.761558284068516,
+      "grad_norm": 0.27286805994349533,
+      "learning_rate": 6.545097307045831e-05,
+      "loss": 0.4795,
+      "step": 363
+    },
+    {
+      "epoch": 1.7664089737759587,
+      "grad_norm": 0.31934491384929364,
+      "learning_rate": 6.534624795133662e-05,
+      "loss": 0.4851,
+      "step": 364
+    },
+    {
+      "epoch": 1.7712596634834017,
+      "grad_norm": 0.35245692182661065,
+      "learning_rate": 6.524123172461711e-05,
+      "loss": 0.4794,
+      "step": 365
+    },
+    {
+      "epoch": 1.7761103531908442,
+      "grad_norm": 0.3396944161317505,
+      "learning_rate": 6.51359255964358e-05,
+      "loss": 0.4774,
+      "step": 366
+    },
+    {
+      "epoch": 1.7809610428982872,
+      "grad_norm": 0.2877754553279699,
+      "learning_rate": 6.503033077625824e-05,
+      "loss": 0.4746,
+      "step": 367
+    },
+    {
+      "epoch": 1.78581173260573,
+      "grad_norm": 0.3326775146217252,
+      "learning_rate": 6.492444847686566e-05,
+      "loss": 0.4849,
+      "step": 368
+    },
+    {
+      "epoch": 1.7906624223131726,
+      "grad_norm": 0.419702080624426,
+      "learning_rate": 6.481827991434111e-05,
+      "loss": 0.4814,
+      "step": 369
+    },
+    {
+      "epoch": 1.7955131120206156,
+      "grad_norm": 0.43780909545325103,
+      "learning_rate": 6.471182630805538e-05,
+      "loss": 0.4813,
+      "step": 370
+    },
+    {
+      "epoch": 1.800363801728058,
+      "grad_norm": 0.5080371772111751,
+      "learning_rate": 6.460508888065314e-05,
+      "loss": 0.4865,
+      "step": 371
+    },
+    {
+      "epoch": 1.805214491435501,
+      "grad_norm": 0.5541771901573003,
+      "learning_rate": 6.449806885803873e-05,
+      "loss": 0.4752,
+      "step": 372
+    },
+    {
+      "epoch": 1.8100651811429438,
+      "grad_norm": 0.5077930985424878,
+      "learning_rate": 6.439076746936219e-05,
+      "loss": 0.4776,
+      "step": 373
+    },
+    {
+      "epoch": 1.8149158708503865,
+      "grad_norm": 0.3974055597915673,
+      "learning_rate": 6.428318594700509e-05,
+      "loss": 0.4833,
+      "step": 374
+    },
+    {
+      "epoch": 1.8197665605578295,
+      "grad_norm": 0.27838793632255576,
+      "learning_rate": 6.417532552656647e-05,
+      "loss": 0.4808,
+      "step": 375
+    },
+    {
+      "epoch": 1.824617250265272,
+      "grad_norm": 0.25531453372215107,
+      "learning_rate": 6.406718744684851e-05,
+      "loss": 0.475,
+      "step": 376
+    },
+    {
+      "epoch": 1.829467939972715,
+      "grad_norm": 0.37619306197372576,
+      "learning_rate": 6.395877294984241e-05,
+      "loss": 0.4718,
+      "step": 377
+    },
+    {
+      "epoch": 1.8343186296801577,
+      "grad_norm": 0.45314622671289495,
+      "learning_rate": 6.385008328071406e-05,
+      "loss": 0.4858,
+      "step": 378
+    },
+    {
+      "epoch": 1.8391693193876004,
+      "grad_norm": 0.4091697946782105,
+      "learning_rate": 6.374111968778982e-05,
+      "loss": 0.4797,
+      "step": 379
+    },
+    {
+      "epoch": 1.8440200090950432,
+      "grad_norm": 0.31229108750141443,
+      "learning_rate": 6.363188342254206e-05,
+      "loss": 0.4819,
+      "step": 380
+    },
+    {
+      "epoch": 1.848870698802486,
+      "grad_norm": 0.29389799920114323,
+      "learning_rate": 6.352237573957488e-05,
+      "loss": 0.4777,
+      "step": 381
+    },
+    {
+      "epoch": 1.8537213885099288,
+      "grad_norm": 0.31830405845533455,
+      "learning_rate": 6.341259789660969e-05,
+      "loss": 0.4812,
+      "step": 382
+    },
+    {
+      "epoch": 1.8585720782173716,
+      "grad_norm": 0.2783419658252336,
+      "learning_rate": 6.330255115447076e-05,
+      "loss": 0.4722,
+      "step": 383
+    },
+    {
+      "epoch": 1.8634227679248143,
+      "grad_norm": 0.252195902565345,
+      "learning_rate": 6.319223677707069e-05,
+      "loss": 0.4786,
+      "step": 384
+    },
+    {
+      "epoch": 1.868273457632257,
+      "grad_norm": 0.24931429544692238,
+      "learning_rate": 6.308165603139598e-05,
+      "loss": 0.4766,
+      "step": 385
+    },
+    {
+      "epoch": 1.8731241473396998,
+      "grad_norm": 0.343056469344131,
+      "learning_rate": 6.29708101874924e-05,
+      "loss": 0.4746,
+      "step": 386
+    },
+    {
+      "epoch": 1.8779748370471427,
+      "grad_norm": 0.39163026769861153,
+      "learning_rate": 6.285970051845045e-05,
+      "loss": 0.4777,
+      "step": 387
+    },
+    {
+      "epoch": 1.8828255267545853,
+      "grad_norm": 0.3925083138992103,
+      "learning_rate": 6.274832830039071e-05,
+      "loss": 0.4762,
+      "step": 388
+    },
+    {
+      "epoch": 1.8876762164620282,
+      "grad_norm": 0.3908015146094751,
+      "learning_rate": 6.26366948124492e-05,
+      "loss": 0.4882,
+      "step": 389
+    },
+    {
+      "epoch": 1.892526906169471,
+      "grad_norm": 0.38768255703534454,
+      "learning_rate": 6.25248013367627e-05,
+      "loss": 0.4746,
+      "step": 390
+    },
+    {
+      "epoch": 1.8973775958769137,
+      "grad_norm": 0.34797592572865116,
+      "learning_rate": 6.241264915845401e-05,
+      "loss": 0.4863,
+      "step": 391
+    },
+    {
+      "epoch": 1.9022282855843566,
+      "grad_norm": 0.2939167601152598,
+      "learning_rate": 6.230023956561716e-05,
+      "loss": 0.4803,
+      "step": 392
+    },
+    {
+      "epoch": 1.9070789752917991,
+      "grad_norm": 0.2683598696324213,
+      "learning_rate": 6.218757384930268e-05,
+      "loss": 0.4769,
+      "step": 393
+    },
+    {
+      "epoch": 1.911929664999242,
+      "grad_norm": 0.2963955328357841,
+      "learning_rate": 6.207465330350273e-05,
+      "loss": 0.4798,
+      "step": 394
+    },
+    {
+      "epoch": 1.9167803547066848,
+      "grad_norm": 0.34936564317737695,
+      "learning_rate": 6.196147922513623e-05,
+      "loss": 0.479,
+      "step": 395
+    },
+    {
+      "epoch": 1.9216310444141276,
+      "grad_norm": 0.3023947037005716,
+      "learning_rate": 6.184805291403402e-05,
+      "loss": 0.4776,
+      "step": 396
+    },
+    {
+      "epoch": 1.9264817341215705,
+      "grad_norm": 0.23472105439779495,
+      "learning_rate": 6.173437567292383e-05,
+      "loss": 0.475,
+      "step": 397
+    },
+    {
+      "epoch": 1.931332423829013,
+      "grad_norm": 0.30151702935680424,
+      "learning_rate": 6.162044880741544e-05,
+      "loss": 0.4719,
+      "step": 398
+    },
+    {
+      "epoch": 1.936183113536456,
+      "grad_norm": 0.4380813843564537,
+      "learning_rate": 6.150627362598557e-05,
+      "loss": 0.4871,
+      "step": 399
+    },
+    {
+      "epoch": 1.9410338032438987,
+      "grad_norm": 0.5034940901090467,
+      "learning_rate": 6.139185143996298e-05,
+      "loss": 0.4806,
+      "step": 400
+    },
+    {
+      "epoch": 1.9458844929513415,
+      "grad_norm": 0.44831770290541656,
+      "learning_rate": 6.127718356351326e-05,
+      "loss": 0.478,
+      "step": 401
+    },
+    {
+      "epoch": 1.9507351826587844,
+      "grad_norm": 0.3881020850233725,
+      "learning_rate": 6.116227131362385e-05,
+      "loss": 0.4714,
+      "step": 402
+    },
+    {
+      "epoch": 1.955585872366227,
+      "grad_norm": 0.3382903208345561,
+      "learning_rate": 6.104711601008888e-05,
+      "loss": 0.4779,
+      "step": 403
+    },
+    {
+      "epoch": 1.96043656207367,
+      "grad_norm": 0.2830781275191087,
+      "learning_rate": 6.0931718975493985e-05,
+      "loss": 0.4846,
+      "step": 404
+    },
+    {
+      "epoch": 1.9652872517811126,
+      "grad_norm": 0.27767803820547865,
+      "learning_rate": 6.081608153520117e-05,
+      "loss": 0.4691,
+      "step": 405
+    },
+    {
+      "epoch": 1.9701379414885554,
+      "grad_norm": 0.35527234014372044,
+      "learning_rate": 6.0700205017333525e-05,
+      "loss": 0.4787,
+      "step": 406
+    },
+    {
+      "epoch": 1.9749886311959983,
+      "grad_norm": 0.34300192919407774,
+      "learning_rate": 6.058409075276002e-05,
+      "loss": 0.4689,
+      "step": 407
+    },
+    {
+      "epoch": 1.9798393209034408,
+      "grad_norm": 0.2964726842271146,
+      "learning_rate": 6.046774007508019e-05,
+      "loss": 0.475,
+      "step": 408
+    },
+    {
+      "epoch": 1.9846900106108838,
+      "grad_norm": 0.2502113324820329,
+      "learning_rate": 6.035115432060883e-05,
+      "loss": 0.4747,
+      "step": 409
+    },
+    {
+      "epoch": 1.9895407003183265,
+      "grad_norm": 0.23471156048166733,
+      "learning_rate": 6.0234334828360655e-05,
+      "loss": 0.4786,
+      "step": 410
+    },
+    {
+      "epoch": 1.9943913900257693,
+      "grad_norm": 0.27137410563019304,
+      "learning_rate": 6.011728294003494e-05,
+      "loss": 0.4802,
+      "step": 411
+    },
+    {
+      "epoch": 1.9992420797332122,
+      "grad_norm": 0.2960970008273601,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 0.4768,
+      "step": 412
+    },
+    {
+      "epoch": 2.004850689707443,
+      "grad_norm": 0.33644070901215145,
+      "learning_rate": 5.988248735527793e-05,
+      "loss": 0.4473,
+      "step": 413
+    },
+    {
+      "epoch": 2.0097013794148855,
+      "grad_norm": 0.39546749871672404,
+      "learning_rate": 5.9764746355528994e-05,
+      "loss": 0.4501,
+      "step": 414
+    },
+    {
+      "epoch": 2.0145520691223284,
+      "grad_norm": 0.47967552460987467,
+      "learning_rate": 5.964677835303615e-05,
+      "loss": 0.4483,
+      "step": 415
+    },
+    {
+      "epoch": 2.019402758829771,
+      "grad_norm": 0.43081443925289625,
+      "learning_rate": 5.952858470268955e-05,
+      "loss": 0.4468,
+      "step": 416
+    },
+    {
+      "epoch": 2.024253448537214,
+      "grad_norm": 0.42692273173821377,
+      "learning_rate": 5.941016676197098e-05,
+      "loss": 0.4499,
+      "step": 417
+    },
+    {
+      "epoch": 2.029104138244657,
+      "grad_norm": 0.5203095933335755,
+      "learning_rate": 5.929152589093825e-05,
+      "loss": 0.4498,
+      "step": 418
+    },
+    {
+      "epoch": 2.0339548279520994,
+      "grad_norm": 0.5390774215298352,
+      "learning_rate": 5.9172663452209554e-05,
+      "loss": 0.449,
+      "step": 419
+    },
+    {
+      "epoch": 2.0388055176595423,
+      "grad_norm": 0.5618164866931807,
+      "learning_rate": 5.9053580810947845e-05,
+      "loss": 0.4555,
+      "step": 420
+    },
+    {
+      "epoch": 2.043656207366985,
+      "grad_norm": 0.6473211411897097,
+      "learning_rate": 5.89342793348452e-05,
+      "loss": 0.4537,
+      "step": 421
+    },
+    {
+      "epoch": 2.048506897074428,
+      "grad_norm": 0.7641174380933736,
+      "learning_rate": 5.881476039410699e-05,
+      "loss": 0.4579,
+      "step": 422
+    },
+    {
+      "epoch": 2.0533575867818707,
+      "grad_norm": 0.6589276143277975,
+      "learning_rate": 5.869502536143629e-05,
+      "loss": 0.4478,
+      "step": 423
+    },
+    {
+      "epoch": 2.0582082764893133,
+      "grad_norm": 0.4055360078631115,
+      "learning_rate": 5.857507561201802e-05,
+      "loss": 0.4501,
+      "step": 424
+    },
+    {
+      "epoch": 2.063058966196756,
+      "grad_norm": 0.3576976481492298,
+      "learning_rate": 5.845491252350312e-05,
+      "loss": 0.4479,
+      "step": 425
+    },
+    {
+      "epoch": 2.0679096559041987,
+      "grad_norm": 0.444390387775824,
+      "learning_rate": 5.833453747599286e-05,
+      "loss": 0.4466,
+      "step": 426
+    },
+    {
+      "epoch": 2.0727603456116417,
+      "grad_norm": 0.38706062123939283,
+      "learning_rate": 5.821395185202285e-05,
+      "loss": 0.449,
+      "step": 427
+    },
+    {
+      "epoch": 2.0776110353190846,
+      "grad_norm": 0.3243650769978332,
+      "learning_rate": 5.809315703654726e-05,
+      "loss": 0.4581,
+      "step": 428
+    },
+    {
+      "epoch": 2.082461725026527,
+      "grad_norm": 0.35862489527477903,
+      "learning_rate": 5.797215441692284e-05,
+      "loss": 0.4534,
+      "step": 429
+    },
+    {
+      "epoch": 2.08731241473397,
+      "grad_norm": 0.3353344776943914,
+      "learning_rate": 5.785094538289304e-05,
+      "loss": 0.4537,
+      "step": 430
+    },
+    {
+      "epoch": 2.0921631044414126,
+      "grad_norm": 0.3620686259692887,
+      "learning_rate": 5.772953132657202e-05,
+      "loss": 0.4553,
+      "step": 431
+    },
+    {
+      "epoch": 2.0970137941488556,
+      "grad_norm": 0.307403523851727,
+      "learning_rate": 5.7607913642428666e-05,
+      "loss": 0.4424,
+      "step": 432
+    },
+    {
+      "epoch": 2.101864483856298,
+      "grad_norm": 0.28710201149568576,
+      "learning_rate": 5.7486093727270606e-05,
+      "loss": 0.4462,
+      "step": 433
+    },
+    {
+      "epoch": 2.106715173563741,
+      "grad_norm": 0.31451691469735704,
+      "learning_rate": 5.736407298022809e-05,
+      "loss": 0.4434,
+      "step": 434
+    },
+    {
+      "epoch": 2.111565863271184,
+      "grad_norm": 0.3105010146819863,
+      "learning_rate": 5.7241852802738e-05,
+      "loss": 0.4533,
+      "step": 435
+    },
+    {
+      "epoch": 2.1164165529786265,
+      "grad_norm": 0.2878408032383936,
+      "learning_rate": 5.711943459852772e-05,
+      "loss": 0.4427,
+      "step": 436
+    },
+    {
+      "epoch": 2.1212672426860695,
+      "grad_norm": 0.3026608573456508,
+      "learning_rate": 5.699681977359902e-05,
+      "loss": 0.4385,
+      "step": 437
+    },
+    {
+      "epoch": 2.1261179323935124,
+      "grad_norm": 0.2857208984253648,
+      "learning_rate": 5.6874009736211896e-05,
+      "loss": 0.4465,
+      "step": 438
+    },
+    {
+      "epoch": 2.130968622100955,
+      "grad_norm": 0.28786201316205207,
+      "learning_rate": 5.675100589686839e-05,
+      "loss": 0.4472,
+      "step": 439
+    },
+    {
+      "epoch": 2.135819311808398,
+      "grad_norm": 0.3397455432854385,
+      "learning_rate": 5.662780966829646e-05,
+      "loss": 0.4486,
+      "step": 440
+    },
+    {
+      "epoch": 2.1406700015158404,
+      "grad_norm": 0.3060713829784068,
+      "learning_rate": 5.650442246543364e-05,
+      "loss": 0.4525,
+      "step": 441
+    },
+    {
+      "epoch": 2.1455206912232834,
+      "grad_norm": 0.2554715156912105,
+      "learning_rate": 5.638084570541088e-05,
+      "loss": 0.4451,
+      "step": 442
+    },
+    {
+      "epoch": 2.150371380930726,
+      "grad_norm": 0.31763468046624377,
+      "learning_rate": 5.625708080753621e-05,
+      "loss": 0.455,
+      "step": 443
+    },
+    {
+      "epoch": 2.155222070638169,
+      "grad_norm": 0.3288505103534938,
+      "learning_rate": 5.6133129193278525e-05,
+      "loss": 0.4453,
+      "step": 444
+    },
+    {
+      "epoch": 2.160072760345612,
+      "grad_norm": 0.27302062189682574,
+      "learning_rate": 5.600899228625112e-05,
+      "loss": 0.4523,
+      "step": 445
+    },
+    {
+      "epoch": 2.1649234500530543,
+      "grad_norm": 0.2367917713116079,
+      "learning_rate": 5.588467151219549e-05,
+      "loss": 0.4481,
+      "step": 446
+    },
+    {
+      "epoch": 2.1697741397604973,
+      "grad_norm": 0.27892881990044693,
+      "learning_rate": 5.5760168298964874e-05,
+      "loss": 0.4397,
+      "step": 447
+    },
+    {
+      "epoch": 2.17462482946794,
+      "grad_norm": 0.3659363982261399,
+      "learning_rate": 5.563548407650782e-05,
+      "loss": 0.4464,
+      "step": 448
+    },
+    {
+      "epoch": 2.1794755191753827,
+      "grad_norm": 0.3139974928184246,
+      "learning_rate": 5.551062027685187e-05,
+      "loss": 0.4487,
+      "step": 449
+    },
+    {
+      "epoch": 2.1843262088828257,
+      "grad_norm": 0.25894377780084493,
+      "learning_rate": 5.5385578334087006e-05,
+      "loss": 0.4481,
+      "step": 450
+    },
+    {
+      "epoch": 2.189176898590268,
+      "grad_norm": 0.17951964246489394,
+      "learning_rate": 5.526035968434927e-05,
+      "loss": 0.4469,
+      "step": 451
+    },
+    {
+      "epoch": 2.194027588297711,
+      "grad_norm": 0.21554515294483917,
+      "learning_rate": 5.513496576580418e-05,
+      "loss": 0.4573,
+      "step": 452
+    },
+    {
+      "epoch": 2.1988782780051537,
+      "grad_norm": 0.24799983544152385,
+      "learning_rate": 5.5009398018630276e-05,
+      "loss": 0.4498,
+      "step": 453
+    },
+    {
+      "epoch": 2.2037289677125966,
+      "grad_norm": 0.186973375547793,
+      "learning_rate": 5.4883657885002575e-05,
+      "loss": 0.4449,
+      "step": 454
+    },
+    {
+      "epoch": 2.2085796574200396,
+      "grad_norm": 0.1888483688795653,
+      "learning_rate": 5.475774680907597e-05,
+      "loss": 0.443,
+      "step": 455
+    },
+    {
+      "epoch": 2.213430347127482,
+      "grad_norm": 0.18914695692387654,
+      "learning_rate": 5.463166623696868e-05,
+      "loss": 0.4434,
+      "step": 456
+    },
+    {
+      "epoch": 2.218281036834925,
+      "grad_norm": 0.19070268199404652,
+      "learning_rate": 5.450541761674562e-05,
+      "loss": 0.4445,
+      "step": 457
+    },
+    {
+      "epoch": 2.2231317265423676,
+      "grad_norm": 0.2259939851217927,
+      "learning_rate": 5.437900239840179e-05,
+      "loss": 0.4465,
+      "step": 458
+    },
+    {
+      "epoch": 2.2279824162498105,
+      "grad_norm": 0.18023421500987896,
+      "learning_rate": 5.42524220338456e-05,
+      "loss": 0.4453,
+      "step": 459
+    },
+    {
+      "epoch": 2.2328331059572535,
+      "grad_norm": 0.16598518940751159,
+      "learning_rate": 5.412567797688219e-05,
+      "loss": 0.4498,
+      "step": 460
+    },
+    {
+      "epoch": 2.237683795664696,
+      "grad_norm": 0.16004882092407235,
+      "learning_rate": 5.3998771683196754e-05,
+      "loss": 0.4527,
+      "step": 461
+    },
+    {
+      "epoch": 2.242534485372139,
+      "grad_norm": 0.17890202722776521,
+      "learning_rate": 5.3871704610337836e-05,
+      "loss": 0.4444,
+      "step": 462
+    },
+    {
+      "epoch": 2.2473851750795815,
+      "grad_norm": 0.182744670257566,
+      "learning_rate": 5.374447821770053e-05,
+      "loss": 0.4431,
+      "step": 463
+    },
+    {
+      "epoch": 2.2522358647870244,
+      "grad_norm": 0.19342699163866056,
+      "learning_rate": 5.361709396650977e-05,
+      "loss": 0.4404,
+      "step": 464
+    },
+    {
+      "epoch": 2.2570865544944674,
+      "grad_norm": 0.20442779398031627,
+      "learning_rate": 5.3489553319803566e-05,
+      "loss": 0.4496,
+      "step": 465
+    },
+    {
+      "epoch": 2.26193724420191,
+      "grad_norm": 0.23359410284964036,
+      "learning_rate": 5.336185774241609e-05,
+      "loss": 0.4469,
+      "step": 466
+    },
+    {
+      "epoch": 2.266787933909353,
+      "grad_norm": 0.23295417695606166,
+      "learning_rate": 5.3234008700961e-05,
+      "loss": 0.4505,
+      "step": 467
+    },
+    {
+      "epoch": 2.2716386236167954,
+      "grad_norm": 0.20207094824496044,
+      "learning_rate": 5.3106007663814505e-05,
+      "loss": 0.4406,
+      "step": 468
+    },
+    {
+      "epoch": 2.2764893133242383,
+      "grad_norm": 0.1850252157104855,
+      "learning_rate": 5.2977856101098484e-05,
+      "loss": 0.4525,
+      "step": 469
+    },
+    {
+      "epoch": 2.281340003031681,
+      "grad_norm": 0.1821206965545461,
+      "learning_rate": 5.284955548466371e-05,
+      "loss": 0.4592,
+      "step": 470
+    },
+    {
+      "epoch": 2.286190692739124,
+      "grad_norm": 0.1913435003815255,
+      "learning_rate": 5.272110728807279e-05,
+      "loss": 0.4459,
+      "step": 471
+    },
+    {
+      "epoch": 2.2910413824465667,
+      "grad_norm": 0.17908151714339782,
+      "learning_rate": 5.25925129865834e-05,
+      "loss": 0.4523,
+      "step": 472
+    },
+    {
+      "epoch": 2.2958920721540093,
+      "grad_norm": 0.17796456682985312,
+      "learning_rate": 5.246377405713121e-05,
+      "loss": 0.4426,
+      "step": 473
+    },
+    {
+      "epoch": 2.300742761861452,
+      "grad_norm": 0.16950491734508644,
+      "learning_rate": 5.2334891978313006e-05,
+      "loss": 0.4426,
+      "step": 474
+    },
+    {
+      "epoch": 2.305593451568895,
+      "grad_norm": 0.18036359667208995,
+      "learning_rate": 5.220586823036966e-05,
+      "loss": 0.4458,
+      "step": 475
+    },
+    {
+      "epoch": 2.3104441412763377,
+      "grad_norm": 0.1873060682555774,
+      "learning_rate": 5.207670429516915e-05,
+      "loss": 0.4433,
+      "step": 476
+    },
+    {
+      "epoch": 2.3152948309837806,
+      "grad_norm": 0.1986452939709168,
+      "learning_rate": 5.1947401656189546e-05,
+      "loss": 0.4593,
+      "step": 477
+    },
+    {
+      "epoch": 2.320145520691223,
+      "grad_norm": 0.20494449288937291,
+      "learning_rate": 5.181796179850197e-05,
+      "loss": 0.4424,
+      "step": 478
+    },
+    {
+      "epoch": 2.324996210398666,
+      "grad_norm": 0.17952910191793728,
+      "learning_rate": 5.168838620875352e-05,
+      "loss": 0.4503,
+      "step": 479
+    },
+    {
+      "epoch": 2.3298469001061086,
+      "grad_norm": 0.1909701609269039,
+      "learning_rate": 5.155867637515019e-05,
+      "loss": 0.4506,
+      "step": 480
+    },
+    {
+      "epoch": 2.3346975898135516,
+      "grad_norm": 0.1838298898954926,
+      "learning_rate": 5.142883378743984e-05,
+      "loss": 0.4513,
+      "step": 481
+    },
+    {
+      "epoch": 2.3395482795209945,
+      "grad_norm": 0.1818267641683358,
+      "learning_rate": 5.129885993689502e-05,
+      "loss": 0.4488,
+      "step": 482
+    },
+    {
+      "epoch": 2.344398969228437,
+      "grad_norm": 0.21197375880432345,
+      "learning_rate": 5.116875631629585e-05,
+      "loss": 0.4456,
+      "step": 483
+    },
+    {
+      "epoch": 2.34924965893588,
+      "grad_norm": 0.21240893965447508,
+      "learning_rate": 5.10385244199129e-05,
+      "loss": 0.4386,
+      "step": 484
+    },
+    {
+      "epoch": 2.354100348643323,
+      "grad_norm": 0.19244612255162405,
+      "learning_rate": 5.0908165743490047e-05,
+      "loss": 0.4482,
+      "step": 485
+    },
+    {
+      "epoch": 2.3589510383507655,
+      "grad_norm": 0.22440529731925618,
+      "learning_rate": 5.0777681784227224e-05,
+      "loss": 0.4496,
+      "step": 486
+    },
+    {
+      "epoch": 2.3638017280582084,
+      "grad_norm": 0.249440062974833,
+      "learning_rate": 5.064707404076327e-05,
+      "loss": 0.4502,
+      "step": 487
+    },
+    {
+      "epoch": 2.368652417765651,
+      "grad_norm": 0.2374206142112278,
+      "learning_rate": 5.051634401315875e-05,
+      "loss": 0.448,
+      "step": 488
+    },
+    {
+      "epoch": 2.373503107473094,
+      "grad_norm": 0.21044332969367502,
+      "learning_rate": 5.0385493202878656e-05,
+      "loss": 0.4416,
+      "step": 489
+    },
+    {
+      "epoch": 2.3783537971805364,
+      "grad_norm": 0.15343545111269605,
+      "learning_rate": 5.025452311277522e-05,
+      "loss": 0.4413,
+      "step": 490
+    },
+    {
+      "epoch": 2.3832044868879794,
+      "grad_norm": 0.17472771019103053,
+      "learning_rate": 5.01234352470706e-05,
+      "loss": 0.4472,
+      "step": 491
+    },
+    {
+      "epoch": 2.3880551765954223,
+      "grad_norm": 0.2225509747823868,
+      "learning_rate": 4.999223111133968e-05,
+      "loss": 0.4405,
+      "step": 492
+    },
+    {
+      "epoch": 2.392905866302865,
+      "grad_norm": 0.27110633436791925,
+      "learning_rate": 4.986091221249269e-05,
+      "loss": 0.44,
+      "step": 493
+    },
+    {
+      "epoch": 2.397756556010308,
+      "grad_norm": 0.24255464597168586,
+      "learning_rate": 4.972948005875796e-05,
+      "loss": 0.4432,
+      "step": 494
+    },
+    {
+      "epoch": 2.4026072457177503,
+      "grad_norm": 0.2496648842091371,
+      "learning_rate": 4.959793615966459e-05,
+      "loss": 0.4401,
+      "step": 495
+    },
+    {
+      "epoch": 2.4074579354251933,
+      "grad_norm": 0.24806426439634907,
+      "learning_rate": 4.946628202602508e-05,
+      "loss": 0.4526,
+      "step": 496
+    },
+    {
+      "epoch": 2.412308625132636,
+      "grad_norm": 0.21808090914084832,
+      "learning_rate": 4.933451916991802e-05,
+      "loss": 0.4474,
+      "step": 497
+    },
+    {
+      "epoch": 2.4171593148400787,
+      "grad_norm": 0.19833835766366836,
+      "learning_rate": 4.920264910467066e-05,
+      "loss": 0.4485,
+      "step": 498
+    },
+    {
+      "epoch": 2.4220100045475217,
+      "grad_norm": 0.1904056029579938,
+      "learning_rate": 4.9070673344841645e-05,
+      "loss": 0.4471,
+      "step": 499
+    },
+    {
+      "epoch": 2.426860694254964,
+      "grad_norm": 0.17821880940044135,
+      "learning_rate": 4.893859340620348e-05,
+      "loss": 0.4518,
+      "step": 500
+    },
+    {
+      "epoch": 2.431711383962407,
+      "grad_norm": 0.16242846601925154,
+      "learning_rate": 4.880641080572522e-05,
+      "loss": 0.4426,
+      "step": 501
+    },
+    {
+      "epoch": 2.43656207366985,
+      "grad_norm": 0.16230843192633562,
+      "learning_rate": 4.8674127061555025e-05,
+      "loss": 0.4492,
+      "step": 502
+    },
+    {
+      "epoch": 2.4414127633772926,
+      "grad_norm": 0.18692985874064466,
+      "learning_rate": 4.8541743693002676e-05,
+      "loss": 0.4576,
+      "step": 503
+    },
+    {
+      "epoch": 2.4462634530847356,
+      "grad_norm": 0.17489245993778632,
+      "learning_rate": 4.8409262220522196e-05,
+      "loss": 0.4476,
+      "step": 504
+    },
+    {
+      "epoch": 2.451114142792178,
+      "grad_norm": 0.14061959670906948,
+      "learning_rate": 4.8276684165694336e-05,
+      "loss": 0.4479,
+      "step": 505
+    },
+    {
+      "epoch": 2.455964832499621,
+      "grad_norm": 0.17289206898304424,
+      "learning_rate": 4.814401105120914e-05,
+      "loss": 0.4479,
+      "step": 506
+    },
+    {
+      "epoch": 2.460815522207064,
+      "grad_norm": 0.21063084112901795,
+      "learning_rate": 4.8011244400848414e-05,
+      "loss": 0.4466,
+      "step": 507
+    },
+    {
+      "epoch": 2.4656662119145065,
+      "grad_norm": 0.2134287283629687,
+      "learning_rate": 4.787838573946825e-05,
+      "loss": 0.4503,
+      "step": 508
+    },
+    {
+      "epoch": 2.4705169016219495,
+      "grad_norm": 0.19387557882251144,
+      "learning_rate": 4.774543659298152e-05,
+      "loss": 0.4419,
+      "step": 509
+    },
+    {
+      "epoch": 2.475367591329392,
+      "grad_norm": 0.1690053079886072,
+      "learning_rate": 4.761239848834031e-05,
+      "loss": 0.4443,
+      "step": 510
+    },
+    {
+      "epoch": 2.480218281036835,
+      "grad_norm": 0.17406180900609755,
+      "learning_rate": 4.747927295351845e-05,
+      "loss": 0.4474,
+      "step": 511
+    },
+    {
+      "epoch": 2.485068970744278,
+      "grad_norm": 0.2024050850623432,
+      "learning_rate": 4.734606151749389e-05,
+      "loss": 0.4473,
+      "step": 512
+    },
+    {
+      "epoch": 2.4899196604517204,
+      "grad_norm": 0.22659357369802574,
+      "learning_rate": 4.7212765710231204e-05,
+      "loss": 0.4481,
+      "step": 513
+    },
+    {
+      "epoch": 2.4947703501591634,
+      "grad_norm": 0.23595309939097722,
+      "learning_rate": 4.707938706266397e-05,
+      "loss": 0.4484,
+      "step": 514
+    },
+    {
+      "epoch": 2.499621039866606,
+      "grad_norm": 0.19918584163751257,
+      "learning_rate": 4.694592710667723e-05,
+      "loss": 0.444,
+      "step": 515
+    },
+    {
+      "epoch": 2.504471729574049,
+      "grad_norm": 0.18418670752131802,
+      "learning_rate": 4.681238737508983e-05,
+      "loss": 0.4424,
+      "step": 516
+    },
+    {
+      "epoch": 2.5093224192814914,
+      "grad_norm": 0.1926237495649244,
+      "learning_rate": 4.6678769401636894e-05,
+      "loss": 0.4444,
+      "step": 517
+    },
+    {
+      "epoch": 2.5141731089889343,
+      "grad_norm": 0.20706125086296728,
+      "learning_rate": 4.6545074720952166e-05,
+      "loss": 0.456,
+      "step": 518
+    },
+    {
+      "epoch": 2.5190237986963773,
+      "grad_norm": 0.17699664563372686,
+      "learning_rate": 4.641130486855038e-05,
+      "loss": 0.4396,
+      "step": 519
+    },
+    {
+      "epoch": 2.52387448840382,
+      "grad_norm": 0.18317752602670304,
+      "learning_rate": 4.627746138080966e-05,
+      "loss": 0.4432,
+      "step": 520
+    },
+    {
+      "epoch": 2.5287251781112627,
+      "grad_norm": 0.2190424482227647,
+      "learning_rate": 4.614354579495379e-05,
+      "loss": 0.4448,
+      "step": 521
+    },
+    {
+      "epoch": 2.5335758678187057,
+      "grad_norm": 0.20135719119048615,
+      "learning_rate": 4.6009559649034695e-05,
+      "loss": 0.4432,
+      "step": 522
+    },
+    {
+      "epoch": 2.538426557526148,
+      "grad_norm": 0.20782420010728125,
+      "learning_rate": 4.587550448191465e-05,
+      "loss": 0.4474,
+      "step": 523
+    },
+    {
+      "epoch": 2.543277247233591,
+      "grad_norm": 0.18668296726800496,
+      "learning_rate": 4.5741381833248655e-05,
+      "loss": 0.455,
+      "step": 524
+    },
+    {
+      "epoch": 2.5481279369410337,
+      "grad_norm": 0.17935132627421838,
+      "learning_rate": 4.560719324346677e-05,
+      "loss": 0.4457,
+      "step": 525
+    },
+    {
+      "epoch": 2.5529786266484766,
+      "grad_norm": 0.16835981784522308,
+      "learning_rate": 4.547294025375641e-05,
+      "loss": 0.4478,
+      "step": 526
+    },
+    {
+      "epoch": 2.557829316355919,
+      "grad_norm": 0.18084487512355504,
+      "learning_rate": 4.533862440604461e-05,
+      "loss": 0.447,
+      "step": 527
+    },
+    {
+      "epoch": 2.562680006063362,
+      "grad_norm": 0.17384784743298828,
+      "learning_rate": 4.520424724298036e-05,
+      "loss": 0.4408,
+      "step": 528
+    },
+    {
+      "epoch": 2.567530695770805,
+      "grad_norm": 0.20150460275113774,
+      "learning_rate": 4.5069810307916874e-05,
+      "loss": 0.4441,
+      "step": 529
+    },
+    {
+      "epoch": 2.5723813854782476,
+      "grad_norm": 0.2240004858996321,
+      "learning_rate": 4.493531514489385e-05,
+      "loss": 0.4425,
+      "step": 530
+    },
+    {
+      "epoch": 2.5772320751856905,
+      "grad_norm": 0.2286831099325836,
+      "learning_rate": 4.480076329861977e-05,
+      "loss": 0.4433,
+      "step": 531
+    },
+    {
+      "epoch": 2.5820827648931335,
+      "grad_norm": 0.2048648110357608,
+      "learning_rate": 4.46661563144541e-05,
+      "loss": 0.4487,
+      "step": 532
+    },
+    {
+      "epoch": 2.586933454600576,
+      "grad_norm": 0.2121488428415987,
+      "learning_rate": 4.453149573838962e-05,
+      "loss": 0.4445,
+      "step": 533
+    },
+    {
+      "epoch": 2.591784144308019,
+      "grad_norm": 0.18541474138380978,
+      "learning_rate": 4.43967831170346e-05,
+      "loss": 0.4494,
+      "step": 534
+    },
+    {
+      "epoch": 2.5966348340154615,
+      "grad_norm": 0.17452762694525445,
+      "learning_rate": 4.426201999759505e-05,
+      "loss": 0.4484,
+      "step": 535
+    },
+    {
+      "epoch": 2.6014855237229044,
+      "grad_norm": 0.21944224068377363,
+      "learning_rate": 4.4127207927857e-05,
+      "loss": 0.4419,
+      "step": 536
+    },
+    {
+      "epoch": 2.606336213430347,
+      "grad_norm": 0.17656272332454842,
+      "learning_rate": 4.3992348456168666e-05,
+      "loss": 0.4568,
+      "step": 537
+    },
+    {
+      "epoch": 2.61118690313779,
+      "grad_norm": 0.1892562653364182,
+      "learning_rate": 4.385744313142267e-05,
+      "loss": 0.4427,
+      "step": 538
+    },
+    {
+      "epoch": 2.616037592845233,
+      "grad_norm": 0.21611454670373548,
+      "learning_rate": 4.372249350303828e-05,
+      "loss": 0.4418,
+      "step": 539
+    },
+    {
+      "epoch": 2.6208882825526754,
+      "grad_norm": 0.17168747953224547,
+      "learning_rate": 4.358750112094363e-05,
+      "loss": 0.4544,
+      "step": 540
+    },
+    {
+      "epoch": 2.6257389722601183,
+      "grad_norm": 0.17941819138400728,
+      "learning_rate": 4.3452467535557846e-05,
+      "loss": 0.4372,
+      "step": 541
+    },
+    {
+      "epoch": 2.6305896619675613,
+      "grad_norm": 0.2025265834742146,
+      "learning_rate": 4.3317394297773304e-05,
+      "loss": 0.4517,
+      "step": 542
+    },
+    {
+      "epoch": 2.635440351675004,
+      "grad_norm": 0.20441246530938206,
+      "learning_rate": 4.3182282958937816e-05,
+      "loss": 0.4333,
+      "step": 543
+    },
+    {
+      "epoch": 2.6402910413824463,
+      "grad_norm": 0.2334105452950634,
+      "learning_rate": 4.304713507083673e-05,
+      "loss": 0.4481,
+      "step": 544
+    },
+    {
+      "epoch": 2.6451417310898893,
+      "grad_norm": 0.26291969340773214,
+      "learning_rate": 4.291195218567523e-05,
+      "loss": 0.4466,
+      "step": 545
+    },
+    {
+      "epoch": 2.649992420797332,
+      "grad_norm": 0.1863631298156993,
+      "learning_rate": 4.277673585606046e-05,
+      "loss": 0.4405,
+      "step": 546
+    },
+    {
+      "epoch": 2.6548431105047747,
+      "grad_norm": 0.23226855973797117,
+      "learning_rate": 4.264148763498364e-05,
+      "loss": 0.4566,
+      "step": 547
+    },
+    {
+      "epoch": 2.6596938002122177,
+      "grad_norm": 0.30482274820740174,
+      "learning_rate": 4.250620907580226e-05,
+      "loss": 0.4407,
+      "step": 548
+    },
+    {
+      "epoch": 2.6645444899196606,
+      "grad_norm": 0.23781311620065457,
+      "learning_rate": 4.237090173222231e-05,
+      "loss": 0.4493,
+      "step": 549
+    },
+    {
+      "epoch": 2.669395179627103,
+      "grad_norm": 0.1808214801234254,
+      "learning_rate": 4.223556715828033e-05,
+      "loss": 0.4511,
+      "step": 550
+    },
+    {
+      "epoch": 2.674245869334546,
+      "grad_norm": 0.26315804734468673,
+      "learning_rate": 4.2100206908325603e-05,
+      "loss": 0.447,
+      "step": 551
+    },
+    {
+      "epoch": 2.6790965590419886,
+      "grad_norm": 0.25781234163394623,
+      "learning_rate": 4.196482253700235e-05,
+      "loss": 0.4415,
+      "step": 552
+    },
+    {
+      "epoch": 2.6839472487494316,
+      "grad_norm": 0.17133762584152984,
+      "learning_rate": 4.182941559923179e-05,
+      "loss": 0.4457,
+      "step": 553
+    },
+    {
+      "epoch": 2.688797938456874,
+      "grad_norm": 0.2266803612041648,
+      "learning_rate": 4.169398765019433e-05,
+      "loss": 0.4422,
+      "step": 554
+    },
+    {
+      "epoch": 2.693648628164317,
+      "grad_norm": 0.23286738752123257,
+      "learning_rate": 4.15585402453117e-05,
+      "loss": 0.4429,
+      "step": 555
+    },
+    {
+      "epoch": 2.69849931787176,
+      "grad_norm": 0.20226496811604636,
+      "learning_rate": 4.14230749402291e-05,
+      "loss": 0.4421,
+      "step": 556
+    },
+    {
+      "epoch": 2.7033500075792025,
+      "grad_norm": 0.21746634743317236,
+      "learning_rate": 4.128759329079732e-05,
+      "loss": 0.4318,
+      "step": 557
+    },
+    {
+      "epoch": 2.7082006972866455,
+      "grad_norm": 0.24285493960537577,
+      "learning_rate": 4.115209685305482e-05,
+      "loss": 0.4374,
+      "step": 558
+    },
+    {
+      "epoch": 2.7130513869940884,
+      "grad_norm": 0.20035101285126697,
+      "learning_rate": 4.101658718320998e-05,
+      "loss": 0.4429,
+      "step": 559
+    },
+    {
+      "epoch": 2.717902076701531,
+      "grad_norm": 0.1733102653989901,
+      "learning_rate": 4.088106583762309e-05,
+      "loss": 0.4456,
+      "step": 560
+    },
+    {
+      "epoch": 2.722752766408974,
+      "grad_norm": 0.25116764609287723,
+      "learning_rate": 4.074553437278857e-05,
+      "loss": 0.4494,
+      "step": 561
+    },
+    {
+      "epoch": 2.7276034561164164,
+      "grad_norm": 0.19896329775589092,
+      "learning_rate": 4.060999434531704e-05,
+      "loss": 0.4449,
+      "step": 562
+    },
+    {
+      "epoch": 2.7324541458238594,
+      "grad_norm": 0.16517536428811208,
+      "learning_rate": 4.047444731191751e-05,
+      "loss": 0.4426,
+      "step": 563
+    },
+    {
+      "epoch": 2.737304835531302,
+      "grad_norm": 0.1656807626865065,
+      "learning_rate": 4.033889482937943e-05,
+      "loss": 0.4445,
+      "step": 564
+    },
+    {
+      "epoch": 2.742155525238745,
+      "grad_norm": 0.15103159619749504,
+      "learning_rate": 4.020333845455478e-05,
+      "loss": 0.4565,
+      "step": 565
+    },
+    {
+      "epoch": 2.747006214946188,
+      "grad_norm": 0.16996820086522443,
+      "learning_rate": 4.0067779744340345e-05,
+      "loss": 0.4459,
+      "step": 566
+    },
+    {
+      "epoch": 2.7518569046536303,
+      "grad_norm": 0.1495970266083701,
+      "learning_rate": 3.993222025565966e-05,
+      "loss": 0.4447,
+      "step": 567
+    },
+    {
+      "epoch": 2.7567075943610733,
+      "grad_norm": 0.15458974892236554,
+      "learning_rate": 3.979666154544522e-05,
+      "loss": 0.4452,
+      "step": 568
+    },
+    {
+      "epoch": 2.7615582840685162,
+      "grad_norm": 0.1741093401099396,
+      "learning_rate": 3.96611051706206e-05,
+      "loss": 0.4421,
+      "step": 569
+    },
+    {
+      "epoch": 2.7664089737759587,
+      "grad_norm": 0.1819530197226333,
+      "learning_rate": 3.9525552688082494e-05,
+      "loss": 0.4509,
+      "step": 570
+    },
+    {
+      "epoch": 2.7712596634834017,
+      "grad_norm": 0.14996389947080183,
+      "learning_rate": 3.939000565468297e-05,
+      "loss": 0.4442,
+      "step": 571
+    },
+    {
+      "epoch": 2.776110353190844,
+      "grad_norm": 0.19015205226216172,
+      "learning_rate": 3.9254465627211444e-05,
+      "loss": 0.4458,
+      "step": 572
+    },
+    {
+      "epoch": 2.780961042898287,
+      "grad_norm": 0.2147271939320094,
+      "learning_rate": 3.911893416237693e-05,
+      "loss": 0.4423,
+      "step": 573
+    },
+    {
+      "epoch": 2.7858117326057297,
+      "grad_norm": 0.17368493707493848,
+      "learning_rate": 3.8983412816790045e-05,
+      "loss": 0.4415,
+      "step": 574
+    },
+    {
+      "epoch": 2.7906624223131726,
+      "grad_norm": 0.18366782229193682,
+      "learning_rate": 3.8847903146945186e-05,
+      "loss": 0.4419,
+      "step": 575
+    },
+    {
+      "epoch": 2.7955131120206156,
+      "grad_norm": 0.1770373720928735,
+      "learning_rate": 3.871240670920269e-05,
+      "loss": 0.4477,
+      "step": 576
+    },
+    {
+      "epoch": 2.800363801728058,
+      "grad_norm": 0.15508137985676013,
+      "learning_rate": 3.85769250597709e-05,
+      "loss": 0.4458,
+      "step": 577
+    },
+    {
+      "epoch": 2.805214491435501,
+      "grad_norm": 0.15518418619016236,
+      "learning_rate": 3.844145975468832e-05,
+      "loss": 0.4403,
+      "step": 578
+    },
+    {
+      "epoch": 2.810065181142944,
+      "grad_norm": 0.13016021632650948,
+      "learning_rate": 3.830601234980569e-05,
+      "loss": 0.4509,
+      "step": 579
+    },
+    {
+      "epoch": 2.8149158708503865,
+      "grad_norm": 0.17125198843453068,
+      "learning_rate": 3.8170584400768224e-05,
+      "loss": 0.4492,
+      "step": 580
+    },
+    {
+      "epoch": 2.8197665605578295,
+      "grad_norm": 0.18182599605587274,
+      "learning_rate": 3.8035177462997664e-05,
+      "loss": 0.4475,
+      "step": 581
+    },
+    {
+      "epoch": 2.824617250265272,
+      "grad_norm": 0.16612208849559923,
+      "learning_rate": 3.7899793091674396e-05,
+      "loss": 0.4419,
+      "step": 582
+    },
+    {
+      "epoch": 2.829467939972715,
+      "grad_norm": 0.15504382692612345,
+      "learning_rate": 3.776443284171969e-05,
+      "loss": 0.4421,
+      "step": 583
+    },
+    {
+      "epoch": 2.8343186296801575,
+      "grad_norm": 0.14158493240403466,
+      "learning_rate": 3.7629098267777706e-05,
+      "loss": 0.4399,
+      "step": 584
+    },
+    {
+      "epoch": 2.8391693193876004,
+      "grad_norm": 0.14521790840725082,
+      "learning_rate": 3.7493790924197746e-05,
+      "loss": 0.4328,
+      "step": 585
+    },
+    {
+      "epoch": 2.8440200090950434,
+      "grad_norm": 0.14223117063886642,
+      "learning_rate": 3.735851236501637e-05,
+      "loss": 0.4403,
+      "step": 586
+    },
+    {
+      "epoch": 2.848870698802486,
+      "grad_norm": 0.13716153813556975,
+      "learning_rate": 3.722326414393954e-05,
+      "loss": 0.4375,
+      "step": 587
+    },
+    {
+      "epoch": 2.853721388509929,
+      "grad_norm": 0.13481835773066367,
+      "learning_rate": 3.708804781432478e-05,
+      "loss": 0.4465,
+      "step": 588
+    },
+    {
+      "epoch": 2.858572078217372,
+      "grad_norm": 0.15098861155914894,
+      "learning_rate": 3.6952864929163286e-05,
+      "loss": 0.4478,
+      "step": 589
+    },
+    {
+      "epoch": 2.8634227679248143,
+      "grad_norm": 0.14428029096945075,
+      "learning_rate": 3.6817717041062204e-05,
+      "loss": 0.4433,
+      "step": 590
+    },
+    {
+      "epoch": 2.868273457632257,
+      "grad_norm": 0.14941685817838865,
+      "learning_rate": 3.66826057022267e-05,
+      "loss": 0.4426,
+      "step": 591
+    },
+    {
+      "epoch": 2.8731241473397,
+      "grad_norm": 0.15814915640950794,
+      "learning_rate": 3.654753246444217e-05,
+      "loss": 0.437,
+      "step": 592
+    },
+    {
+      "epoch": 2.8779748370471427,
+      "grad_norm": 0.1666350361712344,
+      "learning_rate": 3.641249887905638e-05,
+      "loss": 0.4404,
+      "step": 593
+    },
+    {
+      "epoch": 2.8828255267545853,
+      "grad_norm": 0.14761581189993908,
+      "learning_rate": 3.627750649696173e-05,
+      "loss": 0.4418,
+      "step": 594
+    },
+    {
+      "epoch": 2.887676216462028,
+      "grad_norm": 0.16433140815176697,
+      "learning_rate": 3.614255686857734e-05,
+      "loss": 0.4482,
+      "step": 595
+    },
+    {
+      "epoch": 2.892526906169471,
+      "grad_norm": 0.14898117787478918,
+      "learning_rate": 3.600765154383134e-05,
+      "loss": 0.4407,
+      "step": 596
+    },
+    {
+      "epoch": 2.8973775958769137,
+      "grad_norm": 0.1576311183401583,
+      "learning_rate": 3.587279207214301e-05,
+      "loss": 0.4502,
+      "step": 597
+    },
+    {
+      "epoch": 2.9022282855843566,
+      "grad_norm": 0.12713739329535242,
+      "learning_rate": 3.5737980002404965e-05,
+      "loss": 0.4504,
+      "step": 598
+    },
+    {
+      "epoch": 2.907078975291799,
+      "grad_norm": 0.1407162797770253,
+      "learning_rate": 3.5603216882965415e-05,
+      "loss": 0.4444,
+      "step": 599
+    },
+    {
+      "epoch": 2.911929664999242,
+      "grad_norm": 0.15096239269549241,
+      "learning_rate": 3.5468504261610387e-05,
+      "loss": 0.4497,
+      "step": 600
+    },
+    {
+      "epoch": 2.9167803547066846,
+      "grad_norm": 0.1341187523707445,
+      "learning_rate": 3.5333843685545914e-05,
+      "loss": 0.4449,
+      "step": 601
+    },
+    {
+      "epoch": 2.9216310444141276,
+      "grad_norm": 0.1619335943958044,
+      "learning_rate": 3.519923670138025e-05,
+      "loss": 0.4434,
+      "step": 602
+    },
+    {
+      "epoch": 2.9264817341215705,
+      "grad_norm": 0.16211928956809254,
+      "learning_rate": 3.506468485510616e-05,
+      "loss": 0.4394,
+      "step": 603
+    },
+    {
+      "epoch": 2.931332423829013,
+      "grad_norm": 0.13565985896740923,
+      "learning_rate": 3.493018969208314e-05,
+      "loss": 0.4513,
+      "step": 604
+    },
+    {
+      "epoch": 2.936183113536456,
+      "grad_norm": 0.14551792578562125,
+      "learning_rate": 3.479575275701965e-05,
+      "loss": 0.4425,
+      "step": 605
+    },
+    {
+      "epoch": 2.941033803243899,
+      "grad_norm": 0.17680002225930672,
+      "learning_rate": 3.4661375593955405e-05,
+      "loss": 0.4384,
+      "step": 606
+    },
+    {
+      "epoch": 2.9458844929513415,
+      "grad_norm": 0.16447736465809054,
+      "learning_rate": 3.45270597462436e-05,
+      "loss": 0.4451,
+      "step": 607
+    },
+    {
+      "epoch": 2.9507351826587844,
+      "grad_norm": 0.15299942108774958,
+      "learning_rate": 3.4392806756533233e-05,
+      "loss": 0.4383,
+      "step": 608
+    },
+    {
+      "epoch": 2.955585872366227,
+      "grad_norm": 0.1536623395643003,
+      "learning_rate": 3.425861816675135e-05,
+      "loss": 0.4453,
+      "step": 609
+    },
+    {
+      "epoch": 2.96043656207367,
+      "grad_norm": 0.15727509499916784,
+      "learning_rate": 3.4124495518085366e-05,
+      "loss": 0.436,
+      "step": 610
+    },
+    {
+      "epoch": 2.9652872517811124,
+      "grad_norm": 0.17906600408772821,
+      "learning_rate": 3.399044035096532e-05,
+      "loss": 0.4467,
+      "step": 611
+    },
+    {
+      "epoch": 2.9701379414885554,
+      "grad_norm": 0.16243556143131102,
+      "learning_rate": 3.3856454205046223e-05,
+      "loss": 0.4364,
+      "step": 612
+    },
+    {
+      "epoch": 2.9749886311959983,
+      "grad_norm": 0.1608114743453715,
+      "learning_rate": 3.372253861919036e-05,
+      "loss": 0.4517,
+      "step": 613
+    },
+    {
+      "epoch": 2.979839320903441,
+      "grad_norm": 0.16249676695986184,
+      "learning_rate": 3.3588695131449626e-05,
+      "loss": 0.4464,
+      "step": 614
+    },
+    {
+      "epoch": 2.984690010610884,
+      "grad_norm": 0.14682205641070967,
+      "learning_rate": 3.3454925279047854e-05,
+      "loss": 0.4446,
+      "step": 615
+    },
+    {
+      "epoch": 2.9895407003183267,
+      "grad_norm": 0.158750060737996,
+      "learning_rate": 3.3321230598363126e-05,
+      "loss": 0.4449,
+      "step": 616
+    },
+    {
+      "epoch": 2.9943913900257693,
+      "grad_norm": 0.15103569523913019,
+      "learning_rate": 3.3187612624910185e-05,
+      "loss": 0.4457,
+      "step": 617
+    },
+    {
+      "epoch": 2.999242079733212,
+      "grad_norm": 0.13279281722008607,
+      "learning_rate": 3.305407289332279e-05,
+      "loss": 0.4524,
+      "step": 618
+    },
+    {
+      "epoch": 3.004850689707443,
+      "grad_norm": 0.21786343264375474,
+      "learning_rate": 3.2920612937336035e-05,
+      "loss": 0.4186,
+      "step": 619
+    },
+    {
+      "epoch": 3.0097013794148855,
+      "grad_norm": 0.1940657522375173,
+      "learning_rate": 3.2787234289768816e-05,
+      "loss": 0.4198,
+      "step": 620
+    },
+    {
+      "epoch": 3.0145520691223284,
+      "grad_norm": 0.21773031084989108,
+      "learning_rate": 3.2653938482506125e-05,
+      "loss": 0.4257,
+      "step": 621
+    },
+    {
+      "epoch": 3.019402758829771,
+      "grad_norm": 0.20326547030411232,
+      "learning_rate": 3.252072704648157e-05,
+      "loss": 0.4165,
+      "step": 622
+    },
+    {
+      "epoch": 3.024253448537214,
+      "grad_norm": 0.1952882443077655,
+      "learning_rate": 3.2387601511659695e-05,
+      "loss": 0.4099,
+      "step": 623
+    },
+    {
+      "epoch": 3.029104138244657,
+      "grad_norm": 0.1918042514566916,
+      "learning_rate": 3.22545634070185e-05,
+      "loss": 0.4152,
+      "step": 624
+    },
+    {
+      "epoch": 3.0339548279520994,
+      "grad_norm": 0.19705116304049744,
+      "learning_rate": 3.212161426053177e-05,
+      "loss": 0.4128,
+      "step": 625
+    },
+    {
+      "epoch": 3.0388055176595423,
+      "grad_norm": 0.19649678060258355,
+      "learning_rate": 3.19887555991516e-05,
+      "loss": 0.4129,
+      "step": 626
+    },
+    {
+      "epoch": 3.043656207366985,
+      "grad_norm": 0.21708556979200094,
+      "learning_rate": 3.1855988948790866e-05,
+      "loss": 0.419,
+      "step": 627
+    },
+    {
+      "epoch": 3.048506897074428,
+      "grad_norm": 0.1975715363358727,
+      "learning_rate": 3.172331583430567e-05,
+      "loss": 0.4179,
+      "step": 628
+    },
+    {
+      "epoch": 3.0533575867818707,
+      "grad_norm": 0.23335868145268443,
+      "learning_rate": 3.1590737779477825e-05,
+      "loss": 0.4187,
+      "step": 629
+    },
+    {
+      "epoch": 3.0582082764893133,
+      "grad_norm": 0.16312780113553543,
+      "learning_rate": 3.145825630699734e-05,
+      "loss": 0.4145,
+      "step": 630
+    },
+    {
+      "epoch": 3.063058966196756,
+      "grad_norm": 0.1995031684584169,
+      "learning_rate": 3.1325872938444995e-05,
+      "loss": 0.4269,
+      "step": 631
+    },
+    {
+      "epoch": 3.0679096559041987,
+      "grad_norm": 0.16321617015370155,
+      "learning_rate": 3.119358919427478e-05,
+      "loss": 0.4189,
+      "step": 632
+    },
+    {
+      "epoch": 3.0727603456116417,
+      "grad_norm": 0.18410001775256316,
+      "learning_rate": 3.106140659379652e-05,
+      "loss": 0.4218,
+      "step": 633
+    },
+    {
+      "epoch": 3.0776110353190846,
+      "grad_norm": 0.15016856412332835,
+      "learning_rate": 3.092932665515837e-05,
+      "loss": 0.4142,
+      "step": 634
+    },
+    {
+      "epoch": 3.082461725026527,
+      "grad_norm": 0.1502499895895468,
+      "learning_rate": 3.079735089532935e-05,
+      "loss": 0.4092,
+      "step": 635
+    },
+    {
+      "epoch": 3.08731241473397,
+      "grad_norm": 0.1564011362374121,
+      "learning_rate": 3.0665480830082e-05,
+      "loss": 0.4176,
+      "step": 636
+    },
+    {
+      "epoch": 3.0921631044414126,
+      "grad_norm": 0.15490177035198396,
+      "learning_rate": 3.0533717973974924e-05,
+      "loss": 0.4171,
+      "step": 637
+    },
+    {
+      "epoch": 3.0970137941488556,
+      "grad_norm": 0.16842243538179394,
+      "learning_rate": 3.040206384033542e-05,
+      "loss": 0.4142,
+      "step": 638
+    },
+    {
+      "epoch": 3.101864483856298,
+      "grad_norm": 0.15647723657119908,
+      "learning_rate": 3.0270519941242052e-05,
+      "loss": 0.4045,
+      "step": 639
+    },
+    {
+      "epoch": 3.106715173563741,
+      "grad_norm": 0.14187950473293476,
+      "learning_rate": 3.0139087787507323e-05,
+      "loss": 0.4162,
+      "step": 640
+    },
+    {
+      "epoch": 3.111565863271184,
+      "grad_norm": 0.1486994798612613,
+      "learning_rate": 3.0007768888660337e-05,
+      "loss": 0.4162,
+      "step": 641
+    },
+    {
+      "epoch": 3.1164165529786265,
+      "grad_norm": 0.14045982542928215,
+      "learning_rate": 2.9876564752929406e-05,
+      "loss": 0.423,
+      "step": 642
+    },
+    {
+      "epoch": 3.1212672426860695,
+      "grad_norm": 0.14911944405394412,
+      "learning_rate": 2.9745476887224806e-05,
+      "loss": 0.4186,
+      "step": 643
+    },
+    {
+      "epoch": 3.1261179323935124,
+      "grad_norm": 0.130768356118423,
+      "learning_rate": 2.961450679712135e-05,
+      "loss": 0.4149,
+      "step": 644
+    },
+    {
+      "epoch": 3.130968622100955,
+      "grad_norm": 0.15344876789210227,
+      "learning_rate": 2.9483655986841265e-05,
+      "loss": 0.4185,
+      "step": 645
+    },
+    {
+      "epoch": 3.135819311808398,
+      "grad_norm": 0.16373681347456412,
+      "learning_rate": 2.9352925959236732e-05,
+      "loss": 0.4199,
+      "step": 646
+    },
+    {
+      "epoch": 3.1406700015158404,
+      "grad_norm": 0.13681917154319687,
+      "learning_rate": 2.92223182157728e-05,
+      "loss": 0.4187,
+      "step": 647
+    },
+    {
+      "epoch": 3.1455206912232834,
+      "grad_norm": 0.16376916195733163,
+      "learning_rate": 2.909183425650996e-05,
+      "loss": 0.4144,
+      "step": 648
+    },
+    {
+      "epoch": 3.150371380930726,
+      "grad_norm": 0.15231760580025508,
+      "learning_rate": 2.8961475580087108e-05,
+      "loss": 0.4065,
+      "step": 649
+    },
+    {
+      "epoch": 3.155222070638169,
+      "grad_norm": 0.1482250629623317,
+      "learning_rate": 2.8831243683704162e-05,
+      "loss": 0.4167,
+      "step": 650
+    },
+    {
+      "epoch": 3.160072760345612,
+      "grad_norm": 0.1279306947970543,
+      "learning_rate": 2.8701140063104996e-05,
+      "loss": 0.4163,
+      "step": 651
+    },
+    {
+      "epoch": 3.1649234500530543,
+      "grad_norm": 0.14839352063841615,
+      "learning_rate": 2.857116621256018e-05,
+      "loss": 0.4066,
+      "step": 652
+    },
+    {
+      "epoch": 3.1697741397604973,
+      "grad_norm": 0.14182099741632095,
+      "learning_rate": 2.8441323624849827e-05,
+      "loss": 0.4073,
+      "step": 653
+    },
+    {
+      "epoch": 3.17462482946794,
+      "grad_norm": 0.14062989686711488,
+      "learning_rate": 2.83116137912465e-05,
+      "loss": 0.4148,
+      "step": 654
+    },
+    {
+      "epoch": 3.1794755191753827,
+      "grad_norm": 0.13350262763245205,
+      "learning_rate": 2.8182038201498038e-05,
+      "loss": 0.4185,
+      "step": 655
+    },
+    {
+      "epoch": 3.1843262088828257,
+      "grad_norm": 0.14000623741594295,
+      "learning_rate": 2.8052598343810474e-05,
+      "loss": 0.4084,
+      "step": 656
+    },
+    {
+      "epoch": 3.189176898590268,
+      "grad_norm": 0.13922618624994615,
+      "learning_rate": 2.7923295704830868e-05,
+      "loss": 0.4209,
+      "step": 657
+    },
+    {
+      "epoch": 3.194027588297711,
+      "grad_norm": 0.13424360868636123,
+      "learning_rate": 2.7794131769630355e-05,
+      "loss": 0.4203,
+      "step": 658
+    },
+    {
+      "epoch": 3.1988782780051537,
+      "grad_norm": 0.1410971557670894,
+      "learning_rate": 2.7665108021687007e-05,
+      "loss": 0.4229,
+      "step": 659
+    },
+    {
+      "epoch": 3.2037289677125966,
+      "grad_norm": 0.15216131334925778,
+      "learning_rate": 2.753622594286879e-05,
+      "loss": 0.4145,
+      "step": 660
+    },
+    {
+      "epoch": 3.2085796574200396,
+      "grad_norm": 0.1261567069468345,
+      "learning_rate": 2.7407487013416615e-05,
+      "loss": 0.4083,
+      "step": 661
+    },
+    {
+      "epoch": 3.213430347127482,
+      "grad_norm": 0.15853568579534694,
+      "learning_rate": 2.727889271192722e-05,
+      "loss": 0.4187,
+      "step": 662
+    },
+    {
+      "epoch": 3.218281036834925,
+      "grad_norm": 0.1151093763460037,
+      "learning_rate": 2.715044451533631e-05,
+      "loss": 0.4164,
+      "step": 663
+    },
+    {
+      "epoch": 3.2231317265423676,
+      "grad_norm": 0.14296777545925654,
+      "learning_rate": 2.702214389890152e-05,
+      "loss": 0.413,
+      "step": 664
+    },
+    {
+      "epoch": 3.2279824162498105,
+      "grad_norm": 0.12844538251966497,
+      "learning_rate": 2.6893992336185512e-05,
+      "loss": 0.4035,
+      "step": 665
+    },
+    {
+      "epoch": 3.2328331059572535,
+      "grad_norm": 0.13036276585432338,
+      "learning_rate": 2.6765991299039025e-05,
+      "loss": 0.4145,
+      "step": 666
+    },
+    {
+      "epoch": 3.237683795664696,
+      "grad_norm": 0.12243507743746514,
+      "learning_rate": 2.663814225758393e-05,
+      "loss": 0.4117,
+      "step": 667
+    },
+    {
+      "epoch": 3.242534485372139,
+      "grad_norm": 0.14245972122975337,
+      "learning_rate": 2.6510446680196448e-05,
+      "loss": 0.4195,
+      "step": 668
+    },
+    {
+      "epoch": 3.2473851750795815,
+      "grad_norm": 0.12799039988824565,
+      "learning_rate": 2.638290603349023e-05,
+      "loss": 0.4203,
+      "step": 669
+    },
+    {
+      "epoch": 3.2522358647870244,
+      "grad_norm": 0.12961756855915293,
+      "learning_rate": 2.625552178229949e-05,
+      "loss": 0.4159,
+      "step": 670
+    },
+    {
+      "epoch": 3.2570865544944674,
+      "grad_norm": 0.1296944618260485,
+      "learning_rate": 2.612829538966218e-05,
+      "loss": 0.4111,
+      "step": 671
+    },
+    {
+      "epoch": 3.26193724420191,
+      "grad_norm": 0.14508116679367689,
+      "learning_rate": 2.6001228316803256e-05,
+      "loss": 0.4196,
+      "step": 672
+    },
+    {
+      "epoch": 3.266787933909353,
+      "grad_norm": 0.1216509600449271,
+      "learning_rate": 2.5874322023117824e-05,
+      "loss": 0.4162,
+      "step": 673
+    },
+    {
+      "epoch": 3.2716386236167954,
+      "grad_norm": 0.1407698696044201,
+      "learning_rate": 2.5747577966154404e-05,
+      "loss": 0.4165,
+      "step": 674
+    },
+    {
+      "epoch": 3.2764893133242383,
+      "grad_norm": 0.1142728140236527,
+      "learning_rate": 2.5620997601598215e-05,
+      "loss": 0.4076,
+      "step": 675
+    },
+    {
+      "epoch": 3.281340003031681,
+      "grad_norm": 0.14212475123175447,
+      "learning_rate": 2.5494582383254388e-05,
+      "loss": 0.4174,
+      "step": 676
+    },
+    {
+      "epoch": 3.286190692739124,
+      "grad_norm": 0.11798282133422631,
+      "learning_rate": 2.5368333763031324e-05,
+      "loss": 0.4131,
+      "step": 677
+    },
+    {
+      "epoch": 3.2910413824465667,
+      "grad_norm": 0.14618443390836333,
+      "learning_rate": 2.5242253190924034e-05,
+      "loss": 0.4092,
+      "step": 678
+    },
+    {
+      "epoch": 3.2958920721540093,
+      "grad_norm": 0.12815235524330332,
+      "learning_rate": 2.5116342114997442e-05,
+      "loss": 0.409,
+      "step": 679
+    },
+    {
+      "epoch": 3.300742761861452,
+      "grad_norm": 0.1346079173937725,
+      "learning_rate": 2.4990601981369737e-05,
+      "loss": 0.4201,
+      "step": 680
+    },
+    {
+      "epoch": 3.305593451568895,
+      "grad_norm": 0.12020401678661803,
+      "learning_rate": 2.4865034234195834e-05,
+      "loss": 0.4107,
+      "step": 681
+    },
+    {
+      "epoch": 3.3104441412763377,
+      "grad_norm": 0.1342860022847603,
+      "learning_rate": 2.4739640315650747e-05,
+      "loss": 0.4145,
+      "step": 682
+    },
+    {
+      "epoch": 3.3152948309837806,
+      "grad_norm": 0.11394559661374248,
+      "learning_rate": 2.4614421665912997e-05,
+      "loss": 0.4213,
+      "step": 683
+    },
+    {
+      "epoch": 3.320145520691223,
+      "grad_norm": 0.126805017438777,
+      "learning_rate": 2.4489379723148147e-05,
+      "loss": 0.4129,
+      "step": 684
+    },
+    {
+      "epoch": 3.324996210398666,
+      "grad_norm": 0.11817811298525939,
+      "learning_rate": 2.4364515923492187e-05,
+      "loss": 0.4193,
+      "step": 685
+    },
+    {
+      "epoch": 3.3298469001061086,
+      "grad_norm": 0.1342332761111468,
+      "learning_rate": 2.4239831701035143e-05,
+      "loss": 0.418,
+      "step": 686
+    },
+    {
+      "epoch": 3.3346975898135516,
+      "grad_norm": 0.12474527460295737,
+      "learning_rate": 2.411532848780451e-05,
+      "loss": 0.4166,
+      "step": 687
+    },
+    {
+      "epoch": 3.3395482795209945,
+      "grad_norm": 0.1329059048345405,
+      "learning_rate": 2.399100771374888e-05,
+      "loss": 0.4138,
+      "step": 688
+    },
+    {
+      "epoch": 3.344398969228437,
+      "grad_norm": 0.11628441384991241,
+      "learning_rate": 2.3866870806721495e-05,
+      "loss": 0.4111,
+      "step": 689
+    },
+    {
+      "epoch": 3.34924965893588,
+      "grad_norm": 0.1453864432398833,
+      "learning_rate": 2.37429191924638e-05,
+      "loss": 0.42,
+      "step": 690
+    },
+    {
+      "epoch": 3.354100348643323,
+      "grad_norm": 0.11264715608612798,
+      "learning_rate": 2.361915429458913e-05,
+      "loss": 0.417,
+      "step": 691
+    },
+    {
+      "epoch": 3.3589510383507655,
+      "grad_norm": 0.14248256005839127,
+      "learning_rate": 2.349557753456637e-05,
+      "loss": 0.4168,
+      "step": 692
+    },
+    {
+      "epoch": 3.3638017280582084,
+      "grad_norm": 0.12398727462550883,
+      "learning_rate": 2.3372190331703556e-05,
+      "loss": 0.4189,
+      "step": 693
+    },
+    {
+      "epoch": 3.368652417765651,
+      "grad_norm": 0.11869463547338541,
+      "learning_rate": 2.324899410313161e-05,
+      "loss": 0.4125,
+      "step": 694
+    },
+    {
+      "epoch": 3.373503107473094,
+      "grad_norm": 0.12377753865053726,
+      "learning_rate": 2.3125990263788118e-05,
+      "loss": 0.4186,
+      "step": 695
+    },
+    {
+      "epoch": 3.3783537971805364,
+      "grad_norm": 0.13141561076446406,
+      "learning_rate": 2.3003180226400986e-05,
+      "loss": 0.4123,
+      "step": 696
+    },
+    {
+      "epoch": 3.3832044868879794,
+      "grad_norm": 0.11816327910326484,
+      "learning_rate": 2.288056540147229e-05,
+      "loss": 0.4129,
+      "step": 697
+    },
+    {
+      "epoch": 3.3880551765954223,
+      "grad_norm": 0.12260669143866527,
+      "learning_rate": 2.275814719726201e-05,
+      "loss": 0.4133,
+      "step": 698
+    },
+    {
+      "epoch": 3.392905866302865,
+      "grad_norm": 0.133084483131333,
+      "learning_rate": 2.263592701977193e-05,
+      "loss": 0.4219,
+      "step": 699
+    },
+    {
+      "epoch": 3.397756556010308,
+      "grad_norm": 0.11848073628628028,
+      "learning_rate": 2.2513906272729397e-05,
+      "loss": 0.4143,
+      "step": 700
+    },
+    {
+      "epoch": 3.4026072457177503,
+      "grad_norm": 0.12858950370510128,
+      "learning_rate": 2.239208635757133e-05,
+      "loss": 0.4166,
+      "step": 701
+    },
+    {
+      "epoch": 3.4074579354251933,
+      "grad_norm": 0.11722692816596028,
+      "learning_rate": 2.2270468673428004e-05,
+      "loss": 0.4259,
+      "step": 702
+    },
+    {
+      "epoch": 3.412308625132636,
+      "grad_norm": 0.11830608786302087,
+      "learning_rate": 2.2149054617106974e-05,
+      "loss": 0.407,
+      "step": 703
+    },
+    {
+      "epoch": 3.4171593148400787,
+      "grad_norm": 0.12285165179780408,
+      "learning_rate": 2.2027845583077175e-05,
+      "loss": 0.4231,
+      "step": 704
+    },
+    {
+      "epoch": 3.4220100045475217,
+      "grad_norm": 0.12884220459818324,
+      "learning_rate": 2.1906842963452757e-05,
+      "loss": 0.4069,
+      "step": 705
+    },
+    {
+      "epoch": 3.426860694254964,
+      "grad_norm": 0.10700415399515635,
+      "learning_rate": 2.178604814797715e-05,
+      "loss": 0.4149,
+      "step": 706
+    },
+    {
+      "epoch": 3.431711383962407,
+      "grad_norm": 0.11853515304365536,
+      "learning_rate": 2.1665462524007162e-05,
+      "loss": 0.4125,
+      "step": 707
+    },
+    {
+      "epoch": 3.43656207366985,
+      "grad_norm": 0.10772813217204756,
+      "learning_rate": 2.1545087476496903e-05,
+      "loss": 0.4216,
+      "step": 708
+    },
+    {
+      "epoch": 3.4414127633772926,
+      "grad_norm": 0.12946073793938545,
+      "learning_rate": 2.1424924387981996e-05,
+      "loss": 0.4227,
+      "step": 709
+    },
+    {
+      "epoch": 3.4462634530847356,
+      "grad_norm": 0.11256888467579416,
+      "learning_rate": 2.1304974638563715e-05,
+      "loss": 0.4116,
+      "step": 710
+    },
+    {
+      "epoch": 3.451114142792178,
+      "grad_norm": 0.12334394272902899,
+      "learning_rate": 2.1185239605893013e-05,
+      "loss": 0.4217,
+      "step": 711
+    },
+    {
+      "epoch": 3.455964832499621,
+      "grad_norm": 0.11467056354627854,
+      "learning_rate": 2.106572066515482e-05,
+      "loss": 0.4193,
+      "step": 712
+    },
+    {
+      "epoch": 3.460815522207064,
+      "grad_norm": 0.11248104529388973,
+      "learning_rate": 2.0946419189052162e-05,
+      "loss": 0.4147,
+      "step": 713
+    },
+    {
+      "epoch": 3.4656662119145065,
+      "grad_norm": 0.11839620237634582,
+      "learning_rate": 2.0827336547790452e-05,
+      "loss": 0.4214,
+      "step": 714
+    },
+    {
+      "epoch": 3.4705169016219495,
+      "grad_norm": 0.11350680266976185,
+      "learning_rate": 2.0708474109061752e-05,
+      "loss": 0.416,
+      "step": 715
+    },
+    {
+      "epoch": 3.475367591329392,
+      "grad_norm": 0.11310394993568851,
+      "learning_rate": 2.0589833238029032e-05,
+      "loss": 0.4001,
+      "step": 716
+    },
+    {
+      "epoch": 3.480218281036835,
+      "grad_norm": 0.1142115014102716,
+      "learning_rate": 2.0471415297310455e-05,
+      "loss": 0.4158,
+      "step": 717
+    },
+    {
+      "epoch": 3.485068970744278,
+      "grad_norm": 0.12114855144347077,
+      "learning_rate": 2.0353221646963864e-05,
+      "loss": 0.41,
+      "step": 718
+    },
+    {
+      "epoch": 3.4899196604517204,
+      "grad_norm": 0.11155618948058961,
+      "learning_rate": 2.0235253644471012e-05,
+      "loss": 0.4226,
+      "step": 719
+    },
+    {
+      "epoch": 3.4947703501591634,
+      "grad_norm": 0.12346483161916652,
+      "learning_rate": 2.011751264472206e-05,
+      "loss": 0.4163,
+      "step": 720
+    },
+    {
+      "epoch": 3.499621039866606,
+      "grad_norm": 0.1314504891806893,
+      "learning_rate": 2.0000000000000012e-05,
+      "loss": 0.4185,
+      "step": 721
+    },
+    {
+      "epoch": 3.504471729574049,
+      "grad_norm": 0.11587490363769114,
+      "learning_rate": 1.9882717059965086e-05,
+      "loss": 0.4243,
+      "step": 722
+    },
+    {
+      "epoch": 3.5093224192814914,
+      "grad_norm": 0.147492515392888,
+      "learning_rate": 1.9765665171639345e-05,
+      "loss": 0.4139,
+      "step": 723
+    },
+    {
+      "epoch": 3.5141731089889343,
+      "grad_norm": 0.13161876983358525,
+      "learning_rate": 1.964884567939118e-05,
+      "loss": 0.4023,
+      "step": 724
+    },
+    {
+      "epoch": 3.5190237986963773,
+      "grad_norm": 0.1288012773595963,
+      "learning_rate": 1.9532259924919823e-05,
+      "loss": 0.416,
+      "step": 725
+    },
+    {
+      "epoch": 3.52387448840382,
+      "grad_norm": 0.10979902623938809,
+      "learning_rate": 1.9415909247239996e-05,
+      "loss": 0.4198,
+      "step": 726
+    },
+    {
+      "epoch": 3.5287251781112627,
+      "grad_norm": 0.12273321372911124,
+      "learning_rate": 1.9299794982666485e-05,
+      "loss": 0.4084,
+      "step": 727
+    },
+    {
+      "epoch": 3.5335758678187057,
+      "grad_norm": 0.11485189001172974,
+      "learning_rate": 1.9183918464798837e-05,
+      "loss": 0.408,
+      "step": 728
+    },
+    {
+      "epoch": 3.538426557526148,
+      "grad_norm": 0.1142263107632135,
+      "learning_rate": 1.906828102450601e-05,
+      "loss": 0.4131,
+      "step": 729
+    },
+    {
+      "epoch": 3.543277247233591,
+      "grad_norm": 0.11235720950088848,
+      "learning_rate": 1.895288398991114e-05,
+      "loss": 0.4102,
+      "step": 730
+    },
+    {
+      "epoch": 3.5481279369410337,
+      "grad_norm": 0.11476446551397697,
+      "learning_rate": 1.8837728686376158e-05,
+      "loss": 0.4176,
+      "step": 731
+    },
+    {
+      "epoch": 3.5529786266484766,
+      "grad_norm": 0.12252689326129268,
+      "learning_rate": 1.8722816436486754e-05,
+      "loss": 0.4238,
+      "step": 732
+    },
+    {
+      "epoch": 3.557829316355919,
+      "grad_norm": 0.10742367387279236,
+      "learning_rate": 1.8608148560037036e-05,
+      "loss": 0.4248,
+      "step": 733
+    },
+    {
+      "epoch": 3.562680006063362,
+      "grad_norm": 0.11202485484827082,
+      "learning_rate": 1.8493726374014442e-05,
+      "loss": 0.4047,
+      "step": 734
+    },
+    {
+      "epoch": 3.567530695770805,
+      "grad_norm": 0.10619927645163178,
+      "learning_rate": 1.8379551192584588e-05,
+      "loss": 0.4105,
+      "step": 735
+    },
+    {
+      "epoch": 3.5723813854782476,
+      "grad_norm": 0.1104031551739802,
+      "learning_rate": 1.826562432707619e-05,
+      "loss": 0.4158,
+      "step": 736
+    },
+    {
+      "epoch": 3.5772320751856905,
+      "grad_norm": 0.10920848831138874,
+      "learning_rate": 1.8151947085965994e-05,
+      "loss": 0.4157,
+      "step": 737
+    },
+    {
+      "epoch": 3.5820827648931335,
+      "grad_norm": 0.10703865180962287,
+      "learning_rate": 1.803852077486377e-05,
+      "loss": 0.4144,
+      "step": 738
+    },
+    {
+      "epoch": 3.586933454600576,
+      "grad_norm": 0.1043548569854331,
+      "learning_rate": 1.7925346696497295e-05,
+      "loss": 0.4082,
+      "step": 739
+    },
+    {
+      "epoch": 3.591784144308019,
+      "grad_norm": 0.11138455070453787,
+      "learning_rate": 1.781242615069733e-05,
+      "loss": 0.4137,
+      "step": 740
+    },
+    {
+      "epoch": 3.5966348340154615,
+      "grad_norm": 0.11820322046978973,
+      "learning_rate": 1.7699760434382853e-05,
+      "loss": 0.4108,
+      "step": 741
+    },
+    {
+      "epoch": 3.6014855237229044,
+      "grad_norm": 0.11278678223744326,
+      "learning_rate": 1.758735084154601e-05,
+      "loss": 0.4189,
+      "step": 742
+    },
+    {
+      "epoch": 3.606336213430347,
+      "grad_norm": 0.11323272041806605,
+      "learning_rate": 1.7475198663237297e-05,
+      "loss": 0.4123,
+      "step": 743
+    },
+    {
+      "epoch": 3.61118690313779,
+      "grad_norm": 0.10904989882061365,
+      "learning_rate": 1.736330518755082e-05,
+      "loss": 0.4158,
+      "step": 744
+    },
+    {
+      "epoch": 3.616037592845233,
+      "grad_norm": 0.11229772775053595,
+      "learning_rate": 1.7251671699609313e-05,
+      "loss": 0.4182,
+      "step": 745
+    },
+    {
+      "epoch": 3.6208882825526754,
+      "grad_norm": 0.10951334908001022,
+      "learning_rate": 1.7140299481549557e-05,
+      "loss": 0.4213,
+      "step": 746
+    },
+    {
+      "epoch": 3.6257389722601183,
+      "grad_norm": 0.11434493342191,
+      "learning_rate": 1.7029189812507603e-05,
+      "loss": 0.4224,
+      "step": 747
+    },
+    {
+      "epoch": 3.6305896619675613,
+      "grad_norm": 0.10344385368037658,
+      "learning_rate": 1.6918343968604027e-05,
+      "loss": 0.4106,
+      "step": 748
+    },
+    {
+      "epoch": 3.635440351675004,
+      "grad_norm": 0.11168549168378746,
+      "learning_rate": 1.6807763222929315e-05,
+      "loss": 0.408,
+      "step": 749
+    },
+    {
+      "epoch": 3.6402910413824463,
+      "grad_norm": 0.11594040802990377,
+      "learning_rate": 1.669744884552926e-05,
+      "loss": 0.4169,
+      "step": 750
+    },
+    {
+      "epoch": 3.6451417310898893,
+      "grad_norm": 0.11452998735643415,
+      "learning_rate": 1.6587402103390314e-05,
+      "loss": 0.4162,
+      "step": 751
+    },
+    {
+      "epoch": 3.649992420797332,
+      "grad_norm": 0.11033578052460767,
+      "learning_rate": 1.6477624260425137e-05,
+      "loss": 0.4179,
+      "step": 752
+    },
+    {
+      "epoch": 3.6548431105047747,
+      "grad_norm": 0.1198668006175803,
+      "learning_rate": 1.6368116577457973e-05,
+      "loss": 0.4124,
+      "step": 753
+    },
+    {
+      "epoch": 3.6596938002122177,
+      "grad_norm": 0.10884921001547737,
+      "learning_rate": 1.6258880312210195e-05,
+      "loss": 0.4152,
+      "step": 754
+    },
+    {
+      "epoch": 3.6645444899196606,
+      "grad_norm": 0.12009675101549108,
+      "learning_rate": 1.6149916719285942e-05,
+      "loss": 0.4147,
+      "step": 755
+    },
+    {
+      "epoch": 3.669395179627103,
+      "grad_norm": 0.11742205741912104,
+      "learning_rate": 1.6041227050157607e-05,
+      "loss": 0.4096,
+      "step": 756
+    },
+    {
+      "epoch": 3.674245869334546,
+      "grad_norm": 0.1123251375365744,
+      "learning_rate": 1.5932812553151506e-05,
+      "loss": 0.4128,
+      "step": 757
+    },
+    {
+      "epoch": 3.6790965590419886,
+      "grad_norm": 0.12009728599563303,
+      "learning_rate": 1.582467447343355e-05,
+      "loss": 0.4179,
+      "step": 758
+    },
+    {
+      "epoch": 3.6839472487494316,
+      "grad_norm": 0.10679356522304362,
+      "learning_rate": 1.5716814052994928e-05,
+      "loss": 0.4153,
+      "step": 759
+    },
+    {
+      "epoch": 3.688797938456874,
+      "grad_norm": 0.11287526417521035,
+      "learning_rate": 1.5609232530637827e-05,
+      "loss": 0.4065,
+      "step": 760
+    },
+    {
+      "epoch": 3.693648628164317,
+      "grad_norm": 0.10604174975295146,
+      "learning_rate": 1.5501931141961278e-05,
+      "loss": 0.4135,
+      "step": 761
+    },
+    {
+      "epoch": 3.69849931787176,
+      "grad_norm": 0.10895447717123838,
+      "learning_rate": 1.539491111934686e-05,
+      "loss": 0.4102,
+      "step": 762
+    },
+    {
+      "epoch": 3.7033500075792025,
+      "grad_norm": 0.11522106117016057,
+      "learning_rate": 1.5288173691944613e-05,
+      "loss": 0.4193,
+      "step": 763
+    },
+    {
+      "epoch": 3.7082006972866455,
+      "grad_norm": 0.11668799730823959,
+      "learning_rate": 1.5181720085658906e-05,
+      "loss": 0.4131,
+      "step": 764
+    },
+    {
+      "epoch": 3.7130513869940884,
+      "grad_norm": 0.1179757070897769,
+      "learning_rate": 1.5075551523134358e-05,
+      "loss": 0.4107,
+      "step": 765
+    },
+    {
+      "epoch": 3.717902076701531,
+      "grad_norm": 0.11352197320512793,
+      "learning_rate": 1.4969669223741771e-05,
+      "loss": 0.4093,
+      "step": 766
+    },
+    {
+      "epoch": 3.722752766408974,
+      "grad_norm": 0.11945491424386492,
+      "learning_rate": 1.4864074403564216e-05,
+      "loss": 0.4142,
+      "step": 767
+    },
+    {
+      "epoch": 3.7276034561164164,
+      "grad_norm": 0.10892285815783607,
+      "learning_rate": 1.4758768275382887e-05,
+      "loss": 0.4205,
+      "step": 768
+    },
+    {
+      "epoch": 3.7324541458238594,
+      "grad_norm": 0.12765157069597566,
+      "learning_rate": 1.4653752048663394e-05,
+      "loss": 0.412,
+      "step": 769
+    },
+    {
+      "epoch": 3.737304835531302,
+      "grad_norm": 0.10956587187939422,
+      "learning_rate": 1.4549026929541693e-05,
+      "loss": 0.4148,
+      "step": 770
+    },
+    {
+      "epoch": 3.742155525238745,
+      "grad_norm": 0.12138544117757244,
+      "learning_rate": 1.4444594120810326e-05,
+      "loss": 0.4115,
+      "step": 771
+    },
+    {
+      "epoch": 3.747006214946188,
+      "grad_norm": 0.10697996518645103,
+      "learning_rate": 1.4340454821904573e-05,
+      "loss": 0.4194,
+      "step": 772
+    },
+    {
+      "epoch": 3.7518569046536303,
+      "grad_norm": 0.10998535306961314,
+      "learning_rate": 1.4236610228888683e-05,
+      "loss": 0.4143,
+      "step": 773
+    },
+    {
+      "epoch": 3.7567075943610733,
+      "grad_norm": 0.10998459629638649,
+      "learning_rate": 1.4133061534442133e-05,
+      "loss": 0.4121,
+      "step": 774
+    },
+    {
+      "epoch": 3.7615582840685162,
+      "grad_norm": 0.1056241184630642,
+      "learning_rate": 1.4029809927845981e-05,
+      "loss": 0.4146,
+      "step": 775
+    },
+    {
+      "epoch": 3.7664089737759587,
+      "grad_norm": 0.11617390368347923,
+      "learning_rate": 1.3926856594969115e-05,
+      "loss": 0.4125,
+      "step": 776
+    },
+    {
+      "epoch": 3.7712596634834017,
+      "grad_norm": 0.10105741678809248,
+      "learning_rate": 1.3824202718254655e-05,
+      "loss": 0.4081,
+      "step": 777
+    },
+    {
+      "epoch": 3.776110353190844,
+      "grad_norm": 0.12303653861990478,
+      "learning_rate": 1.3721849476706477e-05,
+      "loss": 0.413,
+      "step": 778
+    },
+    {
+      "epoch": 3.780961042898287,
+      "grad_norm": 0.10128855312982177,
+      "learning_rate": 1.3619798045875529e-05,
+      "loss": 0.4117,
+      "step": 779
+    },
+    {
+      "epoch": 3.7858117326057297,
+      "grad_norm": 0.1118697000457396,
+      "learning_rate": 1.3518049597846412e-05,
+      "loss": 0.4097,
+      "step": 780
+    },
+    {
+      "epoch": 3.7906624223131726,
+      "grad_norm": 0.10689143155371568,
+      "learning_rate": 1.3416605301223893e-05,
+      "loss": 0.4146,
+      "step": 781
+    },
+    {
+      "epoch": 3.7955131120206156,
+      "grad_norm": 0.09942573069367382,
+      "learning_rate": 1.3315466321119486e-05,
+      "loss": 0.415,
+      "step": 782
+    },
+    {
+      "epoch": 3.800363801728058,
+      "grad_norm": 0.112223020708749,
+      "learning_rate": 1.3214633819138105e-05,
+      "loss": 0.4187,
+      "step": 783
+    },
+    {
+      "epoch": 3.805214491435501,
+      "grad_norm": 0.09509376603334437,
+      "learning_rate": 1.3114108953364655e-05,
+      "loss": 0.4083,
+      "step": 784
+    },
+    {
+      "epoch": 3.810065181142944,
+      "grad_norm": 0.1039694903593983,
+      "learning_rate": 1.3013892878350771e-05,
+      "loss": 0.415,
+      "step": 785
+    },
+    {
+      "epoch": 3.8149158708503865,
+      "grad_norm": 0.09799457517015436,
+      "learning_rate": 1.2913986745101567e-05,
+      "loss": 0.4082,
+      "step": 786
+    },
+    {
+      "epoch": 3.8197665605578295,
+      "grad_norm": 0.0994744466895985,
+      "learning_rate": 1.2814391701062392e-05,
+      "loss": 0.416,
+      "step": 787
+    },
+    {
+      "epoch": 3.824617250265272,
+      "grad_norm": 0.09937882084294014,
+      "learning_rate": 1.2715108890105663e-05,
+      "loss": 0.4118,
+      "step": 788
+    },
+    {
+      "epoch": 3.829467939972715,
+      "grad_norm": 0.09366936781008306,
+      "learning_rate": 1.2616139452517748e-05,
+      "loss": 0.4202,
+      "step": 789
+    },
+    {
+      "epoch": 3.8343186296801575,
+      "grad_norm": 0.10103247526201467,
+      "learning_rate": 1.2517484524985836e-05,
+      "loss": 0.414,
+      "step": 790
+    },
+    {
+      "epoch": 3.8391693193876004,
+      "grad_norm": 0.10154281976161658,
+      "learning_rate": 1.2419145240584856e-05,
+      "loss": 0.4169,
+      "step": 791
+    },
+    {
+      "epoch": 3.8440200090950434,
+      "grad_norm": 0.09917419975099584,
+      "learning_rate": 1.2321122728764566e-05,
+      "loss": 0.4121,
+      "step": 792
+    },
+    {
+      "epoch": 3.848870698802486,
+      "grad_norm": 0.10067017561953691,
+      "learning_rate": 1.222341811533648e-05,
+      "loss": 0.4177,
+      "step": 793
+    },
+    {
+      "epoch": 3.853721388509929,
+      "grad_norm": 0.10318548830281854,
+      "learning_rate": 1.2126032522460975e-05,
+      "loss": 0.4211,
+      "step": 794
+    },
+    {
+      "epoch": 3.858572078217372,
+      "grad_norm": 0.09616344349182201,
+      "learning_rate": 1.2028967068634417e-05,
+      "loss": 0.4204,
+      "step": 795
+    },
+    {
+      "epoch": 3.8634227679248143,
+      "grad_norm": 0.10489574786705688,
+      "learning_rate": 1.193222286867628e-05,
+      "loss": 0.4119,
+      "step": 796
+    },
+    {
+      "epoch": 3.868273457632257,
+      "grad_norm": 0.1008569871750799,
+      "learning_rate": 1.1835801033716372e-05,
+      "loss": 0.4086,
+      "step": 797
+    },
+    {
+      "epoch": 3.8731241473397,
+      "grad_norm": 0.09122317803788128,
+      "learning_rate": 1.1739702671182083e-05,
+      "loss": 0.4214,
+      "step": 798
+    },
+    {
+      "epoch": 3.8779748370471427,
+      "grad_norm": 0.0991904412555539,
+      "learning_rate": 1.1643928884785618e-05,
+      "loss": 0.4095,
+      "step": 799
+    },
+    {
+      "epoch": 3.8828255267545853,
+      "grad_norm": 0.0959491624589887,
+      "learning_rate": 1.1548480774511353e-05,
+      "loss": 0.4218,
+      "step": 800
+    },
+    {
+      "epoch": 3.887676216462028,
+      "grad_norm": 0.09541302821779325,
+      "learning_rate": 1.1453359436603213e-05,
+      "loss": 0.4218,
+      "step": 801
+    },
+    {
+      "epoch": 3.892526906169471,
+      "grad_norm": 0.09054297930670004,
+      "learning_rate": 1.1358565963552039e-05,
+      "loss": 0.421,
+      "step": 802
+    },
+    {
+      "epoch": 3.8973775958769137,
+      "grad_norm": 0.09175434223796135,
+      "learning_rate": 1.126410144408312e-05,
+      "loss": 0.4088,
+      "step": 803
+    },
+    {
+      "epoch": 3.9022282855843566,
+      "grad_norm": 0.0933968969265613,
+      "learning_rate": 1.1169966963143568e-05,
+      "loss": 0.4105,
+      "step": 804
+    },
+    {
+      "epoch": 3.907078975291799,
+      "grad_norm": 0.09827723734859882,
+      "learning_rate": 1.1076163601889953e-05,
+      "loss": 0.4114,
+      "step": 805
+    },
+    {
+      "epoch": 3.911929664999242,
+      "grad_norm": 0.09185288445531155,
+      "learning_rate": 1.098269243767589e-05,
+      "loss": 0.4099,
+      "step": 806
+    },
+    {
+      "epoch": 3.9167803547066846,
+      "grad_norm": 0.10344811490009215,
+      "learning_rate": 1.0889554544039593e-05,
+      "loss": 0.417,
+      "step": 807
+    },
+    {
+      "epoch": 3.9216310444141276,
+      "grad_norm": 0.09584223413813131,
+      "learning_rate": 1.0796750990691596e-05,
+      "loss": 0.4092,
+      "step": 808
+    },
+    {
+      "epoch": 3.9264817341215705,
+      "grad_norm": 0.10225202685968936,
+      "learning_rate": 1.0704282843502459e-05,
+      "loss": 0.4156,
+      "step": 809
+    },
+    {
+      "epoch": 3.931332423829013,
+      "grad_norm": 0.10087206295646214,
+      "learning_rate": 1.0612151164490525e-05,
+      "loss": 0.4209,
+      "step": 810
+    },
+    {
+      "epoch": 3.936183113536456,
+      "grad_norm": 0.09495839512641314,
+      "learning_rate": 1.0520357011809707e-05,
+      "loss": 0.4193,
+      "step": 811
+    },
+    {
+      "epoch": 3.941033803243899,
+      "grad_norm": 0.10115287872844174,
+      "learning_rate": 1.0428901439737387e-05,
+      "loss": 0.415,
+      "step": 812
+    },
+    {
+      "epoch": 3.9458844929513415,
+      "grad_norm": 0.09538434619439141,
+      "learning_rate": 1.0337785498662223e-05,
+      "loss": 0.4152,
+      "step": 813
+    },
+    {
+      "epoch": 3.9507351826587844,
+      "grad_norm": 0.0955931932252973,
+      "learning_rate": 1.024701023507216e-05,
+      "loss": 0.4153,
+      "step": 814
+    },
+    {
+      "epoch": 3.955585872366227,
+      "grad_norm": 0.10045702643945939,
+      "learning_rate": 1.015657669154237e-05,
+      "loss": 0.4156,
+      "step": 815
+    },
+    {
+      "epoch": 3.96043656207367,
+      "grad_norm": 0.09335473216661304,
+      "learning_rate": 1.00664859067233e-05,
+      "loss": 0.4109,
+      "step": 816
+    },
+    {
+      "epoch": 3.9652872517811124,
+      "grad_norm": 0.09645722262367523,
+      "learning_rate": 9.976738915328719e-06,
+      "loss": 0.4107,
+      "step": 817
+    },
+    {
+      "epoch": 3.9701379414885554,
+      "grad_norm": 0.09446841411423582,
+      "learning_rate": 9.887336748123864e-06,
+      "loss": 0.4178,
+      "step": 818
+    },
+    {
+      "epoch": 3.9749886311959983,
+      "grad_norm": 0.0903235693705833,
+      "learning_rate": 9.798280431913558e-06,
+      "loss": 0.4202,
+      "step": 819
+    },
+    {
+      "epoch": 3.979839320903441,
+      "grad_norm": 0.0948091272844266,
+      "learning_rate": 9.709570989530493e-06,
+      "loss": 0.4123,
+      "step": 820
+    },
+    {
+      "epoch": 3.984690010610884,
+      "grad_norm": 0.10103242493534337,
+      "learning_rate": 9.621209439823388e-06,
+      "loss": 0.4132,
+      "step": 821
+    },
+    {
+      "epoch": 3.9895407003183267,
+      "grad_norm": 0.09199808648298305,
+      "learning_rate": 9.533196797645354e-06,
+      "loss": 0.4101,
+      "step": 822
+    },
+    {
+      "epoch": 3.9943913900257693,
+      "grad_norm": 0.09866422487619428,
+      "learning_rate": 9.44553407384221e-06,
+      "loss": 0.412,
+      "step": 823
+    },
+    {
+      "epoch": 3.999242079733212,
+      "grad_norm": 0.09950215009808663,
+      "learning_rate": 9.358222275240884e-06,
+      "loss": 0.4113,
+      "step": 824
+    },
+    {
+      "epoch": 4.0048506897074425,
+      "grad_norm": 0.16227668213643984,
+      "learning_rate": 9.271262404637835e-06,
+      "loss": 0.4032,
+      "step": 825
+    },
+    {
+      "epoch": 4.009701379414886,
+      "grad_norm": 0.11430363283166092,
+      "learning_rate": 9.184655460787591e-06,
+      "loss": 0.3988,
+      "step": 826
+    },
+    {
+      "epoch": 4.014552069122328,
+      "grad_norm": 0.11751760908320912,
+      "learning_rate": 9.098402438391161e-06,
+      "loss": 0.3943,
+      "step": 827
+    },
+    {
+      "epoch": 4.019402758829771,
+      "grad_norm": 0.13225951402360853,
+      "learning_rate": 9.012504328084724e-06,
+      "loss": 0.4024,
+      "step": 828
+    },
+    {
+      "epoch": 4.0242534485372135,
+      "grad_norm": 0.1269725744791709,
+      "learning_rate": 8.926962116428228e-06,
+      "loss": 0.4,
+      "step": 829
+    },
+    {
+      "epoch": 4.029104138244657,
+      "grad_norm": 0.12436207204015083,
+      "learning_rate": 8.841776785894014e-06,
+      "loss": 0.3994,
+      "step": 830
+    },
+    {
+      "epoch": 4.033954827952099,
+      "grad_norm": 0.12882961013897004,
+      "learning_rate": 8.756949314855565e-06,
+      "loss": 0.3977,
+      "step": 831
+    },
+    {
+      "epoch": 4.038805517659542,
+      "grad_norm": 0.11685196489455994,
+      "learning_rate": 8.672480677576267e-06,
+      "loss": 0.3906,
+      "step": 832
+    },
+    {
+      "epoch": 4.043656207366985,
+      "grad_norm": 0.106568471380297,
+      "learning_rate": 8.58837184419821e-06,
+      "loss": 0.388,
+      "step": 833
+    },
+    {
+      "epoch": 4.048506897074428,
+      "grad_norm": 0.1134401036269532,
+      "learning_rate": 8.504623780731056e-06,
+      "loss": 0.3918,
+      "step": 834
+    },
+    {
+      "epoch": 4.05335758678187,
+      "grad_norm": 0.12093411095557142,
+      "learning_rate": 8.421237449040962e-06,
+      "loss": 0.397,
+      "step": 835
+    },
+    {
+      "epoch": 4.058208276489314,
+      "grad_norm": 0.11674197768223235,
+      "learning_rate": 8.338213806839453e-06,
+      "loss": 0.393,
+      "step": 836
+    },
+    {
+      "epoch": 4.063058966196756,
+      "grad_norm": 0.10082248897884766,
+      "learning_rate": 8.255553807672547e-06,
+      "loss": 0.3897,
+      "step": 837
+    },
+    {
+      "epoch": 4.067909655904199,
+      "grad_norm": 0.11069126863357161,
+      "learning_rate": 8.1732584009097e-06,
+      "loss": 0.3882,
+      "step": 838
+    },
+    {
+      "epoch": 4.072760345611641,
+      "grad_norm": 0.11738423041846735,
+      "learning_rate": 8.091328531732925e-06,
+      "loss": 0.3959,
+      "step": 839
+    },
+    {
+      "epoch": 4.077611035319085,
+      "grad_norm": 0.10387395387726432,
+      "learning_rate": 8.009765141126014e-06,
+      "loss": 0.3891,
+      "step": 840
+    },
+    {
+      "epoch": 4.082461725026527,
+      "grad_norm": 0.10003982048853301,
+      "learning_rate": 7.928569165863584e-06,
+      "loss": 0.3909,
+      "step": 841
+    },
+    {
+      "epoch": 4.08731241473397,
+      "grad_norm": 0.10549262492820612,
+      "learning_rate": 7.847741538500439e-06,
+      "loss": 0.3875,
+      "step": 842
+    },
+    {
+      "epoch": 4.092163104441413,
+      "grad_norm": 0.10986857014715488,
+      "learning_rate": 7.767283187360846e-06,
+      "loss": 0.3929,
+      "step": 843
+    },
+    {
+      "epoch": 4.097013794148856,
+      "grad_norm": 0.09674503227632227,
+      "learning_rate": 7.687195036527813e-06,
+      "loss": 0.3928,
+      "step": 844
+    },
+    {
+      "epoch": 4.101864483856298,
+      "grad_norm": 0.10305617355184506,
+      "learning_rate": 7.60747800583252e-06,
+      "loss": 0.4002,
+      "step": 845
+    },
+    {
+      "epoch": 4.1067151735637415,
+      "grad_norm": 0.09835145173574618,
+      "learning_rate": 7.52813301084375e-06,
+      "loss": 0.3961,
+      "step": 846
+    },
+    {
+      "epoch": 4.111565863271184,
+      "grad_norm": 0.09426867245890704,
+      "learning_rate": 7.449160962857358e-06,
+      "loss": 0.3946,
+      "step": 847
+    },
+    {
+      "epoch": 4.1164165529786265,
+      "grad_norm": 0.0993123791236732,
+      "learning_rate": 7.370562768885823e-06,
+      "loss": 0.3937,
+      "step": 848
+    },
+    {
+      "epoch": 4.121267242686069,
+      "grad_norm": 0.09692548506896993,
+      "learning_rate": 7.292339331647848e-06,
+      "loss": 0.3957,
+      "step": 849
+    },
+    {
+      "epoch": 4.126117932393512,
+      "grad_norm": 0.0977639150148905,
+      "learning_rate": 7.214491549557898e-06,
+      "loss": 0.3969,
+      "step": 850
+    },
+    {
+      "epoch": 4.130968622100955,
+      "grad_norm": 0.10100358672328608,
+      "learning_rate": 7.1370203167160326e-06,
+      "loss": 0.3937,
+      "step": 851
+    },
+    {
+      "epoch": 4.1358193118083975,
+      "grad_norm": 0.0961508995875077,
+      "learning_rate": 7.0599265228975e-06,
+      "loss": 0.3965,
+      "step": 852
+    },
+    {
+      "epoch": 4.140670001515841,
+      "grad_norm": 0.09450992963252156,
+      "learning_rate": 6.983211053542591e-06,
+      "loss": 0.4008,
+      "step": 853
+    },
+    {
+      "epoch": 4.145520691223283,
+      "grad_norm": 0.10411768031595499,
+      "learning_rate": 6.9068747897464535e-06,
+      "loss": 0.4032,
+      "step": 854
+    },
+    {
+      "epoch": 4.150371380930726,
+      "grad_norm": 0.09981896937364662,
+      "learning_rate": 6.830918608248964e-06,
+      "loss": 0.4002,
+      "step": 855
+    },
+    {
+      "epoch": 4.155222070638169,
+      "grad_norm": 0.08754517232013051,
+      "learning_rate": 6.755343381424659e-06,
+      "loss": 0.3976,
+      "step": 856
+    },
+    {
+      "epoch": 4.160072760345612,
+      "grad_norm": 0.09520671649907132,
+      "learning_rate": 6.68014997727275e-06,
+      "loss": 0.3921,
+      "step": 857
+    },
+    {
+      "epoch": 4.164923450053054,
+      "grad_norm": 0.09719121632502135,
+      "learning_rate": 6.605339259407104e-06,
+      "loss": 0.3852,
+      "step": 858
+    },
+    {
+      "epoch": 4.169774139760497,
+      "grad_norm": 0.08870928687089744,
+      "learning_rate": 6.530912087046317e-06,
+      "loss": 0.395,
+      "step": 859
+    },
+    {
+      "epoch": 4.17462482946794,
+      "grad_norm": 0.08685694037296439,
+      "learning_rate": 6.456869315003946e-06,
+      "loss": 0.3941,
+      "step": 860
+    },
+    {
+      "epoch": 4.179475519175383,
+      "grad_norm": 0.08904997313439429,
+      "learning_rate": 6.3832117936785564e-06,
+      "loss": 0.3997,
+      "step": 861
+    },
+    {
+      "epoch": 4.184326208882825,
+      "grad_norm": 0.0910482544729166,
+      "learning_rate": 6.309940369044047e-06,
+      "loss": 0.3945,
+      "step": 862
+    },
+    {
+      "epoch": 4.189176898590269,
+      "grad_norm": 0.0887134263309384,
+      "learning_rate": 6.23705588263992e-06,
+      "loss": 0.3881,
+      "step": 863
+    },
+    {
+      "epoch": 4.194027588297711,
+      "grad_norm": 0.08977043998387933,
+      "learning_rate": 6.164559171561553e-06,
+      "loss": 0.3957,
+      "step": 864
+    },
+    {
+      "epoch": 4.198878278005154,
+      "grad_norm": 0.09216637353138708,
+      "learning_rate": 6.092451068450671e-06,
+      "loss": 0.3969,
+      "step": 865
+    },
+    {
+      "epoch": 4.203728967712596,
+      "grad_norm": 0.08464526503746563,
+      "learning_rate": 6.020732401485751e-06,
+      "loss": 0.3896,
+      "step": 866
+    },
+    {
+      "epoch": 4.20857965742004,
+      "grad_norm": 0.08983668958241023,
+      "learning_rate": 5.9494039943724845e-06,
+      "loss": 0.4025,
+      "step": 867
+    },
+    {
+      "epoch": 4.213430347127482,
+      "grad_norm": 0.08753368121579265,
+      "learning_rate": 5.878466666334341e-06,
+      "loss": 0.3954,
+      "step": 868
+    },
+    {
+      "epoch": 4.218281036834925,
+      "grad_norm": 0.08687937990016806,
+      "learning_rate": 5.80792123210316e-06,
+      "loss": 0.391,
+      "step": 869
+    },
+    {
+      "epoch": 4.223131726542368,
+      "grad_norm": 0.08701962255264487,
+      "learning_rate": 5.737768501909773e-06,
+      "loss": 0.3965,
+      "step": 870
+    },
+    {
+      "epoch": 4.2279824162498105,
+      "grad_norm": 0.08811499746627446,
+      "learning_rate": 5.668009281474751e-06,
+      "loss": 0.3846,
+      "step": 871
+    },
+    {
+      "epoch": 4.232833105957253,
+      "grad_norm": 0.08561633996061273,
+      "learning_rate": 5.598644371999085e-06,
+      "loss": 0.3919,
+      "step": 872
+    },
+    {
+      "epoch": 4.237683795664696,
+      "grad_norm": 0.08321419945310787,
+      "learning_rate": 5.5296745701549906e-06,
+      "loss": 0.394,
+      "step": 873
+    },
+    {
+      "epoch": 4.242534485372139,
+      "grad_norm": 0.08724643182123827,
+      "learning_rate": 5.4611006680768305e-06,
+      "loss": 0.3994,
+      "step": 874
+    },
+    {
+      "epoch": 4.2473851750795815,
+      "grad_norm": 0.08361301460174729,
+      "learning_rate": 5.3929234533519345e-06,
+      "loss": 0.388,
+      "step": 875
+    },
+    {
+      "epoch": 4.252235864787025,
+      "grad_norm": 0.0862576560444418,
+      "learning_rate": 5.325143709011587e-06,
+      "loss": 0.3942,
+      "step": 876
+    },
+    {
+      "epoch": 4.257086554494467,
+      "grad_norm": 0.082606577377262,
+      "learning_rate": 5.257762213522055e-06,
+      "loss": 0.3907,
+      "step": 877
+    },
+    {
+      "epoch": 4.26193724420191,
+      "grad_norm": 0.08601520095290771,
+      "learning_rate": 5.19077974077558e-06,
+      "loss": 0.3927,
+      "step": 878
+    },
+    {
+      "epoch": 4.266787933909352,
+      "grad_norm": 0.08144847498379731,
+      "learning_rate": 5.124197060081564e-06,
+      "loss": 0.3981,
+      "step": 879
+    },
+    {
+      "epoch": 4.271638623616796,
+      "grad_norm": 0.08369648223405692,
+      "learning_rate": 5.058014936157714e-06,
+      "loss": 0.3987,
+      "step": 880
+    },
+    {
+      "epoch": 4.276489313324238,
+      "grad_norm": 0.08487450387282452,
+      "learning_rate": 4.992234129121225e-06,
+      "loss": 0.3993,
+      "step": 881
+    },
+    {
+      "epoch": 4.281340003031681,
+      "grad_norm": 0.0835928002976722,
+      "learning_rate": 4.926855394480079e-06,
+      "loss": 0.394,
+      "step": 882
+    },
+    {
+      "epoch": 4.286190692739124,
+      "grad_norm": 0.0877663910781713,
+      "learning_rate": 4.861879483124372e-06,
+      "loss": 0.3888,
+      "step": 883
+    },
+    {
+      "epoch": 4.291041382446567,
+      "grad_norm": 0.08201902566316537,
+      "learning_rate": 4.797307141317666e-06,
+      "loss": 0.3978,
+      "step": 884
+    },
+    {
+      "epoch": 4.295892072154009,
+      "grad_norm": 0.08252391090570044,
+      "learning_rate": 4.7331391106884364e-06,
+      "loss": 0.3949,
+      "step": 885
+    },
+    {
+      "epoch": 4.300742761861452,
+      "grad_norm": 0.08121494056766716,
+      "learning_rate": 4.6693761282215766e-06,
+      "loss": 0.3922,
+      "step": 886
+    },
+    {
+      "epoch": 4.305593451568895,
+      "grad_norm": 0.08414265476669228,
+      "learning_rate": 4.606018926249851e-06,
+      "loss": 0.389,
+      "step": 887
+    },
+    {
+      "epoch": 4.310444141276338,
+      "grad_norm": 0.08528892500629966,
+      "learning_rate": 4.543068232445596e-06,
+      "loss": 0.3956,
+      "step": 888
+    },
+    {
+      "epoch": 4.31529483098378,
+      "grad_norm": 0.08280199210655267,
+      "learning_rate": 4.480524769812276e-06,
+      "loss": 0.3938,
+      "step": 889
+    },
+    {
+      "epoch": 4.320145520691224,
+      "grad_norm": 0.08884076809121294,
+      "learning_rate": 4.418389256676206e-06,
+      "loss": 0.3947,
+      "step": 890
+    },
+    {
+      "epoch": 4.324996210398666,
+      "grad_norm": 0.08902762031211295,
+      "learning_rate": 4.35666240667834e-06,
+      "loss": 0.3907,
+      "step": 891
+    },
+    {
+      "epoch": 4.329846900106109,
+      "grad_norm": 0.08464410835355753,
+      "learning_rate": 4.295344928765999e-06,
+      "loss": 0.3939,
+      "step": 892
+    },
+    {
+      "epoch": 4.334697589813552,
+      "grad_norm": 0.08347421805827179,
+      "learning_rate": 4.234437527184785e-06,
+      "loss": 0.3985,
+      "step": 893
+    },
+    {
+      "epoch": 4.3395482795209945,
+      "grad_norm": 0.08770395898608876,
+      "learning_rate": 4.173940901470488e-06,
+      "loss": 0.395,
+      "step": 894
+    },
+    {
+      "epoch": 4.344398969228437,
+      "grad_norm": 0.08139356257791996,
+      "learning_rate": 4.11385574644104e-06,
+      "loss": 0.39,
+      "step": 895
+    },
+    {
+      "epoch": 4.34924965893588,
+      "grad_norm": 0.0818047074224124,
+      "learning_rate": 4.054182752188501e-06,
+      "loss": 0.3948,
+      "step": 896
+    },
+    {
+      "epoch": 4.354100348643323,
+      "grad_norm": 0.08370905382784635,
+      "learning_rate": 3.994922604071217e-06,
+      "loss": 0.3941,
+      "step": 897
+    },
+    {
+      "epoch": 4.3589510383507655,
+      "grad_norm": 0.08543140680473625,
+      "learning_rate": 3.936075982705871e-06,
+      "loss": 0.3982,
+      "step": 898
+    },
+    {
+      "epoch": 4.363801728058208,
+      "grad_norm": 0.08330546263387466,
+      "learning_rate": 3.877643563959694e-06,
+      "loss": 0.3986,
+      "step": 899
+    },
+    {
+      "epoch": 4.368652417765651,
+      "grad_norm": 0.08181359063571672,
+      "learning_rate": 3.819626018942732e-06,
+      "loss": 0.3962,
+      "step": 900
+    },
+    {
+      "epoch": 4.373503107473094,
+      "grad_norm": 0.08458157848919085,
+      "learning_rate": 3.762024014000054e-06,
+      "loss": 0.3974,
+      "step": 901
+    },
+    {
+      "epoch": 4.378353797180536,
+      "grad_norm": 0.0797245693813596,
+      "learning_rate": 3.7048382107042113e-06,
+      "loss": 0.3849,
+      "step": 902
+    },
+    {
+      "epoch": 4.38320448688798,
+      "grad_norm": 0.08552325047075819,
+      "learning_rate": 3.6480692658475446e-06,
+      "loss": 0.3908,
+      "step": 903
+    },
+    {
+      "epoch": 4.388055176595422,
+      "grad_norm": 0.08191614537028945,
+      "learning_rate": 3.5917178314346955e-06,
+      "loss": 0.398,
+      "step": 904
+    },
+    {
+      "epoch": 4.392905866302865,
+      "grad_norm": 0.08324472646577967,
+      "learning_rate": 3.535784554675088e-06,
+      "loss": 0.3941,
+      "step": 905
+    },
+    {
+      "epoch": 4.397756556010307,
+      "grad_norm": 0.08386476178169076,
+      "learning_rate": 3.480270077975525e-06,
+      "loss": 0.395,
+      "step": 906
+    },
+    {
+      "epoch": 4.402607245717751,
+      "grad_norm": 0.08759720428686872,
+      "learning_rate": 3.42517503893276e-06,
+      "loss": 0.3879,
+      "step": 907
+    },
+    {
+      "epoch": 4.407457935425193,
+      "grad_norm": 0.0834695533465509,
+      "learning_rate": 3.370500070326257e-06,
+      "loss": 0.3832,
+      "step": 908
+    },
+    {
+      "epoch": 4.412308625132636,
+      "grad_norm": 0.080547419306159,
+      "learning_rate": 3.3162458001108332e-06,
+      "loss": 0.3858,
+      "step": 909
+    },
+    {
+      "epoch": 4.417159314840079,
+      "grad_norm": 0.08406690651169581,
+      "learning_rate": 3.2624128514094778e-06,
+      "loss": 0.3923,
+      "step": 910
+    },
+    {
+      "epoch": 4.422010004547522,
+      "grad_norm": 0.08417275645743631,
+      "learning_rate": 3.20900184250625e-06,
+      "loss": 0.3933,
+      "step": 911
+    },
+    {
+      "epoch": 4.426860694254964,
+      "grad_norm": 0.08104238775596306,
+      "learning_rate": 3.1560133868390895e-06,
+      "loss": 0.4023,
+      "step": 912
+    },
+    {
+      "epoch": 4.431711383962407,
+      "grad_norm": 0.08376361425082632,
+      "learning_rate": 3.1034480929928333e-06,
+      "loss": 0.399,
+      "step": 913
+    },
+    {
+      "epoch": 4.43656207366985,
+      "grad_norm": 0.08058350379685782,
+      "learning_rate": 3.0513065646921957e-06,
+      "loss": 0.3946,
+      "step": 914
+    },
+    {
+      "epoch": 4.441412763377293,
+      "grad_norm": 0.08013760556674378,
+      "learning_rate": 2.999589400794851e-06,
+      "loss": 0.392,
+      "step": 915
+    },
+    {
+      "epoch": 4.446263453084735,
+      "grad_norm": 0.08082216615145961,
+      "learning_rate": 2.948297195284546e-06,
+      "loss": 0.3916,
+      "step": 916
+    },
+    {
+      "epoch": 4.4511141427921785,
+      "grad_norm": 0.09007717527915819,
+      "learning_rate": 2.897430537264283e-06,
+      "loss": 0.3947,
+      "step": 917
+    },
+    {
+      "epoch": 4.455964832499621,
+      "grad_norm": 0.0866390993026153,
+      "learning_rate": 2.8469900109495553e-06,
+      "loss": 0.3942,
+      "step": 918
+    },
+    {
+      "epoch": 4.460815522207064,
+      "grad_norm": 0.08034663815912857,
+      "learning_rate": 2.79697619566162e-06,
+      "loss": 0.3888,
+      "step": 919
+    },
+    {
+      "epoch": 4.465666211914507,
+      "grad_norm": 0.08423465438674188,
+      "learning_rate": 2.7473896658208743e-06,
+      "loss": 0.391,
+      "step": 920
+    },
+    {
+      "epoch": 4.4705169016219495,
+      "grad_norm": 0.08769412223500794,
+      "learning_rate": 2.6982309909402293e-06,
+      "loss": 0.3936,
+      "step": 921
+    },
+    {
+      "epoch": 4.475367591329392,
+      "grad_norm": 0.0794092004599977,
+      "learning_rate": 2.649500735618582e-06,
+      "loss": 0.3993,
+      "step": 922
+    },
+    {
+      "epoch": 4.480218281036835,
+      "grad_norm": 0.08321318361659834,
+      "learning_rate": 2.6011994595343516e-06,
+      "loss": 0.3965,
+      "step": 923
+    },
+    {
+      "epoch": 4.485068970744278,
+      "grad_norm": 0.08400144661884966,
+      "learning_rate": 2.5533277174389916e-06,
+      "loss": 0.3927,
+      "step": 924
+    },
+    {
+      "epoch": 4.48991966045172,
+      "grad_norm": 0.08039773750467258,
+      "learning_rate": 2.5058860591506973e-06,
+      "loss": 0.3927,
+      "step": 925
+    },
+    {
+      "epoch": 4.494770350159163,
+      "grad_norm": 0.08132175209524008,
+      "learning_rate": 2.4588750295480246e-06,
+      "loss": 0.3888,
+      "step": 926
+    },
+    {
+      "epoch": 4.499621039866606,
+      "grad_norm": 0.07920562392368859,
+      "learning_rate": 2.4122951685636674e-06,
+      "loss": 0.3896,
+      "step": 927
+    },
+    {
+      "epoch": 4.504471729574049,
+      "grad_norm": 0.07833571502263627,
+      "learning_rate": 2.366147011178246e-06,
+      "loss": 0.398,
+      "step": 928
+    },
+    {
+      "epoch": 4.509322419281491,
+      "grad_norm": 0.08232920047142565,
+      "learning_rate": 2.320431087414159e-06,
+      "loss": 0.3838,
+      "step": 929
+    },
+    {
+      "epoch": 4.514173108988935,
+      "grad_norm": 0.08136833742803433,
+      "learning_rate": 2.275147922329506e-06,
+      "loss": 0.3935,
+      "step": 930
+    },
+    {
+      "epoch": 4.519023798696377,
+      "grad_norm": 0.08383396678636225,
+      "learning_rate": 2.230298036012055e-06,
+      "loss": 0.3913,
+      "step": 931
+    },
+    {
+      "epoch": 4.52387448840382,
+      "grad_norm": 0.08000391439211,
+      "learning_rate": 2.1858819435732583e-06,
+      "loss": 0.395,
+      "step": 932
+    },
+    {
+      "epoch": 4.528725178111262,
+      "grad_norm": 0.080147749223563,
+      "learning_rate": 2.141900155142351e-06,
+      "loss": 0.3951,
+      "step": 933
+    },
+    {
+      "epoch": 4.533575867818706,
+      "grad_norm": 0.07978655829531454,
+      "learning_rate": 2.0983531758604726e-06,
+      "loss": 0.4011,
+      "step": 934
+    },
+    {
+      "epoch": 4.538426557526148,
+      "grad_norm": 0.08386421814730027,
+      "learning_rate": 2.055241505874892e-06,
+      "loss": 0.3999,
+      "step": 935
+    },
+    {
+      "epoch": 4.543277247233591,
+      "grad_norm": 0.08174904946083562,
+      "learning_rate": 2.0125656403332396e-06,
+      "loss": 0.3968,
+      "step": 936
+    },
+    {
+      "epoch": 4.548127936941034,
+      "grad_norm": 0.07927025601942647,
+      "learning_rate": 1.970326069377828e-06,
+      "loss": 0.4001,
+      "step": 937
+    },
+    {
+      "epoch": 4.552978626648477,
+      "grad_norm": 0.0783638905343862,
+      "learning_rate": 1.928523278140033e-06,
+      "loss": 0.3943,
+      "step": 938
+    },
+    {
+      "epoch": 4.557829316355919,
+      "grad_norm": 0.08012821453349714,
+      "learning_rate": 1.887157746734718e-06,
+      "loss": 0.3972,
+      "step": 939
+    },
+    {
+      "epoch": 4.562680006063362,
+      "grad_norm": 0.08022389886437123,
+      "learning_rate": 1.846229950254692e-06,
+      "loss": 0.3888,
+      "step": 940
+    },
+    {
+      "epoch": 4.567530695770805,
+      "grad_norm": 0.07938938767749412,
+      "learning_rate": 1.8057403587652977e-06,
+      "loss": 0.3892,
+      "step": 941
+    },
+    {
+      "epoch": 4.572381385478248,
+      "grad_norm": 0.07979264917062728,
+      "learning_rate": 1.7656894372989785e-06,
+      "loss": 0.3991,
+      "step": 942
+    },
+    {
+      "epoch": 4.57723207518569,
+      "grad_norm": 0.07900416574598519,
+      "learning_rate": 1.726077645849955e-06,
+      "loss": 0.4033,
+      "step": 943
+    },
+    {
+      "epoch": 4.5820827648931335,
+      "grad_norm": 0.0816244017270053,
+      "learning_rate": 1.6869054393689265e-06,
+      "loss": 0.394,
+      "step": 944
+    },
+    {
+      "epoch": 4.586933454600576,
+      "grad_norm": 0.07678354405433441,
+      "learning_rate": 1.6481732677578798e-06,
+      "loss": 0.4026,
+      "step": 945
+    },
+    {
+      "epoch": 4.5917841443080185,
+      "grad_norm": 0.07913436605637802,
+      "learning_rate": 1.60988157586488e-06,
+      "loss": 0.378,
+      "step": 946
+    },
+    {
+      "epoch": 4.596634834015462,
+      "grad_norm": 0.0793297266554538,
+      "learning_rate": 1.5720308034789721e-06,
+      "loss": 0.391,
+      "step": 947
+    },
+    {
+      "epoch": 4.601485523722904,
+      "grad_norm": 0.07767980761029898,
+      "learning_rate": 1.5346213853251546e-06,
+      "loss": 0.3978,
+      "step": 948
+    },
+    {
+      "epoch": 4.606336213430347,
+      "grad_norm": 0.08494507410525762,
+      "learning_rate": 1.4976537510593646e-06,
+      "loss": 0.3995,
+      "step": 949
+    },
+    {
+      "epoch": 4.61118690313779,
+      "grad_norm": 0.08237577571806884,
+      "learning_rate": 1.4611283252635412e-06,
+      "loss": 0.4038,
+      "step": 950
+    },
+    {
+      "epoch": 4.616037592845233,
+      "grad_norm": 0.07590194142960284,
+      "learning_rate": 1.425045527440756e-06,
+      "loss": 0.3956,
+      "step": 951
+    },
+    {
+      "epoch": 4.620888282552675,
+      "grad_norm": 0.07954982455169565,
+      "learning_rate": 1.3894057720104104e-06,
+      "loss": 0.399,
+      "step": 952
+    },
+    {
+      "epoch": 4.625738972260118,
+      "grad_norm": 0.07731849776688487,
+      "learning_rate": 1.354209468303429e-06,
+      "loss": 0.3828,
+      "step": 953
+    },
+    {
+      "epoch": 4.630589661967561,
+      "grad_norm": 0.07970299041030604,
+      "learning_rate": 1.3194570205576284e-06,
+      "loss": 0.3954,
+      "step": 954
+    },
+    {
+      "epoch": 4.635440351675004,
+      "grad_norm": 0.07874124868840192,
+      "learning_rate": 1.2851488279130053e-06,
+      "loss": 0.3876,
+      "step": 955
+    },
+    {
+      "epoch": 4.640291041382446,
+      "grad_norm": 0.07691612719760402,
+      "learning_rate": 1.2512852844071933e-06,
+      "loss": 0.3949,
+      "step": 956
+    },
+    {
+      "epoch": 4.64514173108989,
+      "grad_norm": 0.08126016720695953,
+      "learning_rate": 1.2178667789709287e-06,
+      "loss": 0.3919,
+      "step": 957
+    },
+    {
+      "epoch": 4.649992420797332,
+      "grad_norm": 0.08002643782323322,
+      "learning_rate": 1.1848936954235702e-06,
+      "loss": 0.395,
+      "step": 958
+    },
+    {
+      "epoch": 4.654843110504775,
+      "grad_norm": 0.08026194265591635,
+      "learning_rate": 1.1523664124687284e-06,
+      "loss": 0.3997,
+      "step": 959
+    },
+    {
+      "epoch": 4.659693800212217,
+      "grad_norm": 0.07660352848559206,
+      "learning_rate": 1.1202853036898476e-06,
+      "loss": 0.3974,
+      "step": 960
+    },
+    {
+      "epoch": 4.664544489919661,
+      "grad_norm": 0.08113236962163348,
+      "learning_rate": 1.0886507375459908e-06,
+      "loss": 0.3981,
+      "step": 961
+    },
+    {
+      "epoch": 4.669395179627103,
+      "grad_norm": 0.07701309756204706,
+      "learning_rate": 1.0574630773675687e-06,
+      "loss": 0.3839,
+      "step": 962
+    },
+    {
+      "epoch": 4.674245869334546,
+      "grad_norm": 0.07891978078549244,
+      "learning_rate": 1.0267226813521635e-06,
+      "loss": 0.3877,
+      "step": 963
+    },
+    {
+      "epoch": 4.679096559041989,
+      "grad_norm": 0.07852321868608765,
+      "learning_rate": 9.964299025604274e-07,
+      "loss": 0.3921,
+      "step": 964
+    },
+    {
+      "epoch": 4.683947248749432,
+      "grad_norm": 0.07818107376466682,
+      "learning_rate": 9.66585088912022e-07,
+      "loss": 0.3967,
+      "step": 965
+    },
+    {
+      "epoch": 4.688797938456874,
+      "grad_norm": 0.076600770628096,
+      "learning_rate": 9.371885831816319e-07,
+      "loss": 0.395,
+      "step": 966
+    },
+    {
+      "epoch": 4.6936486281643175,
+      "grad_norm": 0.07627250485843899,
+      "learning_rate": 9.082407229950018e-07,
+      "loss": 0.3976,
+      "step": 967
+    },
+    {
+      "epoch": 4.69849931787176,
+      "grad_norm": 0.07905168555654667,
+      "learning_rate": 8.797418408251101e-07,
+      "loss": 0.3918,
+      "step": 968
+    },
+    {
+      "epoch": 4.7033500075792025,
+      "grad_norm": 0.07625038396328788,
+      "learning_rate": 8.516922639882819e-07,
+      "loss": 0.3897,
+      "step": 969
+    },
+    {
+      "epoch": 4.708200697286646,
+      "grad_norm": 0.08055085830124646,
+      "learning_rate": 8.2409231464049e-07,
+      "loss": 0.3901,
+      "step": 970
+    },
+    {
+      "epoch": 4.713051386994088,
+      "grad_norm": 0.07599467665769233,
+      "learning_rate": 7.969423097736162e-07,
+      "loss": 0.3931,
+      "step": 971
+    },
+    {
+      "epoch": 4.717902076701531,
+      "grad_norm": 0.08049982038815076,
+      "learning_rate": 7.702425612118269e-07,
+      "loss": 0.3962,
+      "step": 972
+    },
+    {
+      "epoch": 4.7227527664089735,
+      "grad_norm": 0.07817439712988589,
+      "learning_rate": 7.439933756079942e-07,
+      "loss": 0.3913,
+      "step": 973
+    },
+    {
+      "epoch": 4.727603456116417,
+      "grad_norm": 0.08001209525722262,
+      "learning_rate": 7.181950544401695e-07,
+      "loss": 0.3935,
+      "step": 974
+    },
+    {
+      "epoch": 4.732454145823859,
+      "grad_norm": 0.07544329449164154,
+      "learning_rate": 6.928478940081107e-07,
+      "loss": 0.3984,
+      "step": 975
+    },
+    {
+      "epoch": 4.737304835531302,
+      "grad_norm": 0.07792056514995602,
+      "learning_rate": 6.679521854299032e-07,
+      "loss": 0.3914,
+      "step": 976
+    },
+    {
+      "epoch": 4.742155525238745,
+      "grad_norm": 0.07777159770613211,
+      "learning_rate": 6.435082146385885e-07,
+      "loss": 0.3901,
+      "step": 977
+    },
+    {
+      "epoch": 4.747006214946188,
+      "grad_norm": 0.07684966968716175,
+      "learning_rate": 6.195162623789052e-07,
+      "loss": 0.393,
+      "step": 978
+    },
+    {
+      "epoch": 4.75185690465363,
+      "grad_norm": 0.07663626226452194,
+      "learning_rate": 5.959766042040426e-07,
+      "loss": 0.3879,
+      "step": 979
+    },
+    {
+      "epoch": 4.756707594361073,
+      "grad_norm": 0.07653692258850123,
+      "learning_rate": 5.728895104724963e-07,
+      "loss": 0.3959,
+      "step": 980
+    },
+    {
+      "epoch": 4.761558284068516,
+      "grad_norm": 0.07731590848782262,
+      "learning_rate": 5.502552463449418e-07,
+      "loss": 0.3984,
+      "step": 981
+    },
+    {
+      "epoch": 4.766408973775959,
+      "grad_norm": 0.07715742201671594,
+      "learning_rate": 5.280740717812149e-07,
+      "loss": 0.3977,
+      "step": 982
+    },
+    {
+      "epoch": 4.771259663483401,
+      "grad_norm": 0.07733908081214975,
+      "learning_rate": 5.063462415372967e-07,
+      "loss": 0.3948,
+      "step": 983
+    },
+    {
+      "epoch": 4.776110353190845,
+      "grad_norm": 0.07797117300936077,
+      "learning_rate": 4.850720051624124e-07,
+      "loss": 0.3914,
+      "step": 984
+    },
+    {
+      "epoch": 4.780961042898287,
+      "grad_norm": 0.07721734718928677,
+      "learning_rate": 4.642516069961556e-07,
+      "loss": 0.392,
+      "step": 985
+    },
+    {
+      "epoch": 4.78581173260573,
+      "grad_norm": 0.07648551851063208,
+      "learning_rate": 4.438852861656751e-07,
+      "loss": 0.3951,
+      "step": 986
+    },
+    {
+      "epoch": 4.790662422313172,
+      "grad_norm": 0.07521119037010907,
+      "learning_rate": 4.2397327658294076e-07,
+      "loss": 0.3899,
+      "step": 987
+    },
+    {
+      "epoch": 4.795513112020616,
+      "grad_norm": 0.07655879860272995,
+      "learning_rate": 4.045158069420474e-07,
+      "loss": 0.3963,
+      "step": 988
+    },
+    {
+      "epoch": 4.800363801728058,
+      "grad_norm": 0.07622760847546149,
+      "learning_rate": 3.8551310071659023e-07,
+      "loss": 0.3975,
+      "step": 989
+    },
+    {
+      "epoch": 4.805214491435501,
+      "grad_norm": 0.0750428400819878,
+      "learning_rate": 3.6696537615711124e-07,
+      "loss": 0.3968,
+      "step": 990
+    },
+    {
+      "epoch": 4.810065181142944,
+      "grad_norm": 0.07852243689578567,
+      "learning_rate": 3.4887284628857266e-07,
+      "loss": 0.3932,
+      "step": 991
+    },
+    {
+      "epoch": 4.8149158708503865,
+      "grad_norm": 0.0771348195411035,
+      "learning_rate": 3.3123571890791405e-07,
+      "loss": 0.3887,
+      "step": 992
+    },
+    {
+      "epoch": 4.819766560557829,
+      "grad_norm": 0.07614331045752255,
+      "learning_rate": 3.1405419658168125e-07,
+      "loss": 0.394,
+      "step": 993
+    },
+    {
+      "epoch": 4.824617250265272,
+      "grad_norm": 0.07563944264490313,
+      "learning_rate": 2.973284766436857e-07,
+      "loss": 0.3917,
+      "step": 994
+    },
+    {
+      "epoch": 4.829467939972715,
+      "grad_norm": 0.07503164664083585,
+      "learning_rate": 2.810587511927354e-07,
+      "loss": 0.3901,
+      "step": 995
+    },
+    {
+      "epoch": 4.8343186296801575,
+      "grad_norm": 0.07663595734772072,
+      "learning_rate": 2.652452070904499e-07,
+      "loss": 0.3923,
+      "step": 996
+    },
+    {
+      "epoch": 4.839169319387601,
+      "grad_norm": 0.07606961706547127,
+      "learning_rate": 2.498880259590797e-07,
+      "loss": 0.3944,
+      "step": 997
+    },
+    {
+      "epoch": 4.844020009095043,
+      "grad_norm": 0.0761182148197196,
+      "learning_rate": 2.3498738417945034e-07,
+      "loss": 0.3975,
+      "step": 998
+    },
+    {
+      "epoch": 4.848870698802486,
+      "grad_norm": 0.08039104985634041,
+      "learning_rate": 2.205434528889283e-07,
+      "loss": 0.3971,
+      "step": 999
+    },
+    {
+      "epoch": 4.853721388509928,
+      "grad_norm": 0.07775606427920397,
+      "learning_rate": 2.0655639797944937e-07,
+      "loss": 0.3903,
+      "step": 1000
+    },
+    {
+      "epoch": 4.858572078217372,
+      "grad_norm": 0.07610940999590161,
+      "learning_rate": 1.9302638009561782e-07,
+      "loss": 0.396,
+      "step": 1001
+    },
+    {
+      "epoch": 4.863422767924814,
+      "grad_norm": 0.07569450912275462,
+      "learning_rate": 1.7995355463285457e-07,
+      "loss": 0.3965,
+      "step": 1002
+    },
+    {
+      "epoch": 4.868273457632257,
+      "grad_norm": 0.07643720956773195,
+      "learning_rate": 1.6733807173562988e-07,
+      "loss": 0.3913,
+      "step": 1003
+    },
+    {
+      "epoch": 4.8731241473397,
+      "grad_norm": 0.07571887859535044,
+      "learning_rate": 1.5518007629571342e-07,
+      "loss": 0.3995,
+      "step": 1004
+    },
+    {
+      "epoch": 4.877974837047143,
+      "grad_norm": 0.07657090064859913,
+      "learning_rate": 1.4347970795054456e-07,
+      "loss": 0.3967,
+      "step": 1005
+    },
+    {
+      "epoch": 4.882825526754585,
+      "grad_norm": 0.07475277823538463,
+      "learning_rate": 1.3223710108158483e-07,
+      "loss": 0.3965,
+      "step": 1006
+    },
+    {
+      "epoch": 4.887676216462028,
+      "grad_norm": 0.07465351831333361,
+      "learning_rate": 1.214523848128124e-07,
+      "loss": 0.3971,
+      "step": 1007
+    },
+    {
+      "epoch": 4.892526906169471,
+      "grad_norm": 0.07637388169909817,
+      "learning_rate": 1.111256830092211e-07,
+      "loss": 0.3996,
+      "step": 1008
+    },
+    {
+      "epoch": 4.897377595876914,
+      "grad_norm": 0.07639665552172381,
+      "learning_rate": 1.0125711427540374e-07,
+      "loss": 0.3949,
+      "step": 1009
+    },
+    {
+      "epoch": 4.902228285584356,
+      "grad_norm": 0.07373092012101537,
+      "learning_rate": 9.184679195417989e-08,
+      "loss": 0.3889,
+      "step": 1010
+    },
+    {
+      "epoch": 4.9070789752918,
+      "grad_norm": 0.07583698163385665,
+      "learning_rate": 8.289482412531246e-08,
+      "loss": 0.3984,
+      "step": 1011
+    },
+    {
+      "epoch": 4.911929664999242,
+      "grad_norm": 0.07647787343235872,
+      "learning_rate": 7.440131360424652e-08,
+      "loss": 0.3887,
+      "step": 1012
+    },
+    {
+      "epoch": 4.916780354706685,
+      "grad_norm": 0.07751142464425709,
+      "learning_rate": 6.636635794094126e-08,
+      "loss": 0.3908,
+      "step": 1013
+    },
+    {
+      "epoch": 4.921631044414128,
+      "grad_norm": 0.07884353160777417,
+      "learning_rate": 5.879004941874655e-08,
+      "loss": 0.3951,
+      "step": 1014
+    },
+    {
+      "epoch": 4.9264817341215705,
+      "grad_norm": 0.07661585400143268,
+      "learning_rate": 5.16724750533415e-08,
+      "loss": 0.398,
+      "step": 1015
+    },
+    {
+      "epoch": 4.931332423829013,
+      "grad_norm": 0.07506936516416932,
+      "learning_rate": 4.5013716591730815e-08,
+      "loss": 0.3944,
+      "step": 1016
+    },
+    {
+      "epoch": 4.9361831135364564,
+      "grad_norm": 0.07625120541429038,
+      "learning_rate": 3.881385051132114e-08,
+      "loss": 0.3997,
+      "step": 1017
+    },
+    {
+      "epoch": 4.941033803243899,
+      "grad_norm": 0.0754548388472603,
+      "learning_rate": 3.307294801902838e-08,
+      "loss": 0.3968,
+      "step": 1018
+    },
+    {
+      "epoch": 4.9458844929513415,
+      "grad_norm": 0.07563871272528049,
+      "learning_rate": 2.7791075050460636e-08,
+      "loss": 0.3977,
+      "step": 1019
+    },
+    {
+      "epoch": 4.950735182658784,
+      "grad_norm": 0.0760975342061979,
+      "learning_rate": 2.2968292269167637e-08,
+      "loss": 0.3978,
+      "step": 1020
+    },
+    {
+      "epoch": 4.955585872366227,
+      "grad_norm": 0.07484229156933196,
+      "learning_rate": 1.8604655065939116e-08,
+      "loss": 0.3888,
+      "step": 1021
+    },
+    {
+      "epoch": 4.96043656207367,
+      "grad_norm": 0.07489129192172529,
+      "learning_rate": 1.470021355816975e-08,
+      "loss": 0.3952,
+      "step": 1022
+    },
+    {
+      "epoch": 4.965287251781112,
+      "grad_norm": 0.07561234415729748,
+      "learning_rate": 1.1255012589286297e-08,
+      "loss": 0.3988,
+      "step": 1023
+    },
+    {
+      "epoch": 4.970137941488556,
+      "grad_norm": 0.07589712959511802,
+      "learning_rate": 8.269091728232426e-09,
+      "loss": 0.3858,
+      "step": 1024
+    },
+    {
+      "epoch": 4.974988631195998,
+      "grad_norm": 0.07684966203710471,
+      "learning_rate": 5.742485269006892e-09,
+      "loss": 0.3976,
+      "step": 1025
+    },
+    {
+      "epoch": 4.979839320903441,
+      "grad_norm": 0.07450707523151534,
+      "learning_rate": 3.6752222302727238e-09,
+      "loss": 0.3922,
+      "step": 1026
+    },
+    {
+      "epoch": 4.984690010610883,
+      "grad_norm": 0.0754634711281639,
+      "learning_rate": 2.06732635503748e-09,
+      "loss": 0.3894,
+      "step": 1027
+    },
+    {
+      "epoch": 4.989540700318327,
+      "grad_norm": 0.07460842756020492,
+      "learning_rate": 9.188161103557136e-10,
+      "loss": 0.3893,
+      "step": 1028
+    },
+    {
+      "epoch": 4.994391390025769,
+      "grad_norm": 0.07411367814575705,
+      "learning_rate": 2.2970468714245132e-10,
+      "loss": 0.3934,
+      "step": 1029
+    },
+    {
+      "epoch": 4.999242079733212,
+      "grad_norm": 0.07533391242414708,
+      "learning_rate": 0.0,
+      "loss": 0.3877,
+      "step": 1030
+    },
+    {
+      "epoch": 4.999242079733212,
+      "step": 1030,
+      "total_flos": 2.739131934768418e+19,
+      "train_loss": 0.07882811409755817,
+      "train_runtime": 48155.7887,
+      "train_samples_per_second": 10.958,
+      "train_steps_per_second": 0.021
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1030,
+  "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": 2.739131934768418e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}