diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17445 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2486,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004023335345001006,
+      "grad_norm": 27.82010551409744,
+      "learning_rate": 0.0,
+      "loss": 1.7886,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008046670690002012,
+      "grad_norm": 19.872529474831897,
+      "learning_rate": 2.008032128514056e-08,
+      "loss": 1.8136,
+      "step": 2
+    },
+    {
+      "epoch": 0.0012070006035003018,
+      "grad_norm": 24.180123303383308,
+      "learning_rate": 4.016064257028112e-08,
+      "loss": 1.9299,
+      "step": 3
+    },
+    {
+      "epoch": 0.0016093341380004024,
+      "grad_norm": 21.489391381684396,
+      "learning_rate": 6.02409638554217e-08,
+      "loss": 1.92,
+      "step": 4
+    },
+    {
+      "epoch": 0.0020116676725005027,
+      "grad_norm": 17.558745295130834,
+      "learning_rate": 8.032128514056224e-08,
+      "loss": 1.8227,
+      "step": 5
+    },
+    {
+      "epoch": 0.0024140012070006035,
+      "grad_norm": 21.46699607887353,
+      "learning_rate": 1.0040160642570281e-07,
+      "loss": 1.8824,
+      "step": 6
+    },
+    {
+      "epoch": 0.002816334741500704,
+      "grad_norm": 21.298475273313784,
+      "learning_rate": 1.204819277108434e-07,
+      "loss": 1.8179,
+      "step": 7
+    },
+    {
+      "epoch": 0.0032186682760008047,
+      "grad_norm": 21.372007205869878,
+      "learning_rate": 1.4056224899598394e-07,
+      "loss": 1.6859,
+      "step": 8
+    },
+    {
+      "epoch": 0.003621001810500905,
+      "grad_norm": 23.292357598058903,
+      "learning_rate": 1.6064257028112448e-07,
+      "loss": 2.0856,
+      "step": 9
+    },
+    {
+      "epoch": 0.0040233353450010055,
+      "grad_norm": 23.233045584039576,
+      "learning_rate": 1.8072289156626505e-07,
+      "loss": 2.1395,
+      "step": 10
+    },
+    {
+      "epoch": 0.004425668879501106,
+      "grad_norm": 19.4337570491022,
+      "learning_rate": 2.0080321285140563e-07,
+      "loss": 1.8688,
+      "step": 11
+    },
+    {
+      "epoch": 0.004828002414001207,
+      "grad_norm": 24.82725637763247,
+      "learning_rate": 2.208835341365462e-07,
+      "loss": 1.8054,
+      "step": 12
+    },
+    {
+      "epoch": 0.005230335948501308,
+      "grad_norm": 27.369177322875704,
+      "learning_rate": 2.409638554216868e-07,
+      "loss": 1.8794,
+      "step": 13
+    },
+    {
+      "epoch": 0.005632669483001408,
+      "grad_norm": 20.854134960964704,
+      "learning_rate": 2.610441767068273e-07,
+      "loss": 1.8873,
+      "step": 14
+    },
+    {
+      "epoch": 0.006035003017501509,
+      "grad_norm": 25.685971689153487,
+      "learning_rate": 2.811244979919679e-07,
+      "loss": 2.0743,
+      "step": 15
+    },
+    {
+      "epoch": 0.0064373365520016095,
+      "grad_norm": 21.501547984115398,
+      "learning_rate": 3.0120481927710845e-07,
+      "loss": 1.8415,
+      "step": 16
+    },
+    {
+      "epoch": 0.00683967008650171,
+      "grad_norm": 22.75991867094887,
+      "learning_rate": 3.2128514056224897e-07,
+      "loss": 1.9089,
+      "step": 17
+    },
+    {
+      "epoch": 0.00724200362100181,
+      "grad_norm": 22.23601729043514,
+      "learning_rate": 3.413654618473896e-07,
+      "loss": 1.9641,
+      "step": 18
+    },
+    {
+      "epoch": 0.007644337155501911,
+      "grad_norm": 26.35207563502954,
+      "learning_rate": 3.614457831325301e-07,
+      "loss": 2.0458,
+      "step": 19
+    },
+    {
+      "epoch": 0.008046670690002011,
+      "grad_norm": 20.57786492934256,
+      "learning_rate": 3.8152610441767073e-07,
+      "loss": 1.7877,
+      "step": 20
+    },
+    {
+      "epoch": 0.008449004224502113,
+      "grad_norm": 25.563694935501676,
+      "learning_rate": 4.0160642570281125e-07,
+      "loss": 2.1414,
+      "step": 21
+    },
+    {
+      "epoch": 0.008851337759002213,
+      "grad_norm": 18.98280746175437,
+      "learning_rate": 4.216867469879518e-07,
+      "loss": 1.7995,
+      "step": 22
+    },
+    {
+      "epoch": 0.009253671293502314,
+      "grad_norm": 22.278384224532818,
+      "learning_rate": 4.417670682730924e-07,
+      "loss": 2.0412,
+      "step": 23
+    },
+    {
+      "epoch": 0.009656004828002414,
+      "grad_norm": 18.256806226009903,
+      "learning_rate": 4.6184738955823296e-07,
+      "loss": 1.7538,
+      "step": 24
+    },
+    {
+      "epoch": 0.010058338362502514,
+      "grad_norm": 19.087838303092653,
+      "learning_rate": 4.819277108433736e-07,
+      "loss": 1.708,
+      "step": 25
+    },
+    {
+      "epoch": 0.010460671897002616,
+      "grad_norm": 18.961034656693933,
+      "learning_rate": 5.020080321285141e-07,
+      "loss": 1.6661,
+      "step": 26
+    },
+    {
+      "epoch": 0.010863005431502716,
+      "grad_norm": 18.11814523029599,
+      "learning_rate": 5.220883534136546e-07,
+      "loss": 1.753,
+      "step": 27
+    },
+    {
+      "epoch": 0.011265338966002816,
+      "grad_norm": 19.63691340845534,
+      "learning_rate": 5.421686746987952e-07,
+      "loss": 1.7005,
+      "step": 28
+    },
+    {
+      "epoch": 0.011667672500502917,
+      "grad_norm": 16.293048438980087,
+      "learning_rate": 5.622489959839358e-07,
+      "loss": 1.7269,
+      "step": 29
+    },
+    {
+      "epoch": 0.012070006035003017,
+      "grad_norm": 15.86827584036101,
+      "learning_rate": 5.823293172690764e-07,
+      "loss": 1.5401,
+      "step": 30
+    },
+    {
+      "epoch": 0.012472339569503117,
+      "grad_norm": 17.38302283506762,
+      "learning_rate": 6.024096385542169e-07,
+      "loss": 1.5943,
+      "step": 31
+    },
+    {
+      "epoch": 0.012874673104003219,
+      "grad_norm": 15.374566189246599,
+      "learning_rate": 6.224899598393574e-07,
+      "loss": 1.6221,
+      "step": 32
+    },
+    {
+      "epoch": 0.013277006638503319,
+      "grad_norm": 14.904500940732934,
+      "learning_rate": 6.425702811244979e-07,
+      "loss": 1.5452,
+      "step": 33
+    },
+    {
+      "epoch": 0.01367934017300342,
+      "grad_norm": 16.29526851000047,
+      "learning_rate": 6.626506024096387e-07,
+      "loss": 1.2262,
+      "step": 34
+    },
+    {
+      "epoch": 0.01408167370750352,
+      "grad_norm": 12.819998198911712,
+      "learning_rate": 6.827309236947792e-07,
+      "loss": 1.3476,
+      "step": 35
+    },
+    {
+      "epoch": 0.01448400724200362,
+      "grad_norm": 13.25182925491938,
+      "learning_rate": 7.028112449799197e-07,
+      "loss": 1.4371,
+      "step": 36
+    },
+    {
+      "epoch": 0.014886340776503722,
+      "grad_norm": 14.477871111934059,
+      "learning_rate": 7.228915662650602e-07,
+      "loss": 1.3476,
+      "step": 37
+    },
+    {
+      "epoch": 0.015288674311003822,
+      "grad_norm": 14.6196272151239,
+      "learning_rate": 7.429718875502008e-07,
+      "loss": 1.5309,
+      "step": 38
+    },
+    {
+      "epoch": 0.015691007845503924,
+      "grad_norm": 11.90353169885529,
+      "learning_rate": 7.630522088353415e-07,
+      "loss": 1.3026,
+      "step": 39
+    },
+    {
+      "epoch": 0.016093341380004022,
+      "grad_norm": 13.271008324770724,
+      "learning_rate": 7.83132530120482e-07,
+      "loss": 1.4655,
+      "step": 40
+    },
+    {
+      "epoch": 0.016495674914504124,
+      "grad_norm": 11.2349199407351,
+      "learning_rate": 8.032128514056225e-07,
+      "loss": 1.2782,
+      "step": 41
+    },
+    {
+      "epoch": 0.016898008449004225,
+      "grad_norm": 9.663111453936613,
+      "learning_rate": 8.232931726907631e-07,
+      "loss": 1.0628,
+      "step": 42
+    },
+    {
+      "epoch": 0.017300341983504323,
+      "grad_norm": 10.290302552273262,
+      "learning_rate": 8.433734939759036e-07,
+      "loss": 1.2398,
+      "step": 43
+    },
+    {
+      "epoch": 0.017702675518004425,
+      "grad_norm": 9.437964794637868,
+      "learning_rate": 8.634538152610443e-07,
+      "loss": 1.1087,
+      "step": 44
+    },
+    {
+      "epoch": 0.018105009052504527,
+      "grad_norm": 9.928972956598479,
+      "learning_rate": 8.835341365461848e-07,
+      "loss": 1.1231,
+      "step": 45
+    },
+    {
+      "epoch": 0.01850734258700463,
+      "grad_norm": 7.193942700746645,
+      "learning_rate": 9.036144578313254e-07,
+      "loss": 0.9053,
+      "step": 46
+    },
+    {
+      "epoch": 0.018909676121504727,
+      "grad_norm": 8.545887733782381,
+      "learning_rate": 9.236947791164659e-07,
+      "loss": 1.0188,
+      "step": 47
+    },
+    {
+      "epoch": 0.01931200965600483,
+      "grad_norm": 9.964310833592453,
+      "learning_rate": 9.437751004016064e-07,
+      "loss": 1.2336,
+      "step": 48
+    },
+    {
+      "epoch": 0.01971434319050493,
+      "grad_norm": 8.465102281501581,
+      "learning_rate": 9.638554216867472e-07,
+      "loss": 0.978,
+      "step": 49
+    },
+    {
+      "epoch": 0.020116676725005028,
+      "grad_norm": 8.387761909932218,
+      "learning_rate": 9.839357429718876e-07,
+      "loss": 1.0664,
+      "step": 50
+    },
+    {
+      "epoch": 0.02051901025950513,
+      "grad_norm": 8.853387211734997,
+      "learning_rate": 1.0040160642570282e-06,
+      "loss": 1.1316,
+      "step": 51
+    },
+    {
+      "epoch": 0.02092134379400523,
+      "grad_norm": 8.36811488486395,
+      "learning_rate": 1.0240963855421688e-06,
+      "loss": 0.9175,
+      "step": 52
+    },
+    {
+      "epoch": 0.02132367732850533,
+      "grad_norm": 8.394630361238342,
+      "learning_rate": 1.0441767068273092e-06,
+      "loss": 0.9371,
+      "step": 53
+    },
+    {
+      "epoch": 0.02172601086300543,
+      "grad_norm": 7.411500707300452,
+      "learning_rate": 1.0642570281124499e-06,
+      "loss": 0.9567,
+      "step": 54
+    },
+    {
+      "epoch": 0.022128344397505533,
+      "grad_norm": 8.194667792436356,
+      "learning_rate": 1.0843373493975905e-06,
+      "loss": 0.9994,
+      "step": 55
+    },
+    {
+      "epoch": 0.02253067793200563,
+      "grad_norm": 7.141458407534788,
+      "learning_rate": 1.1044176706827311e-06,
+      "loss": 0.8714,
+      "step": 56
+    },
+    {
+      "epoch": 0.022933011466505733,
+      "grad_norm": 5.765708921674402,
+      "learning_rate": 1.1244979919678715e-06,
+      "loss": 0.9238,
+      "step": 57
+    },
+    {
+      "epoch": 0.023335345001005835,
+      "grad_norm": 7.924253998059311,
+      "learning_rate": 1.1445783132530121e-06,
+      "loss": 0.9884,
+      "step": 58
+    },
+    {
+      "epoch": 0.023737678535505933,
+      "grad_norm": 6.597672306780267,
+      "learning_rate": 1.1646586345381528e-06,
+      "loss": 0.9904,
+      "step": 59
+    },
+    {
+      "epoch": 0.024140012070006035,
+      "grad_norm": 6.937688487718752,
+      "learning_rate": 1.1847389558232934e-06,
+      "loss": 0.971,
+      "step": 60
+    },
+    {
+      "epoch": 0.024542345604506136,
+      "grad_norm": 6.148469030855221,
+      "learning_rate": 1.2048192771084338e-06,
+      "loss": 0.7868,
+      "step": 61
+    },
+    {
+      "epoch": 0.024944679139006234,
+      "grad_norm": 6.388105517333996,
+      "learning_rate": 1.2248995983935744e-06,
+      "loss": 0.8567,
+      "step": 62
+    },
+    {
+      "epoch": 0.025347012673506336,
+      "grad_norm": 6.105324559387419,
+      "learning_rate": 1.2449799196787148e-06,
+      "loss": 0.9989,
+      "step": 63
+    },
+    {
+      "epoch": 0.025749346208006438,
+      "grad_norm": 6.565241933640869,
+      "learning_rate": 1.2650602409638555e-06,
+      "loss": 0.8641,
+      "step": 64
+    },
+    {
+      "epoch": 0.02615167974250654,
+      "grad_norm": 5.869907162685047,
+      "learning_rate": 1.2851405622489959e-06,
+      "loss": 0.9526,
+      "step": 65
+    },
+    {
+      "epoch": 0.026554013277006638,
+      "grad_norm": 5.851371649159624,
+      "learning_rate": 1.3052208835341367e-06,
+      "loss": 0.8387,
+      "step": 66
+    },
+    {
+      "epoch": 0.02695634681150674,
+      "grad_norm": 6.3493989766408765,
+      "learning_rate": 1.3253012048192773e-06,
+      "loss": 1.0458,
+      "step": 67
+    },
+    {
+      "epoch": 0.02735868034600684,
+      "grad_norm": 5.9283912382403114,
+      "learning_rate": 1.345381526104418e-06,
+      "loss": 1.0158,
+      "step": 68
+    },
+    {
+      "epoch": 0.02776101388050694,
+      "grad_norm": 6.184126721660312,
+      "learning_rate": 1.3654618473895584e-06,
+      "loss": 0.8218,
+      "step": 69
+    },
+    {
+      "epoch": 0.02816334741500704,
+      "grad_norm": 6.0169493477739975,
+      "learning_rate": 1.385542168674699e-06,
+      "loss": 0.7966,
+      "step": 70
+    },
+    {
+      "epoch": 0.028565680949507143,
+      "grad_norm": 6.084096016864222,
+      "learning_rate": 1.4056224899598394e-06,
+      "loss": 0.8221,
+      "step": 71
+    },
+    {
+      "epoch": 0.02896801448400724,
+      "grad_norm": 6.441697966234361,
+      "learning_rate": 1.42570281124498e-06,
+      "loss": 0.7189,
+      "step": 72
+    },
+    {
+      "epoch": 0.029370348018507342,
+      "grad_norm": 5.684758471405106,
+      "learning_rate": 1.4457831325301204e-06,
+      "loss": 0.811,
+      "step": 73
+    },
+    {
+      "epoch": 0.029772681553007444,
+      "grad_norm": 6.066554340342954,
+      "learning_rate": 1.465863453815261e-06,
+      "loss": 0.8128,
+      "step": 74
+    },
+    {
+      "epoch": 0.030175015087507542,
+      "grad_norm": 5.398586406654046,
+      "learning_rate": 1.4859437751004017e-06,
+      "loss": 0.8521,
+      "step": 75
+    },
+    {
+      "epoch": 0.030577348622007644,
+      "grad_norm": 5.72164636854203,
+      "learning_rate": 1.5060240963855425e-06,
+      "loss": 0.8733,
+      "step": 76
+    },
+    {
+      "epoch": 0.030979682156507746,
+      "grad_norm": 5.317662321058184,
+      "learning_rate": 1.526104417670683e-06,
+      "loss": 0.6758,
+      "step": 77
+    },
+    {
+      "epoch": 0.03138201569100785,
+      "grad_norm": 5.008046012529329,
+      "learning_rate": 1.5461847389558236e-06,
+      "loss": 0.77,
+      "step": 78
+    },
+    {
+      "epoch": 0.03178434922550795,
+      "grad_norm": 5.806938383916589,
+      "learning_rate": 1.566265060240964e-06,
+      "loss": 0.9218,
+      "step": 79
+    },
+    {
+      "epoch": 0.032186682760008044,
+      "grad_norm": 5.687427641764964,
+      "learning_rate": 1.5863453815261046e-06,
+      "loss": 0.873,
+      "step": 80
+    },
+    {
+      "epoch": 0.032589016294508145,
+      "grad_norm": 5.877177229539529,
+      "learning_rate": 1.606425702811245e-06,
+      "loss": 0.8347,
+      "step": 81
+    },
+    {
+      "epoch": 0.03299134982900825,
+      "grad_norm": 5.661277204998088,
+      "learning_rate": 1.6265060240963856e-06,
+      "loss": 0.7537,
+      "step": 82
+    },
+    {
+      "epoch": 0.03339368336350835,
+      "grad_norm": 6.0004706518847675,
+      "learning_rate": 1.6465863453815263e-06,
+      "loss": 0.9341,
+      "step": 83
+    },
+    {
+      "epoch": 0.03379601689800845,
+      "grad_norm": 5.383372715468371,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.716,
+      "step": 84
+    },
+    {
+      "epoch": 0.03419835043250855,
+      "grad_norm": 5.870492937673813,
+      "learning_rate": 1.6867469879518073e-06,
+      "loss": 0.8515,
+      "step": 85
+    },
+    {
+      "epoch": 0.03460068396700865,
+      "grad_norm": 5.40282247472889,
+      "learning_rate": 1.7068273092369481e-06,
+      "loss": 0.7575,
+      "step": 86
+    },
+    {
+      "epoch": 0.03500301750150875,
+      "grad_norm": 5.244298542730251,
+      "learning_rate": 1.7269076305220885e-06,
+      "loss": 0.7342,
+      "step": 87
+    },
+    {
+      "epoch": 0.03540535103600885,
+      "grad_norm": 5.118324863903489,
+      "learning_rate": 1.7469879518072292e-06,
+      "loss": 0.7633,
+      "step": 88
+    },
+    {
+      "epoch": 0.03580768457050895,
+      "grad_norm": 5.559590746485852,
+      "learning_rate": 1.7670682730923696e-06,
+      "loss": 0.915,
+      "step": 89
+    },
+    {
+      "epoch": 0.036210018105009054,
+      "grad_norm": 5.093580932305843,
+      "learning_rate": 1.7871485943775102e-06,
+      "loss": 0.7234,
+      "step": 90
+    },
+    {
+      "epoch": 0.036612351639509155,
+      "grad_norm": 5.949076037097858,
+      "learning_rate": 1.8072289156626508e-06,
+      "loss": 0.9392,
+      "step": 91
+    },
+    {
+      "epoch": 0.03701468517400926,
+      "grad_norm": 6.126962098905458,
+      "learning_rate": 1.8273092369477912e-06,
+      "loss": 0.7973,
+      "step": 92
+    },
+    {
+      "epoch": 0.03741701870850935,
+      "grad_norm": 5.418925049178253,
+      "learning_rate": 1.8473895582329318e-06,
+      "loss": 0.7386,
+      "step": 93
+    },
+    {
+      "epoch": 0.03781935224300945,
+      "grad_norm": 5.30622581323819,
+      "learning_rate": 1.8674698795180723e-06,
+      "loss": 0.7712,
+      "step": 94
+    },
+    {
+      "epoch": 0.038221685777509555,
+      "grad_norm": 5.203905538553713,
+      "learning_rate": 1.8875502008032129e-06,
+      "loss": 0.7403,
+      "step": 95
+    },
+    {
+      "epoch": 0.03862401931200966,
+      "grad_norm": 5.497752744670633,
+      "learning_rate": 1.9076305220883537e-06,
+      "loss": 0.8343,
+      "step": 96
+    },
+    {
+      "epoch": 0.03902635284650976,
+      "grad_norm": 5.792666990034754,
+      "learning_rate": 1.9277108433734943e-06,
+      "loss": 0.9855,
+      "step": 97
+    },
+    {
+      "epoch": 0.03942868638100986,
+      "grad_norm": 4.95103734609948,
+      "learning_rate": 1.947791164658635e-06,
+      "loss": 0.7676,
+      "step": 98
+    },
+    {
+      "epoch": 0.039831019915509955,
+      "grad_norm": 5.313239350915902,
+      "learning_rate": 1.967871485943775e-06,
+      "loss": 0.8112,
+      "step": 99
+    },
+    {
+      "epoch": 0.040233353450010056,
+      "grad_norm": 5.262581677945773,
+      "learning_rate": 1.987951807228916e-06,
+      "loss": 0.7726,
+      "step": 100
+    },
+    {
+      "epoch": 0.04063568698451016,
+      "grad_norm": 5.7284063119812965,
+      "learning_rate": 2.0080321285140564e-06,
+      "loss": 0.7143,
+      "step": 101
+    },
+    {
+      "epoch": 0.04103802051901026,
+      "grad_norm": 5.61074467493987,
+      "learning_rate": 2.028112449799197e-06,
+      "loss": 0.7821,
+      "step": 102
+    },
+    {
+      "epoch": 0.04144035405351036,
+      "grad_norm": 5.668818940572837,
+      "learning_rate": 2.0481927710843377e-06,
+      "loss": 0.6217,
+      "step": 103
+    },
+    {
+      "epoch": 0.04184268758801046,
+      "grad_norm": 6.102827563222031,
+      "learning_rate": 2.068273092369478e-06,
+      "loss": 0.7094,
+      "step": 104
+    },
+    {
+      "epoch": 0.04224502112251056,
+      "grad_norm": 5.106901986680839,
+      "learning_rate": 2.0883534136546185e-06,
+      "loss": 0.7452,
+      "step": 105
+    },
+    {
+      "epoch": 0.04264735465701066,
+      "grad_norm": 5.320689458556449,
+      "learning_rate": 2.1084337349397595e-06,
+      "loss": 0.8121,
+      "step": 106
+    },
+    {
+      "epoch": 0.04304968819151076,
+      "grad_norm": 4.851575241509713,
+      "learning_rate": 2.1285140562248997e-06,
+      "loss": 0.7695,
+      "step": 107
+    },
+    {
+      "epoch": 0.04345202172601086,
+      "grad_norm": 5.7294050539399315,
+      "learning_rate": 2.1485943775100404e-06,
+      "loss": 0.8343,
+      "step": 108
+    },
+    {
+      "epoch": 0.043854355260510965,
+      "grad_norm": 6.2102432153860265,
+      "learning_rate": 2.168674698795181e-06,
+      "loss": 0.8478,
+      "step": 109
+    },
+    {
+      "epoch": 0.044256688795011066,
+      "grad_norm": 5.8055227032618415,
+      "learning_rate": 2.1887550200803216e-06,
+      "loss": 0.7647,
+      "step": 110
+    },
+    {
+      "epoch": 0.04465902232951117,
+      "grad_norm": 5.59165747057277,
+      "learning_rate": 2.2088353413654622e-06,
+      "loss": 0.6905,
+      "step": 111
+    },
+    {
+      "epoch": 0.04506135586401126,
+      "grad_norm": 4.682897930842293,
+      "learning_rate": 2.2289156626506024e-06,
+      "loss": 0.6875,
+      "step": 112
+    },
+    {
+      "epoch": 0.045463689398511364,
+      "grad_norm": 4.320466219380915,
+      "learning_rate": 2.248995983935743e-06,
+      "loss": 0.696,
+      "step": 113
+    },
+    {
+      "epoch": 0.045866022933011466,
+      "grad_norm": 5.050663115560423,
+      "learning_rate": 2.2690763052208837e-06,
+      "loss": 0.6224,
+      "step": 114
+    },
+    {
+      "epoch": 0.04626835646751157,
+      "grad_norm": 4.681780822596801,
+      "learning_rate": 2.2891566265060243e-06,
+      "loss": 0.6462,
+      "step": 115
+    },
+    {
+      "epoch": 0.04667069000201167,
+      "grad_norm": 5.954135299109462,
+      "learning_rate": 2.309236947791165e-06,
+      "loss": 0.8848,
+      "step": 116
+    },
+    {
+      "epoch": 0.04707302353651177,
+      "grad_norm": 5.573998592787985,
+      "learning_rate": 2.3293172690763055e-06,
+      "loss": 0.8088,
+      "step": 117
+    },
+    {
+      "epoch": 0.047475357071011866,
+      "grad_norm": 5.070007694615057,
+      "learning_rate": 2.349397590361446e-06,
+      "loss": 0.7173,
+      "step": 118
+    },
+    {
+      "epoch": 0.04787769060551197,
+      "grad_norm": 5.671554086319649,
+      "learning_rate": 2.3694779116465868e-06,
+      "loss": 0.8136,
+      "step": 119
+    },
+    {
+      "epoch": 0.04828002414001207,
+      "grad_norm": 5.769697685018851,
+      "learning_rate": 2.389558232931727e-06,
+      "loss": 0.7733,
+      "step": 120
+    },
+    {
+      "epoch": 0.04868235767451217,
+      "grad_norm": 4.920298621449247,
+      "learning_rate": 2.4096385542168676e-06,
+      "loss": 0.5931,
+      "step": 121
+    },
+    {
+      "epoch": 0.04908469120901227,
+      "grad_norm": 5.49690667652572,
+      "learning_rate": 2.4297188755020082e-06,
+      "loss": 0.8079,
+      "step": 122
+    },
+    {
+      "epoch": 0.049487024743512374,
+      "grad_norm": 5.369187249008034,
+      "learning_rate": 2.449799196787149e-06,
+      "loss": 0.726,
+      "step": 123
+    },
+    {
+      "epoch": 0.04988935827801247,
+      "grad_norm": 4.8515152911462796,
+      "learning_rate": 2.469879518072289e-06,
+      "loss": 0.7691,
+      "step": 124
+    },
+    {
+      "epoch": 0.05029169181251257,
+      "grad_norm": 5.099204794104859,
+      "learning_rate": 2.4899598393574297e-06,
+      "loss": 0.8361,
+      "step": 125
+    },
+    {
+      "epoch": 0.05069402534701267,
+      "grad_norm": 5.390609640239935,
+      "learning_rate": 2.5100401606425707e-06,
+      "loss": 0.6527,
+      "step": 126
+    },
+    {
+      "epoch": 0.051096358881512774,
+      "grad_norm": 5.053936099287885,
+      "learning_rate": 2.530120481927711e-06,
+      "loss": 0.7326,
+      "step": 127
+    },
+    {
+      "epoch": 0.051498692416012876,
+      "grad_norm": 5.876719841742455,
+      "learning_rate": 2.5502008032128516e-06,
+      "loss": 0.8571,
+      "step": 128
+    },
+    {
+      "epoch": 0.05190102595051298,
+      "grad_norm": 5.291120710287588,
+      "learning_rate": 2.5702811244979918e-06,
+      "loss": 0.7641,
+      "step": 129
+    },
+    {
+      "epoch": 0.05230335948501308,
+      "grad_norm": 5.2648612797895336,
+      "learning_rate": 2.590361445783133e-06,
+      "loss": 0.768,
+      "step": 130
+    },
+    {
+      "epoch": 0.052705693019513174,
+      "grad_norm": 5.146667666976231,
+      "learning_rate": 2.6104417670682734e-06,
+      "loss": 0.729,
+      "step": 131
+    },
+    {
+      "epoch": 0.053108026554013275,
+      "grad_norm": 5.607251668752912,
+      "learning_rate": 2.6305220883534136e-06,
+      "loss": 0.7812,
+      "step": 132
+    },
+    {
+      "epoch": 0.05351036008851338,
+      "grad_norm": 4.774953294338218,
+      "learning_rate": 2.6506024096385547e-06,
+      "loss": 0.6996,
+      "step": 133
+    },
+    {
+      "epoch": 0.05391269362301348,
+      "grad_norm": 5.187652607031923,
+      "learning_rate": 2.670682730923695e-06,
+      "loss": 0.7029,
+      "step": 134
+    },
+    {
+      "epoch": 0.05431502715751358,
+      "grad_norm": 5.631076339839678,
+      "learning_rate": 2.690763052208836e-06,
+      "loss": 0.7618,
+      "step": 135
+    },
+    {
+      "epoch": 0.05471736069201368,
+      "grad_norm": 5.1080370644891735,
+      "learning_rate": 2.710843373493976e-06,
+      "loss": 0.6746,
+      "step": 136
+    },
+    {
+      "epoch": 0.05511969422651378,
+      "grad_norm": 5.177774314908037,
+      "learning_rate": 2.7309236947791167e-06,
+      "loss": 0.6576,
+      "step": 137
+    },
+    {
+      "epoch": 0.05552202776101388,
+      "grad_norm": 5.4428731816372355,
+      "learning_rate": 2.751004016064257e-06,
+      "loss": 0.8666,
+      "step": 138
+    },
+    {
+      "epoch": 0.05592436129551398,
+      "grad_norm": 4.737577892176393,
+      "learning_rate": 2.771084337349398e-06,
+      "loss": 0.6303,
+      "step": 139
+    },
+    {
+      "epoch": 0.05632669483001408,
+      "grad_norm": 5.958074686629733,
+      "learning_rate": 2.791164658634538e-06,
+      "loss": 0.7375,
+      "step": 140
+    },
+    {
+      "epoch": 0.056729028364514184,
+      "grad_norm": 5.347638262573924,
+      "learning_rate": 2.811244979919679e-06,
+      "loss": 0.7496,
+      "step": 141
+    },
+    {
+      "epoch": 0.057131361899014285,
+      "grad_norm": 5.25633214854894,
+      "learning_rate": 2.83132530120482e-06,
+      "loss": 0.7703,
+      "step": 142
+    },
+    {
+      "epoch": 0.05753369543351439,
+      "grad_norm": 6.3493191578507355,
+      "learning_rate": 2.85140562248996e-06,
+      "loss": 0.7297,
+      "step": 143
+    },
+    {
+      "epoch": 0.05793602896801448,
+      "grad_norm": 5.291536986822124,
+      "learning_rate": 2.8714859437751007e-06,
+      "loss": 0.7848,
+      "step": 144
+    },
+    {
+      "epoch": 0.05833836250251458,
+      "grad_norm": 4.327776932092466,
+      "learning_rate": 2.891566265060241e-06,
+      "loss": 0.5537,
+      "step": 145
+    },
+    {
+      "epoch": 0.058740696037014685,
+      "grad_norm": 4.827406966450388,
+      "learning_rate": 2.911646586345382e-06,
+      "loss": 0.6614,
+      "step": 146
+    },
+    {
+      "epoch": 0.05914302957151479,
+      "grad_norm": 5.4512595989034605,
+      "learning_rate": 2.931726907630522e-06,
+      "loss": 0.5105,
+      "step": 147
+    },
+    {
+      "epoch": 0.05954536310601489,
+      "grad_norm": 5.349806231832712,
+      "learning_rate": 2.9518072289156627e-06,
+      "loss": 0.7398,
+      "step": 148
+    },
+    {
+      "epoch": 0.05994769664051499,
+      "grad_norm": 5.376464953972205,
+      "learning_rate": 2.9718875502008034e-06,
+      "loss": 0.832,
+      "step": 149
+    },
+    {
+      "epoch": 0.060350030175015085,
+      "grad_norm": 4.4973784287811,
+      "learning_rate": 2.991967871485944e-06,
+      "loss": 0.6631,
+      "step": 150
+    },
+    {
+      "epoch": 0.060752363709515186,
+      "grad_norm": 4.1574904576313925,
+      "learning_rate": 3.012048192771085e-06,
+      "loss": 0.5736,
+      "step": 151
+    },
+    {
+      "epoch": 0.06115469724401529,
+      "grad_norm": 4.932454920014607,
+      "learning_rate": 3.0321285140562252e-06,
+      "loss": 0.6843,
+      "step": 152
+    },
+    {
+      "epoch": 0.06155703077851539,
+      "grad_norm": 4.920072821237323,
+      "learning_rate": 3.052208835341366e-06,
+      "loss": 0.6115,
+      "step": 153
+    },
+    {
+      "epoch": 0.06195936431301549,
+      "grad_norm": 5.069151288790082,
+      "learning_rate": 3.072289156626506e-06,
+      "loss": 0.6592,
+      "step": 154
+    },
+    {
+      "epoch": 0.06236169784751559,
+      "grad_norm": 4.940734045454529,
+      "learning_rate": 3.092369477911647e-06,
+      "loss": 0.6823,
+      "step": 155
+    },
+    {
+      "epoch": 0.0627640313820157,
+      "grad_norm": 4.7329970403721635,
+      "learning_rate": 3.1124497991967873e-06,
+      "loss": 0.6183,
+      "step": 156
+    },
+    {
+      "epoch": 0.0631663649165158,
+      "grad_norm": 5.584324231160161,
+      "learning_rate": 3.132530120481928e-06,
+      "loss": 0.8407,
+      "step": 157
+    },
+    {
+      "epoch": 0.0635686984510159,
+      "grad_norm": 4.9494988025514095,
+      "learning_rate": 3.152610441767068e-06,
+      "loss": 0.7362,
+      "step": 158
+    },
+    {
+      "epoch": 0.06397103198551599,
+      "grad_norm": 4.826336881024485,
+      "learning_rate": 3.172690763052209e-06,
+      "loss": 0.6756,
+      "step": 159
+    },
+    {
+      "epoch": 0.06437336552001609,
+      "grad_norm": 4.663584106854257,
+      "learning_rate": 3.1927710843373494e-06,
+      "loss": 0.6618,
+      "step": 160
+    },
+    {
+      "epoch": 0.06477569905451619,
+      "grad_norm": 5.358113823802763,
+      "learning_rate": 3.21285140562249e-06,
+      "loss": 0.7845,
+      "step": 161
+    },
+    {
+      "epoch": 0.06517803258901629,
+      "grad_norm": 4.393966562734061,
+      "learning_rate": 3.232931726907631e-06,
+      "loss": 0.6643,
+      "step": 162
+    },
+    {
+      "epoch": 0.06558036612351639,
+      "grad_norm": 5.269005574739929,
+      "learning_rate": 3.2530120481927713e-06,
+      "loss": 0.7892,
+      "step": 163
+    },
+    {
+      "epoch": 0.0659826996580165,
+      "grad_norm": 5.191377211307637,
+      "learning_rate": 3.2730923694779123e-06,
+      "loss": 0.6834,
+      "step": 164
+    },
+    {
+      "epoch": 0.0663850331925166,
+      "grad_norm": 5.9411717854682875,
+      "learning_rate": 3.2931726907630525e-06,
+      "loss": 0.682,
+      "step": 165
+    },
+    {
+      "epoch": 0.0667873667270167,
+      "grad_norm": 4.657895689681778,
+      "learning_rate": 3.313253012048193e-06,
+      "loss": 0.6533,
+      "step": 166
+    },
+    {
+      "epoch": 0.0671897002615168,
+      "grad_norm": 4.689805393308291,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5851,
+      "step": 167
+    },
+    {
+      "epoch": 0.0675920337960169,
+      "grad_norm": 4.934046837424603,
+      "learning_rate": 3.3534136546184744e-06,
+      "loss": 0.7834,
+      "step": 168
+    },
+    {
+      "epoch": 0.067994367330517,
+      "grad_norm": 5.295304511873193,
+      "learning_rate": 3.3734939759036146e-06,
+      "loss": 0.6978,
+      "step": 169
+    },
+    {
+      "epoch": 0.0683967008650171,
+      "grad_norm": 5.8805644086916935,
+      "learning_rate": 3.393574297188755e-06,
+      "loss": 0.7022,
+      "step": 170
+    },
+    {
+      "epoch": 0.0687990343995172,
+      "grad_norm": 4.986410631963467,
+      "learning_rate": 3.4136546184738962e-06,
+      "loss": 0.8589,
+      "step": 171
+    },
+    {
+      "epoch": 0.0692013679340173,
+      "grad_norm": 4.587536055949948,
+      "learning_rate": 3.4337349397590364e-06,
+      "loss": 0.5962,
+      "step": 172
+    },
+    {
+      "epoch": 0.0696037014685174,
+      "grad_norm": 4.542521557186459,
+      "learning_rate": 3.453815261044177e-06,
+      "loss": 0.5778,
+      "step": 173
+    },
+    {
+      "epoch": 0.0700060350030175,
+      "grad_norm": 5.006237220238328,
+      "learning_rate": 3.4738955823293173e-06,
+      "loss": 0.7066,
+      "step": 174
+    },
+    {
+      "epoch": 0.0704083685375176,
+      "grad_norm": 5.03127022104005,
+      "learning_rate": 3.4939759036144583e-06,
+      "loss": 0.7143,
+      "step": 175
+    },
+    {
+      "epoch": 0.0708107020720177,
+      "grad_norm": 4.978976609084576,
+      "learning_rate": 3.5140562248995985e-06,
+      "loss": 0.5806,
+      "step": 176
+    },
+    {
+      "epoch": 0.0712130356065178,
+      "grad_norm": 4.953415145134539,
+      "learning_rate": 3.534136546184739e-06,
+      "loss": 0.6169,
+      "step": 177
+    },
+    {
+      "epoch": 0.0716153691410179,
+      "grad_norm": 5.224274445585009,
+      "learning_rate": 3.5542168674698798e-06,
+      "loss": 0.7055,
+      "step": 178
+    },
+    {
+      "epoch": 0.072017702675518,
+      "grad_norm": 4.793231833957601,
+      "learning_rate": 3.5742971887550204e-06,
+      "loss": 0.6589,
+      "step": 179
+    },
+    {
+      "epoch": 0.07242003621001811,
+      "grad_norm": 4.897332398174339,
+      "learning_rate": 3.5943775100401606e-06,
+      "loss": 0.5835,
+      "step": 180
+    },
+    {
+      "epoch": 0.07282236974451821,
+      "grad_norm": 5.461164108231627,
+      "learning_rate": 3.6144578313253016e-06,
+      "loss": 0.6153,
+      "step": 181
+    },
+    {
+      "epoch": 0.07322470327901831,
+      "grad_norm": 4.750713371821981,
+      "learning_rate": 3.6345381526104423e-06,
+      "loss": 0.5576,
+      "step": 182
+    },
+    {
+      "epoch": 0.07362703681351841,
+      "grad_norm": 4.612815212986164,
+      "learning_rate": 3.6546184738955825e-06,
+      "loss": 0.6695,
+      "step": 183
+    },
+    {
+      "epoch": 0.07402937034801851,
+      "grad_norm": 5.035137988160493,
+      "learning_rate": 3.6746987951807235e-06,
+      "loss": 0.6451,
+      "step": 184
+    },
+    {
+      "epoch": 0.0744317038825186,
+      "grad_norm": 5.189253752749726,
+      "learning_rate": 3.6947791164658637e-06,
+      "loss": 0.6897,
+      "step": 185
+    },
+    {
+      "epoch": 0.0748340374170187,
+      "grad_norm": 5.061038391462879,
+      "learning_rate": 3.7148594377510043e-06,
+      "loss": 0.7363,
+      "step": 186
+    },
+    {
+      "epoch": 0.0752363709515188,
+      "grad_norm": 5.12457801565824,
+      "learning_rate": 3.7349397590361445e-06,
+      "loss": 0.6662,
+      "step": 187
+    },
+    {
+      "epoch": 0.0756387044860189,
+      "grad_norm": 5.468566358022457,
+      "learning_rate": 3.7550200803212856e-06,
+      "loss": 0.7333,
+      "step": 188
+    },
+    {
+      "epoch": 0.07604103802051901,
+      "grad_norm": 4.83323388980342,
+      "learning_rate": 3.7751004016064258e-06,
+      "loss": 0.6179,
+      "step": 189
+    },
+    {
+      "epoch": 0.07644337155501911,
+      "grad_norm": 5.618759075616741,
+      "learning_rate": 3.7951807228915664e-06,
+      "loss": 0.7861,
+      "step": 190
+    },
+    {
+      "epoch": 0.07684570508951921,
+      "grad_norm": 5.571721878536531,
+      "learning_rate": 3.8152610441767074e-06,
+      "loss": 0.6481,
+      "step": 191
+    },
+    {
+      "epoch": 0.07724803862401931,
+      "grad_norm": 5.067103682358002,
+      "learning_rate": 3.835341365461848e-06,
+      "loss": 0.7295,
+      "step": 192
+    },
+    {
+      "epoch": 0.07765037215851942,
+      "grad_norm": 4.945134722999282,
+      "learning_rate": 3.855421686746989e-06,
+      "loss": 0.6275,
+      "step": 193
+    },
+    {
+      "epoch": 0.07805270569301952,
+      "grad_norm": 4.995120702157204,
+      "learning_rate": 3.875502008032129e-06,
+      "loss": 0.6156,
+      "step": 194
+    },
+    {
+      "epoch": 0.07845503922751962,
+      "grad_norm": 4.936376571967402,
+      "learning_rate": 3.89558232931727e-06,
+      "loss": 0.6141,
+      "step": 195
+    },
+    {
+      "epoch": 0.07885737276201972,
+      "grad_norm": 5.014965363677265,
+      "learning_rate": 3.91566265060241e-06,
+      "loss": 0.6706,
+      "step": 196
+    },
+    {
+      "epoch": 0.07925970629651982,
+      "grad_norm": 4.379077775855287,
+      "learning_rate": 3.93574297188755e-06,
+      "loss": 0.625,
+      "step": 197
+    },
+    {
+      "epoch": 0.07966203983101991,
+      "grad_norm": 5.213065143955326,
+      "learning_rate": 3.9558232931726905e-06,
+      "loss": 0.6259,
+      "step": 198
+    },
+    {
+      "epoch": 0.08006437336552001,
+      "grad_norm": 4.943005393732397,
+      "learning_rate": 3.975903614457832e-06,
+      "loss": 0.6867,
+      "step": 199
+    },
+    {
+      "epoch": 0.08046670690002011,
+      "grad_norm": 6.1486973779873795,
+      "learning_rate": 3.995983935742972e-06,
+      "loss": 0.8347,
+      "step": 200
+    },
+    {
+      "epoch": 0.08086904043452021,
+      "grad_norm": 4.733552314219692,
+      "learning_rate": 4.016064257028113e-06,
+      "loss": 0.6611,
+      "step": 201
+    },
+    {
+      "epoch": 0.08127137396902032,
+      "grad_norm": 4.8269187112270195,
+      "learning_rate": 4.036144578313254e-06,
+      "loss": 0.6939,
+      "step": 202
+    },
+    {
+      "epoch": 0.08167370750352042,
+      "grad_norm": 5.3439700963806125,
+      "learning_rate": 4.056224899598394e-06,
+      "loss": 0.7133,
+      "step": 203
+    },
+    {
+      "epoch": 0.08207604103802052,
+      "grad_norm": 5.29998702305394,
+      "learning_rate": 4.076305220883534e-06,
+      "loss": 0.7792,
+      "step": 204
+    },
+    {
+      "epoch": 0.08247837457252062,
+      "grad_norm": 5.408216152846704,
+      "learning_rate": 4.096385542168675e-06,
+      "loss": 0.6411,
+      "step": 205
+    },
+    {
+      "epoch": 0.08288070810702072,
+      "grad_norm": 4.419630999712613,
+      "learning_rate": 4.1164658634538155e-06,
+      "loss": 0.615,
+      "step": 206
+    },
+    {
+      "epoch": 0.08328304164152082,
+      "grad_norm": 4.543564292966689,
+      "learning_rate": 4.136546184738956e-06,
+      "loss": 0.5462,
+      "step": 207
+    },
+    {
+      "epoch": 0.08368537517602093,
+      "grad_norm": 4.814136944508732,
+      "learning_rate": 4.156626506024097e-06,
+      "loss": 0.6407,
+      "step": 208
+    },
+    {
+      "epoch": 0.08408770871052103,
+      "grad_norm": 5.118958059055448,
+      "learning_rate": 4.176706827309237e-06,
+      "loss": 0.6922,
+      "step": 209
+    },
+    {
+      "epoch": 0.08449004224502112,
+      "grad_norm": 5.761968321529592,
+      "learning_rate": 4.196787148594378e-06,
+      "loss": 0.7785,
+      "step": 210
+    },
+    {
+      "epoch": 0.08489237577952122,
+      "grad_norm": 4.7973887668760185,
+      "learning_rate": 4.216867469879519e-06,
+      "loss": 0.5855,
+      "step": 211
+    },
+    {
+      "epoch": 0.08529470931402132,
+      "grad_norm": 5.063119851943765,
+      "learning_rate": 4.236947791164659e-06,
+      "loss": 0.5061,
+      "step": 212
+    },
+    {
+      "epoch": 0.08569704284852142,
+      "grad_norm": 5.185457141038217,
+      "learning_rate": 4.2570281124497995e-06,
+      "loss": 0.6547,
+      "step": 213
+    },
+    {
+      "epoch": 0.08609937638302152,
+      "grad_norm": 4.908063923355144,
+      "learning_rate": 4.27710843373494e-06,
+      "loss": 0.6325,
+      "step": 214
+    },
+    {
+      "epoch": 0.08650170991752162,
+      "grad_norm": 4.902236509892525,
+      "learning_rate": 4.297188755020081e-06,
+      "loss": 0.771,
+      "step": 215
+    },
+    {
+      "epoch": 0.08690404345202173,
+      "grad_norm": 4.939421470208826,
+      "learning_rate": 4.317269076305221e-06,
+      "loss": 0.6009,
+      "step": 216
+    },
+    {
+      "epoch": 0.08730637698652183,
+      "grad_norm": 5.585627020639388,
+      "learning_rate": 4.337349397590362e-06,
+      "loss": 0.75,
+      "step": 217
+    },
+    {
+      "epoch": 0.08770871052102193,
+      "grad_norm": 5.768733340993033,
+      "learning_rate": 4.357429718875502e-06,
+      "loss": 0.8889,
+      "step": 218
+    },
+    {
+      "epoch": 0.08811104405552203,
+      "grad_norm": 4.756073859984964,
+      "learning_rate": 4.377510040160643e-06,
+      "loss": 0.5845,
+      "step": 219
+    },
+    {
+      "epoch": 0.08851337759002213,
+      "grad_norm": 5.602957070495176,
+      "learning_rate": 4.397590361445783e-06,
+      "loss": 0.5766,
+      "step": 220
+    },
+    {
+      "epoch": 0.08891571112452223,
+      "grad_norm": 5.51373108651177,
+      "learning_rate": 4.4176706827309244e-06,
+      "loss": 0.667,
+      "step": 221
+    },
+    {
+      "epoch": 0.08931804465902234,
+      "grad_norm": 5.2818421493270895,
+      "learning_rate": 4.437751004016065e-06,
+      "loss": 0.6132,
+      "step": 222
+    },
+    {
+      "epoch": 0.08972037819352242,
+      "grad_norm": 4.687481374933495,
+      "learning_rate": 4.457831325301205e-06,
+      "loss": 0.69,
+      "step": 223
+    },
+    {
+      "epoch": 0.09012271172802253,
+      "grad_norm": 5.045402036516882,
+      "learning_rate": 4.477911646586346e-06,
+      "loss": 0.5496,
+      "step": 224
+    },
+    {
+      "epoch": 0.09052504526252263,
+      "grad_norm": 4.949655367217769,
+      "learning_rate": 4.497991967871486e-06,
+      "loss": 0.6777,
+      "step": 225
+    },
+    {
+      "epoch": 0.09092737879702273,
+      "grad_norm": 4.725446380096972,
+      "learning_rate": 4.518072289156627e-06,
+      "loss": 0.5211,
+      "step": 226
+    },
+    {
+      "epoch": 0.09132971233152283,
+      "grad_norm": 5.627747083860127,
+      "learning_rate": 4.538152610441767e-06,
+      "loss": 0.7335,
+      "step": 227
+    },
+    {
+      "epoch": 0.09173204586602293,
+      "grad_norm": 4.658915831199445,
+      "learning_rate": 4.558232931726908e-06,
+      "loss": 0.618,
+      "step": 228
+    },
+    {
+      "epoch": 0.09213437940052303,
+      "grad_norm": 5.169380307261355,
+      "learning_rate": 4.578313253012049e-06,
+      "loss": 0.8083,
+      "step": 229
+    },
+    {
+      "epoch": 0.09253671293502314,
+      "grad_norm": 4.462280637097748,
+      "learning_rate": 4.598393574297189e-06,
+      "loss": 0.5873,
+      "step": 230
+    },
+    {
+      "epoch": 0.09293904646952324,
+      "grad_norm": 5.088180514844276,
+      "learning_rate": 4.61847389558233e-06,
+      "loss": 0.6948,
+      "step": 231
+    },
+    {
+      "epoch": 0.09334138000402334,
+      "grad_norm": 4.911645340461795,
+      "learning_rate": 4.63855421686747e-06,
+      "loss": 0.5014,
+      "step": 232
+    },
+    {
+      "epoch": 0.09374371353852344,
+      "grad_norm": 5.183650867648958,
+      "learning_rate": 4.658634538152611e-06,
+      "loss": 0.6731,
+      "step": 233
+    },
+    {
+      "epoch": 0.09414604707302354,
+      "grad_norm": 4.942647346081943,
+      "learning_rate": 4.678714859437751e-06,
+      "loss": 0.796,
+      "step": 234
+    },
+    {
+      "epoch": 0.09454838060752364,
+      "grad_norm": 5.413754367435916,
+      "learning_rate": 4.698795180722892e-06,
+      "loss": 0.7029,
+      "step": 235
+    },
+    {
+      "epoch": 0.09495071414202373,
+      "grad_norm": 5.394996509930947,
+      "learning_rate": 4.7188755020080325e-06,
+      "loss": 0.7448,
+      "step": 236
+    },
+    {
+      "epoch": 0.09535304767652383,
+      "grad_norm": 4.807099705141704,
+      "learning_rate": 4.7389558232931736e-06,
+      "loss": 0.6608,
+      "step": 237
+    },
+    {
+      "epoch": 0.09575538121102393,
+      "grad_norm": 5.204336077501797,
+      "learning_rate": 4.759036144578314e-06,
+      "loss": 0.585,
+      "step": 238
+    },
+    {
+      "epoch": 0.09615771474552404,
+      "grad_norm": 4.845053158029385,
+      "learning_rate": 4.779116465863454e-06,
+      "loss": 0.6299,
+      "step": 239
+    },
+    {
+      "epoch": 0.09656004828002414,
+      "grad_norm": 5.037832345926416,
+      "learning_rate": 4.799196787148594e-06,
+      "loss": 0.6598,
+      "step": 240
+    },
+    {
+      "epoch": 0.09696238181452424,
+      "grad_norm": 4.930571962391263,
+      "learning_rate": 4.819277108433735e-06,
+      "loss": 0.5918,
+      "step": 241
+    },
+    {
+      "epoch": 0.09736471534902434,
+      "grad_norm": 4.749158619882368,
+      "learning_rate": 4.839357429718876e-06,
+      "loss": 0.5785,
+      "step": 242
+    },
+    {
+      "epoch": 0.09776704888352444,
+      "grad_norm": 5.774299138017609,
+      "learning_rate": 4.8594377510040165e-06,
+      "loss": 0.7465,
+      "step": 243
+    },
+    {
+      "epoch": 0.09816938241802455,
+      "grad_norm": 4.582929546310413,
+      "learning_rate": 4.8795180722891575e-06,
+      "loss": 0.5464,
+      "step": 244
+    },
+    {
+      "epoch": 0.09857171595252465,
+      "grad_norm": 4.604283671045787,
+      "learning_rate": 4.899598393574298e-06,
+      "loss": 0.6183,
+      "step": 245
+    },
+    {
+      "epoch": 0.09897404948702475,
+      "grad_norm": 5.126092092509994,
+      "learning_rate": 4.919678714859438e-06,
+      "loss": 0.7267,
+      "step": 246
+    },
+    {
+      "epoch": 0.09937638302152485,
+      "grad_norm": 4.6087281951772345,
+      "learning_rate": 4.939759036144578e-06,
+      "loss": 0.5911,
+      "step": 247
+    },
+    {
+      "epoch": 0.09977871655602494,
+      "grad_norm": 4.4308954136706165,
+      "learning_rate": 4.959839357429719e-06,
+      "loss": 0.6525,
+      "step": 248
+    },
+    {
+      "epoch": 0.10018105009052504,
+      "grad_norm": 4.704767612087696,
+      "learning_rate": 4.979919678714859e-06,
+      "loss": 0.7537,
+      "step": 249
+    },
+    {
+      "epoch": 0.10058338362502514,
+      "grad_norm": 5.171569719861641,
+      "learning_rate": 5e-06,
+      "loss": 0.5941,
+      "step": 250
+    },
+    {
+      "epoch": 0.10098571715952524,
+      "grad_norm": 5.7449160884679245,
+      "learning_rate": 4.99999753465491e-06,
+      "loss": 0.6686,
+      "step": 251
+    },
+    {
+      "epoch": 0.10138805069402534,
+      "grad_norm": 4.648308764039487,
+      "learning_rate": 4.999990138624503e-06,
+      "loss": 0.6264,
+      "step": 252
+    },
+    {
+      "epoch": 0.10179038422852545,
+      "grad_norm": 5.064339952308565,
+      "learning_rate": 4.999977811923365e-06,
+      "loss": 0.6921,
+      "step": 253
+    },
+    {
+      "epoch": 0.10219271776302555,
+      "grad_norm": 4.810326779178174,
+      "learning_rate": 4.9999605545758065e-06,
+      "loss": 0.6119,
+      "step": 254
+    },
+    {
+      "epoch": 0.10259505129752565,
+      "grad_norm": 4.805314994215823,
+      "learning_rate": 4.999938366615867e-06,
+      "loss": 0.6364,
+      "step": 255
+    },
+    {
+      "epoch": 0.10299738483202575,
+      "grad_norm": 4.473992853756316,
+      "learning_rate": 4.999911248087305e-06,
+      "loss": 0.5508,
+      "step": 256
+    },
+    {
+      "epoch": 0.10339971836652585,
+      "grad_norm": 5.452559581429947,
+      "learning_rate": 4.999879199043606e-06,
+      "loss": 0.7831,
+      "step": 257
+    },
+    {
+      "epoch": 0.10380205190102595,
+      "grad_norm": 5.061524793864594,
+      "learning_rate": 4.99984221954798e-06,
+      "loss": 0.5644,
+      "step": 258
+    },
+    {
+      "epoch": 0.10420438543552606,
+      "grad_norm": 4.673975167806026,
+      "learning_rate": 4.999800309673361e-06,
+      "loss": 0.649,
+      "step": 259
+    },
+    {
+      "epoch": 0.10460671897002616,
+      "grad_norm": 4.8181768852216855,
+      "learning_rate": 4.999753469502406e-06,
+      "loss": 0.6528,
+      "step": 260
+    },
+    {
+      "epoch": 0.10500905250452625,
+      "grad_norm": 5.3922298074237345,
+      "learning_rate": 4.999701699127497e-06,
+      "loss": 0.8216,
+      "step": 261
+    },
+    {
+      "epoch": 0.10541138603902635,
+      "grad_norm": 4.463138014998754,
+      "learning_rate": 4.99964499865074e-06,
+      "loss": 0.6824,
+      "step": 262
+    },
+    {
+      "epoch": 0.10581371957352645,
+      "grad_norm": 4.937246753810877,
+      "learning_rate": 4.999583368183965e-06,
+      "loss": 0.5256,
+      "step": 263
+    },
+    {
+      "epoch": 0.10621605310802655,
+      "grad_norm": 5.02716535200141,
+      "learning_rate": 4.999516807848721e-06,
+      "loss": 0.6658,
+      "step": 264
+    },
+    {
+      "epoch": 0.10661838664252665,
+      "grad_norm": 4.719903057858775,
+      "learning_rate": 4.999445317776286e-06,
+      "loss": 0.5943,
+      "step": 265
+    },
+    {
+      "epoch": 0.10702072017702675,
+      "grad_norm": 4.327891289019763,
+      "learning_rate": 4.9993688981076574e-06,
+      "loss": 0.6267,
+      "step": 266
+    },
+    {
+      "epoch": 0.10742305371152686,
+      "grad_norm": 5.04792717283726,
+      "learning_rate": 4.999287548993555e-06,
+      "loss": 0.7649,
+      "step": 267
+    },
+    {
+      "epoch": 0.10782538724602696,
+      "grad_norm": 4.398384896034094,
+      "learning_rate": 4.999201270594424e-06,
+      "loss": 0.6237,
+      "step": 268
+    },
+    {
+      "epoch": 0.10822772078052706,
+      "grad_norm": 4.513204332172601,
+      "learning_rate": 4.999110063080428e-06,
+      "loss": 0.685,
+      "step": 269
+    },
+    {
+      "epoch": 0.10863005431502716,
+      "grad_norm": 4.910504259364484,
+      "learning_rate": 4.999013926631453e-06,
+      "loss": 0.6393,
+      "step": 270
+    },
+    {
+      "epoch": 0.10903238784952726,
+      "grad_norm": 3.8051932823856927,
+      "learning_rate": 4.998912861437106e-06,
+      "loss": 0.557,
+      "step": 271
+    },
+    {
+      "epoch": 0.10943472138402736,
+      "grad_norm": 4.261798228362777,
+      "learning_rate": 4.9988068676967176e-06,
+      "loss": 0.6685,
+      "step": 272
+    },
+    {
+      "epoch": 0.10983705491852747,
+      "grad_norm": 5.199991754679721,
+      "learning_rate": 4.9986959456193345e-06,
+      "loss": 0.6926,
+      "step": 273
+    },
+    {
+      "epoch": 0.11023938845302755,
+      "grad_norm": 4.800143443642004,
+      "learning_rate": 4.998580095423728e-06,
+      "loss": 0.6262,
+      "step": 274
+    },
+    {
+      "epoch": 0.11064172198752766,
+      "grad_norm": 5.703676474195908,
+      "learning_rate": 4.998459317338383e-06,
+      "loss": 0.8102,
+      "step": 275
+    },
+    {
+      "epoch": 0.11104405552202776,
+      "grad_norm": 5.554521472398849,
+      "learning_rate": 4.998333611601511e-06,
+      "loss": 0.8502,
+      "step": 276
+    },
+    {
+      "epoch": 0.11144638905652786,
+      "grad_norm": 5.223628882218702,
+      "learning_rate": 4.998202978461038e-06,
+      "loss": 0.6968,
+      "step": 277
+    },
+    {
+      "epoch": 0.11184872259102796,
+      "grad_norm": 5.028503778823592,
+      "learning_rate": 4.9980674181746055e-06,
+      "loss": 0.6237,
+      "step": 278
+    },
+    {
+      "epoch": 0.11225105612552806,
+      "grad_norm": 4.810062581233454,
+      "learning_rate": 4.997926931009579e-06,
+      "loss": 0.6566,
+      "step": 279
+    },
+    {
+      "epoch": 0.11265338966002816,
+      "grad_norm": 4.895795240240787,
+      "learning_rate": 4.997781517243037e-06,
+      "loss": 0.6358,
+      "step": 280
+    },
+    {
+      "epoch": 0.11305572319452827,
+      "grad_norm": 4.678689813671165,
+      "learning_rate": 4.997631177161775e-06,
+      "loss": 0.7098,
+      "step": 281
+    },
+    {
+      "epoch": 0.11345805672902837,
+      "grad_norm": 4.8363206866892465,
+      "learning_rate": 4.9974759110623065e-06,
+      "loss": 0.7234,
+      "step": 282
+    },
+    {
+      "epoch": 0.11386039026352847,
+      "grad_norm": 4.872936830415585,
+      "learning_rate": 4.997315719250857e-06,
+      "loss": 0.5937,
+      "step": 283
+    },
+    {
+      "epoch": 0.11426272379802857,
+      "grad_norm": 4.28046131366056,
+      "learning_rate": 4.997150602043371e-06,
+      "loss": 0.6211,
+      "step": 284
+    },
+    {
+      "epoch": 0.11466505733252867,
+      "grad_norm": 4.837073628599811,
+      "learning_rate": 4.996980559765504e-06,
+      "loss": 0.5828,
+      "step": 285
+    },
+    {
+      "epoch": 0.11506739086702877,
+      "grad_norm": 4.889286615585772,
+      "learning_rate": 4.996805592752626e-06,
+      "loss": 0.7301,
+      "step": 286
+    },
+    {
+      "epoch": 0.11546972440152886,
+      "grad_norm": 3.885336153971243,
+      "learning_rate": 4.996625701349822e-06,
+      "loss": 0.5091,
+      "step": 287
+    },
+    {
+      "epoch": 0.11587205793602896,
+      "grad_norm": 4.738737155243486,
+      "learning_rate": 4.996440885911886e-06,
+      "loss": 0.5519,
+      "step": 288
+    },
+    {
+      "epoch": 0.11627439147052906,
+      "grad_norm": 4.600896924297193,
+      "learning_rate": 4.996251146803326e-06,
+      "loss": 0.5357,
+      "step": 289
+    },
+    {
+      "epoch": 0.11667672500502917,
+      "grad_norm": 4.348185410245171,
+      "learning_rate": 4.996056484398359e-06,
+      "loss": 0.606,
+      "step": 290
+    },
+    {
+      "epoch": 0.11707905853952927,
+      "grad_norm": 4.609344088685658,
+      "learning_rate": 4.995856899080914e-06,
+      "loss": 0.7079,
+      "step": 291
+    },
+    {
+      "epoch": 0.11748139207402937,
+      "grad_norm": 4.982538761307109,
+      "learning_rate": 4.9956523912446274e-06,
+      "loss": 0.6626,
+      "step": 292
+    },
+    {
+      "epoch": 0.11788372560852947,
+      "grad_norm": 4.958292330122082,
+      "learning_rate": 4.9954429612928455e-06,
+      "loss": 0.5589,
+      "step": 293
+    },
+    {
+      "epoch": 0.11828605914302957,
+      "grad_norm": 5.3063992889234015,
+      "learning_rate": 4.995228609638623e-06,
+      "loss": 0.6761,
+      "step": 294
+    },
+    {
+      "epoch": 0.11868839267752967,
+      "grad_norm": 4.787491132159558,
+      "learning_rate": 4.995009336704718e-06,
+      "loss": 0.6907,
+      "step": 295
+    },
+    {
+      "epoch": 0.11909072621202978,
+      "grad_norm": 5.042532894966667,
+      "learning_rate": 4.994785142923599e-06,
+      "loss": 0.7259,
+      "step": 296
+    },
+    {
+      "epoch": 0.11949305974652988,
+      "grad_norm": 5.24654264942669,
+      "learning_rate": 4.994556028737439e-06,
+      "loss": 0.7376,
+      "step": 297
+    },
+    {
+      "epoch": 0.11989539328102998,
+      "grad_norm": 4.7381264053196475,
+      "learning_rate": 4.9943219945981126e-06,
+      "loss": 0.6419,
+      "step": 298
+    },
+    {
+      "epoch": 0.12029772681553007,
+      "grad_norm": 4.174895491105691,
+      "learning_rate": 4.9940830409672e-06,
+      "loss": 0.5846,
+      "step": 299
+    },
+    {
+      "epoch": 0.12070006035003017,
+      "grad_norm": 5.265350639824991,
+      "learning_rate": 4.993839168315985e-06,
+      "loss": 0.7269,
+      "step": 300
+    },
+    {
+      "epoch": 0.12110239388453027,
+      "grad_norm": 5.293670654130999,
+      "learning_rate": 4.99359037712545e-06,
+      "loss": 0.7539,
+      "step": 301
+    },
+    {
+      "epoch": 0.12150472741903037,
+      "grad_norm": 4.832726818218027,
+      "learning_rate": 4.993336667886281e-06,
+      "loss": 0.6617,
+      "step": 302
+    },
+    {
+      "epoch": 0.12190706095353047,
+      "grad_norm": 4.982756198499471,
+      "learning_rate": 4.993078041098862e-06,
+      "loss": 0.6517,
+      "step": 303
+    },
+    {
+      "epoch": 0.12230939448803058,
+      "grad_norm": 4.787146336142149,
+      "learning_rate": 4.9928144972732785e-06,
+      "loss": 0.5943,
+      "step": 304
+    },
+    {
+      "epoch": 0.12271172802253068,
+      "grad_norm": 5.725699716607812,
+      "learning_rate": 4.992546036929309e-06,
+      "loss": 0.7385,
+      "step": 305
+    },
+    {
+      "epoch": 0.12311406155703078,
+      "grad_norm": 4.388246603675133,
+      "learning_rate": 4.992272660596432e-06,
+      "loss": 0.5268,
+      "step": 306
+    },
+    {
+      "epoch": 0.12351639509153088,
+      "grad_norm": 4.323516644235707,
+      "learning_rate": 4.991994368813823e-06,
+      "loss": 0.5069,
+      "step": 307
+    },
+    {
+      "epoch": 0.12391872862603098,
+      "grad_norm": 4.968139714274168,
+      "learning_rate": 4.991711162130347e-06,
+      "loss": 0.6089,
+      "step": 308
+    },
+    {
+      "epoch": 0.12432106216053108,
+      "grad_norm": 5.393689160165955,
+      "learning_rate": 4.99142304110457e-06,
+      "loss": 0.5749,
+      "step": 309
+    },
+    {
+      "epoch": 0.12472339569503119,
+      "grad_norm": 5.010889684729688,
+      "learning_rate": 4.991130006304742e-06,
+      "loss": 0.6868,
+      "step": 310
+    },
+    {
+      "epoch": 0.12512572922953127,
+      "grad_norm": 4.77584804740267,
+      "learning_rate": 4.99083205830881e-06,
+      "loss": 0.7153,
+      "step": 311
+    },
+    {
+      "epoch": 0.1255280627640314,
+      "grad_norm": 3.8671259573453343,
+      "learning_rate": 4.990529197704411e-06,
+      "loss": 0.4938,
+      "step": 312
+    },
+    {
+      "epoch": 0.12593039629853148,
+      "grad_norm": 4.878545701273808,
+      "learning_rate": 4.990221425088868e-06,
+      "loss": 0.5954,
+      "step": 313
+    },
+    {
+      "epoch": 0.1263327298330316,
+      "grad_norm": 4.481594062259529,
+      "learning_rate": 4.989908741069195e-06,
+      "loss": 0.6397,
+      "step": 314
+    },
+    {
+      "epoch": 0.12673506336753168,
+      "grad_norm": 4.317626894827104,
+      "learning_rate": 4.98959114626209e-06,
+      "loss": 0.609,
+      "step": 315
+    },
+    {
+      "epoch": 0.1271373969020318,
+      "grad_norm": 4.415809282581258,
+      "learning_rate": 4.989268641293939e-06,
+      "loss": 0.5092,
+      "step": 316
+    },
+    {
+      "epoch": 0.12753973043653188,
+      "grad_norm": 4.4424914565481375,
+      "learning_rate": 4.9889412268008096e-06,
+      "loss": 0.6316,
+      "step": 317
+    },
+    {
+      "epoch": 0.12794206397103197,
+      "grad_norm": 4.431277408902586,
+      "learning_rate": 4.988608903428454e-06,
+      "loss": 0.6015,
+      "step": 318
+    },
+    {
+      "epoch": 0.1283443975055321,
+      "grad_norm": 4.618861534462686,
+      "learning_rate": 4.988271671832305e-06,
+      "loss": 0.5562,
+      "step": 319
+    },
+    {
+      "epoch": 0.12874673104003218,
+      "grad_norm": 5.078731601531793,
+      "learning_rate": 4.987929532677478e-06,
+      "loss": 0.7308,
+      "step": 320
+    },
+    {
+      "epoch": 0.1291490645745323,
+      "grad_norm": 4.814095780713649,
+      "learning_rate": 4.987582486638763e-06,
+      "loss": 0.6856,
+      "step": 321
+    },
+    {
+      "epoch": 0.12955139810903238,
+      "grad_norm": 5.171953766801144,
+      "learning_rate": 4.987230534400634e-06,
+      "loss": 0.7413,
+      "step": 322
+    },
+    {
+      "epoch": 0.1299537316435325,
+      "grad_norm": 4.527962518284751,
+      "learning_rate": 4.986873676657237e-06,
+      "loss": 0.5227,
+      "step": 323
+    },
+    {
+      "epoch": 0.13035606517803258,
+      "grad_norm": 4.9194193843235405,
+      "learning_rate": 4.986511914112392e-06,
+      "loss": 0.6038,
+      "step": 324
+    },
+    {
+      "epoch": 0.1307583987125327,
+      "grad_norm": 4.403379209157847,
+      "learning_rate": 4.986145247479597e-06,
+      "loss": 0.6624,
+      "step": 325
+    },
+    {
+      "epoch": 0.13116073224703279,
+      "grad_norm": 4.784759384702467,
+      "learning_rate": 4.985773677482018e-06,
+      "loss": 0.7109,
+      "step": 326
+    },
+    {
+      "epoch": 0.1315630657815329,
+      "grad_norm": 4.928746115672455,
+      "learning_rate": 4.985397204852496e-06,
+      "loss": 0.6241,
+      "step": 327
+    },
+    {
+      "epoch": 0.131965399316033,
+      "grad_norm": 5.6685753103245755,
+      "learning_rate": 4.985015830333536e-06,
+      "loss": 0.7121,
+      "step": 328
+    },
+    {
+      "epoch": 0.1323677328505331,
+      "grad_norm": 4.723811520101061,
+      "learning_rate": 4.984629554677316e-06,
+      "loss": 0.5683,
+      "step": 329
+    },
+    {
+      "epoch": 0.1327700663850332,
+      "grad_norm": 4.106210405752034,
+      "learning_rate": 4.984238378645677e-06,
+      "loss": 0.6128,
+      "step": 330
+    },
+    {
+      "epoch": 0.13317239991953328,
+      "grad_norm": 5.439987884236508,
+      "learning_rate": 4.983842303010127e-06,
+      "loss": 0.7294,
+      "step": 331
+    },
+    {
+      "epoch": 0.1335747334540334,
+      "grad_norm": 4.76281391948892,
+      "learning_rate": 4.9834413285518365e-06,
+      "loss": 0.6116,
+      "step": 332
+    },
+    {
+      "epoch": 0.13397706698853348,
+      "grad_norm": 4.594103270680001,
+      "learning_rate": 4.983035456061637e-06,
+      "loss": 0.5509,
+      "step": 333
+    },
+    {
+      "epoch": 0.1343794005230336,
+      "grad_norm": 4.8229418474364785,
+      "learning_rate": 4.982624686340022e-06,
+      "loss": 0.6565,
+      "step": 334
+    },
+    {
+      "epoch": 0.1347817340575337,
+      "grad_norm": 4.572125613909981,
+      "learning_rate": 4.9822090201971414e-06,
+      "loss": 0.5739,
+      "step": 335
+    },
+    {
+      "epoch": 0.1351840675920338,
+      "grad_norm": 4.657722588184461,
+      "learning_rate": 4.981788458452806e-06,
+      "loss": 0.716,
+      "step": 336
+    },
+    {
+      "epoch": 0.1355864011265339,
+      "grad_norm": 4.485714819813409,
+      "learning_rate": 4.9813630019364765e-06,
+      "loss": 0.5809,
+      "step": 337
+    },
+    {
+      "epoch": 0.135988734661034,
+      "grad_norm": 4.730520981461843,
+      "learning_rate": 4.9809326514872735e-06,
+      "loss": 0.6006,
+      "step": 338
+    },
+    {
+      "epoch": 0.1363910681955341,
+      "grad_norm": 4.407751381335739,
+      "learning_rate": 4.980497407953966e-06,
+      "loss": 0.6979,
+      "step": 339
+    },
+    {
+      "epoch": 0.1367934017300342,
+      "grad_norm": 5.036488553382197,
+      "learning_rate": 4.980057272194973e-06,
+      "loss": 0.6719,
+      "step": 340
+    },
+    {
+      "epoch": 0.1371957352645343,
+      "grad_norm": 4.131020183485464,
+      "learning_rate": 4.979612245078366e-06,
+      "loss": 0.656,
+      "step": 341
+    },
+    {
+      "epoch": 0.1375980687990344,
+      "grad_norm": 4.494385344326802,
+      "learning_rate": 4.979162327481859e-06,
+      "loss": 0.5824,
+      "step": 342
+    },
+    {
+      "epoch": 0.1380004023335345,
+      "grad_norm": 4.887767115382313,
+      "learning_rate": 4.9787075202928155e-06,
+      "loss": 0.7065,
+      "step": 343
+    },
+    {
+      "epoch": 0.1384027358680346,
+      "grad_norm": 4.929080555596624,
+      "learning_rate": 4.97824782440824e-06,
+      "loss": 0.6462,
+      "step": 344
+    },
+    {
+      "epoch": 0.1388050694025347,
+      "grad_norm": 4.768200287000246,
+      "learning_rate": 4.97778324073478e-06,
+      "loss": 0.7052,
+      "step": 345
+    },
+    {
+      "epoch": 0.1392074029370348,
+      "grad_norm": 4.824866676754677,
+      "learning_rate": 4.977313770188723e-06,
+      "loss": 0.6169,
+      "step": 346
+    },
+    {
+      "epoch": 0.1396097364715349,
+      "grad_norm": 5.0730067342381515,
+      "learning_rate": 4.976839413695994e-06,
+      "loss": 0.6907,
+      "step": 347
+    },
+    {
+      "epoch": 0.140012070006035,
+      "grad_norm": 4.400048281159567,
+      "learning_rate": 4.976360172192156e-06,
+      "loss": 0.6843,
+      "step": 348
+    },
+    {
+      "epoch": 0.1404144035405351,
+      "grad_norm": 4.826940942258961,
+      "learning_rate": 4.975876046622404e-06,
+      "loss": 0.6425,
+      "step": 349
+    },
+    {
+      "epoch": 0.1408167370750352,
+      "grad_norm": 4.1980516292894645,
+      "learning_rate": 4.975387037941568e-06,
+      "loss": 0.513,
+      "step": 350
+    },
+    {
+      "epoch": 0.1412190706095353,
+      "grad_norm": 4.599176289887721,
+      "learning_rate": 4.974893147114108e-06,
+      "loss": 0.6138,
+      "step": 351
+    },
+    {
+      "epoch": 0.1416214041440354,
+      "grad_norm": 5.224094683564219,
+      "learning_rate": 4.974394375114113e-06,
+      "loss": 0.7688,
+      "step": 352
+    },
+    {
+      "epoch": 0.14202373767853552,
+      "grad_norm": 4.886795665659055,
+      "learning_rate": 4.9738907229253e-06,
+      "loss": 0.6447,
+      "step": 353
+    },
+    {
+      "epoch": 0.1424260712130356,
+      "grad_norm": 4.686162997153853,
+      "learning_rate": 4.973382191541008e-06,
+      "loss": 0.6489,
+      "step": 354
+    },
+    {
+      "epoch": 0.14282840474753572,
+      "grad_norm": 4.633889069385792,
+      "learning_rate": 4.972868781964204e-06,
+      "loss": 0.6841,
+      "step": 355
+    },
+    {
+      "epoch": 0.1432307382820358,
+      "grad_norm": 4.714123893258144,
+      "learning_rate": 4.972350495207472e-06,
+      "loss": 0.7282,
+      "step": 356
+    },
+    {
+      "epoch": 0.1436330718165359,
+      "grad_norm": 4.620909154154889,
+      "learning_rate": 4.971827332293017e-06,
+      "loss": 0.6267,
+      "step": 357
+    },
+    {
+      "epoch": 0.144035405351036,
+      "grad_norm": 5.112805770554144,
+      "learning_rate": 4.971299294252659e-06,
+      "loss": 0.7187,
+      "step": 358
+    },
+    {
+      "epoch": 0.1444377388855361,
+      "grad_norm": 4.491917871078975,
+      "learning_rate": 4.970766382127838e-06,
+      "loss": 0.6243,
+      "step": 359
+    },
+    {
+      "epoch": 0.14484007242003621,
+      "grad_norm": 4.91569774829319,
+      "learning_rate": 4.9702285969696006e-06,
+      "loss": 0.6044,
+      "step": 360
+    },
+    {
+      "epoch": 0.1452424059545363,
+      "grad_norm": 4.347791813311231,
+      "learning_rate": 4.96968593983861e-06,
+      "loss": 0.6496,
+      "step": 361
+    },
+    {
+      "epoch": 0.14564473948903642,
+      "grad_norm": 4.6534811500518645,
+      "learning_rate": 4.9691384118051346e-06,
+      "loss": 0.7414,
+      "step": 362
+    },
+    {
+      "epoch": 0.1460470730235365,
+      "grad_norm": 4.5918592667861455,
+      "learning_rate": 4.968586013949051e-06,
+      "loss": 0.5802,
+      "step": 363
+    },
+    {
+      "epoch": 0.14644940655803662,
+      "grad_norm": 4.435482175102624,
+      "learning_rate": 4.96802874735984e-06,
+      "loss": 0.5162,
+      "step": 364
+    },
+    {
+      "epoch": 0.1468517400925367,
+      "grad_norm": 4.407001515831722,
+      "learning_rate": 4.967466613136586e-06,
+      "loss": 0.5361,
+      "step": 365
+    },
+    {
+      "epoch": 0.14725407362703682,
+      "grad_norm": 4.369142207445966,
+      "learning_rate": 4.966899612387972e-06,
+      "loss": 0.664,
+      "step": 366
+    },
+    {
+      "epoch": 0.1476564071615369,
+      "grad_norm": 4.607767758241303,
+      "learning_rate": 4.966327746232281e-06,
+      "loss": 0.5882,
+      "step": 367
+    },
+    {
+      "epoch": 0.14805874069603703,
+      "grad_norm": 4.938226519089643,
+      "learning_rate": 4.96575101579739e-06,
+      "loss": 0.5841,
+      "step": 368
+    },
+    {
+      "epoch": 0.14846107423053712,
+      "grad_norm": 4.096479564143721,
+      "learning_rate": 4.965169422220771e-06,
+      "loss": 0.5691,
+      "step": 369
+    },
+    {
+      "epoch": 0.1488634077650372,
+      "grad_norm": 4.775093512712085,
+      "learning_rate": 4.964582966649488e-06,
+      "loss": 0.6063,
+      "step": 370
+    },
+    {
+      "epoch": 0.14926574129953732,
+      "grad_norm": 4.211658294897708,
+      "learning_rate": 4.963991650240192e-06,
+      "loss": 0.5215,
+      "step": 371
+    },
+    {
+      "epoch": 0.1496680748340374,
+      "grad_norm": 4.364777540143213,
+      "learning_rate": 4.963395474159122e-06,
+      "loss": 0.4866,
+      "step": 372
+    },
+    {
+      "epoch": 0.15007040836853752,
+      "grad_norm": 4.681087632312734,
+      "learning_rate": 4.962794439582102e-06,
+      "loss": 0.6122,
+      "step": 373
+    },
+    {
+      "epoch": 0.1504727419030376,
+      "grad_norm": 5.077820479708522,
+      "learning_rate": 4.9621885476945405e-06,
+      "loss": 0.7555,
+      "step": 374
+    },
+    {
+      "epoch": 0.15087507543753773,
+      "grad_norm": 5.978600203961224,
+      "learning_rate": 4.961577799691421e-06,
+      "loss": 0.7113,
+      "step": 375
+    },
+    {
+      "epoch": 0.1512774089720378,
+      "grad_norm": 5.376960173895862,
+      "learning_rate": 4.960962196777307e-06,
+      "loss": 0.577,
+      "step": 376
+    },
+    {
+      "epoch": 0.15167974250653793,
+      "grad_norm": 5.18055770132837,
+      "learning_rate": 4.960341740166338e-06,
+      "loss": 0.6625,
+      "step": 377
+    },
+    {
+      "epoch": 0.15208207604103802,
+      "grad_norm": 4.460009104960063,
+      "learning_rate": 4.959716431082227e-06,
+      "loss": 0.5306,
+      "step": 378
+    },
+    {
+      "epoch": 0.15248440957553813,
+      "grad_norm": 4.303191111673615,
+      "learning_rate": 4.959086270758255e-06,
+      "loss": 0.5604,
+      "step": 379
+    },
+    {
+      "epoch": 0.15288674311003822,
+      "grad_norm": 4.8364861459301185,
+      "learning_rate": 4.9584512604372704e-06,
+      "loss": 0.6154,
+      "step": 380
+    },
+    {
+      "epoch": 0.15328907664453834,
+      "grad_norm": 4.638041078223287,
+      "learning_rate": 4.957811401371692e-06,
+      "loss": 0.5852,
+      "step": 381
+    },
+    {
+      "epoch": 0.15369141017903842,
+      "grad_norm": 4.8155087041398295,
+      "learning_rate": 4.9571666948234975e-06,
+      "loss": 0.591,
+      "step": 382
+    },
+    {
+      "epoch": 0.1540937437135385,
+      "grad_norm": 4.365724907978,
+      "learning_rate": 4.956517142064226e-06,
+      "loss": 0.5645,
+      "step": 383
+    },
+    {
+      "epoch": 0.15449607724803863,
+      "grad_norm": 4.800750875035706,
+      "learning_rate": 4.955862744374974e-06,
+      "loss": 0.6958,
+      "step": 384
+    },
+    {
+      "epoch": 0.15489841078253871,
+      "grad_norm": 4.218913109898912,
+      "learning_rate": 4.9552035030463955e-06,
+      "loss": 0.5981,
+      "step": 385
+    },
+    {
+      "epoch": 0.15530074431703883,
+      "grad_norm": 3.8608056472996775,
+      "learning_rate": 4.954539419378695e-06,
+      "loss": 0.5516,
+      "step": 386
+    },
+    {
+      "epoch": 0.15570307785153892,
+      "grad_norm": 4.422539247772956,
+      "learning_rate": 4.953870494681631e-06,
+      "loss": 0.466,
+      "step": 387
+    },
+    {
+      "epoch": 0.15610541138603903,
+      "grad_norm": 5.389694964679017,
+      "learning_rate": 4.9531967302745065e-06,
+      "loss": 0.7187,
+      "step": 388
+    },
+    {
+      "epoch": 0.15650774492053912,
+      "grad_norm": 4.641539036731872,
+      "learning_rate": 4.952518127486171e-06,
+      "loss": 0.6118,
+      "step": 389
+    },
+    {
+      "epoch": 0.15691007845503924,
+      "grad_norm": 4.38792439545412,
+      "learning_rate": 4.951834687655016e-06,
+      "loss": 0.6125,
+      "step": 390
+    },
+    {
+      "epoch": 0.15731241198953932,
+      "grad_norm": 4.859276712096591,
+      "learning_rate": 4.9511464121289745e-06,
+      "loss": 0.789,
+      "step": 391
+    },
+    {
+      "epoch": 0.15771474552403944,
+      "grad_norm": 4.123781252290954,
+      "learning_rate": 4.950453302265516e-06,
+      "loss": 0.5189,
+      "step": 392
+    },
+    {
+      "epoch": 0.15811707905853953,
+      "grad_norm": 4.5665024165035915,
+      "learning_rate": 4.949755359431643e-06,
+      "loss": 0.5843,
+      "step": 393
+    },
+    {
+      "epoch": 0.15851941259303964,
+      "grad_norm": 4.554644389766188,
+      "learning_rate": 4.949052585003892e-06,
+      "loss": 0.5165,
+      "step": 394
+    },
+    {
+      "epoch": 0.15892174612753973,
+      "grad_norm": 4.652445776343059,
+      "learning_rate": 4.9483449803683295e-06,
+      "loss": 0.6496,
+      "step": 395
+    },
+    {
+      "epoch": 0.15932407966203982,
+      "grad_norm": 4.155340040403773,
+      "learning_rate": 4.947632546920545e-06,
+      "loss": 0.5823,
+      "step": 396
+    },
+    {
+      "epoch": 0.15972641319653993,
+      "grad_norm": 5.685373741993176,
+      "learning_rate": 4.946915286065656e-06,
+      "loss": 0.7257,
+      "step": 397
+    },
+    {
+      "epoch": 0.16012874673104002,
+      "grad_norm": 3.794107938623536,
+      "learning_rate": 4.946193199218298e-06,
+      "loss": 0.5539,
+      "step": 398
+    },
+    {
+      "epoch": 0.16053108026554014,
+      "grad_norm": 3.345825002647911,
+      "learning_rate": 4.945466287802625e-06,
+      "loss": 0.4522,
+      "step": 399
+    },
+    {
+      "epoch": 0.16093341380004023,
+      "grad_norm": 4.499068362412442,
+      "learning_rate": 4.944734553252308e-06,
+      "loss": 0.7294,
+      "step": 400
+    },
+    {
+      "epoch": 0.16133574733454034,
+      "grad_norm": 4.710035321209727,
+      "learning_rate": 4.94399799701053e-06,
+      "loss": 0.6495,
+      "step": 401
+    },
+    {
+      "epoch": 0.16173808086904043,
+      "grad_norm": 4.665060195725978,
+      "learning_rate": 4.943256620529983e-06,
+      "loss": 0.5549,
+      "step": 402
+    },
+    {
+      "epoch": 0.16214041440354054,
+      "grad_norm": 4.526568640183799,
+      "learning_rate": 4.942510425272864e-06,
+      "loss": 0.5342,
+      "step": 403
+    },
+    {
+      "epoch": 0.16254274793804063,
+      "grad_norm": 4.500912340574114,
+      "learning_rate": 4.941759412710878e-06,
+      "loss": 0.5154,
+      "step": 404
+    },
+    {
+      "epoch": 0.16294508147254075,
+      "grad_norm": 4.871395743184042,
+      "learning_rate": 4.941003584325229e-06,
+      "loss": 0.7273,
+      "step": 405
+    },
+    {
+      "epoch": 0.16334741500704084,
+      "grad_norm": 5.226443625458406,
+      "learning_rate": 4.940242941606619e-06,
+      "loss": 0.5695,
+      "step": 406
+    },
+    {
+      "epoch": 0.16374974854154092,
+      "grad_norm": 4.3424277680190215,
+      "learning_rate": 4.939477486055245e-06,
+      "loss": 0.6903,
+      "step": 407
+    },
+    {
+      "epoch": 0.16415208207604104,
+      "grad_norm": 4.63055317361875,
+      "learning_rate": 4.938707219180797e-06,
+      "loss": 0.7724,
+      "step": 408
+    },
+    {
+      "epoch": 0.16455441561054113,
+      "grad_norm": 4.632025607743007,
+      "learning_rate": 4.9379321425024545e-06,
+      "loss": 0.5579,
+      "step": 409
+    },
+    {
+      "epoch": 0.16495674914504124,
+      "grad_norm": 4.949408477476733,
+      "learning_rate": 4.9371522575488814e-06,
+      "loss": 0.5585,
+      "step": 410
+    },
+    {
+      "epoch": 0.16535908267954133,
+      "grad_norm": 4.82833151193623,
+      "learning_rate": 4.936367565858228e-06,
+      "loss": 0.6281,
+      "step": 411
+    },
+    {
+      "epoch": 0.16576141621404145,
+      "grad_norm": 4.135170899195267,
+      "learning_rate": 4.935578068978121e-06,
+      "loss": 0.5236,
+      "step": 412
+    },
+    {
+      "epoch": 0.16616374974854153,
+      "grad_norm": 4.848865820547502,
+      "learning_rate": 4.9347837684656675e-06,
+      "loss": 0.7572,
+      "step": 413
+    },
+    {
+      "epoch": 0.16656608328304165,
+      "grad_norm": 4.717705787081838,
+      "learning_rate": 4.933984665887447e-06,
+      "loss": 0.7282,
+      "step": 414
+    },
+    {
+      "epoch": 0.16696841681754174,
+      "grad_norm": 4.528215697412672,
+      "learning_rate": 4.93318076281951e-06,
+      "loss": 0.5974,
+      "step": 415
+    },
+    {
+      "epoch": 0.16737075035204185,
+      "grad_norm": 5.165943035201723,
+      "learning_rate": 4.9323720608473754e-06,
+      "loss": 0.6622,
+      "step": 416
+    },
+    {
+      "epoch": 0.16777308388654194,
+      "grad_norm": 4.244718678666535,
+      "learning_rate": 4.9315585615660275e-06,
+      "loss": 0.5827,
+      "step": 417
+    },
+    {
+      "epoch": 0.16817541742104206,
+      "grad_norm": 5.152052375848677,
+      "learning_rate": 4.930740266579911e-06,
+      "loss": 0.6625,
+      "step": 418
+    },
+    {
+      "epoch": 0.16857775095554214,
+      "grad_norm": 4.55196067766976,
+      "learning_rate": 4.92991717750293e-06,
+      "loss": 0.5734,
+      "step": 419
+    },
+    {
+      "epoch": 0.16898008449004223,
+      "grad_norm": 5.130317188995626,
+      "learning_rate": 4.929089295958442e-06,
+      "loss": 0.6448,
+      "step": 420
+    },
+    {
+      "epoch": 0.16938241802454235,
+      "grad_norm": 4.221455100939359,
+      "learning_rate": 4.928256623579259e-06,
+      "loss": 0.5736,
+      "step": 421
+    },
+    {
+      "epoch": 0.16978475155904243,
+      "grad_norm": 5.042585074077865,
+      "learning_rate": 4.9274191620076405e-06,
+      "loss": 0.6948,
+      "step": 422
+    },
+    {
+      "epoch": 0.17018708509354255,
+      "grad_norm": 4.475716088239217,
+      "learning_rate": 4.926576912895292e-06,
+      "loss": 0.5762,
+      "step": 423
+    },
+    {
+      "epoch": 0.17058941862804264,
+      "grad_norm": 4.629681308629434,
+      "learning_rate": 4.9257298779033615e-06,
+      "loss": 0.7584,
+      "step": 424
+    },
+    {
+      "epoch": 0.17099175216254275,
+      "grad_norm": 4.624861162695507,
+      "learning_rate": 4.924878058702436e-06,
+      "loss": 0.6725,
+      "step": 425
+    },
+    {
+      "epoch": 0.17139408569704284,
+      "grad_norm": 4.729048209294549,
+      "learning_rate": 4.9240214569725375e-06,
+      "loss": 0.7722,
+      "step": 426
+    },
+    {
+      "epoch": 0.17179641923154296,
+      "grad_norm": 4.21954665313362,
+      "learning_rate": 4.923160074403122e-06,
+      "loss": 0.6253,
+      "step": 427
+    },
+    {
+      "epoch": 0.17219875276604305,
+      "grad_norm": 3.755869045635417,
+      "learning_rate": 4.922293912693072e-06,
+      "loss": 0.4626,
+      "step": 428
+    },
+    {
+      "epoch": 0.17260108630054316,
+      "grad_norm": 4.263780577574945,
+      "learning_rate": 4.921422973550699e-06,
+      "loss": 0.695,
+      "step": 429
+    },
+    {
+      "epoch": 0.17300341983504325,
+      "grad_norm": 5.108523767367576,
+      "learning_rate": 4.920547258693735e-06,
+      "loss": 0.7092,
+      "step": 430
+    },
+    {
+      "epoch": 0.17340575336954336,
+      "grad_norm": 4.317384116136861,
+      "learning_rate": 4.919666769849332e-06,
+      "loss": 0.4564,
+      "step": 431
+    },
+    {
+      "epoch": 0.17380808690404345,
+      "grad_norm": 5.268931042465572,
+      "learning_rate": 4.918781508754057e-06,
+      "loss": 0.7064,
+      "step": 432
+    },
+    {
+      "epoch": 0.17421042043854354,
+      "grad_norm": 4.295948103968784,
+      "learning_rate": 4.9178914771538875e-06,
+      "loss": 0.6892,
+      "step": 433
+    },
+    {
+      "epoch": 0.17461275397304366,
+      "grad_norm": 4.214776194153066,
+      "learning_rate": 4.916996676804214e-06,
+      "loss": 0.5877,
+      "step": 434
+    },
+    {
+      "epoch": 0.17501508750754374,
+      "grad_norm": 4.187051473349894,
+      "learning_rate": 4.916097109469829e-06,
+      "loss": 0.5542,
+      "step": 435
+    },
+    {
+      "epoch": 0.17541742104204386,
+      "grad_norm": 4.7370251117595235,
+      "learning_rate": 4.915192776924926e-06,
+      "loss": 0.6131,
+      "step": 436
+    },
+    {
+      "epoch": 0.17581975457654395,
+      "grad_norm": 4.790572516699059,
+      "learning_rate": 4.9142836809531e-06,
+      "loss": 0.6847,
+      "step": 437
+    },
+    {
+      "epoch": 0.17622208811104406,
+      "grad_norm": 4.024460804862858,
+      "learning_rate": 4.91336982334734e-06,
+      "loss": 0.4475,
+      "step": 438
+    },
+    {
+      "epoch": 0.17662442164554415,
+      "grad_norm": 4.675530435693111,
+      "learning_rate": 4.912451205910024e-06,
+      "loss": 0.641,
+      "step": 439
+    },
+    {
+      "epoch": 0.17702675518004427,
+      "grad_norm": 4.070435859439258,
+      "learning_rate": 4.91152783045292e-06,
+      "loss": 0.5715,
+      "step": 440
+    },
+    {
+      "epoch": 0.17742908871454435,
+      "grad_norm": 5.014878172443035,
+      "learning_rate": 4.910599698797179e-06,
+      "loss": 0.6227,
+      "step": 441
+    },
+    {
+      "epoch": 0.17783142224904447,
+      "grad_norm": 4.471540054012383,
+      "learning_rate": 4.909666812773333e-06,
+      "loss": 0.5145,
+      "step": 442
+    },
+    {
+      "epoch": 0.17823375578354456,
+      "grad_norm": 4.442414191534445,
+      "learning_rate": 4.908729174221289e-06,
+      "loss": 0.6688,
+      "step": 443
+    },
+    {
+      "epoch": 0.17863608931804467,
+      "grad_norm": 4.624231827213014,
+      "learning_rate": 4.9077867849903325e-06,
+      "loss": 0.6327,
+      "step": 444
+    },
+    {
+      "epoch": 0.17903842285254476,
+      "grad_norm": 4.743489707184984,
+      "learning_rate": 4.906839646939113e-06,
+      "loss": 0.7075,
+      "step": 445
+    },
+    {
+      "epoch": 0.17944075638704485,
+      "grad_norm": 4.200250704275894,
+      "learning_rate": 4.905887761935649e-06,
+      "loss": 0.5171,
+      "step": 446
+    },
+    {
+      "epoch": 0.17984308992154496,
+      "grad_norm": 4.0039914177118545,
+      "learning_rate": 4.90493113185732e-06,
+      "loss": 0.5334,
+      "step": 447
+    },
+    {
+      "epoch": 0.18024542345604505,
+      "grad_norm": 4.437438466205263,
+      "learning_rate": 4.903969758590865e-06,
+      "loss": 0.584,
+      "step": 448
+    },
+    {
+      "epoch": 0.18064775699054517,
+      "grad_norm": 4.668639275979892,
+      "learning_rate": 4.9030036440323766e-06,
+      "loss": 0.669,
+      "step": 449
+    },
+    {
+      "epoch": 0.18105009052504525,
+      "grad_norm": 4.262713251256841,
+      "learning_rate": 4.902032790087301e-06,
+      "loss": 0.518,
+      "step": 450
+    },
+    {
+      "epoch": 0.18145242405954537,
+      "grad_norm": 4.58562990715973,
+      "learning_rate": 4.9010571986704295e-06,
+      "loss": 0.5627,
+      "step": 451
+    },
+    {
+      "epoch": 0.18185475759404546,
+      "grad_norm": 4.2511574164879615,
+      "learning_rate": 4.900076871705897e-06,
+      "loss": 0.6189,
+      "step": 452
+    },
+    {
+      "epoch": 0.18225709112854557,
+      "grad_norm": 3.879096119034437,
+      "learning_rate": 4.89909181112718e-06,
+      "loss": 0.4688,
+      "step": 453
+    },
+    {
+      "epoch": 0.18265942466304566,
+      "grad_norm": 4.4698856213971645,
+      "learning_rate": 4.898102018877088e-06,
+      "loss": 0.6272,
+      "step": 454
+    },
+    {
+      "epoch": 0.18306175819754578,
+      "grad_norm": 4.25349966912043,
+      "learning_rate": 4.897107496907767e-06,
+      "loss": 0.4844,
+      "step": 455
+    },
+    {
+      "epoch": 0.18346409173204586,
+      "grad_norm": 5.180942020863308,
+      "learning_rate": 4.896108247180688e-06,
+      "loss": 0.7363,
+      "step": 456
+    },
+    {
+      "epoch": 0.18386642526654598,
+      "grad_norm": 4.464543366415725,
+      "learning_rate": 4.895104271666647e-06,
+      "loss": 0.507,
+      "step": 457
+    },
+    {
+      "epoch": 0.18426875880104607,
+      "grad_norm": 4.35656943725317,
+      "learning_rate": 4.8940955723457604e-06,
+      "loss": 0.5626,
+      "step": 458
+    },
+    {
+      "epoch": 0.18467109233554616,
+      "grad_norm": 4.46990880672421,
+      "learning_rate": 4.893082151207464e-06,
+      "loss": 0.5986,
+      "step": 459
+    },
+    {
+      "epoch": 0.18507342587004627,
+      "grad_norm": 4.315333956337568,
+      "learning_rate": 4.8920640102505005e-06,
+      "loss": 0.5399,
+      "step": 460
+    },
+    {
+      "epoch": 0.18547575940454636,
+      "grad_norm": 4.414262914713829,
+      "learning_rate": 4.891041151482928e-06,
+      "loss": 0.6866,
+      "step": 461
+    },
+    {
+      "epoch": 0.18587809293904647,
+      "grad_norm": 4.68767919716049,
+      "learning_rate": 4.8900135769221045e-06,
+      "loss": 0.6205,
+      "step": 462
+    },
+    {
+      "epoch": 0.18628042647354656,
+      "grad_norm": 4.55495109723687,
+      "learning_rate": 4.888981288594692e-06,
+      "loss": 0.6232,
+      "step": 463
+    },
+    {
+      "epoch": 0.18668276000804668,
+      "grad_norm": 5.072112114859536,
+      "learning_rate": 4.887944288536648e-06,
+      "loss": 0.6246,
+      "step": 464
+    },
+    {
+      "epoch": 0.18708509354254677,
+      "grad_norm": 4.598620261784645,
+      "learning_rate": 4.886902578793221e-06,
+      "loss": 0.6688,
+      "step": 465
+    },
+    {
+      "epoch": 0.18748742707704688,
+      "grad_norm": 4.156182800147685,
+      "learning_rate": 4.885856161418953e-06,
+      "loss": 0.6185,
+      "step": 466
+    },
+    {
+      "epoch": 0.18788976061154697,
+      "grad_norm": 5.123876789008787,
+      "learning_rate": 4.8848050384776665e-06,
+      "loss": 0.6485,
+      "step": 467
+    },
+    {
+      "epoch": 0.18829209414604708,
+      "grad_norm": 4.7877986741742555,
+      "learning_rate": 4.883749212042466e-06,
+      "loss": 0.6336,
+      "step": 468
+    },
+    {
+      "epoch": 0.18869442768054717,
+      "grad_norm": 4.170455407715718,
+      "learning_rate": 4.882688684195732e-06,
+      "loss": 0.5708,
+      "step": 469
+    },
+    {
+      "epoch": 0.1890967612150473,
+      "grad_norm": 4.394557070936156,
+      "learning_rate": 4.881623457029121e-06,
+      "loss": 0.6584,
+      "step": 470
+    },
+    {
+      "epoch": 0.18949909474954738,
+      "grad_norm": 4.4866621521500205,
+      "learning_rate": 4.880553532643552e-06,
+      "loss": 0.5816,
+      "step": 471
+    },
+    {
+      "epoch": 0.18990142828404746,
+      "grad_norm": 4.263089971509693,
+      "learning_rate": 4.879478913149212e-06,
+      "loss": 0.6563,
+      "step": 472
+    },
+    {
+      "epoch": 0.19030376181854758,
+      "grad_norm": 3.897184614846541,
+      "learning_rate": 4.8783996006655485e-06,
+      "loss": 0.5816,
+      "step": 473
+    },
+    {
+      "epoch": 0.19070609535304767,
+      "grad_norm": 4.759719975025391,
+      "learning_rate": 4.877315597321263e-06,
+      "loss": 0.7219,
+      "step": 474
+    },
+    {
+      "epoch": 0.19110842888754778,
+      "grad_norm": 4.561889671156211,
+      "learning_rate": 4.876226905254309e-06,
+      "loss": 0.6307,
+      "step": 475
+    },
+    {
+      "epoch": 0.19151076242204787,
+      "grad_norm": 4.622879218120913,
+      "learning_rate": 4.875133526611888e-06,
+      "loss": 0.5739,
+      "step": 476
+    },
+    {
+      "epoch": 0.19191309595654799,
+      "grad_norm": 4.183836330008031,
+      "learning_rate": 4.874035463550445e-06,
+      "loss": 0.4165,
+      "step": 477
+    },
+    {
+      "epoch": 0.19231542949104807,
+      "grad_norm": 4.431780943086593,
+      "learning_rate": 4.872932718235663e-06,
+      "loss": 0.5768,
+      "step": 478
+    },
+    {
+      "epoch": 0.1927177630255482,
+      "grad_norm": 4.700772397419956,
+      "learning_rate": 4.87182529284246e-06,
+      "loss": 0.6635,
+      "step": 479
+    },
+    {
+      "epoch": 0.19312009656004828,
+      "grad_norm": 4.0384190687435915,
+      "learning_rate": 4.870713189554985e-06,
+      "loss": 0.4994,
+      "step": 480
+    },
+    {
+      "epoch": 0.1935224300945484,
+      "grad_norm": 4.16066989214431,
+      "learning_rate": 4.869596410566614e-06,
+      "loss": 0.523,
+      "step": 481
+    },
+    {
+      "epoch": 0.19392476362904848,
+      "grad_norm": 3.7758221004395565,
+      "learning_rate": 4.8684749580799405e-06,
+      "loss": 0.5106,
+      "step": 482
+    },
+    {
+      "epoch": 0.1943270971635486,
+      "grad_norm": 4.792256321316127,
+      "learning_rate": 4.867348834306781e-06,
+      "loss": 0.5955,
+      "step": 483
+    },
+    {
+      "epoch": 0.19472943069804868,
+      "grad_norm": 4.408528934091589,
+      "learning_rate": 4.866218041468161e-06,
+      "loss": 0.6032,
+      "step": 484
+    },
+    {
+      "epoch": 0.19513176423254877,
+      "grad_norm": 4.425894822386934,
+      "learning_rate": 4.865082581794317e-06,
+      "loss": 0.5298,
+      "step": 485
+    },
+    {
+      "epoch": 0.1955340977670489,
+      "grad_norm": 4.356368057153947,
+      "learning_rate": 4.863942457524689e-06,
+      "loss": 0.5381,
+      "step": 486
+    },
+    {
+      "epoch": 0.19593643130154897,
+      "grad_norm": 3.990737447917161,
+      "learning_rate": 4.862797670907915e-06,
+      "loss": 0.5049,
+      "step": 487
+    },
+    {
+      "epoch": 0.1963387648360491,
+      "grad_norm": 4.229237889406895,
+      "learning_rate": 4.8616482242018325e-06,
+      "loss": 0.5467,
+      "step": 488
+    },
+    {
+      "epoch": 0.19674109837054918,
+      "grad_norm": 4.126506367371427,
+      "learning_rate": 4.860494119673466e-06,
+      "loss": 0.5455,
+      "step": 489
+    },
+    {
+      "epoch": 0.1971434319050493,
+      "grad_norm": 5.231977983505604,
+      "learning_rate": 4.859335359599031e-06,
+      "loss": 0.7528,
+      "step": 490
+    },
+    {
+      "epoch": 0.19754576543954938,
+      "grad_norm": 5.1291616835145115,
+      "learning_rate": 4.85817194626392e-06,
+      "loss": 0.679,
+      "step": 491
+    },
+    {
+      "epoch": 0.1979480989740495,
+      "grad_norm": 3.737681461647685,
+      "learning_rate": 4.857003881962705e-06,
+      "loss": 0.4926,
+      "step": 492
+    },
+    {
+      "epoch": 0.19835043250854958,
+      "grad_norm": 4.703925829656442,
+      "learning_rate": 4.855831168999132e-06,
+      "loss": 0.6439,
+      "step": 493
+    },
+    {
+      "epoch": 0.1987527660430497,
+      "grad_norm": 4.627856862526262,
+      "learning_rate": 4.854653809686115e-06,
+      "loss": 0.6698,
+      "step": 494
+    },
+    {
+      "epoch": 0.1991550995775498,
+      "grad_norm": 4.364999176797817,
+      "learning_rate": 4.853471806345732e-06,
+      "loss": 0.6015,
+      "step": 495
+    },
+    {
+      "epoch": 0.19955743311204988,
+      "grad_norm": 4.602684651491329,
+      "learning_rate": 4.852285161309218e-06,
+      "loss": 0.5809,
+      "step": 496
+    },
+    {
+      "epoch": 0.19995976664655,
+      "grad_norm": 4.325677574896615,
+      "learning_rate": 4.851093876916967e-06,
+      "loss": 0.526,
+      "step": 497
+    },
+    {
+      "epoch": 0.20036210018105008,
+      "grad_norm": 4.81776828160967,
+      "learning_rate": 4.849897955518518e-06,
+      "loss": 0.6654,
+      "step": 498
+    },
+    {
+      "epoch": 0.2007644337155502,
+      "grad_norm": 5.49681992539213,
+      "learning_rate": 4.848697399472561e-06,
+      "loss": 0.8468,
+      "step": 499
+    },
+    {
+      "epoch": 0.20116676725005028,
+      "grad_norm": 4.782602085058219,
+      "learning_rate": 4.8474922111469225e-06,
+      "loss": 0.6339,
+      "step": 500
+    },
+    {
+      "epoch": 0.2015691007845504,
+      "grad_norm": 4.705465090280715,
+      "learning_rate": 4.846282392918566e-06,
+      "loss": 0.5111,
+      "step": 501
+    },
+    {
+      "epoch": 0.20197143431905049,
+      "grad_norm": 4.044280178788859,
+      "learning_rate": 4.845067947173589e-06,
+      "loss": 0.4453,
+      "step": 502
+    },
+    {
+      "epoch": 0.2023737678535506,
+      "grad_norm": 5.299453770363048,
+      "learning_rate": 4.843848876307211e-06,
+      "loss": 0.7287,
+      "step": 503
+    },
+    {
+      "epoch": 0.2027761013880507,
+      "grad_norm": 3.9060416143674,
+      "learning_rate": 4.842625182723779e-06,
+      "loss": 0.4903,
+      "step": 504
+    },
+    {
+      "epoch": 0.2031784349225508,
+      "grad_norm": 4.320966869452432,
+      "learning_rate": 4.841396868836753e-06,
+      "loss": 0.5312,
+      "step": 505
+    },
+    {
+      "epoch": 0.2035807684570509,
+      "grad_norm": 4.059590568863906,
+      "learning_rate": 4.840163937068707e-06,
+      "loss": 0.5953,
+      "step": 506
+    },
+    {
+      "epoch": 0.203983101991551,
+      "grad_norm": 4.544560777009811,
+      "learning_rate": 4.838926389851324e-06,
+      "loss": 0.5274,
+      "step": 507
+    },
+    {
+      "epoch": 0.2043854355260511,
+      "grad_norm": 4.643239965625325,
+      "learning_rate": 4.837684229625389e-06,
+      "loss": 0.6284,
+      "step": 508
+    },
+    {
+      "epoch": 0.20478776906055118,
+      "grad_norm": 3.848260216211789,
+      "learning_rate": 4.836437458840783e-06,
+      "loss": 0.5052,
+      "step": 509
+    },
+    {
+      "epoch": 0.2051901025950513,
+      "grad_norm": 4.65542258301039,
+      "learning_rate": 4.835186079956483e-06,
+      "loss": 0.5641,
+      "step": 510
+    },
+    {
+      "epoch": 0.2055924361295514,
+      "grad_norm": 3.7048624001601738,
+      "learning_rate": 4.8339300954405545e-06,
+      "loss": 0.5507,
+      "step": 511
+    },
+    {
+      "epoch": 0.2059947696640515,
+      "grad_norm": 4.721724963544143,
+      "learning_rate": 4.832669507770144e-06,
+      "loss": 0.6945,
+      "step": 512
+    },
+    {
+      "epoch": 0.2063971031985516,
+      "grad_norm": 3.9937965024761892,
+      "learning_rate": 4.83140431943148e-06,
+      "loss": 0.5444,
+      "step": 513
+    },
+    {
+      "epoch": 0.2067994367330517,
+      "grad_norm": 4.150283320138307,
+      "learning_rate": 4.830134532919863e-06,
+      "loss": 0.5835,
+      "step": 514
+    },
+    {
+      "epoch": 0.2072017702675518,
+      "grad_norm": 5.2365671694488904,
+      "learning_rate": 4.828860150739662e-06,
+      "loss": 0.697,
+      "step": 515
+    },
+    {
+      "epoch": 0.2076041038020519,
+      "grad_norm": 3.98085660862504,
+      "learning_rate": 4.827581175404311e-06,
+      "loss": 0.4872,
+      "step": 516
+    },
+    {
+      "epoch": 0.208006437336552,
+      "grad_norm": 4.574435055815776,
+      "learning_rate": 4.8262976094363016e-06,
+      "loss": 0.5925,
+      "step": 517
+    },
+    {
+      "epoch": 0.2084087708710521,
+      "grad_norm": 4.459834553682444,
+      "learning_rate": 4.825009455367181e-06,
+      "loss": 0.6426,
+      "step": 518
+    },
+    {
+      "epoch": 0.2088111044055522,
+      "grad_norm": 3.8557681901771037,
+      "learning_rate": 4.823716715737544e-06,
+      "loss": 0.5719,
+      "step": 519
+    },
+    {
+      "epoch": 0.20921343794005232,
+      "grad_norm": 4.776729631302629,
+      "learning_rate": 4.82241939309703e-06,
+      "loss": 0.6118,
+      "step": 520
+    },
+    {
+      "epoch": 0.2096157714745524,
+      "grad_norm": 4.701717061474551,
+      "learning_rate": 4.821117490004319e-06,
+      "loss": 0.4766,
+      "step": 521
+    },
+    {
+      "epoch": 0.2100181050090525,
+      "grad_norm": 4.486472816066567,
+      "learning_rate": 4.819811009027122e-06,
+      "loss": 0.6134,
+      "step": 522
+    },
+    {
+      "epoch": 0.2104204385435526,
+      "grad_norm": 4.784573013198364,
+      "learning_rate": 4.818499952742179e-06,
+      "loss": 0.5982,
+      "step": 523
+    },
+    {
+      "epoch": 0.2108227720780527,
+      "grad_norm": 4.953599282635229,
+      "learning_rate": 4.817184323735258e-06,
+      "loss": 0.5732,
+      "step": 524
+    },
+    {
+      "epoch": 0.2112251056125528,
+      "grad_norm": 4.808483380881421,
+      "learning_rate": 4.815864124601139e-06,
+      "loss": 0.6771,
+      "step": 525
+    },
+    {
+      "epoch": 0.2116274391470529,
+      "grad_norm": 4.024625165836511,
+      "learning_rate": 4.814539357943622e-06,
+      "loss": 0.6355,
+      "step": 526
+    },
+    {
+      "epoch": 0.212029772681553,
+      "grad_norm": 3.943315026550887,
+      "learning_rate": 4.813210026375513e-06,
+      "loss": 0.521,
+      "step": 527
+    },
+    {
+      "epoch": 0.2124321062160531,
+      "grad_norm": 3.463894710975418,
+      "learning_rate": 4.811876132518618e-06,
+      "loss": 0.4542,
+      "step": 528
+    },
+    {
+      "epoch": 0.21283443975055322,
+      "grad_norm": 4.513349087640872,
+      "learning_rate": 4.810537679003746e-06,
+      "loss": 0.6093,
+      "step": 529
+    },
+    {
+      "epoch": 0.2132367732850533,
+      "grad_norm": 4.649205790356446,
+      "learning_rate": 4.8091946684706956e-06,
+      "loss": 0.6726,
+      "step": 530
+    },
+    {
+      "epoch": 0.21363910681955342,
+      "grad_norm": 4.2037442337203315,
+      "learning_rate": 4.807847103568255e-06,
+      "loss": 0.6265,
+      "step": 531
+    },
+    {
+      "epoch": 0.2140414403540535,
+      "grad_norm": 4.893988522032171,
+      "learning_rate": 4.806494986954195e-06,
+      "loss": 0.7373,
+      "step": 532
+    },
+    {
+      "epoch": 0.21444377388855362,
+      "grad_norm": 4.455715486715626,
+      "learning_rate": 4.805138321295262e-06,
+      "loss": 0.5457,
+      "step": 533
+    },
+    {
+      "epoch": 0.2148461074230537,
+      "grad_norm": 4.649691607815862,
+      "learning_rate": 4.803777109267174e-06,
+      "loss": 0.5872,
+      "step": 534
+    },
+    {
+      "epoch": 0.2152484409575538,
+      "grad_norm": 4.606972921739361,
+      "learning_rate": 4.802411353554619e-06,
+      "loss": 0.6556,
+      "step": 535
+    },
+    {
+      "epoch": 0.21565077449205391,
+      "grad_norm": 4.262679120822924,
+      "learning_rate": 4.801041056851244e-06,
+      "loss": 0.5814,
+      "step": 536
+    },
+    {
+      "epoch": 0.216053108026554,
+      "grad_norm": 4.734408703763602,
+      "learning_rate": 4.7996662218596505e-06,
+      "loss": 0.5975,
+      "step": 537
+    },
+    {
+      "epoch": 0.21645544156105412,
+      "grad_norm": 4.0966775674750515,
+      "learning_rate": 4.798286851291395e-06,
+      "loss": 0.5373,
+      "step": 538
+    },
+    {
+      "epoch": 0.2168577750955542,
+      "grad_norm": 4.794723705583196,
+      "learning_rate": 4.796902947866976e-06,
+      "loss": 0.6411,
+      "step": 539
+    },
+    {
+      "epoch": 0.21726010863005432,
+      "grad_norm": 5.177706885118282,
+      "learning_rate": 4.795514514315833e-06,
+      "loss": 0.7129,
+      "step": 540
+    },
+    {
+      "epoch": 0.2176624421645544,
+      "grad_norm": 4.235334038988407,
+      "learning_rate": 4.794121553376341e-06,
+      "loss": 0.6368,
+      "step": 541
+    },
+    {
+      "epoch": 0.21806477569905452,
+      "grad_norm": 4.250039615427904,
+      "learning_rate": 4.792724067795802e-06,
+      "loss": 0.5897,
+      "step": 542
+    },
+    {
+      "epoch": 0.2184671092335546,
+      "grad_norm": 4.248095551159398,
+      "learning_rate": 4.7913220603304455e-06,
+      "loss": 0.6119,
+      "step": 543
+    },
+    {
+      "epoch": 0.21886944276805473,
+      "grad_norm": 4.7784528939759126,
+      "learning_rate": 4.789915533745415e-06,
+      "loss": 0.6428,
+      "step": 544
+    },
+    {
+      "epoch": 0.21927177630255482,
+      "grad_norm": 3.7433210513797857,
+      "learning_rate": 4.788504490814771e-06,
+      "loss": 0.4616,
+      "step": 545
+    },
+    {
+      "epoch": 0.21967410983705493,
+      "grad_norm": 4.477510756273104,
+      "learning_rate": 4.78708893432148e-06,
+      "loss": 0.5564,
+      "step": 546
+    },
+    {
+      "epoch": 0.22007644337155502,
+      "grad_norm": 4.159189423897508,
+      "learning_rate": 4.785668867057408e-06,
+      "loss": 0.5005,
+      "step": 547
+    },
+    {
+      "epoch": 0.2204787769060551,
+      "grad_norm": 4.801923291900065,
+      "learning_rate": 4.784244291823321e-06,
+      "loss": 0.6703,
+      "step": 548
+    },
+    {
+      "epoch": 0.22088111044055522,
+      "grad_norm": 4.9705885588356065,
+      "learning_rate": 4.782815211428875e-06,
+      "loss": 0.6892,
+      "step": 549
+    },
+    {
+      "epoch": 0.2212834439750553,
+      "grad_norm": 4.689425225044057,
+      "learning_rate": 4.7813816286926116e-06,
+      "loss": 0.6438,
+      "step": 550
+    },
+    {
+      "epoch": 0.22168577750955543,
+      "grad_norm": 3.9964370496914836,
+      "learning_rate": 4.7799435464419496e-06,
+      "loss": 0.5283,
+      "step": 551
+    },
+    {
+      "epoch": 0.22208811104405551,
+      "grad_norm": 3.9063906383285523,
+      "learning_rate": 4.778500967513186e-06,
+      "loss": 0.5477,
+      "step": 552
+    },
+    {
+      "epoch": 0.22249044457855563,
+      "grad_norm": 4.972674347366186,
+      "learning_rate": 4.777053894751484e-06,
+      "loss": 0.6762,
+      "step": 553
+    },
+    {
+      "epoch": 0.22289277811305572,
+      "grad_norm": 4.605207302352324,
+      "learning_rate": 4.77560233101087e-06,
+      "loss": 0.5823,
+      "step": 554
+    },
+    {
+      "epoch": 0.22329511164755583,
+      "grad_norm": 5.023738846112498,
+      "learning_rate": 4.774146279154231e-06,
+      "loss": 0.645,
+      "step": 555
+    },
+    {
+      "epoch": 0.22369744518205592,
+      "grad_norm": 4.21478405630217,
+      "learning_rate": 4.772685742053299e-06,
+      "loss": 0.6361,
+      "step": 556
+    },
+    {
+      "epoch": 0.22409977871655604,
+      "grad_norm": 4.443333059344436,
+      "learning_rate": 4.7712207225886605e-06,
+      "loss": 0.5808,
+      "step": 557
+    },
+    {
+      "epoch": 0.22450211225105612,
+      "grad_norm": 4.136335273179504,
+      "learning_rate": 4.7697512236497355e-06,
+      "loss": 0.4543,
+      "step": 558
+    },
+    {
+      "epoch": 0.22490444578555624,
+      "grad_norm": 5.228702785570883,
+      "learning_rate": 4.7682772481347835e-06,
+      "loss": 0.6677,
+      "step": 559
+    },
+    {
+      "epoch": 0.22530677932005633,
+      "grad_norm": 4.607023738968648,
+      "learning_rate": 4.76679879895089e-06,
+      "loss": 0.6972,
+      "step": 560
+    },
+    {
+      "epoch": 0.22570911285455642,
+      "grad_norm": 3.9623979182055646,
+      "learning_rate": 4.7653158790139655e-06,
+      "loss": 0.562,
+      "step": 561
+    },
+    {
+      "epoch": 0.22611144638905653,
+      "grad_norm": 4.785540664618103,
+      "learning_rate": 4.763828491248737e-06,
+      "loss": 0.6667,
+      "step": 562
+    },
+    {
+      "epoch": 0.22651377992355662,
+      "grad_norm": 4.626975611158776,
+      "learning_rate": 4.762336638588745e-06,
+      "loss": 0.5222,
+      "step": 563
+    },
+    {
+      "epoch": 0.22691611345805673,
+      "grad_norm": 4.148716543735861,
+      "learning_rate": 4.760840323976333e-06,
+      "loss": 0.5792,
+      "step": 564
+    },
+    {
+      "epoch": 0.22731844699255682,
+      "grad_norm": 3.7576481851339785,
+      "learning_rate": 4.759339550362647e-06,
+      "loss": 0.4171,
+      "step": 565
+    },
+    {
+      "epoch": 0.22772078052705694,
+      "grad_norm": 4.056873981914918,
+      "learning_rate": 4.757834320707629e-06,
+      "loss": 0.5823,
+      "step": 566
+    },
+    {
+      "epoch": 0.22812311406155703,
+      "grad_norm": 4.485956120607346,
+      "learning_rate": 4.7563246379800036e-06,
+      "loss": 0.6038,
+      "step": 567
+    },
+    {
+      "epoch": 0.22852544759605714,
+      "grad_norm": 4.095512060377993,
+      "learning_rate": 4.754810505157285e-06,
+      "loss": 0.5807,
+      "step": 568
+    },
+    {
+      "epoch": 0.22892778113055723,
+      "grad_norm": 4.316692633895088,
+      "learning_rate": 4.753291925225761e-06,
+      "loss": 0.5689,
+      "step": 569
+    },
+    {
+      "epoch": 0.22933011466505734,
+      "grad_norm": 4.0964431049127095,
+      "learning_rate": 4.751768901180488e-06,
+      "loss": 0.4724,
+      "step": 570
+    },
+    {
+      "epoch": 0.22973244819955743,
+      "grad_norm": 3.8505086953071683,
+      "learning_rate": 4.750241436025292e-06,
+      "loss": 0.5599,
+      "step": 571
+    },
+    {
+      "epoch": 0.23013478173405755,
+      "grad_norm": 4.97127913813276,
+      "learning_rate": 4.748709532772756e-06,
+      "loss": 0.7189,
+      "step": 572
+    },
+    {
+      "epoch": 0.23053711526855764,
+      "grad_norm": 4.376792824386442,
+      "learning_rate": 4.7471731944442154e-06,
+      "loss": 0.6149,
+      "step": 573
+    },
+    {
+      "epoch": 0.23093944880305772,
+      "grad_norm": 4.633754764886852,
+      "learning_rate": 4.745632424069755e-06,
+      "loss": 0.6118,
+      "step": 574
+    },
+    {
+      "epoch": 0.23134178233755784,
+      "grad_norm": 4.682951429282118,
+      "learning_rate": 4.744087224688197e-06,
+      "loss": 0.6089,
+      "step": 575
+    },
+    {
+      "epoch": 0.23174411587205793,
+      "grad_norm": 4.504821455509545,
+      "learning_rate": 4.742537599347101e-06,
+      "loss": 0.5747,
+      "step": 576
+    },
+    {
+      "epoch": 0.23214644940655804,
+      "grad_norm": 4.922801626757808,
+      "learning_rate": 4.740983551102759e-06,
+      "loss": 0.7267,
+      "step": 577
+    },
+    {
+      "epoch": 0.23254878294105813,
+      "grad_norm": 4.128104128218885,
+      "learning_rate": 4.73942508302018e-06,
+      "loss": 0.5344,
+      "step": 578
+    },
+    {
+      "epoch": 0.23295111647555825,
+      "grad_norm": 4.502862129743137,
+      "learning_rate": 4.7378621981730966e-06,
+      "loss": 0.4465,
+      "step": 579
+    },
+    {
+      "epoch": 0.23335345001005833,
+      "grad_norm": 4.476573377559871,
+      "learning_rate": 4.736294899643946e-06,
+      "loss": 0.5327,
+      "step": 580
+    },
+    {
+      "epoch": 0.23375578354455845,
+      "grad_norm": 4.525722183672701,
+      "learning_rate": 4.734723190523875e-06,
+      "loss": 0.6182,
+      "step": 581
+    },
+    {
+      "epoch": 0.23415811707905854,
+      "grad_norm": 4.4174724143850455,
+      "learning_rate": 4.7331470739127284e-06,
+      "loss": 0.582,
+      "step": 582
+    },
+    {
+      "epoch": 0.23456045061355865,
+      "grad_norm": 4.087588153474757,
+      "learning_rate": 4.731566552919042e-06,
+      "loss": 0.5626,
+      "step": 583
+    },
+    {
+      "epoch": 0.23496278414805874,
+      "grad_norm": 4.793168855698109,
+      "learning_rate": 4.7299816306600395e-06,
+      "loss": 0.712,
+      "step": 584
+    },
+    {
+      "epoch": 0.23536511768255883,
+      "grad_norm": 4.088862969423477,
+      "learning_rate": 4.728392310261628e-06,
+      "loss": 0.5867,
+      "step": 585
+    },
+    {
+      "epoch": 0.23576745121705894,
+      "grad_norm": 4.31442875545057,
+      "learning_rate": 4.726798594858382e-06,
+      "loss": 0.5255,
+      "step": 586
+    },
+    {
+      "epoch": 0.23616978475155903,
+      "grad_norm": 4.2694537336500495,
+      "learning_rate": 4.7252004875935506e-06,
+      "loss": 0.6822,
+      "step": 587
+    },
+    {
+      "epoch": 0.23657211828605915,
+      "grad_norm": 4.179846752454285,
+      "learning_rate": 4.723597991619043e-06,
+      "loss": 0.5433,
+      "step": 588
+    },
+    {
+      "epoch": 0.23697445182055923,
+      "grad_norm": 4.1764050880208,
+      "learning_rate": 4.721991110095422e-06,
+      "loss": 0.586,
+      "step": 589
+    },
+    {
+      "epoch": 0.23737678535505935,
+      "grad_norm": 3.699033328030597,
+      "learning_rate": 4.720379846191903e-06,
+      "loss": 0.4942,
+      "step": 590
+    },
+    {
+      "epoch": 0.23777911888955944,
+      "grad_norm": 4.798015125682991,
+      "learning_rate": 4.718764203086342e-06,
+      "loss": 0.7269,
+      "step": 591
+    },
+    {
+      "epoch": 0.23818145242405955,
+      "grad_norm": 4.974938507339555,
+      "learning_rate": 4.717144183965233e-06,
+      "loss": 0.6753,
+      "step": 592
+    },
+    {
+      "epoch": 0.23858378595855964,
+      "grad_norm": 5.08563513422606,
+      "learning_rate": 4.715519792023703e-06,
+      "loss": 0.6613,
+      "step": 593
+    },
+    {
+      "epoch": 0.23898611949305976,
+      "grad_norm": 4.6691045707182655,
+      "learning_rate": 4.7138910304655e-06,
+      "loss": 0.5635,
+      "step": 594
+    },
+    {
+      "epoch": 0.23938845302755984,
+      "grad_norm": 4.366016787794075,
+      "learning_rate": 4.712257902502992e-06,
+      "loss": 0.6603,
+      "step": 595
+    },
+    {
+      "epoch": 0.23979078656205996,
+      "grad_norm": 4.98238528037948,
+      "learning_rate": 4.710620411357157e-06,
+      "loss": 0.6241,
+      "step": 596
+    },
+    {
+      "epoch": 0.24019312009656005,
+      "grad_norm": 4.023463566077988,
+      "learning_rate": 4.7089785602575805e-06,
+      "loss": 0.5202,
+      "step": 597
+    },
+    {
+      "epoch": 0.24059545363106014,
+      "grad_norm": 4.040098710840788,
+      "learning_rate": 4.707332352442446e-06,
+      "loss": 0.5843,
+      "step": 598
+    },
+    {
+      "epoch": 0.24099778716556025,
+      "grad_norm": 4.151014624312927,
+      "learning_rate": 4.705681791158531e-06,
+      "loss": 0.6206,
+      "step": 599
+    },
+    {
+      "epoch": 0.24140012070006034,
+      "grad_norm": 4.0875213661566665,
+      "learning_rate": 4.704026879661196e-06,
+      "loss": 0.5614,
+      "step": 600
+    },
+    {
+      "epoch": 0.24180245423456045,
+      "grad_norm": 4.553877858262055,
+      "learning_rate": 4.7023676212143834e-06,
+      "loss": 0.6649,
+      "step": 601
+    },
+    {
+      "epoch": 0.24220478776906054,
+      "grad_norm": 4.674650194014434,
+      "learning_rate": 4.70070401909061e-06,
+      "loss": 0.606,
+      "step": 602
+    },
+    {
+      "epoch": 0.24260712130356066,
+      "grad_norm": 4.63353917939836,
+      "learning_rate": 4.699036076570959e-06,
+      "loss": 0.6668,
+      "step": 603
+    },
+    {
+      "epoch": 0.24300945483806075,
+      "grad_norm": 4.723870743698028,
+      "learning_rate": 4.697363796945072e-06,
+      "loss": 0.6728,
+      "step": 604
+    },
+    {
+      "epoch": 0.24341178837256086,
+      "grad_norm": 4.386624516452244,
+      "learning_rate": 4.6956871835111475e-06,
+      "loss": 0.5139,
+      "step": 605
+    },
+    {
+      "epoch": 0.24381412190706095,
+      "grad_norm": 4.244611479893601,
+      "learning_rate": 4.694006239575929e-06,
+      "loss": 0.609,
+      "step": 606
+    },
+    {
+      "epoch": 0.24421645544156106,
+      "grad_norm": 4.072589096296677,
+      "learning_rate": 4.692320968454702e-06,
+      "loss": 0.5657,
+      "step": 607
+    },
+    {
+      "epoch": 0.24461878897606115,
+      "grad_norm": 4.88910796894103,
+      "learning_rate": 4.690631373471287e-06,
+      "loss": 0.701,
+      "step": 608
+    },
+    {
+      "epoch": 0.24502112251056127,
+      "grad_norm": 4.686738823395271,
+      "learning_rate": 4.6889374579580315e-06,
+      "loss": 0.637,
+      "step": 609
+    },
+    {
+      "epoch": 0.24542345604506136,
+      "grad_norm": 4.905698840133413,
+      "learning_rate": 4.687239225255805e-06,
+      "loss": 0.6003,
+      "step": 610
+    },
+    {
+      "epoch": 0.24582578957956144,
+      "grad_norm": 4.338740241210239,
+      "learning_rate": 4.68553667871399e-06,
+      "loss": 0.5906,
+      "step": 611
+    },
+    {
+      "epoch": 0.24622812311406156,
+      "grad_norm": 4.119688719772878,
+      "learning_rate": 4.68382982169048e-06,
+      "loss": 0.4954,
+      "step": 612
+    },
+    {
+      "epoch": 0.24663045664856165,
+      "grad_norm": 3.96691281411716,
+      "learning_rate": 4.6821186575516665e-06,
+      "loss": 0.6126,
+      "step": 613
+    },
+    {
+      "epoch": 0.24703279018306176,
+      "grad_norm": 4.760926273551105,
+      "learning_rate": 4.680403189672439e-06,
+      "loss": 0.6613,
+      "step": 614
+    },
+    {
+      "epoch": 0.24743512371756185,
+      "grad_norm": 4.4384316824035075,
+      "learning_rate": 4.678683421436173e-06,
+      "loss": 0.5687,
+      "step": 615
+    },
+    {
+      "epoch": 0.24783745725206197,
+      "grad_norm": 3.931664070509481,
+      "learning_rate": 4.676959356234726e-06,
+      "loss": 0.5857,
+      "step": 616
+    },
+    {
+      "epoch": 0.24823979078656205,
+      "grad_norm": 5.119336252733733,
+      "learning_rate": 4.6752309974684315e-06,
+      "loss": 0.6312,
+      "step": 617
+    },
+    {
+      "epoch": 0.24864212432106217,
+      "grad_norm": 4.229793646214086,
+      "learning_rate": 4.6734983485460895e-06,
+      "loss": 0.5093,
+      "step": 618
+    },
+    {
+      "epoch": 0.24904445785556226,
+      "grad_norm": 4.9794376894703944,
+      "learning_rate": 4.671761412884962e-06,
+      "loss": 0.7092,
+      "step": 619
+    },
+    {
+      "epoch": 0.24944679139006237,
+      "grad_norm": 4.36204998657697,
+      "learning_rate": 4.670020193910766e-06,
+      "loss": 0.559,
+      "step": 620
+    },
+    {
+      "epoch": 0.24984912492456246,
+      "grad_norm": 4.297208174823204,
+      "learning_rate": 4.668274695057666e-06,
+      "loss": 0.4965,
+      "step": 621
+    },
+    {
+      "epoch": 0.25025145845906255,
+      "grad_norm": 4.235510447453734,
+      "learning_rate": 4.666524919768267e-06,
+      "loss": 0.5434,
+      "step": 622
+    },
+    {
+      "epoch": 0.25065379199356264,
+      "grad_norm": 4.297045599708538,
+      "learning_rate": 4.66477087149361e-06,
+      "loss": 0.5559,
+      "step": 623
+    },
+    {
+      "epoch": 0.2510561255280628,
+      "grad_norm": 5.638228122643803,
+      "learning_rate": 4.663012553693161e-06,
+      "loss": 0.6393,
+      "step": 624
+    },
+    {
+      "epoch": 0.25145845906256287,
+      "grad_norm": 4.307369262356312,
+      "learning_rate": 4.661249969834809e-06,
+      "loss": 0.5516,
+      "step": 625
+    },
+    {
+      "epoch": 0.25186079259706295,
+      "grad_norm": 4.348008968073508,
+      "learning_rate": 4.659483123394855e-06,
+      "loss": 0.5858,
+      "step": 626
+    },
+    {
+      "epoch": 0.25226312613156304,
+      "grad_norm": 4.245775793874246,
+      "learning_rate": 4.657712017858011e-06,
+      "loss": 0.5985,
+      "step": 627
+    },
+    {
+      "epoch": 0.2526654596660632,
+      "grad_norm": 4.27241102613026,
+      "learning_rate": 4.6559366567173824e-06,
+      "loss": 0.5247,
+      "step": 628
+    },
+    {
+      "epoch": 0.2530677932005633,
+      "grad_norm": 3.9143192748019384,
+      "learning_rate": 4.6541570434744735e-06,
+      "loss": 0.4936,
+      "step": 629
+    },
+    {
+      "epoch": 0.25347012673506336,
+      "grad_norm": 4.550429412487192,
+      "learning_rate": 4.6523731816391725e-06,
+      "loss": 0.7499,
+      "step": 630
+    },
+    {
+      "epoch": 0.25387246026956345,
+      "grad_norm": 4.202697728155226,
+      "learning_rate": 4.650585074729747e-06,
+      "loss": 0.5184,
+      "step": 631
+    },
+    {
+      "epoch": 0.2542747938040636,
+      "grad_norm": 4.083559252407811,
+      "learning_rate": 4.648792726272838e-06,
+      "loss": 0.5242,
+      "step": 632
+    },
+    {
+      "epoch": 0.2546771273385637,
+      "grad_norm": 3.9861255665329267,
+      "learning_rate": 4.646996139803452e-06,
+      "loss": 0.5269,
+      "step": 633
+    },
+    {
+      "epoch": 0.25507946087306377,
+      "grad_norm": 4.559490860669659,
+      "learning_rate": 4.645195318864951e-06,
+      "loss": 0.6372,
+      "step": 634
+    },
+    {
+      "epoch": 0.25548179440756386,
+      "grad_norm": 4.394457869705605,
+      "learning_rate": 4.643390267009054e-06,
+      "loss": 0.5105,
+      "step": 635
+    },
+    {
+      "epoch": 0.25588412794206394,
+      "grad_norm": 5.018376167864149,
+      "learning_rate": 4.641580987795821e-06,
+      "loss": 0.6628,
+      "step": 636
+    },
+    {
+      "epoch": 0.2562864614765641,
+      "grad_norm": 3.7893948824739465,
+      "learning_rate": 4.639767484793648e-06,
+      "loss": 0.508,
+      "step": 637
+    },
+    {
+      "epoch": 0.2566887950110642,
+      "grad_norm": 4.596240270523842,
+      "learning_rate": 4.637949761579266e-06,
+      "loss": 0.6529,
+      "step": 638
+    },
+    {
+      "epoch": 0.25709112854556426,
+      "grad_norm": 5.140583964722019,
+      "learning_rate": 4.636127821737726e-06,
+      "loss": 0.6489,
+      "step": 639
+    },
+    {
+      "epoch": 0.25749346208006435,
+      "grad_norm": 3.8394758085160823,
+      "learning_rate": 4.634301668862397e-06,
+      "loss": 0.4736,
+      "step": 640
+    },
+    {
+      "epoch": 0.2578957956145645,
+      "grad_norm": 4.38356346084736,
+      "learning_rate": 4.632471306554955e-06,
+      "loss": 0.5353,
+      "step": 641
+    },
+    {
+      "epoch": 0.2582981291490646,
+      "grad_norm": 3.8889360472386185,
+      "learning_rate": 4.630636738425381e-06,
+      "loss": 0.5511,
+      "step": 642
+    },
+    {
+      "epoch": 0.25870046268356467,
+      "grad_norm": 4.817925477512709,
+      "learning_rate": 4.62879796809195e-06,
+      "loss": 0.6963,
+      "step": 643
+    },
+    {
+      "epoch": 0.25910279621806476,
+      "grad_norm": 4.742191877421797,
+      "learning_rate": 4.626954999181224e-06,
+      "loss": 0.7752,
+      "step": 644
+    },
+    {
+      "epoch": 0.2595051297525649,
+      "grad_norm": 4.245951623762571,
+      "learning_rate": 4.625107835328048e-06,
+      "loss": 0.5754,
+      "step": 645
+    },
+    {
+      "epoch": 0.259907463287065,
+      "grad_norm": 4.836575633325676,
+      "learning_rate": 4.623256480175537e-06,
+      "loss": 0.6639,
+      "step": 646
+    },
+    {
+      "epoch": 0.2603097968215651,
+      "grad_norm": 3.806176284035789,
+      "learning_rate": 4.621400937375075e-06,
+      "loss": 0.5292,
+      "step": 647
+    },
+    {
+      "epoch": 0.26071213035606516,
+      "grad_norm": 4.025115976057339,
+      "learning_rate": 4.619541210586307e-06,
+      "loss": 0.4815,
+      "step": 648
+    },
+    {
+      "epoch": 0.26111446389056525,
+      "grad_norm": 4.1347392050793825,
+      "learning_rate": 4.617677303477124e-06,
+      "loss": 0.5681,
+      "step": 649
+    },
+    {
+      "epoch": 0.2615167974250654,
+      "grad_norm": 4.371554358928095,
+      "learning_rate": 4.615809219723667e-06,
+      "loss": 0.592,
+      "step": 650
+    },
+    {
+      "epoch": 0.2619191309595655,
+      "grad_norm": 4.976927834726334,
+      "learning_rate": 4.613936963010314e-06,
+      "loss": 0.7898,
+      "step": 651
+    },
+    {
+      "epoch": 0.26232146449406557,
+      "grad_norm": 4.80195477309382,
+      "learning_rate": 4.612060537029671e-06,
+      "loss": 0.6565,
+      "step": 652
+    },
+    {
+      "epoch": 0.26272379802856566,
+      "grad_norm": 4.2203256747490805,
+      "learning_rate": 4.610179945482568e-06,
+      "loss": 0.5597,
+      "step": 653
+    },
+    {
+      "epoch": 0.2631261315630658,
+      "grad_norm": 4.3119037789418035,
+      "learning_rate": 4.608295192078051e-06,
+      "loss": 0.5773,
+      "step": 654
+    },
+    {
+      "epoch": 0.2635284650975659,
+      "grad_norm": 3.938767904065124,
+      "learning_rate": 4.606406280533373e-06,
+      "loss": 0.4691,
+      "step": 655
+    },
+    {
+      "epoch": 0.263930798632066,
+      "grad_norm": 3.873313542196617,
+      "learning_rate": 4.6045132145739914e-06,
+      "loss": 0.5446,
+      "step": 656
+    },
+    {
+      "epoch": 0.26433313216656606,
+      "grad_norm": 4.720535373828625,
+      "learning_rate": 4.602615997933552e-06,
+      "loss": 0.6451,
+      "step": 657
+    },
+    {
+      "epoch": 0.2647354657010662,
+      "grad_norm": 5.319338434241136,
+      "learning_rate": 4.600714634353893e-06,
+      "loss": 0.6786,
+      "step": 658
+    },
+    {
+      "epoch": 0.2651377992355663,
+      "grad_norm": 3.986441538874652,
+      "learning_rate": 4.598809127585026e-06,
+      "loss": 0.5815,
+      "step": 659
+    },
+    {
+      "epoch": 0.2655401327700664,
+      "grad_norm": 4.190229942704305,
+      "learning_rate": 4.596899481385137e-06,
+      "loss": 0.5933,
+      "step": 660
+    },
+    {
+      "epoch": 0.26594246630456647,
+      "grad_norm": 4.211284711325467,
+      "learning_rate": 4.5949856995205745e-06,
+      "loss": 0.6062,
+      "step": 661
+    },
+    {
+      "epoch": 0.26634479983906656,
+      "grad_norm": 5.022372015405014,
+      "learning_rate": 4.593067785765846e-06,
+      "loss": 0.6632,
+      "step": 662
+    },
+    {
+      "epoch": 0.2667471333735667,
+      "grad_norm": 3.9365356585193148,
+      "learning_rate": 4.5911457439036075e-06,
+      "loss": 0.5324,
+      "step": 663
+    },
+    {
+      "epoch": 0.2671494669080668,
+      "grad_norm": 4.891303145874928,
+      "learning_rate": 4.589219577724654e-06,
+      "loss": 0.7504,
+      "step": 664
+    },
+    {
+      "epoch": 0.2675518004425669,
+      "grad_norm": 3.855940902006332,
+      "learning_rate": 4.5872892910279185e-06,
+      "loss": 0.5476,
+      "step": 665
+    },
+    {
+      "epoch": 0.26795413397706697,
+      "grad_norm": 4.37750980680622,
+      "learning_rate": 4.58535488762046e-06,
+      "loss": 0.5902,
+      "step": 666
+    },
+    {
+      "epoch": 0.2683564675115671,
+      "grad_norm": 4.450877429194335,
+      "learning_rate": 4.583416371317454e-06,
+      "loss": 0.6345,
+      "step": 667
+    },
+    {
+      "epoch": 0.2687588010460672,
+      "grad_norm": 4.162519879691083,
+      "learning_rate": 4.581473745942191e-06,
+      "loss": 0.444,
+      "step": 668
+    },
+    {
+      "epoch": 0.2691611345805673,
+      "grad_norm": 4.2236499578643025,
+      "learning_rate": 4.579527015326065e-06,
+      "loss": 0.492,
+      "step": 669
+    },
+    {
+      "epoch": 0.2695634681150674,
+      "grad_norm": 4.482147273633291,
+      "learning_rate": 4.5775761833085645e-06,
+      "loss": 0.5648,
+      "step": 670
+    },
+    {
+      "epoch": 0.2699658016495675,
+      "grad_norm": 4.760060597047131,
+      "learning_rate": 4.57562125373727e-06,
+      "loss": 0.6674,
+      "step": 671
+    },
+    {
+      "epoch": 0.2703681351840676,
+      "grad_norm": 4.362389446472747,
+      "learning_rate": 4.573662230467844e-06,
+      "loss": 0.577,
+      "step": 672
+    },
+    {
+      "epoch": 0.2707704687185677,
+      "grad_norm": 3.8159558241349987,
+      "learning_rate": 4.5716991173640165e-06,
+      "loss": 0.5185,
+      "step": 673
+    },
+    {
+      "epoch": 0.2711728022530678,
+      "grad_norm": 4.005051687655319,
+      "learning_rate": 4.5697319182975944e-06,
+      "loss": 0.5922,
+      "step": 674
+    },
+    {
+      "epoch": 0.27157513578756787,
+      "grad_norm": 4.667139467518304,
+      "learning_rate": 4.567760637148432e-06,
+      "loss": 0.6032,
+      "step": 675
+    },
+    {
+      "epoch": 0.271977469322068,
+      "grad_norm": 3.9606776473436405,
+      "learning_rate": 4.5657852778044435e-06,
+      "loss": 0.5174,
+      "step": 676
+    },
+    {
+      "epoch": 0.2723798028565681,
+      "grad_norm": 4.322780347285095,
+      "learning_rate": 4.5638058441615815e-06,
+      "loss": 0.6089,
+      "step": 677
+    },
+    {
+      "epoch": 0.2727821363910682,
+      "grad_norm": 4.433317399280547,
+      "learning_rate": 4.561822340123836e-06,
+      "loss": 0.6047,
+      "step": 678
+    },
+    {
+      "epoch": 0.2731844699255683,
+      "grad_norm": 4.552805703827102,
+      "learning_rate": 4.559834769603224e-06,
+      "loss": 0.6281,
+      "step": 679
+    },
+    {
+      "epoch": 0.2735868034600684,
+      "grad_norm": 5.128182512733822,
+      "learning_rate": 4.557843136519784e-06,
+      "loss": 0.6697,
+      "step": 680
+    },
+    {
+      "epoch": 0.2739891369945685,
+      "grad_norm": 4.063920951512369,
+      "learning_rate": 4.555847444801565e-06,
+      "loss": 0.5885,
+      "step": 681
+    },
+    {
+      "epoch": 0.2743914705290686,
+      "grad_norm": 4.908841427141714,
+      "learning_rate": 4.5538476983846245e-06,
+      "loss": 0.6771,
+      "step": 682
+    },
+    {
+      "epoch": 0.2747938040635687,
+      "grad_norm": 4.405724720337232,
+      "learning_rate": 4.551843901213012e-06,
+      "loss": 0.6479,
+      "step": 683
+    },
+    {
+      "epoch": 0.2751961375980688,
+      "grad_norm": 3.945129587052982,
+      "learning_rate": 4.549836057238769e-06,
+      "loss": 0.6039,
+      "step": 684
+    },
+    {
+      "epoch": 0.2755984711325689,
+      "grad_norm": 4.96376428039534,
+      "learning_rate": 4.547824170421921e-06,
+      "loss": 0.7191,
+      "step": 685
+    },
+    {
+      "epoch": 0.276000804667069,
+      "grad_norm": 4.378595779287015,
+      "learning_rate": 4.54580824473046e-06,
+      "loss": 0.5549,
+      "step": 686
+    },
+    {
+      "epoch": 0.2764031382015691,
+      "grad_norm": 4.262426880188478,
+      "learning_rate": 4.54378828414035e-06,
+      "loss": 0.5024,
+      "step": 687
+    },
+    {
+      "epoch": 0.2768054717360692,
+      "grad_norm": 4.426695285396216,
+      "learning_rate": 4.541764292635512e-06,
+      "loss": 0.7052,
+      "step": 688
+    },
+    {
+      "epoch": 0.2772078052705693,
+      "grad_norm": 4.473404410086587,
+      "learning_rate": 4.5397362742078145e-06,
+      "loss": 0.5557,
+      "step": 689
+    },
+    {
+      "epoch": 0.2776101388050694,
+      "grad_norm": 4.052844689460026,
+      "learning_rate": 4.537704232857069e-06,
+      "loss": 0.4541,
+      "step": 690
+    },
+    {
+      "epoch": 0.2780124723395695,
+      "grad_norm": 3.89050386632389,
+      "learning_rate": 4.5356681725910255e-06,
+      "loss": 0.549,
+      "step": 691
+    },
+    {
+      "epoch": 0.2784148058740696,
+      "grad_norm": 4.349538143222671,
+      "learning_rate": 4.5336280974253535e-06,
+      "loss": 0.6076,
+      "step": 692
+    },
+    {
+      "epoch": 0.2788171394085697,
+      "grad_norm": 4.361849160433631,
+      "learning_rate": 4.5315840113836454e-06,
+      "loss": 0.6524,
+      "step": 693
+    },
+    {
+      "epoch": 0.2792194729430698,
+      "grad_norm": 4.617234754063093,
+      "learning_rate": 4.529535918497403e-06,
+      "loss": 0.5616,
+      "step": 694
+    },
+    {
+      "epoch": 0.2796218064775699,
+      "grad_norm": 4.09023083083644,
+      "learning_rate": 4.5274838228060326e-06,
+      "loss": 0.5407,
+      "step": 695
+    },
+    {
+      "epoch": 0.28002414001207,
+      "grad_norm": 3.7157295921866718,
+      "learning_rate": 4.525427728356832e-06,
+      "loss": 0.5361,
+      "step": 696
+    },
+    {
+      "epoch": 0.28042647354657013,
+      "grad_norm": 4.457227461962157,
+      "learning_rate": 4.523367639204985e-06,
+      "loss": 0.5593,
+      "step": 697
+    },
+    {
+      "epoch": 0.2808288070810702,
+      "grad_norm": 4.348610370546784,
+      "learning_rate": 4.52130355941356e-06,
+      "loss": 0.5121,
+      "step": 698
+    },
+    {
+      "epoch": 0.2812311406155703,
+      "grad_norm": 3.9173059552199194,
+      "learning_rate": 4.519235493053491e-06,
+      "loss": 0.4927,
+      "step": 699
+    },
+    {
+      "epoch": 0.2816334741500704,
+      "grad_norm": 4.208431911956715,
+      "learning_rate": 4.517163444203575e-06,
+      "loss": 0.5504,
+      "step": 700
+    },
+    {
+      "epoch": 0.2820358076845705,
+      "grad_norm": 4.404532536673193,
+      "learning_rate": 4.515087416950464e-06,
+      "loss": 0.6057,
+      "step": 701
+    },
+    {
+      "epoch": 0.2824381412190706,
+      "grad_norm": 4.163697909062566,
+      "learning_rate": 4.513007415388659e-06,
+      "loss": 0.4682,
+      "step": 702
+    },
+    {
+      "epoch": 0.2828404747535707,
+      "grad_norm": 5.152927142247097,
+      "learning_rate": 4.510923443620494e-06,
+      "loss": 0.6601,
+      "step": 703
+    },
+    {
+      "epoch": 0.2832428082880708,
+      "grad_norm": 4.544361331270338,
+      "learning_rate": 4.50883550575614e-06,
+      "loss": 0.5968,
+      "step": 704
+    },
+    {
+      "epoch": 0.2836451418225709,
+      "grad_norm": 4.134394917706537,
+      "learning_rate": 4.5067436059135836e-06,
+      "loss": 0.6647,
+      "step": 705
+    },
+    {
+      "epoch": 0.28404747535707103,
+      "grad_norm": 4.734085611831448,
+      "learning_rate": 4.504647748218633e-06,
+      "loss": 0.601,
+      "step": 706
+    },
+    {
+      "epoch": 0.2844498088915711,
+      "grad_norm": 4.681267518422373,
+      "learning_rate": 4.502547936804894e-06,
+      "loss": 0.5539,
+      "step": 707
+    },
+    {
+      "epoch": 0.2848521424260712,
+      "grad_norm": 4.325088947732379,
+      "learning_rate": 4.500444175813776e-06,
+      "loss": 0.5872,
+      "step": 708
+    },
+    {
+      "epoch": 0.2852544759605713,
+      "grad_norm": 3.850534498836246,
+      "learning_rate": 4.4983364693944775e-06,
+      "loss": 0.5061,
+      "step": 709
+    },
+    {
+      "epoch": 0.28565680949507144,
+      "grad_norm": 4.373584617561936,
+      "learning_rate": 4.496224821703977e-06,
+      "loss": 0.6693,
+      "step": 710
+    },
+    {
+      "epoch": 0.2860591430295715,
+      "grad_norm": 4.377477636040013,
+      "learning_rate": 4.494109236907026e-06,
+      "loss": 0.5966,
+      "step": 711
+    },
+    {
+      "epoch": 0.2864614765640716,
+      "grad_norm": 4.271381874115667,
+      "learning_rate": 4.491989719176142e-06,
+      "loss": 0.6221,
+      "step": 712
+    },
+    {
+      "epoch": 0.2868638100985717,
+      "grad_norm": 3.7259627257148202,
+      "learning_rate": 4.489866272691599e-06,
+      "loss": 0.4313,
+      "step": 713
+    },
+    {
+      "epoch": 0.2872661436330718,
+      "grad_norm": 4.0336320558123155,
+      "learning_rate": 4.48773890164142e-06,
+      "loss": 0.5305,
+      "step": 714
+    },
+    {
+      "epoch": 0.28766847716757193,
+      "grad_norm": 4.585992205034134,
+      "learning_rate": 4.485607610221367e-06,
+      "loss": 0.7636,
+      "step": 715
+    },
+    {
+      "epoch": 0.288070810702072,
+      "grad_norm": 4.67820441036977,
+      "learning_rate": 4.4834724026349376e-06,
+      "loss": 0.5866,
+      "step": 716
+    },
+    {
+      "epoch": 0.2884731442365721,
+      "grad_norm": 5.012556667522281,
+      "learning_rate": 4.4813332830933484e-06,
+      "loss": 0.6498,
+      "step": 717
+    },
+    {
+      "epoch": 0.2888754777710722,
+      "grad_norm": 4.19440277086636,
+      "learning_rate": 4.4791902558155345e-06,
+      "loss": 0.5434,
+      "step": 718
+    },
+    {
+      "epoch": 0.28927781130557234,
+      "grad_norm": 4.330716534353544,
+      "learning_rate": 4.477043325028137e-06,
+      "loss": 0.5507,
+      "step": 719
+    },
+    {
+      "epoch": 0.28968014484007243,
+      "grad_norm": 4.097976140286037,
+      "learning_rate": 4.474892494965495e-06,
+      "loss": 0.4591,
+      "step": 720
+    },
+    {
+      "epoch": 0.2900824783745725,
+      "grad_norm": 4.132010732456553,
+      "learning_rate": 4.472737769869642e-06,
+      "loss": 0.5217,
+      "step": 721
+    },
+    {
+      "epoch": 0.2904848119090726,
+      "grad_norm": 3.7815121295908343,
+      "learning_rate": 4.470579153990288e-06,
+      "loss": 0.4932,
+      "step": 722
+    },
+    {
+      "epoch": 0.29088714544357275,
+      "grad_norm": 3.724905186650606,
+      "learning_rate": 4.468416651584822e-06,
+      "loss": 0.5448,
+      "step": 723
+    },
+    {
+      "epoch": 0.29128947897807284,
+      "grad_norm": 3.82443773314746,
+      "learning_rate": 4.4662502669182935e-06,
+      "loss": 0.5298,
+      "step": 724
+    },
+    {
+      "epoch": 0.2916918125125729,
+      "grad_norm": 4.343371987708737,
+      "learning_rate": 4.464080004263411e-06,
+      "loss": 0.606,
+      "step": 725
+    },
+    {
+      "epoch": 0.292094146047073,
+      "grad_norm": 4.009570342579401,
+      "learning_rate": 4.461905867900535e-06,
+      "loss": 0.5916,
+      "step": 726
+    },
+    {
+      "epoch": 0.2924964795815731,
+      "grad_norm": 4.148496022814315,
+      "learning_rate": 4.459727862117658e-06,
+      "loss": 0.5155,
+      "step": 727
+    },
+    {
+      "epoch": 0.29289881311607324,
+      "grad_norm": 4.568599931304069,
+      "learning_rate": 4.457545991210412e-06,
+      "loss": 0.555,
+      "step": 728
+    },
+    {
+      "epoch": 0.29330114665057333,
+      "grad_norm": 4.026645287125634,
+      "learning_rate": 4.455360259482047e-06,
+      "loss": 0.5265,
+      "step": 729
+    },
+    {
+      "epoch": 0.2937034801850734,
+      "grad_norm": 4.4466778657917265,
+      "learning_rate": 4.4531706712434305e-06,
+      "loss": 0.6318,
+      "step": 730
+    },
+    {
+      "epoch": 0.2941058137195735,
+      "grad_norm": 4.444852619715758,
+      "learning_rate": 4.450977230813035e-06,
+      "loss": 0.671,
+      "step": 731
+    },
+    {
+      "epoch": 0.29450814725407365,
+      "grad_norm": 5.136550532204147,
+      "learning_rate": 4.44877994251693e-06,
+      "loss": 0.611,
+      "step": 732
+    },
+    {
+      "epoch": 0.29491048078857374,
+      "grad_norm": 4.472272090985854,
+      "learning_rate": 4.446578810688774e-06,
+      "loss": 0.595,
+      "step": 733
+    },
+    {
+      "epoch": 0.2953128143230738,
+      "grad_norm": 4.755288406312762,
+      "learning_rate": 4.444373839669808e-06,
+      "loss": 0.801,
+      "step": 734
+    },
+    {
+      "epoch": 0.2957151478575739,
+      "grad_norm": 4.964978523542175,
+      "learning_rate": 4.442165033808843e-06,
+      "loss": 0.6712,
+      "step": 735
+    },
+    {
+      "epoch": 0.29611748139207406,
+      "grad_norm": 4.590633696444609,
+      "learning_rate": 4.439952397462254e-06,
+      "loss": 0.6214,
+      "step": 736
+    },
+    {
+      "epoch": 0.29651981492657414,
+      "grad_norm": 4.05753322012328,
+      "learning_rate": 4.43773593499397e-06,
+      "loss": 0.4477,
+      "step": 737
+    },
+    {
+      "epoch": 0.29692214846107423,
+      "grad_norm": 4.354711360779599,
+      "learning_rate": 4.435515650775468e-06,
+      "loss": 0.5618,
+      "step": 738
+    },
+    {
+      "epoch": 0.2973244819955743,
+      "grad_norm": 4.543128550216277,
+      "learning_rate": 4.433291549185761e-06,
+      "loss": 0.4807,
+      "step": 739
+    },
+    {
+      "epoch": 0.2977268155300744,
+      "grad_norm": 3.9444140759236737,
+      "learning_rate": 4.431063634611392e-06,
+      "loss": 0.5393,
+      "step": 740
+    },
+    {
+      "epoch": 0.29812914906457455,
+      "grad_norm": 4.516456335270593,
+      "learning_rate": 4.428831911446422e-06,
+      "loss": 0.5839,
+      "step": 741
+    },
+    {
+      "epoch": 0.29853148259907464,
+      "grad_norm": 4.4607761036583815,
+      "learning_rate": 4.426596384092426e-06,
+      "loss": 0.6538,
+      "step": 742
+    },
+    {
+      "epoch": 0.2989338161335747,
+      "grad_norm": 6.196021389574422,
+      "learning_rate": 4.424357056958483e-06,
+      "loss": 0.6484,
+      "step": 743
+    },
+    {
+      "epoch": 0.2993361496680748,
+      "grad_norm": 4.251042887568661,
+      "learning_rate": 4.422113934461161e-06,
+      "loss": 0.5523,
+      "step": 744
+    },
+    {
+      "epoch": 0.29973848320257496,
+      "grad_norm": 4.697763472346534,
+      "learning_rate": 4.4198670210245186e-06,
+      "loss": 0.6053,
+      "step": 745
+    },
+    {
+      "epoch": 0.30014081673707504,
+      "grad_norm": 4.665389980976664,
+      "learning_rate": 4.41761632108009e-06,
+      "loss": 0.6618,
+      "step": 746
+    },
+    {
+      "epoch": 0.30054315027157513,
+      "grad_norm": 4.2980125151698525,
+      "learning_rate": 4.415361839066874e-06,
+      "loss": 0.4863,
+      "step": 747
+    },
+    {
+      "epoch": 0.3009454838060752,
+      "grad_norm": 4.25422933800907,
+      "learning_rate": 4.413103579431335e-06,
+      "loss": 0.6285,
+      "step": 748
+    },
+    {
+      "epoch": 0.30134781734057536,
+      "grad_norm": 4.408647829220997,
+      "learning_rate": 4.410841546627383e-06,
+      "loss": 0.6214,
+      "step": 749
+    },
+    {
+      "epoch": 0.30175015087507545,
+      "grad_norm": 3.5964308822983626,
+      "learning_rate": 4.408575745116371e-06,
+      "loss": 0.4709,
+      "step": 750
+    },
+    {
+      "epoch": 0.30215248440957554,
+      "grad_norm": 4.177190728366978,
+      "learning_rate": 4.406306179367086e-06,
+      "loss": 0.6584,
+      "step": 751
+    },
+    {
+      "epoch": 0.3025548179440756,
+      "grad_norm": 3.9436659867851,
+      "learning_rate": 4.404032853855736e-06,
+      "loss": 0.5804,
+      "step": 752
+    },
+    {
+      "epoch": 0.3029571514785757,
+      "grad_norm": 4.5139405502647465,
+      "learning_rate": 4.40175577306595e-06,
+      "loss": 0.6131,
+      "step": 753
+    },
+    {
+      "epoch": 0.30335948501307586,
+      "grad_norm": 4.656972033870911,
+      "learning_rate": 4.399474941488756e-06,
+      "loss": 0.5378,
+      "step": 754
+    },
+    {
+      "epoch": 0.30376181854757595,
+      "grad_norm": 4.2754391925078785,
+      "learning_rate": 4.397190363622588e-06,
+      "loss": 0.5117,
+      "step": 755
+    },
+    {
+      "epoch": 0.30416415208207603,
+      "grad_norm": 4.057091286795055,
+      "learning_rate": 4.3949020439732594e-06,
+      "loss": 0.4969,
+      "step": 756
+    },
+    {
+      "epoch": 0.3045664856165761,
+      "grad_norm": 3.914038182134245,
+      "learning_rate": 4.392609987053972e-06,
+      "loss": 0.4484,
+      "step": 757
+    },
+    {
+      "epoch": 0.30496881915107626,
+      "grad_norm": 4.364733148673897,
+      "learning_rate": 4.390314197385292e-06,
+      "loss": 0.662,
+      "step": 758
+    },
+    {
+      "epoch": 0.30537115268557635,
+      "grad_norm": 4.866819672714999,
+      "learning_rate": 4.388014679495154e-06,
+      "loss": 0.6832,
+      "step": 759
+    },
+    {
+      "epoch": 0.30577348622007644,
+      "grad_norm": 4.563239673265389,
+      "learning_rate": 4.385711437918839e-06,
+      "loss": 0.633,
+      "step": 760
+    },
+    {
+      "epoch": 0.30617581975457653,
+      "grad_norm": 3.952293727810083,
+      "learning_rate": 4.383404477198976e-06,
+      "loss": 0.4319,
+      "step": 761
+    },
+    {
+      "epoch": 0.30657815328907667,
+      "grad_norm": 3.8982970475676866,
+      "learning_rate": 4.38109380188553e-06,
+      "loss": 0.5412,
+      "step": 762
+    },
+    {
+      "epoch": 0.30698048682357676,
+      "grad_norm": 4.767355411844817,
+      "learning_rate": 4.3787794165357875e-06,
+      "loss": 0.6688,
+      "step": 763
+    },
+    {
+      "epoch": 0.30738282035807685,
+      "grad_norm": 5.490594878274815,
+      "learning_rate": 4.3764613257143585e-06,
+      "loss": 0.8315,
+      "step": 764
+    },
+    {
+      "epoch": 0.30778515389257693,
+      "grad_norm": 4.4753569929038965,
+      "learning_rate": 4.374139533993157e-06,
+      "loss": 0.5577,
+      "step": 765
+    },
+    {
+      "epoch": 0.308187487427077,
+      "grad_norm": 4.234373998218113,
+      "learning_rate": 4.371814045951396e-06,
+      "loss": 0.6195,
+      "step": 766
+    },
+    {
+      "epoch": 0.30858982096157717,
+      "grad_norm": 3.8789979804624277,
+      "learning_rate": 4.369484866175581e-06,
+      "loss": 0.5812,
+      "step": 767
+    },
+    {
+      "epoch": 0.30899215449607725,
+      "grad_norm": 4.3774657256617635,
+      "learning_rate": 4.367151999259498e-06,
+      "loss": 0.5334,
+      "step": 768
+    },
+    {
+      "epoch": 0.30939448803057734,
+      "grad_norm": 4.413468965671534,
+      "learning_rate": 4.364815449804204e-06,
+      "loss": 0.6043,
+      "step": 769
+    },
+    {
+      "epoch": 0.30979682156507743,
+      "grad_norm": 4.977248443532888,
+      "learning_rate": 4.362475222418019e-06,
+      "loss": 0.7427,
+      "step": 770
+    },
+    {
+      "epoch": 0.3101991550995776,
+      "grad_norm": 4.449181903975735,
+      "learning_rate": 4.360131321716518e-06,
+      "loss": 0.5336,
+      "step": 771
+    },
+    {
+      "epoch": 0.31060148863407766,
+      "grad_norm": 4.203059807958007,
+      "learning_rate": 4.357783752322522e-06,
+      "loss": 0.5362,
+      "step": 772
+    },
+    {
+      "epoch": 0.31100382216857775,
+      "grad_norm": 3.8919997316821004,
+      "learning_rate": 4.355432518866084e-06,
+      "loss": 0.5202,
+      "step": 773
+    },
+    {
+      "epoch": 0.31140615570307784,
+      "grad_norm": 4.125644261476737,
+      "learning_rate": 4.353077625984484e-06,
+      "loss": 0.5136,
+      "step": 774
+    },
+    {
+      "epoch": 0.311808489237578,
+      "grad_norm": 3.7230845852814687,
+      "learning_rate": 4.3507190783222245e-06,
+      "loss": 0.4731,
+      "step": 775
+    },
+    {
+      "epoch": 0.31221082277207807,
+      "grad_norm": 4.619811557245368,
+      "learning_rate": 4.348356880531011e-06,
+      "loss": 0.698,
+      "step": 776
+    },
+    {
+      "epoch": 0.31261315630657815,
+      "grad_norm": 4.064918048601741,
+      "learning_rate": 4.345991037269748e-06,
+      "loss": 0.5594,
+      "step": 777
+    },
+    {
+      "epoch": 0.31301548984107824,
+      "grad_norm": 4.28030910141664,
+      "learning_rate": 4.343621553204533e-06,
+      "loss": 0.6272,
+      "step": 778
+    },
+    {
+      "epoch": 0.31341782337557833,
+      "grad_norm": 4.3485146730819,
+      "learning_rate": 4.341248433008645e-06,
+      "loss": 0.6078,
+      "step": 779
+    },
+    {
+      "epoch": 0.3138201569100785,
+      "grad_norm": 4.211683329853448,
+      "learning_rate": 4.338871681362528e-06,
+      "loss": 0.5247,
+      "step": 780
+    },
+    {
+      "epoch": 0.31422249044457856,
+      "grad_norm": 3.8752448583148453,
+      "learning_rate": 4.3364913029537946e-06,
+      "loss": 0.5274,
+      "step": 781
+    },
+    {
+      "epoch": 0.31462482397907865,
+      "grad_norm": 3.999853155987491,
+      "learning_rate": 4.334107302477208e-06,
+      "loss": 0.5421,
+      "step": 782
+    },
+    {
+      "epoch": 0.31502715751357874,
+      "grad_norm": 4.961501923709214,
+      "learning_rate": 4.331719684634676e-06,
+      "loss": 0.7151,
+      "step": 783
+    },
+    {
+      "epoch": 0.3154294910480789,
+      "grad_norm": 4.148580969978077,
+      "learning_rate": 4.3293284541352384e-06,
+      "loss": 0.5084,
+      "step": 784
+    },
+    {
+      "epoch": 0.31583182458257897,
+      "grad_norm": 3.623114769478068,
+      "learning_rate": 4.326933615695064e-06,
+      "loss": 0.4912,
+      "step": 785
+    },
+    {
+      "epoch": 0.31623415811707906,
+      "grad_norm": 4.179951168808134,
+      "learning_rate": 4.324535174037433e-06,
+      "loss": 0.4943,
+      "step": 786
+    },
+    {
+      "epoch": 0.31663649165157914,
+      "grad_norm": 4.206035486231768,
+      "learning_rate": 4.322133133892737e-06,
+      "loss": 0.6175,
+      "step": 787
+    },
+    {
+      "epoch": 0.3170388251860793,
+      "grad_norm": 4.029670988995613,
+      "learning_rate": 4.31972749999846e-06,
+      "loss": 0.5329,
+      "step": 788
+    },
+    {
+      "epoch": 0.3174411587205794,
+      "grad_norm": 4.024639871547482,
+      "learning_rate": 4.317318277099178e-06,
+      "loss": 0.5121,
+      "step": 789
+    },
+    {
+      "epoch": 0.31784349225507946,
+      "grad_norm": 4.04730827392964,
+      "learning_rate": 4.314905469946542e-06,
+      "loss": 0.4979,
+      "step": 790
+    },
+    {
+      "epoch": 0.31824582578957955,
+      "grad_norm": 4.440133410808506,
+      "learning_rate": 4.312489083299275e-06,
+      "loss": 0.6333,
+      "step": 791
+    },
+    {
+      "epoch": 0.31864815932407964,
+      "grad_norm": 4.079663513764819,
+      "learning_rate": 4.310069121923158e-06,
+      "loss": 0.5987,
+      "step": 792
+    },
+    {
+      "epoch": 0.3190504928585798,
+      "grad_norm": 4.17692958034715,
+      "learning_rate": 4.307645590591024e-06,
+      "loss": 0.4798,
+      "step": 793
+    },
+    {
+      "epoch": 0.31945282639307987,
+      "grad_norm": 4.417314019145798,
+      "learning_rate": 4.305218494082744e-06,
+      "loss": 0.4832,
+      "step": 794
+    },
+    {
+      "epoch": 0.31985515992757996,
+      "grad_norm": 4.920563992053252,
+      "learning_rate": 4.302787837185224e-06,
+      "loss": 0.6288,
+      "step": 795
+    },
+    {
+      "epoch": 0.32025749346208005,
+      "grad_norm": 4.067969775601857,
+      "learning_rate": 4.300353624692389e-06,
+      "loss": 0.52,
+      "step": 796
+    },
+    {
+      "epoch": 0.3206598269965802,
+      "grad_norm": 4.232889302389981,
+      "learning_rate": 4.2979158614051805e-06,
+      "loss": 0.6205,
+      "step": 797
+    },
+    {
+      "epoch": 0.3210621605310803,
+      "grad_norm": 4.335177482388414,
+      "learning_rate": 4.295474552131538e-06,
+      "loss": 0.6153,
+      "step": 798
+    },
+    {
+      "epoch": 0.32146449406558036,
+      "grad_norm": 4.698659170570413,
+      "learning_rate": 4.2930297016863985e-06,
+      "loss": 0.6404,
+      "step": 799
+    },
+    {
+      "epoch": 0.32186682760008045,
+      "grad_norm": 3.931383036386479,
+      "learning_rate": 4.2905813148916816e-06,
+      "loss": 0.599,
+      "step": 800
+    },
+    {
+      "epoch": 0.32226916113458054,
+      "grad_norm": 5.048444025824699,
+      "learning_rate": 4.288129396576284e-06,
+      "loss": 0.7259,
+      "step": 801
+    },
+    {
+      "epoch": 0.3226714946690807,
+      "grad_norm": 4.851397090992265,
+      "learning_rate": 4.285673951576062e-06,
+      "loss": 0.642,
+      "step": 802
+    },
+    {
+      "epoch": 0.32307382820358077,
+      "grad_norm": 4.28645665558046,
+      "learning_rate": 4.283214984733833e-06,
+      "loss": 0.5315,
+      "step": 803
+    },
+    {
+      "epoch": 0.32347616173808086,
+      "grad_norm": 4.265366711136778,
+      "learning_rate": 4.28075250089936e-06,
+      "loss": 0.5529,
+      "step": 804
+    },
+    {
+      "epoch": 0.32387849527258095,
+      "grad_norm": 3.848835311590686,
+      "learning_rate": 4.278286504929338e-06,
+      "loss": 0.5323,
+      "step": 805
+    },
+    {
+      "epoch": 0.3242808288070811,
+      "grad_norm": 4.627843495385828,
+      "learning_rate": 4.2758170016873934e-06,
+      "loss": 0.5561,
+      "step": 806
+    },
+    {
+      "epoch": 0.3246831623415812,
+      "grad_norm": 4.363493591898399,
+      "learning_rate": 4.273343996044068e-06,
+      "loss": 0.4805,
+      "step": 807
+    },
+    {
+      "epoch": 0.32508549587608127,
+      "grad_norm": 3.9805290188441043,
+      "learning_rate": 4.270867492876812e-06,
+      "loss": 0.5204,
+      "step": 808
+    },
+    {
+      "epoch": 0.32548782941058135,
+      "grad_norm": 3.9644340799634588,
+      "learning_rate": 4.268387497069974e-06,
+      "loss": 0.4642,
+      "step": 809
+    },
+    {
+      "epoch": 0.3258901629450815,
+      "grad_norm": 5.269711583475643,
+      "learning_rate": 4.265904013514788e-06,
+      "loss": 0.6183,
+      "step": 810
+    },
+    {
+      "epoch": 0.3262924964795816,
+      "grad_norm": 4.6945618708498165,
+      "learning_rate": 4.263417047109371e-06,
+      "loss": 0.6567,
+      "step": 811
+    },
+    {
+      "epoch": 0.32669483001408167,
+      "grad_norm": 4.413878483629632,
+      "learning_rate": 4.260926602758707e-06,
+      "loss": 0.5518,
+      "step": 812
+    },
+    {
+      "epoch": 0.32709716354858176,
+      "grad_norm": 4.355371114130468,
+      "learning_rate": 4.258432685374641e-06,
+      "loss": 0.478,
+      "step": 813
+    },
+    {
+      "epoch": 0.32749949708308185,
+      "grad_norm": 4.625144876838143,
+      "learning_rate": 4.255935299875864e-06,
+      "loss": 0.5918,
+      "step": 814
+    },
+    {
+      "epoch": 0.327901830617582,
+      "grad_norm": 3.8722185488332994,
+      "learning_rate": 4.253434451187911e-06,
+      "loss": 0.4698,
+      "step": 815
+    },
+    {
+      "epoch": 0.3283041641520821,
+      "grad_norm": 3.8155166378719403,
+      "learning_rate": 4.250930144243147e-06,
+      "loss": 0.43,
+      "step": 816
+    },
+    {
+      "epoch": 0.32870649768658217,
+      "grad_norm": 4.373317134651141,
+      "learning_rate": 4.248422383980756e-06,
+      "loss": 0.581,
+      "step": 817
+    },
+    {
+      "epoch": 0.32910883122108225,
+      "grad_norm": 4.316760308023842,
+      "learning_rate": 4.245911175346733e-06,
+      "loss": 0.525,
+      "step": 818
+    },
+    {
+      "epoch": 0.3295111647555824,
+      "grad_norm": 4.1806833930239895,
+      "learning_rate": 4.243396523293875e-06,
+      "loss": 0.6027,
+      "step": 819
+    },
+    {
+      "epoch": 0.3299134982900825,
+      "grad_norm": 4.011770765526375,
+      "learning_rate": 4.240878432781769e-06,
+      "loss": 0.5642,
+      "step": 820
+    },
+    {
+      "epoch": 0.3303158318245826,
+      "grad_norm": 4.106735076414742,
+      "learning_rate": 4.238356908776786e-06,
+      "loss": 0.5576,
+      "step": 821
+    },
+    {
+      "epoch": 0.33071816535908266,
+      "grad_norm": 4.7608485825878955,
+      "learning_rate": 4.235831956252069e-06,
+      "loss": 0.7088,
+      "step": 822
+    },
+    {
+      "epoch": 0.3311204988935828,
+      "grad_norm": 4.8786607685170305,
+      "learning_rate": 4.2333035801875175e-06,
+      "loss": 0.6513,
+      "step": 823
+    },
+    {
+      "epoch": 0.3315228324280829,
+      "grad_norm": 4.288963783688971,
+      "learning_rate": 4.230771785569791e-06,
+      "loss": 0.5073,
+      "step": 824
+    },
+    {
+      "epoch": 0.331925165962583,
+      "grad_norm": 4.014903400829248,
+      "learning_rate": 4.228236577392285e-06,
+      "loss": 0.5901,
+      "step": 825
+    },
+    {
+      "epoch": 0.33232749949708307,
+      "grad_norm": 4.154812074980563,
+      "learning_rate": 4.225697960655131e-06,
+      "loss": 0.5568,
+      "step": 826
+    },
+    {
+      "epoch": 0.33272983303158316,
+      "grad_norm": 3.827755329803438,
+      "learning_rate": 4.223155940365181e-06,
+      "loss": 0.4738,
+      "step": 827
+    },
+    {
+      "epoch": 0.3331321665660833,
+      "grad_norm": 3.856077516315061,
+      "learning_rate": 4.220610521536002e-06,
+      "loss": 0.4963,
+      "step": 828
+    },
+    {
+      "epoch": 0.3335345001005834,
+      "grad_norm": 4.615466783911023,
+      "learning_rate": 4.218061709187862e-06,
+      "loss": 0.5656,
+      "step": 829
+    },
+    {
+      "epoch": 0.3339368336350835,
+      "grad_norm": 4.176966772191371,
+      "learning_rate": 4.215509508347723e-06,
+      "loss": 0.5398,
+      "step": 830
+    },
+    {
+      "epoch": 0.33433916716958356,
+      "grad_norm": 4.49327690032652,
+      "learning_rate": 4.212953924049229e-06,
+      "loss": 0.6772,
+      "step": 831
+    },
+    {
+      "epoch": 0.3347415007040837,
+      "grad_norm": 4.067266904294674,
+      "learning_rate": 4.210394961332698e-06,
+      "loss": 0.5311,
+      "step": 832
+    },
+    {
+      "epoch": 0.3351438342385838,
+      "grad_norm": 3.7375221543142128,
+      "learning_rate": 4.207832625245112e-06,
+      "loss": 0.4573,
+      "step": 833
+    },
+    {
+      "epoch": 0.3355461677730839,
+      "grad_norm": 4.056435688935082,
+      "learning_rate": 4.205266920840103e-06,
+      "loss": 0.543,
+      "step": 834
+    },
+    {
+      "epoch": 0.33594850130758397,
+      "grad_norm": 3.9887060748088254,
+      "learning_rate": 4.202697853177951e-06,
+      "loss": 0.4613,
+      "step": 835
+    },
+    {
+      "epoch": 0.3363508348420841,
+      "grad_norm": 3.5896338953258473,
+      "learning_rate": 4.2001254273255646e-06,
+      "loss": 0.4031,
+      "step": 836
+    },
+    {
+      "epoch": 0.3367531683765842,
+      "grad_norm": 4.094188015061212,
+      "learning_rate": 4.197549648356478e-06,
+      "loss": 0.5933,
+      "step": 837
+    },
+    {
+      "epoch": 0.3371555019110843,
+      "grad_norm": 3.6605198066957216,
+      "learning_rate": 4.19497052135084e-06,
+      "loss": 0.417,
+      "step": 838
+    },
+    {
+      "epoch": 0.3375578354455844,
+      "grad_norm": 5.538220960178179,
+      "learning_rate": 4.192388051395398e-06,
+      "loss": 0.7984,
+      "step": 839
+    },
+    {
+      "epoch": 0.33796016898008446,
+      "grad_norm": 4.543985654558515,
+      "learning_rate": 4.1898022435835e-06,
+      "loss": 0.5526,
+      "step": 840
+    },
+    {
+      "epoch": 0.3383625025145846,
+      "grad_norm": 3.719487495093868,
+      "learning_rate": 4.187213103015069e-06,
+      "loss": 0.4399,
+      "step": 841
+    },
+    {
+      "epoch": 0.3387648360490847,
+      "grad_norm": 3.958292027050775,
+      "learning_rate": 4.184620634796608e-06,
+      "loss": 0.4891,
+      "step": 842
+    },
+    {
+      "epoch": 0.3391671695835848,
+      "grad_norm": 4.219443448956219,
+      "learning_rate": 4.182024844041177e-06,
+      "loss": 0.6417,
+      "step": 843
+    },
+    {
+      "epoch": 0.33956950311808487,
+      "grad_norm": 4.637518704467857,
+      "learning_rate": 4.179425735868395e-06,
+      "loss": 0.5912,
+      "step": 844
+    },
+    {
+      "epoch": 0.339971836652585,
+      "grad_norm": 4.725084492132898,
+      "learning_rate": 4.176823315404419e-06,
+      "loss": 0.705,
+      "step": 845
+    },
+    {
+      "epoch": 0.3403741701870851,
+      "grad_norm": 4.081067015807228,
+      "learning_rate": 4.17421758778194e-06,
+      "loss": 0.5006,
+      "step": 846
+    },
+    {
+      "epoch": 0.3407765037215852,
+      "grad_norm": 4.431343567792434,
+      "learning_rate": 4.1716085581401746e-06,
+      "loss": 0.5081,
+      "step": 847
+    },
+    {
+      "epoch": 0.3411788372560853,
+      "grad_norm": 4.173696495740008,
+      "learning_rate": 4.1689962316248475e-06,
+      "loss": 0.497,
+      "step": 848
+    },
+    {
+      "epoch": 0.3415811707905854,
+      "grad_norm": 4.125681881515539,
+      "learning_rate": 4.166380613388189e-06,
+      "loss": 0.5498,
+      "step": 849
+    },
+    {
+      "epoch": 0.3419835043250855,
+      "grad_norm": 4.333217037647524,
+      "learning_rate": 4.163761708588919e-06,
+      "loss": 0.5597,
+      "step": 850
+    },
+    {
+      "epoch": 0.3423858378595856,
+      "grad_norm": 4.191848923537687,
+      "learning_rate": 4.161139522392243e-06,
+      "loss": 0.7259,
+      "step": 851
+    },
+    {
+      "epoch": 0.3427881713940857,
+      "grad_norm": 3.1772920188403164,
+      "learning_rate": 4.158514059969834e-06,
+      "loss": 0.4225,
+      "step": 852
+    },
+    {
+      "epoch": 0.34319050492858577,
+      "grad_norm": 4.166291911511059,
+      "learning_rate": 4.15588532649983e-06,
+      "loss": 0.6061,
+      "step": 853
+    },
+    {
+      "epoch": 0.3435928384630859,
+      "grad_norm": 3.5991395036430447,
+      "learning_rate": 4.1532533271668175e-06,
+      "loss": 0.461,
+      "step": 854
+    },
+    {
+      "epoch": 0.343995171997586,
+      "grad_norm": 4.29116994161306,
+      "learning_rate": 4.150618067161828e-06,
+      "loss": 0.567,
+      "step": 855
+    },
+    {
+      "epoch": 0.3443975055320861,
+      "grad_norm": 4.403897532608296,
+      "learning_rate": 4.14797955168232e-06,
+      "loss": 0.6118,
+      "step": 856
+    },
+    {
+      "epoch": 0.3447998390665862,
+      "grad_norm": 4.145689255622535,
+      "learning_rate": 4.145337785932174e-06,
+      "loss": 0.5709,
+      "step": 857
+    },
+    {
+      "epoch": 0.3452021726010863,
+      "grad_norm": 4.540600097537597,
+      "learning_rate": 4.142692775121684e-06,
+      "loss": 0.5506,
+      "step": 858
+    },
+    {
+      "epoch": 0.3456045061355864,
+      "grad_norm": 4.730037786980354,
+      "learning_rate": 4.14004452446754e-06,
+      "loss": 0.6088,
+      "step": 859
+    },
+    {
+      "epoch": 0.3460068396700865,
+      "grad_norm": 3.9409862740254487,
+      "learning_rate": 4.137393039192822e-06,
+      "loss": 0.5775,
+      "step": 860
+    },
+    {
+      "epoch": 0.3464091732045866,
+      "grad_norm": 4.113320460512886,
+      "learning_rate": 4.1347383245269935e-06,
+      "loss": 0.5695,
+      "step": 861
+    },
+    {
+      "epoch": 0.34681150673908673,
+      "grad_norm": 4.6445735426137364,
+      "learning_rate": 4.1320803857058835e-06,
+      "loss": 0.6158,
+      "step": 862
+    },
+    {
+      "epoch": 0.3472138402735868,
+      "grad_norm": 4.634455536740089,
+      "learning_rate": 4.129419227971681e-06,
+      "loss": 0.6799,
+      "step": 863
+    },
+    {
+      "epoch": 0.3476161738080869,
+      "grad_norm": 4.767706536948578,
+      "learning_rate": 4.1267548565729235e-06,
+      "loss": 0.5473,
+      "step": 864
+    },
+    {
+      "epoch": 0.348018507342587,
+      "grad_norm": 4.4478746327105405,
+      "learning_rate": 4.124087276764488e-06,
+      "loss": 0.6564,
+      "step": 865
+    },
+    {
+      "epoch": 0.3484208408770871,
+      "grad_norm": 5.1282545461378195,
+      "learning_rate": 4.121416493807577e-06,
+      "loss": 0.617,
+      "step": 866
+    },
+    {
+      "epoch": 0.3488231744115872,
+      "grad_norm": 3.850370414902996,
+      "learning_rate": 4.118742512969713e-06,
+      "loss": 0.5775,
+      "step": 867
+    },
+    {
+      "epoch": 0.3492255079460873,
+      "grad_norm": 4.66805214245803,
+      "learning_rate": 4.116065339524724e-06,
+      "loss": 0.6724,
+      "step": 868
+    },
+    {
+      "epoch": 0.3496278414805874,
+      "grad_norm": 4.7651508619199765,
+      "learning_rate": 4.113384978752734e-06,
+      "loss": 0.6151,
+      "step": 869
+    },
+    {
+      "epoch": 0.3500301750150875,
+      "grad_norm": 4.207593613220341,
+      "learning_rate": 4.110701435940157e-06,
+      "loss": 0.641,
+      "step": 870
+    },
+    {
+      "epoch": 0.35043250854958763,
+      "grad_norm": 4.3034195107515965,
+      "learning_rate": 4.108014716379679e-06,
+      "loss": 0.6542,
+      "step": 871
+    },
+    {
+      "epoch": 0.3508348420840877,
+      "grad_norm": 3.987719667248183,
+      "learning_rate": 4.105324825370251e-06,
+      "loss": 0.4996,
+      "step": 872
+    },
+    {
+      "epoch": 0.3512371756185878,
+      "grad_norm": 3.9071320373239824,
+      "learning_rate": 4.102631768217083e-06,
+      "loss": 0.5168,
+      "step": 873
+    },
+    {
+      "epoch": 0.3516395091530879,
+      "grad_norm": 4.2277706570506295,
+      "learning_rate": 4.099935550231626e-06,
+      "loss": 0.5379,
+      "step": 874
+    },
+    {
+      "epoch": 0.35204184268758804,
+      "grad_norm": 3.9154962483786373,
+      "learning_rate": 4.097236176731567e-06,
+      "loss": 0.5523,
+      "step": 875
+    },
+    {
+      "epoch": 0.3524441762220881,
+      "grad_norm": 3.465039425796171,
+      "learning_rate": 4.094533653040814e-06,
+      "loss": 0.4001,
+      "step": 876
+    },
+    {
+      "epoch": 0.3528465097565882,
+      "grad_norm": 3.4706674601181167,
+      "learning_rate": 4.091827984489493e-06,
+      "loss": 0.4629,
+      "step": 877
+    },
+    {
+      "epoch": 0.3532488432910883,
+      "grad_norm": 4.340590023749821,
+      "learning_rate": 4.089119176413926e-06,
+      "loss": 0.6815,
+      "step": 878
+    },
+    {
+      "epoch": 0.3536511768255884,
+      "grad_norm": 4.471466388864396,
+      "learning_rate": 4.086407234156633e-06,
+      "loss": 0.6459,
+      "step": 879
+    },
+    {
+      "epoch": 0.35405351036008853,
+      "grad_norm": 4.449019271914241,
+      "learning_rate": 4.08369216306631e-06,
+      "loss": 0.5592,
+      "step": 880
+    },
+    {
+      "epoch": 0.3544558438945886,
+      "grad_norm": 4.125369265434845,
+      "learning_rate": 4.080973968497829e-06,
+      "loss": 0.5138,
+      "step": 881
+    },
+    {
+      "epoch": 0.3548581774290887,
+      "grad_norm": 4.523305588736027,
+      "learning_rate": 4.07825265581222e-06,
+      "loss": 0.5929,
+      "step": 882
+    },
+    {
+      "epoch": 0.3552605109635888,
+      "grad_norm": 4.077469504247697,
+      "learning_rate": 4.075528230376662e-06,
+      "loss": 0.5288,
+      "step": 883
+    },
+    {
+      "epoch": 0.35566284449808894,
+      "grad_norm": 5.092003038039258,
+      "learning_rate": 4.072800697564474e-06,
+      "loss": 0.6624,
+      "step": 884
+    },
+    {
+      "epoch": 0.356065178032589,
+      "grad_norm": 3.7756634610021935,
+      "learning_rate": 4.070070062755105e-06,
+      "loss": 0.5265,
+      "step": 885
+    },
+    {
+      "epoch": 0.3564675115670891,
+      "grad_norm": 4.106660405022582,
+      "learning_rate": 4.0673363313341195e-06,
+      "loss": 0.4853,
+      "step": 886
+    },
+    {
+      "epoch": 0.3568698451015892,
+      "grad_norm": 4.215215996153993,
+      "learning_rate": 4.064599508693191e-06,
+      "loss": 0.5721,
+      "step": 887
+    },
+    {
+      "epoch": 0.35727217863608934,
+      "grad_norm": 4.171756065223583,
+      "learning_rate": 4.06185960023009e-06,
+      "loss": 0.4649,
+      "step": 888
+    },
+    {
+      "epoch": 0.35767451217058943,
+      "grad_norm": 5.138546885500691,
+      "learning_rate": 4.05911661134867e-06,
+      "loss": 0.5972,
+      "step": 889
+    },
+    {
+      "epoch": 0.3580768457050895,
+      "grad_norm": 4.136963664338033,
+      "learning_rate": 4.0563705474588655e-06,
+      "loss": 0.5027,
+      "step": 890
+    },
+    {
+      "epoch": 0.3584791792395896,
+      "grad_norm": 3.940446094705443,
+      "learning_rate": 4.05362141397667e-06,
+      "loss": 0.5627,
+      "step": 891
+    },
+    {
+      "epoch": 0.3588815127740897,
+      "grad_norm": 4.024692229149252,
+      "learning_rate": 4.0508692163241356e-06,
+      "loss": 0.5521,
+      "step": 892
+    },
+    {
+      "epoch": 0.35928384630858984,
+      "grad_norm": 3.866043251441755,
+      "learning_rate": 4.048113959929354e-06,
+      "loss": 0.5125,
+      "step": 893
+    },
+    {
+      "epoch": 0.3596861798430899,
+      "grad_norm": 4.786779622248551,
+      "learning_rate": 4.0453556502264535e-06,
+      "loss": 0.6475,
+      "step": 894
+    },
+    {
+      "epoch": 0.36008851337759,
+      "grad_norm": 4.127247200504385,
+      "learning_rate": 4.042594292655581e-06,
+      "loss": 0.4877,
+      "step": 895
+    },
+    {
+      "epoch": 0.3604908469120901,
+      "grad_norm": 3.3408217784719634,
+      "learning_rate": 4.039829892662897e-06,
+      "loss": 0.4337,
+      "step": 896
+    },
+    {
+      "epoch": 0.36089318044659024,
+      "grad_norm": 4.010232394860985,
+      "learning_rate": 4.037062455700559e-06,
+      "loss": 0.493,
+      "step": 897
+    },
+    {
+      "epoch": 0.36129551398109033,
+      "grad_norm": 4.3695516591433785,
+      "learning_rate": 4.03429198722672e-06,
+      "loss": 0.6051,
+      "step": 898
+    },
+    {
+      "epoch": 0.3616978475155904,
+      "grad_norm": 4.766410636581824,
+      "learning_rate": 4.0315184927055065e-06,
+      "loss": 0.6602,
+      "step": 899
+    },
+    {
+      "epoch": 0.3621001810500905,
+      "grad_norm": 4.1149436048907955,
+      "learning_rate": 4.028741977607016e-06,
+      "loss": 0.599,
+      "step": 900
+    },
+    {
+      "epoch": 0.36250251458459065,
+      "grad_norm": 3.4871089312720045,
+      "learning_rate": 4.0259624474073025e-06,
+      "loss": 0.4779,
+      "step": 901
+    },
+    {
+      "epoch": 0.36290484811909074,
+      "grad_norm": 3.6367752409912266,
+      "learning_rate": 4.023179907588367e-06,
+      "loss": 0.4541,
+      "step": 902
+    },
+    {
+      "epoch": 0.3633071816535908,
+      "grad_norm": 3.3961433962608685,
+      "learning_rate": 4.020394363638147e-06,
+      "loss": 0.4265,
+      "step": 903
+    },
+    {
+      "epoch": 0.3637095151880909,
+      "grad_norm": 3.8352646417492524,
+      "learning_rate": 4.0176058210505045e-06,
+      "loss": 0.5351,
+      "step": 904
+    },
+    {
+      "epoch": 0.364111848722591,
+      "grad_norm": 4.242220553627046,
+      "learning_rate": 4.0148142853252125e-06,
+      "loss": 0.5792,
+      "step": 905
+    },
+    {
+      "epoch": 0.36451418225709115,
+      "grad_norm": 4.096283695840529,
+      "learning_rate": 4.0120197619679536e-06,
+      "loss": 0.5412,
+      "step": 906
+    },
+    {
+      "epoch": 0.36491651579159123,
+      "grad_norm": 4.729251903541468,
+      "learning_rate": 4.009222256490297e-06,
+      "loss": 0.6285,
+      "step": 907
+    },
+    {
+      "epoch": 0.3653188493260913,
+      "grad_norm": 4.084180186714877,
+      "learning_rate": 4.006421774409697e-06,
+      "loss": 0.6304,
+      "step": 908
+    },
+    {
+      "epoch": 0.3657211828605914,
+      "grad_norm": 4.280731987602122,
+      "learning_rate": 4.003618321249476e-06,
+      "loss": 0.5632,
+      "step": 909
+    },
+    {
+      "epoch": 0.36612351639509155,
+      "grad_norm": 3.868764716611033,
+      "learning_rate": 4.000811902538821e-06,
+      "loss": 0.4715,
+      "step": 910
+    },
+    {
+      "epoch": 0.36652584992959164,
+      "grad_norm": 3.6445778686982155,
+      "learning_rate": 3.99800252381276e-06,
+      "loss": 0.4084,
+      "step": 911
+    },
+    {
+      "epoch": 0.36692818346409173,
+      "grad_norm": 4.600601862430362,
+      "learning_rate": 3.995190190612165e-06,
+      "loss": 0.7218,
+      "step": 912
+    },
+    {
+      "epoch": 0.3673305169985918,
+      "grad_norm": 4.09418361850337,
+      "learning_rate": 3.992374908483735e-06,
+      "loss": 0.5913,
+      "step": 913
+    },
+    {
+      "epoch": 0.36773285053309196,
+      "grad_norm": 4.495329343052196,
+      "learning_rate": 3.9895566829799825e-06,
+      "loss": 0.6735,
+      "step": 914
+    },
+    {
+      "epoch": 0.36813518406759205,
+      "grad_norm": 4.408452152812246,
+      "learning_rate": 3.986735519659226e-06,
+      "loss": 0.6896,
+      "step": 915
+    },
+    {
+      "epoch": 0.36853751760209214,
+      "grad_norm": 3.7084757585295276,
+      "learning_rate": 3.983911424085578e-06,
+      "loss": 0.552,
+      "step": 916
+    },
+    {
+      "epoch": 0.3689398511365922,
+      "grad_norm": 3.8368900456075985,
+      "learning_rate": 3.981084401828937e-06,
+      "loss": 0.4944,
+      "step": 917
+    },
+    {
+      "epoch": 0.3693421846710923,
+      "grad_norm": 5.696133375208026,
+      "learning_rate": 3.978254458464969e-06,
+      "loss": 0.6338,
+      "step": 918
+    },
+    {
+      "epoch": 0.36974451820559245,
+      "grad_norm": 4.144365741395642,
+      "learning_rate": 3.975421599575103e-06,
+      "loss": 0.6463,
+      "step": 919
+    },
+    {
+      "epoch": 0.37014685174009254,
+      "grad_norm": 3.7665583063095682,
+      "learning_rate": 3.972585830746522e-06,
+      "loss": 0.4442,
+      "step": 920
+    },
+    {
+      "epoch": 0.37054918527459263,
+      "grad_norm": 4.341604001350467,
+      "learning_rate": 3.969747157572142e-06,
+      "loss": 0.7191,
+      "step": 921
+    },
+    {
+      "epoch": 0.3709515188090927,
+      "grad_norm": 4.724527719836307,
+      "learning_rate": 3.966905585650611e-06,
+      "loss": 0.7508,
+      "step": 922
+    },
+    {
+      "epoch": 0.37135385234359286,
+      "grad_norm": 6.112644943798556,
+      "learning_rate": 3.964061120586294e-06,
+      "loss": 0.6744,
+      "step": 923
+    },
+    {
+      "epoch": 0.37175618587809295,
+      "grad_norm": 4.630742134072846,
+      "learning_rate": 3.961213767989261e-06,
+      "loss": 0.5703,
+      "step": 924
+    },
+    {
+      "epoch": 0.37215851941259304,
+      "grad_norm": 3.5499854666112802,
+      "learning_rate": 3.958363533475277e-06,
+      "loss": 0.4563,
+      "step": 925
+    },
+    {
+      "epoch": 0.3725608529470931,
+      "grad_norm": 3.926859330076321,
+      "learning_rate": 3.955510422665791e-06,
+      "loss": 0.5159,
+      "step": 926
+    },
+    {
+      "epoch": 0.37296318648159327,
+      "grad_norm": 4.029967110899967,
+      "learning_rate": 3.952654441187927e-06,
+      "loss": 0.5267,
+      "step": 927
+    },
+    {
+      "epoch": 0.37336552001609336,
+      "grad_norm": 3.4483709639926414,
+      "learning_rate": 3.9497955946744675e-06,
+      "loss": 0.4584,
+      "step": 928
+    },
+    {
+      "epoch": 0.37376785355059344,
+      "grad_norm": 4.752285945278346,
+      "learning_rate": 3.946933888763847e-06,
+      "loss": 0.5743,
+      "step": 929
+    },
+    {
+      "epoch": 0.37417018708509353,
+      "grad_norm": 3.985801543832948,
+      "learning_rate": 3.9440693291001406e-06,
+      "loss": 0.5319,
+      "step": 930
+    },
+    {
+      "epoch": 0.3745725206195936,
+      "grad_norm": 3.5946794819227903,
+      "learning_rate": 3.941201921333048e-06,
+      "loss": 0.4217,
+      "step": 931
+    },
+    {
+      "epoch": 0.37497485415409376,
+      "grad_norm": 4.035357970865375,
+      "learning_rate": 3.9383316711178946e-06,
+      "loss": 0.503,
+      "step": 932
+    },
+    {
+      "epoch": 0.37537718768859385,
+      "grad_norm": 4.034839036975055,
+      "learning_rate": 3.935458584115599e-06,
+      "loss": 0.4698,
+      "step": 933
+    },
+    {
+      "epoch": 0.37577952122309394,
+      "grad_norm": 4.286963186422666,
+      "learning_rate": 3.932582665992688e-06,
+      "loss": 0.5901,
+      "step": 934
+    },
+    {
+      "epoch": 0.376181854757594,
+      "grad_norm": 4.021909925329209,
+      "learning_rate": 3.929703922421263e-06,
+      "loss": 0.5111,
+      "step": 935
+    },
+    {
+      "epoch": 0.37658418829209417,
+      "grad_norm": 4.0942098752342035,
+      "learning_rate": 3.926822359079001e-06,
+      "loss": 0.5472,
+      "step": 936
+    },
+    {
+      "epoch": 0.37698652182659426,
+      "grad_norm": 4.261504012145757,
+      "learning_rate": 3.923937981649143e-06,
+      "loss": 0.6453,
+      "step": 937
+    },
+    {
+      "epoch": 0.37738885536109434,
+      "grad_norm": 4.014117574115026,
+      "learning_rate": 3.921050795820474e-06,
+      "loss": 0.5056,
+      "step": 938
+    },
+    {
+      "epoch": 0.37779118889559443,
+      "grad_norm": 3.904726379027022,
+      "learning_rate": 3.9181608072873244e-06,
+      "loss": 0.5304,
+      "step": 939
+    },
+    {
+      "epoch": 0.3781935224300946,
+      "grad_norm": 3.881019901522325,
+      "learning_rate": 3.9152680217495475e-06,
+      "loss": 0.5298,
+      "step": 940
+    },
+    {
+      "epoch": 0.37859585596459466,
+      "grad_norm": 4.058485666284913,
+      "learning_rate": 3.912372444912517e-06,
+      "loss": 0.574,
+      "step": 941
+    },
+    {
+      "epoch": 0.37899818949909475,
+      "grad_norm": 3.9630383581476205,
+      "learning_rate": 3.909474082487107e-06,
+      "loss": 0.5804,
+      "step": 942
+    },
+    {
+      "epoch": 0.37940052303359484,
+      "grad_norm": 3.8676063595588683,
+      "learning_rate": 3.906572940189691e-06,
+      "loss": 0.4804,
+      "step": 943
+    },
+    {
+      "epoch": 0.3798028565680949,
+      "grad_norm": 4.32834774261345,
+      "learning_rate": 3.9036690237421215e-06,
+      "loss": 0.6161,
+      "step": 944
+    },
+    {
+      "epoch": 0.38020519010259507,
+      "grad_norm": 4.726002477400136,
+      "learning_rate": 3.900762338871723e-06,
+      "loss": 0.6898,
+      "step": 945
+    },
+    {
+      "epoch": 0.38060752363709516,
+      "grad_norm": 4.577558255257823,
+      "learning_rate": 3.897852891311282e-06,
+      "loss": 0.5504,
+      "step": 946
+    },
+    {
+      "epoch": 0.38100985717159525,
+      "grad_norm": 4.300878005170684,
+      "learning_rate": 3.89494068679903e-06,
+      "loss": 0.6734,
+      "step": 947
+    },
+    {
+      "epoch": 0.38141219070609533,
+      "grad_norm": 4.519778162808581,
+      "learning_rate": 3.892025731078641e-06,
+      "loss": 0.5694,
+      "step": 948
+    },
+    {
+      "epoch": 0.3818145242405955,
+      "grad_norm": 3.8176275411987683,
+      "learning_rate": 3.88910802989921e-06,
+      "loss": 0.4444,
+      "step": 949
+    },
+    {
+      "epoch": 0.38221685777509556,
+      "grad_norm": 4.014909553832162,
+      "learning_rate": 3.886187589015251e-06,
+      "loss": 0.6054,
+      "step": 950
+    },
+    {
+      "epoch": 0.38261919130959565,
+      "grad_norm": 4.267781377287406,
+      "learning_rate": 3.883264414186677e-06,
+      "loss": 0.6257,
+      "step": 951
+    },
+    {
+      "epoch": 0.38302152484409574,
+      "grad_norm": 4.345044782459552,
+      "learning_rate": 3.8803385111788e-06,
+      "loss": 0.5543,
+      "step": 952
+    },
+    {
+      "epoch": 0.3834238583785959,
+      "grad_norm": 4.221128544535545,
+      "learning_rate": 3.877409885762305e-06,
+      "loss": 0.5544,
+      "step": 953
+    },
+    {
+      "epoch": 0.38382619191309597,
+      "grad_norm": 4.000060887768021,
+      "learning_rate": 3.8744785437132506e-06,
+      "loss": 0.5878,
+      "step": 954
+    },
+    {
+      "epoch": 0.38422852544759606,
+      "grad_norm": 3.8372993108948488,
+      "learning_rate": 3.871544490813054e-06,
+      "loss": 0.5076,
+      "step": 955
+    },
+    {
+      "epoch": 0.38463085898209615,
+      "grad_norm": 4.059706126497868,
+      "learning_rate": 3.868607732848475e-06,
+      "loss": 0.5353,
+      "step": 956
+    },
+    {
+      "epoch": 0.38503319251659623,
+      "grad_norm": 4.408924474844376,
+      "learning_rate": 3.865668275611614e-06,
+      "loss": 0.519,
+      "step": 957
+    },
+    {
+      "epoch": 0.3854355260510964,
+      "grad_norm": 4.010089079135815,
+      "learning_rate": 3.862726124899889e-06,
+      "loss": 0.5952,
+      "step": 958
+    },
+    {
+      "epoch": 0.38583785958559647,
+      "grad_norm": 5.459071377748317,
+      "learning_rate": 3.859781286516037e-06,
+      "loss": 0.7175,
+      "step": 959
+    },
+    {
+      "epoch": 0.38624019312009655,
+      "grad_norm": 4.542741280511702,
+      "learning_rate": 3.85683376626809e-06,
+      "loss": 0.5667,
+      "step": 960
+    },
+    {
+      "epoch": 0.38664252665459664,
+      "grad_norm": 3.99184341470565,
+      "learning_rate": 3.853883569969372e-06,
+      "loss": 0.6148,
+      "step": 961
+    },
+    {
+      "epoch": 0.3870448601890968,
+      "grad_norm": 4.078873510656604,
+      "learning_rate": 3.850930703438484e-06,
+      "loss": 0.5383,
+      "step": 962
+    },
+    {
+      "epoch": 0.38744719372359687,
+      "grad_norm": 3.896065226196885,
+      "learning_rate": 3.847975172499295e-06,
+      "loss": 0.6785,
+      "step": 963
+    },
+    {
+      "epoch": 0.38784952725809696,
+      "grad_norm": 4.328062559714918,
+      "learning_rate": 3.845016982980929e-06,
+      "loss": 0.5543,
+      "step": 964
+    },
+    {
+      "epoch": 0.38825186079259705,
+      "grad_norm": 4.2981177827507935,
+      "learning_rate": 3.842056140717749e-06,
+      "loss": 0.6935,
+      "step": 965
+    },
+    {
+      "epoch": 0.3886541943270972,
+      "grad_norm": 4.489409385825919,
+      "learning_rate": 3.839092651549357e-06,
+      "loss": 0.6412,
+      "step": 966
+    },
+    {
+      "epoch": 0.3890565278615973,
+      "grad_norm": 4.162311148181462,
+      "learning_rate": 3.836126521320569e-06,
+      "loss": 0.5133,
+      "step": 967
+    },
+    {
+      "epoch": 0.38945886139609737,
+      "grad_norm": 3.7564756341509464,
+      "learning_rate": 3.833157755881414e-06,
+      "loss": 0.5198,
+      "step": 968
+    },
+    {
+      "epoch": 0.38986119493059745,
+      "grad_norm": 4.2353823981014385,
+      "learning_rate": 3.830186361087117e-06,
+      "loss": 0.4271,
+      "step": 969
+    },
+    {
+      "epoch": 0.39026352846509754,
+      "grad_norm": 3.4858668216855437,
+      "learning_rate": 3.827212342798089e-06,
+      "loss": 0.4654,
+      "step": 970
+    },
+    {
+      "epoch": 0.3906658619995977,
+      "grad_norm": 3.920259549785761,
+      "learning_rate": 3.824235706879915e-06,
+      "loss": 0.5822,
+      "step": 971
+    },
+    {
+      "epoch": 0.3910681955340978,
+      "grad_norm": 4.239371514940848,
+      "learning_rate": 3.821256459203343e-06,
+      "loss": 0.5093,
+      "step": 972
+    },
+    {
+      "epoch": 0.39147052906859786,
+      "grad_norm": 4.037277711642943,
+      "learning_rate": 3.81827460564427e-06,
+      "loss": 0.5222,
+      "step": 973
+    },
+    {
+      "epoch": 0.39187286260309795,
+      "grad_norm": 4.456827976048942,
+      "learning_rate": 3.815290152083737e-06,
+      "loss": 0.6783,
+      "step": 974
+    },
+    {
+      "epoch": 0.3922751961375981,
+      "grad_norm": 4.246220187755789,
+      "learning_rate": 3.8123031044079084e-06,
+      "loss": 0.5888,
+      "step": 975
+    },
+    {
+      "epoch": 0.3926775296720982,
+      "grad_norm": 4.3203808668287635,
+      "learning_rate": 3.809313468508068e-06,
+      "loss": 0.5569,
+      "step": 976
+    },
+    {
+      "epoch": 0.39307986320659827,
+      "grad_norm": 4.860143257226567,
+      "learning_rate": 3.8063212502806035e-06,
+      "loss": 0.5953,
+      "step": 977
+    },
+    {
+      "epoch": 0.39348219674109836,
+      "grad_norm": 4.233378915892834,
+      "learning_rate": 3.8033264556269954e-06,
+      "loss": 0.6167,
+      "step": 978
+    },
+    {
+      "epoch": 0.39388453027559844,
+      "grad_norm": 4.3534789683754305,
+      "learning_rate": 3.8003290904538035e-06,
+      "loss": 0.6133,
+      "step": 979
+    },
+    {
+      "epoch": 0.3942868638100986,
+      "grad_norm": 3.9519882332598257,
+      "learning_rate": 3.7973291606726614e-06,
+      "loss": 0.4515,
+      "step": 980
+    },
+    {
+      "epoch": 0.3946891973445987,
+      "grad_norm": 4.138408229253922,
+      "learning_rate": 3.794326672200258e-06,
+      "loss": 0.5607,
+      "step": 981
+    },
+    {
+      "epoch": 0.39509153087909876,
+      "grad_norm": 4.3761985221008635,
+      "learning_rate": 3.79132163095833e-06,
+      "loss": 0.5507,
+      "step": 982
+    },
+    {
+      "epoch": 0.39549386441359885,
+      "grad_norm": 4.023180334056979,
+      "learning_rate": 3.7883140428736477e-06,
+      "loss": 0.5314,
+      "step": 983
+    },
+    {
+      "epoch": 0.395896197948099,
+      "grad_norm": 4.151232898341051,
+      "learning_rate": 3.7853039138780057e-06,
+      "loss": 0.5392,
+      "step": 984
+    },
+    {
+      "epoch": 0.3962985314825991,
+      "grad_norm": 4.539638233125882,
+      "learning_rate": 3.7822912499082087e-06,
+      "loss": 0.5791,
+      "step": 985
+    },
+    {
+      "epoch": 0.39670086501709917,
+      "grad_norm": 4.335330132423181,
+      "learning_rate": 3.7792760569060626e-06,
+      "loss": 0.5178,
+      "step": 986
+    },
+    {
+      "epoch": 0.39710319855159926,
+      "grad_norm": 4.143033413053681,
+      "learning_rate": 3.7762583408183594e-06,
+      "loss": 0.5815,
+      "step": 987
+    },
+    {
+      "epoch": 0.3975055320860994,
+      "grad_norm": 4.915299915525361,
+      "learning_rate": 3.7732381075968694e-06,
+      "loss": 0.6269,
+      "step": 988
+    },
+    {
+      "epoch": 0.3979078656205995,
+      "grad_norm": 3.7212476195736506,
+      "learning_rate": 3.770215363198325e-06,
+      "loss": 0.4599,
+      "step": 989
+    },
+    {
+      "epoch": 0.3983101991550996,
+      "grad_norm": 3.9977745943956884,
+      "learning_rate": 3.7671901135844148e-06,
+      "loss": 0.5973,
+      "step": 990
+    },
+    {
+      "epoch": 0.39871253268959966,
+      "grad_norm": 4.3088468446821615,
+      "learning_rate": 3.764162364721764e-06,
+      "loss": 0.5963,
+      "step": 991
+    },
+    {
+      "epoch": 0.39911486622409975,
+      "grad_norm": 4.2991674025596955,
+      "learning_rate": 3.761132122581931e-06,
+      "loss": 0.6923,
+      "step": 992
+    },
+    {
+      "epoch": 0.3995171997585999,
+      "grad_norm": 3.952746214125807,
+      "learning_rate": 3.758099393141388e-06,
+      "loss": 0.5384,
+      "step": 993
+    },
+    {
+      "epoch": 0.3999195332931,
+      "grad_norm": 4.327315800519712,
+      "learning_rate": 3.755064182381516e-06,
+      "loss": 0.5144,
+      "step": 994
+    },
+    {
+      "epoch": 0.40032186682760007,
+      "grad_norm": 4.713412610694241,
+      "learning_rate": 3.752026496288588e-06,
+      "loss": 0.641,
+      "step": 995
+    },
+    {
+      "epoch": 0.40072420036210016,
+      "grad_norm": 3.9912240405323507,
+      "learning_rate": 3.74898634085376e-06,
+      "loss": 0.5646,
+      "step": 996
+    },
+    {
+      "epoch": 0.4011265338966003,
+      "grad_norm": 4.530973641905437,
+      "learning_rate": 3.7459437220730583e-06,
+      "loss": 0.5614,
+      "step": 997
+    },
+    {
+      "epoch": 0.4015288674311004,
+      "grad_norm": 6.864543852090396,
+      "learning_rate": 3.742898645947366e-06,
+      "loss": 0.5336,
+      "step": 998
+    },
+    {
+      "epoch": 0.4019312009656005,
+      "grad_norm": 4.429956976671279,
+      "learning_rate": 3.7398511184824144e-06,
+      "loss": 0.7004,
+      "step": 999
+    },
+    {
+      "epoch": 0.40233353450010056,
+      "grad_norm": 3.164463326468864,
+      "learning_rate": 3.7368011456887693e-06,
+      "loss": 0.4163,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4027358680346007,
+      "grad_norm": 3.7984416296897456,
+      "learning_rate": 3.7337487335818185e-06,
+      "loss": 0.5342,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4031382015691008,
+      "grad_norm": 4.186189188641626,
+      "learning_rate": 3.730693888181761e-06,
+      "loss": 0.6125,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4035405351036009,
+      "grad_norm": 3.8782636909404506,
+      "learning_rate": 3.7276366155135968e-06,
+      "loss": 0.5669,
+      "step": 1003
+    },
+    {
+      "epoch": 0.40394286863810097,
+      "grad_norm": 4.222464351010827,
+      "learning_rate": 3.7245769216071104e-06,
+      "loss": 0.5574,
+      "step": 1004
+    },
+    {
+      "epoch": 0.40434520217260106,
+      "grad_norm": 4.0429197887930135,
+      "learning_rate": 3.721514812496863e-06,
+      "loss": 0.5031,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4047475357071012,
+      "grad_norm": 4.015964628315664,
+      "learning_rate": 3.718450294222179e-06,
+      "loss": 0.5994,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4051498692416013,
+      "grad_norm": 3.846521589103569,
+      "learning_rate": 3.7153833728271356e-06,
+      "loss": 0.4928,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4055522027761014,
+      "grad_norm": 4.138445980201307,
+      "learning_rate": 3.712314054360547e-06,
+      "loss": 0.5722,
+      "step": 1008
+    },
+    {
+      "epoch": 0.40595453631060147,
+      "grad_norm": 4.476195845673172,
+      "learning_rate": 3.7092423448759577e-06,
+      "loss": 0.563,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4063568698451016,
+      "grad_norm": 3.786638148470317,
+      "learning_rate": 3.7061682504316264e-06,
+      "loss": 0.5181,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4067592033796017,
+      "grad_norm": 4.563364096910093,
+      "learning_rate": 3.7030917770905155e-06,
+      "loss": 0.5247,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4071615369141018,
+      "grad_norm": 4.276417493626447,
+      "learning_rate": 3.7000129309202804e-06,
+      "loss": 0.5815,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4075638704486019,
+      "grad_norm": 4.495474983317033,
+      "learning_rate": 3.696931717993256e-06,
+      "loss": 0.7269,
+      "step": 1013
+    },
+    {
+      "epoch": 0.407966203983102,
+      "grad_norm": 3.6238415974427527,
+      "learning_rate": 3.693848144386444e-06,
+      "loss": 0.5462,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4083685375176021,
+      "grad_norm": 4.262996632479673,
+      "learning_rate": 3.6907622161815027e-06,
+      "loss": 0.64,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4087708710521022,
+      "grad_norm": 3.8744140465414745,
+      "learning_rate": 3.687673939464736e-06,
+      "loss": 0.5281,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4091732045866023,
+      "grad_norm": 4.021193749407921,
+      "learning_rate": 3.6845833203270766e-06,
+      "loss": 0.5238,
+      "step": 1017
+    },
+    {
+      "epoch": 0.40957553812110237,
+      "grad_norm": 3.7942462225606084,
+      "learning_rate": 3.6814903648640787e-06,
+      "loss": 0.5274,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4099778716556025,
+      "grad_norm": 3.8584763272892193,
+      "learning_rate": 3.6783950791759053e-06,
+      "loss": 0.6652,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4103802051901026,
+      "grad_norm": 4.224914273954211,
+      "learning_rate": 3.6752974693673144e-06,
+      "loss": 0.6418,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4107825387246027,
+      "grad_norm": 3.611777243329413,
+      "learning_rate": 3.6721975415476475e-06,
+      "loss": 0.4478,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4111848722591028,
+      "grad_norm": 4.492308490256186,
+      "learning_rate": 3.669095301830816e-06,
+      "loss": 0.6383,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4115872057936029,
+      "grad_norm": 3.966328790708119,
+      "learning_rate": 3.6659907563352964e-06,
+      "loss": 0.5496,
+      "step": 1023
+    },
+    {
+      "epoch": 0.411989539328103,
+      "grad_norm": 4.331090241917638,
+      "learning_rate": 3.6628839111841067e-06,
+      "loss": 0.6138,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4123918728626031,
+      "grad_norm": 4.039380276956274,
+      "learning_rate": 3.6597747725048054e-06,
+      "loss": 0.514,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4127942063971032,
+      "grad_norm": 4.2678553912438515,
+      "learning_rate": 3.656663346429471e-06,
+      "loss": 0.5304,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4131965399316033,
+      "grad_norm": 3.6311274537700444,
+      "learning_rate": 3.6535496390946958e-06,
+      "loss": 0.5324,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4135988734661034,
+      "grad_norm": 4.179608207864191,
+      "learning_rate": 3.650433656641569e-06,
+      "loss": 0.5913,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4140012070006035,
+      "grad_norm": 3.8042489733382308,
+      "learning_rate": 3.647315405215668e-06,
+      "loss": 0.5007,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4144035405351036,
+      "grad_norm": 3.4526666984193404,
+      "learning_rate": 3.644194890967047e-06,
+      "loss": 0.4141,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4148058740696037,
+      "grad_norm": 4.657500083163389,
+      "learning_rate": 3.641072120050221e-06,
+      "loss": 0.6175,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4152082076041038,
+      "grad_norm": 3.5394692246958934,
+      "learning_rate": 3.637947098624156e-06,
+      "loss": 0.493,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4156105411386039,
+      "grad_norm": 3.6382852897066003,
+      "learning_rate": 3.6348198328522565e-06,
+      "loss": 0.4733,
+      "step": 1033
+    },
+    {
+      "epoch": 0.416012874673104,
+      "grad_norm": 4.701593231983809,
+      "learning_rate": 3.631690328902355e-06,
+      "loss": 0.6759,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4164152082076041,
+      "grad_norm": 4.126519542614592,
+      "learning_rate": 3.6285585929466982e-06,
+      "loss": 0.6065,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4168175417421042,
+      "grad_norm": 4.174303064525279,
+      "learning_rate": 3.6254246311619317e-06,
+      "loss": 0.6019,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4172198752766043,
+      "grad_norm": 4.369071505943041,
+      "learning_rate": 3.6222884497290937e-06,
+      "loss": 0.47,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4176222088111044,
+      "grad_norm": 4.547283701461554,
+      "learning_rate": 3.619150054833601e-06,
+      "loss": 0.5666,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4180245423456045,
+      "grad_norm": 6.593645962994248,
+      "learning_rate": 3.616009452665234e-06,
+      "loss": 0.4854,
+      "step": 1039
+    },
+    {
+      "epoch": 0.41842687588010463,
+      "grad_norm": 4.3350300541777145,
+      "learning_rate": 3.6128666494181274e-06,
+      "loss": 0.6522,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4188292094146047,
+      "grad_norm": 3.4916298319626233,
+      "learning_rate": 3.6097216512907574e-06,
+      "loss": 0.5449,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4192315429491048,
+      "grad_norm": 3.721480691979447,
+      "learning_rate": 3.6065744644859276e-06,
+      "loss": 0.4152,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4196338764836049,
+      "grad_norm": 4.419075622535455,
+      "learning_rate": 3.6034250952107598e-06,
+      "loss": 0.5998,
+      "step": 1043
+    },
+    {
+      "epoch": 0.420036210018105,
+      "grad_norm": 4.066379360835346,
+      "learning_rate": 3.6002735496766787e-06,
+      "loss": 0.4443,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4204385435526051,
+      "grad_norm": 5.277931637633872,
+      "learning_rate": 3.5971198340994035e-06,
+      "loss": 0.6578,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4208408770871052,
+      "grad_norm": 4.171317951485179,
+      "learning_rate": 3.5939639546989315e-06,
+      "loss": 0.5362,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4212432106216053,
+      "grad_norm": 3.6581061080480786,
+      "learning_rate": 3.5908059176995274e-06,
+      "loss": 0.4868,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4216455441561054,
+      "grad_norm": 4.089304060454962,
+      "learning_rate": 3.587645729329713e-06,
+      "loss": 0.4962,
+      "step": 1048
+    },
+    {
+      "epoch": 0.42204787769060553,
+      "grad_norm": 4.193929975116655,
+      "learning_rate": 3.584483395822252e-06,
+      "loss": 0.585,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4224502112251056,
+      "grad_norm": 4.4833727752476324,
+      "learning_rate": 3.5813189234141386e-06,
+      "loss": 0.5295,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4228525447596057,
+      "grad_norm": 4.131740789977084,
+      "learning_rate": 3.5781523183465868e-06,
+      "loss": 0.5408,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4232548782941058,
+      "grad_norm": 3.972093255632944,
+      "learning_rate": 3.574983586865015e-06,
+      "loss": 0.5256,
+      "step": 1052
+    },
+    {
+      "epoch": 0.42365721182860594,
+      "grad_norm": 4.619320510735413,
+      "learning_rate": 3.571812735219037e-06,
+      "loss": 0.6365,
+      "step": 1053
+    },
+    {
+      "epoch": 0.424059545363106,
+      "grad_norm": 4.543758819054478,
+      "learning_rate": 3.5686397696624486e-06,
+      "loss": 0.5358,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4244618788976061,
+      "grad_norm": 3.9488894694458305,
+      "learning_rate": 3.565464696453212e-06,
+      "loss": 0.5101,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4248642124321062,
+      "grad_norm": 3.9008766737779217,
+      "learning_rate": 3.5622875218534493e-06,
+      "loss": 0.6157,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4252665459666063,
+      "grad_norm": 4.072195384389267,
+      "learning_rate": 3.5591082521294264e-06,
+      "loss": 0.572,
+      "step": 1057
+    },
+    {
+      "epoch": 0.42566887950110643,
+      "grad_norm": 3.7407260188563622,
+      "learning_rate": 3.555926893551539e-06,
+      "loss": 0.4571,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4260712130356065,
+      "grad_norm": 4.030808733877317,
+      "learning_rate": 3.552743452394306e-06,
+      "loss": 0.5615,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4264735465701066,
+      "grad_norm": 3.833659387273303,
+      "learning_rate": 3.5495579349363517e-06,
+      "loss": 0.4978,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4268758801046067,
+      "grad_norm": 4.304988237070821,
+      "learning_rate": 3.5463703474603964e-06,
+      "loss": 0.6441,
+      "step": 1061
+    },
+    {
+      "epoch": 0.42727821363910684,
+      "grad_norm": 4.68914366889499,
+      "learning_rate": 3.5431806962532426e-06,
+      "loss": 0.5009,
+      "step": 1062
+    },
+    {
+      "epoch": 0.42768054717360693,
+      "grad_norm": 4.208417431491987,
+      "learning_rate": 3.539988987605763e-06,
+      "loss": 0.5556,
+      "step": 1063
+    },
+    {
+      "epoch": 0.428082880708107,
+      "grad_norm": 3.913021062753838,
+      "learning_rate": 3.5367952278128877e-06,
+      "loss": 0.3997,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4284852142426071,
+      "grad_norm": 3.975035687293758,
+      "learning_rate": 3.5335994231735927e-06,
+      "loss": 0.5394,
+      "step": 1065
+    },
+    {
+      "epoch": 0.42888754777710725,
+      "grad_norm": 3.969954958131227,
+      "learning_rate": 3.5304015799908875e-06,
+      "loss": 0.579,
+      "step": 1066
+    },
+    {
+      "epoch": 0.42928988131160734,
+      "grad_norm": 4.942462188783348,
+      "learning_rate": 3.5272017045718016e-06,
+      "loss": 0.6723,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4296922148461074,
+      "grad_norm": 4.354492013681113,
+      "learning_rate": 3.5239998032273727e-06,
+      "loss": 0.5724,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4300945483806075,
+      "grad_norm": 4.275789917585218,
+      "learning_rate": 3.520795882272634e-06,
+      "loss": 0.5592,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4304968819151076,
+      "grad_norm": 4.0220453706013535,
+      "learning_rate": 3.5175899480266023e-06,
+      "loss": 0.6004,
+      "step": 1070
+    },
+    {
+      "epoch": 0.43089921544960774,
+      "grad_norm": 3.956820759857449,
+      "learning_rate": 3.5143820068122647e-06,
+      "loss": 0.4667,
+      "step": 1071
+    },
+    {
+      "epoch": 0.43130154898410783,
+      "grad_norm": 3.9359268522062996,
+      "learning_rate": 3.5111720649565685e-06,
+      "loss": 0.4766,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4317038825186079,
+      "grad_norm": 3.9665074015294555,
+      "learning_rate": 3.507960128790402e-06,
+      "loss": 0.5496,
+      "step": 1073
+    },
+    {
+      "epoch": 0.432106216053108,
+      "grad_norm": 4.086635346156316,
+      "learning_rate": 3.5047462046485934e-06,
+      "loss": 0.4863,
+      "step": 1074
+    },
+    {
+      "epoch": 0.43250854958760815,
+      "grad_norm": 4.234886651798155,
+      "learning_rate": 3.501530298869886e-06,
+      "loss": 0.5373,
+      "step": 1075
+    },
+    {
+      "epoch": 0.43291088312210824,
+      "grad_norm": 4.578190826029045,
+      "learning_rate": 3.4983124177969353e-06,
+      "loss": 0.6952,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4333132166566083,
+      "grad_norm": 4.198352580671215,
+      "learning_rate": 3.495092567776291e-06,
+      "loss": 0.6753,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4337155501911084,
+      "grad_norm": 4.25314346947828,
+      "learning_rate": 3.4918707551583853e-06,
+      "loss": 0.6139,
+      "step": 1078
+    },
+    {
+      "epoch": 0.43411788372560856,
+      "grad_norm": 3.7809774120139084,
+      "learning_rate": 3.4886469862975224e-06,
+      "loss": 0.5401,
+      "step": 1079
+    },
+    {
+      "epoch": 0.43452021726010864,
+      "grad_norm": 3.861747118430867,
+      "learning_rate": 3.485421267551865e-06,
+      "loss": 0.4343,
+      "step": 1080
+    },
+    {
+      "epoch": 0.43492255079460873,
+      "grad_norm": 4.1971689225974655,
+      "learning_rate": 3.482193605283421e-06,
+      "loss": 0.4651,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4353248843291088,
+      "grad_norm": 4.719733540576249,
+      "learning_rate": 3.478964005858031e-06,
+      "loss": 0.6265,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4357272178636089,
+      "grad_norm": 4.199121895188385,
+      "learning_rate": 3.475732475645357e-06,
+      "loss": 0.5915,
+      "step": 1083
+    },
+    {
+      "epoch": 0.43612955139810905,
+      "grad_norm": 4.117901413415391,
+      "learning_rate": 3.4724990210188693e-06,
+      "loss": 0.5859,
+      "step": 1084
+    },
+    {
+      "epoch": 0.43653188493260914,
+      "grad_norm": 4.063314519073365,
+      "learning_rate": 3.4692636483558316e-06,
+      "loss": 0.5799,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4369342184671092,
+      "grad_norm": 4.291259902841789,
+      "learning_rate": 3.4660263640372926e-06,
+      "loss": 0.5972,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4373365520016093,
+      "grad_norm": 4.317671430467203,
+      "learning_rate": 3.462787174448071e-06,
+      "loss": 0.5818,
+      "step": 1087
+    },
+    {
+      "epoch": 0.43773888553610946,
+      "grad_norm": 4.041367856828536,
+      "learning_rate": 3.459546085976743e-06,
+      "loss": 0.5635,
+      "step": 1088
+    },
+    {
+      "epoch": 0.43814121907060954,
+      "grad_norm": 5.295724094479603,
+      "learning_rate": 3.45630310501563e-06,
+      "loss": 0.6712,
+      "step": 1089
+    },
+    {
+      "epoch": 0.43854355260510963,
+      "grad_norm": 4.415113705137869,
+      "learning_rate": 3.453058237960785e-06,
+      "loss": 0.578,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4389458861396097,
+      "grad_norm": 4.107191353105501,
+      "learning_rate": 3.4498114912119817e-06,
+      "loss": 0.5199,
+      "step": 1091
+    },
+    {
+      "epoch": 0.43934821967410986,
+      "grad_norm": 4.1779462968909336,
+      "learning_rate": 3.4465628711727022e-06,
+      "loss": 0.577,
+      "step": 1092
+    },
+    {
+      "epoch": 0.43975055320860995,
+      "grad_norm": 3.779499327289251,
+      "learning_rate": 3.4433123842501205e-06,
+      "loss": 0.4508,
+      "step": 1093
+    },
+    {
+      "epoch": 0.44015288674311004,
+      "grad_norm": 3.9872787051249077,
+      "learning_rate": 3.440060036855095e-06,
+      "loss": 0.5583,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4405552202776101,
+      "grad_norm": 4.317372927534622,
+      "learning_rate": 3.4368058354021526e-06,
+      "loss": 0.5136,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4409575538121102,
+      "grad_norm": 3.7148972800031053,
+      "learning_rate": 3.4335497863094773e-06,
+      "loss": 0.514,
+      "step": 1096
+    },
+    {
+      "epoch": 0.44135988734661036,
+      "grad_norm": 4.342233656787418,
+      "learning_rate": 3.430291895998896e-06,
+      "loss": 0.6742,
+      "step": 1097
+    },
+    {
+      "epoch": 0.44176222088111045,
+      "grad_norm": 4.799719138290185,
+      "learning_rate": 3.4270321708958686e-06,
+      "loss": 0.6306,
+      "step": 1098
+    },
+    {
+      "epoch": 0.44216455441561053,
+      "grad_norm": 3.9209860725209382,
+      "learning_rate": 3.4237706174294726e-06,
+      "loss": 0.5836,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4425668879501106,
+      "grad_norm": 4.127404724187905,
+      "learning_rate": 3.420507242032392e-06,
+      "loss": 0.5628,
+      "step": 1100
+    },
+    {
+      "epoch": 0.44296922148461076,
+      "grad_norm": 4.956274683013535,
+      "learning_rate": 3.4172420511409033e-06,
+      "loss": 0.5366,
+      "step": 1101
+    },
+    {
+      "epoch": 0.44337155501911085,
+      "grad_norm": 3.9527811079754285,
+      "learning_rate": 3.413975051194865e-06,
+      "loss": 0.4531,
+      "step": 1102
+    },
+    {
+      "epoch": 0.44377388855361094,
+      "grad_norm": 4.364141896413706,
+      "learning_rate": 3.4107062486377028e-06,
+      "loss": 0.4914,
+      "step": 1103
+    },
+    {
+      "epoch": 0.44417622208811103,
+      "grad_norm": 4.349108898092637,
+      "learning_rate": 3.4074356499163976e-06,
+      "loss": 0.7239,
+      "step": 1104
+    },
+    {
+      "epoch": 0.44457855562261117,
+      "grad_norm": 4.038974541606429,
+      "learning_rate": 3.4041632614814734e-06,
+      "loss": 0.6375,
+      "step": 1105
+    },
+    {
+      "epoch": 0.44498088915711126,
+      "grad_norm": 4.811612078810641,
+      "learning_rate": 3.400889089786984e-06,
+      "loss": 0.4963,
+      "step": 1106
+    },
+    {
+      "epoch": 0.44538322269161135,
+      "grad_norm": 4.07473806896573,
+      "learning_rate": 3.397613141290499e-06,
+      "loss": 0.575,
+      "step": 1107
+    },
+    {
+      "epoch": 0.44578555622611143,
+      "grad_norm": 4.160302353185681,
+      "learning_rate": 3.3943354224530934e-06,
+      "loss": 0.4766,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4461878897606115,
+      "grad_norm": 3.6900007830747934,
+      "learning_rate": 3.3910559397393335e-06,
+      "loss": 0.5544,
+      "step": 1109
+    },
+    {
+      "epoch": 0.44659022329511167,
+      "grad_norm": 4.333978444807443,
+      "learning_rate": 3.3877746996172645e-06,
+      "loss": 0.5637,
+      "step": 1110
+    },
+    {
+      "epoch": 0.44699255682961175,
+      "grad_norm": 3.768061403793603,
+      "learning_rate": 3.384491708558399e-06,
+      "loss": 0.4659,
+      "step": 1111
+    },
+    {
+      "epoch": 0.44739489036411184,
+      "grad_norm": 3.9713683928936208,
+      "learning_rate": 3.3812069730377007e-06,
+      "loss": 0.539,
+      "step": 1112
+    },
+    {
+      "epoch": 0.44779722389861193,
+      "grad_norm": 4.95129840549997,
+      "learning_rate": 3.3779204995335747e-06,
+      "loss": 0.6069,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4481995574331121,
+      "grad_norm": 4.316287155035261,
+      "learning_rate": 3.3746322945278543e-06,
+      "loss": 0.66,
+      "step": 1114
+    },
+    {
+      "epoch": 0.44860189096761216,
+      "grad_norm": 4.697629623928124,
+      "learning_rate": 3.3713423645057873e-06,
+      "loss": 0.4885,
+      "step": 1115
+    },
+    {
+      "epoch": 0.44900422450211225,
+      "grad_norm": 3.920242841924085,
+      "learning_rate": 3.368050715956025e-06,
+      "loss": 0.5213,
+      "step": 1116
+    },
+    {
+      "epoch": 0.44940655803661234,
+      "grad_norm": 4.521566910606868,
+      "learning_rate": 3.3647573553706052e-06,
+      "loss": 0.5743,
+      "step": 1117
+    },
+    {
+      "epoch": 0.4498088915711125,
+      "grad_norm": 4.082210637179168,
+      "learning_rate": 3.3614622892449465e-06,
+      "loss": 0.5361,
+      "step": 1118
+    },
+    {
+      "epoch": 0.45021122510561257,
+      "grad_norm": 4.017855379880608,
+      "learning_rate": 3.3581655240778277e-06,
+      "loss": 0.5492,
+      "step": 1119
+    },
+    {
+      "epoch": 0.45061355864011265,
+      "grad_norm": 3.3069156613524577,
+      "learning_rate": 3.35486706637138e-06,
+      "loss": 0.4142,
+      "step": 1120
+    },
+    {
+      "epoch": 0.45101589217461274,
+      "grad_norm": 4.230766496789279,
+      "learning_rate": 3.3515669226310716e-06,
+      "loss": 0.5883,
+      "step": 1121
+    },
+    {
+      "epoch": 0.45141822570911283,
+      "grad_norm": 3.932193721134766,
+      "learning_rate": 3.3482650993656983e-06,
+      "loss": 0.4732,
+      "step": 1122
+    },
+    {
+      "epoch": 0.451820559243613,
+      "grad_norm": 3.9297564952360577,
+      "learning_rate": 3.344961603087367e-06,
+      "loss": 0.5852,
+      "step": 1123
+    },
+    {
+      "epoch": 0.45222289277811306,
+      "grad_norm": 4.909401223605511,
+      "learning_rate": 3.341656440311484e-06,
+      "loss": 0.5709,
+      "step": 1124
+    },
+    {
+      "epoch": 0.45262522631261315,
+      "grad_norm": 3.6889862040375525,
+      "learning_rate": 3.3383496175567435e-06,
+      "loss": 0.5966,
+      "step": 1125
+    },
+    {
+      "epoch": 0.45302755984711324,
+      "grad_norm": 3.4699341807873694,
+      "learning_rate": 3.3350411413451125e-06,
+      "loss": 0.4692,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4534298933816134,
+      "grad_norm": 4.481843121210536,
+      "learning_rate": 3.3317310182018186e-06,
+      "loss": 0.6374,
+      "step": 1127
+    },
+    {
+      "epoch": 0.45383222691611347,
+      "grad_norm": 4.674277827776204,
+      "learning_rate": 3.328419254655339e-06,
+      "loss": 0.7174,
+      "step": 1128
+    },
+    {
+      "epoch": 0.45423456045061356,
+      "grad_norm": 3.8262305960434575,
+      "learning_rate": 3.325105857237386e-06,
+      "loss": 0.4618,
+      "step": 1129
+    },
+    {
+      "epoch": 0.45463689398511364,
+      "grad_norm": 3.9899678016482305,
+      "learning_rate": 3.3217908324828942e-06,
+      "loss": 0.5839,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4550392275196138,
+      "grad_norm": 3.39198458807276,
+      "learning_rate": 3.3184741869300073e-06,
+      "loss": 0.5191,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4554415610541139,
+      "grad_norm": 4.032286065452944,
+      "learning_rate": 3.3151559271200656e-06,
+      "loss": 0.4835,
+      "step": 1132
+    },
+    {
+      "epoch": 0.45584389458861396,
+      "grad_norm": 4.381591989574539,
+      "learning_rate": 3.3118360595975947e-06,
+      "loss": 0.5592,
+      "step": 1133
+    },
+    {
+      "epoch": 0.45624622812311405,
+      "grad_norm": 4.022213508423109,
+      "learning_rate": 3.3085145909102885e-06,
+      "loss": 0.5304,
+      "step": 1134
+    },
+    {
+      "epoch": 0.45664856165761414,
+      "grad_norm": 3.834367329162151,
+      "learning_rate": 3.305191527609e-06,
+      "loss": 0.4678,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4570508951921143,
+      "grad_norm": 4.448203020700801,
+      "learning_rate": 3.3018668762477294e-06,
+      "loss": 0.5172,
+      "step": 1136
+    },
+    {
+      "epoch": 0.45745322872661437,
+      "grad_norm": 4.128965375687235,
+      "learning_rate": 3.2985406433836055e-06,
+      "loss": 0.4742,
+      "step": 1137
+    },
+    {
+      "epoch": 0.45785556226111446,
+      "grad_norm": 4.129504251009841,
+      "learning_rate": 3.295212835576878e-06,
+      "loss": 0.5873,
+      "step": 1138
+    },
+    {
+      "epoch": 0.45825789579561454,
+      "grad_norm": 4.422224105540513,
+      "learning_rate": 3.2918834593909028e-06,
+      "loss": 0.6049,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4586602293301147,
+      "grad_norm": 4.434937349809798,
+      "learning_rate": 3.2885525213921286e-06,
+      "loss": 0.5708,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4590625628646148,
+      "grad_norm": 3.668465429492197,
+      "learning_rate": 3.2852200281500855e-06,
+      "loss": 0.4559,
+      "step": 1141
+    },
+    {
+      "epoch": 0.45946489639911486,
+      "grad_norm": 4.232532968956556,
+      "learning_rate": 3.2818859862373696e-06,
+      "loss": 0.5721,
+      "step": 1142
+    },
+    {
+      "epoch": 0.45986722993361495,
+      "grad_norm": 4.7338258325238085,
+      "learning_rate": 3.2785504022296316e-06,
+      "loss": 0.6157,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4602695634681151,
+      "grad_norm": 3.7791426452155723,
+      "learning_rate": 3.2752132827055646e-06,
+      "loss": 0.5458,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4606718970026152,
+      "grad_norm": 4.1019637445172314,
+      "learning_rate": 3.2718746342468894e-06,
+      "loss": 0.5972,
+      "step": 1145
+    },
+    {
+      "epoch": 0.46107423053711527,
+      "grad_norm": 5.277495798462077,
+      "learning_rate": 3.268534463438342e-06,
+      "loss": 0.8266,
+      "step": 1146
+    },
+    {
+      "epoch": 0.46147656407161536,
+      "grad_norm": 4.184453273932426,
+      "learning_rate": 3.2651927768676626e-06,
+      "loss": 0.4562,
+      "step": 1147
+    },
+    {
+      "epoch": 0.46187889760611545,
+      "grad_norm": 4.45907273350243,
+      "learning_rate": 3.261849581125579e-06,
+      "loss": 0.6532,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4622812311406156,
+      "grad_norm": 3.965076229787784,
+      "learning_rate": 3.258504882805796e-06,
+      "loss": 0.5247,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4626835646751157,
+      "grad_norm": 4.461055711184354,
+      "learning_rate": 3.2551586885049814e-06,
+      "loss": 0.4588,
+      "step": 1150
+    },
+    {
+      "epoch": 0.46308589820961576,
+      "grad_norm": 4.3124706302870255,
+      "learning_rate": 3.2518110048227553e-06,
+      "loss": 0.7004,
+      "step": 1151
+    },
+    {
+      "epoch": 0.46348823174411585,
+      "grad_norm": 4.326890541788323,
+      "learning_rate": 3.2484618383616735e-06,
+      "loss": 0.5989,
+      "step": 1152
+    },
+    {
+      "epoch": 0.463890565278616,
+      "grad_norm": 4.323105374837098,
+      "learning_rate": 3.2451111957272173e-06,
+      "loss": 0.5451,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4642928988131161,
+      "grad_norm": 4.134824677878268,
+      "learning_rate": 3.241759083527779e-06,
+      "loss": 0.5589,
+      "step": 1154
+    },
+    {
+      "epoch": 0.46469523234761617,
+      "grad_norm": 4.0981640141993365,
+      "learning_rate": 3.238405508374649e-06,
+      "loss": 0.5771,
+      "step": 1155
+    },
+    {
+      "epoch": 0.46509756588211626,
+      "grad_norm": 3.7455178072585644,
+      "learning_rate": 3.2350504768820036e-06,
+      "loss": 0.5125,
+      "step": 1156
+    },
+    {
+      "epoch": 0.46549989941661635,
+      "grad_norm": 3.72359460364635,
+      "learning_rate": 3.23169399566689e-06,
+      "loss": 0.4213,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4659022329511165,
+      "grad_norm": 3.509972924609966,
+      "learning_rate": 3.228336071349218e-06,
+      "loss": 0.4757,
+      "step": 1158
+    },
+    {
+      "epoch": 0.4663045664856166,
+      "grad_norm": 4.028400275439829,
+      "learning_rate": 3.2249767105517395e-06,
+      "loss": 0.5022,
+      "step": 1159
+    },
+    {
+      "epoch": 0.46670690002011667,
+      "grad_norm": 4.150972255373417,
+      "learning_rate": 3.221615919900042e-06,
+      "loss": 0.6257,
+      "step": 1160
+    },
+    {
+      "epoch": 0.46710923355461675,
+      "grad_norm": 4.494189407683215,
+      "learning_rate": 3.218253706022533e-06,
+      "loss": 0.5914,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4675115670891169,
+      "grad_norm": 3.558180537366331,
+      "learning_rate": 3.2148900755504253e-06,
+      "loss": 0.5102,
+      "step": 1162
+    },
+    {
+      "epoch": 0.467913900623617,
+      "grad_norm": 3.438406963531958,
+      "learning_rate": 3.2115250351177274e-06,
+      "loss": 0.515,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4683162341581171,
+      "grad_norm": 4.323925078750027,
+      "learning_rate": 3.208158591361228e-06,
+      "loss": 0.5109,
+      "step": 1164
+    },
+    {
+      "epoch": 0.46871856769261716,
+      "grad_norm": 4.1534707963201285,
+      "learning_rate": 3.204790750920484e-06,
+      "loss": 0.5736,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4691209012271173,
+      "grad_norm": 3.6871720181954406,
+      "learning_rate": 3.2014215204378064e-06,
+      "loss": 0.5765,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4695232347616174,
+      "grad_norm": 3.9220478852176264,
+      "learning_rate": 3.1980509065582476e-06,
+      "loss": 0.5545,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4699255682961175,
+      "grad_norm": 4.603512689826395,
+      "learning_rate": 3.194678915929589e-06,
+      "loss": 0.5587,
+      "step": 1168
+    },
+    {
+      "epoch": 0.47032790183061757,
+      "grad_norm": 4.0356910482849075,
+      "learning_rate": 3.1913055552023263e-06,
+      "loss": 0.4642,
+      "step": 1169
+    },
+    {
+      "epoch": 0.47073023536511766,
+      "grad_norm": 4.251943286279792,
+      "learning_rate": 3.1879308310296586e-06,
+      "loss": 0.5496,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4711325688996178,
+      "grad_norm": 4.189042957707438,
+      "learning_rate": 3.1845547500674737e-06,
+      "loss": 0.5646,
+      "step": 1171
+    },
+    {
+      "epoch": 0.4715349024341179,
+      "grad_norm": 3.3326084801089353,
+      "learning_rate": 3.1811773189743355e-06,
+      "loss": 0.4837,
+      "step": 1172
+    },
+    {
+      "epoch": 0.471937235968618,
+      "grad_norm": 4.523273563587766,
+      "learning_rate": 3.17779854441147e-06,
+      "loss": 0.6807,
+      "step": 1173
+    },
+    {
+      "epoch": 0.47233956950311806,
+      "grad_norm": 3.7232216855517315,
+      "learning_rate": 3.1744184330427543e-06,
+      "loss": 0.4492,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4727419030376182,
+      "grad_norm": 3.8162014027986437,
+      "learning_rate": 3.1710369915347e-06,
+      "loss": 0.5626,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4731442365721183,
+      "grad_norm": 4.575379469711398,
+      "learning_rate": 3.1676542265564443e-06,
+      "loss": 0.6358,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4735465701066184,
+      "grad_norm": 4.563047737373793,
+      "learning_rate": 3.1642701447797324e-06,
+      "loss": 0.6282,
+      "step": 1177
+    },
+    {
+      "epoch": 0.47394890364111847,
+      "grad_norm": 4.254463762469985,
+      "learning_rate": 3.1608847528789096e-06,
+      "loss": 0.4974,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4743512371756186,
+      "grad_norm": 4.290561801295637,
+      "learning_rate": 3.1574980575309015e-06,
+      "loss": 0.6116,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4747535707101187,
+      "grad_norm": 4.276362615412023,
+      "learning_rate": 3.154110065415208e-06,
+      "loss": 0.5274,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4751559042446188,
+      "grad_norm": 3.8389354369931374,
+      "learning_rate": 3.150720783213883e-06,
+      "loss": 0.5338,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4755582377791189,
+      "grad_norm": 4.448041403696648,
+      "learning_rate": 3.1473302176115283e-06,
+      "loss": 0.5265,
+      "step": 1182
+    },
+    {
+      "epoch": 0.47596057131361896,
+      "grad_norm": 4.46050378901124,
+      "learning_rate": 3.1439383752952753e-06,
+      "loss": 0.6299,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4763629048481191,
+      "grad_norm": 4.111077775261664,
+      "learning_rate": 3.140545262954772e-06,
+      "loss": 0.5723,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4767652383826192,
+      "grad_norm": 3.678492409328739,
+      "learning_rate": 3.137150887282174e-06,
+      "loss": 0.4178,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4771675719171193,
+      "grad_norm": 3.661689522078919,
+      "learning_rate": 3.1337552549721262e-06,
+      "loss": 0.4658,
+      "step": 1186
+    },
+    {
+      "epoch": 0.47756990545161937,
+      "grad_norm": 4.250351645694971,
+      "learning_rate": 3.130358372721754e-06,
+      "loss": 0.5614,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4779722389861195,
+      "grad_norm": 3.599583256588181,
+      "learning_rate": 3.126960247230646e-06,
+      "loss": 0.4569,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4783745725206196,
+      "grad_norm": 3.821896657809243,
+      "learning_rate": 3.1235608852008446e-06,
+      "loss": 0.4834,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4787769060551197,
+      "grad_norm": 3.9335780159969906,
+      "learning_rate": 3.1201602933368308e-06,
+      "loss": 0.5152,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4791792395896198,
+      "grad_norm": 4.770540317196732,
+      "learning_rate": 3.116758478345509e-06,
+      "loss": 0.5948,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4795815731241199,
+      "grad_norm": 3.998481220511283,
+      "learning_rate": 3.1133554469361976e-06,
+      "loss": 0.4881,
+      "step": 1192
+    },
+    {
+      "epoch": 0.47998390665862,
+      "grad_norm": 4.208648289802026,
+      "learning_rate": 3.109951205820615e-06,
+      "loss": 0.4933,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4803862401931201,
+      "grad_norm": 3.752607346638518,
+      "learning_rate": 3.1065457617128635e-06,
+      "loss": 0.547,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4807885737276202,
+      "grad_norm": 4.151251812002077,
+      "learning_rate": 3.1031391213294203e-06,
+      "loss": 0.5015,
+      "step": 1195
+    },
+    {
+      "epoch": 0.48119090726212027,
+      "grad_norm": 4.7037289072786095,
+      "learning_rate": 3.0997312913891196e-06,
+      "loss": 0.6795,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4815932407966204,
+      "grad_norm": 4.364147422750159,
+      "learning_rate": 3.0963222786131435e-06,
+      "loss": 0.5422,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4819955743311205,
+      "grad_norm": 3.301076213262608,
+      "learning_rate": 3.0929120897250066e-06,
+      "loss": 0.4646,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4823979078656206,
+      "grad_norm": 3.803770968997955,
+      "learning_rate": 3.0895007314505415e-06,
+      "loss": 0.4579,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4828002414001207,
+      "grad_norm": 3.6319529462360816,
+      "learning_rate": 3.0860882105178897e-06,
+      "loss": 0.5152,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4832025749346208,
+      "grad_norm": 3.7950010562493453,
+      "learning_rate": 3.082674533657484e-06,
+      "loss": 0.588,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4836049084691209,
+      "grad_norm": 3.9284108228666796,
+      "learning_rate": 3.079259707602038e-06,
+      "loss": 0.5565,
+      "step": 1202
+    },
+    {
+      "epoch": 0.484007242003621,
+      "grad_norm": 4.764319789283065,
+      "learning_rate": 3.0758437390865314e-06,
+      "loss": 0.6115,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4844095755381211,
+      "grad_norm": 3.781006793375389,
+      "learning_rate": 3.0724266348481966e-06,
+      "loss": 0.5239,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4848119090726212,
+      "grad_norm": 4.259150542468107,
+      "learning_rate": 3.069008401626507e-06,
+      "loss": 0.5505,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4852142426071213,
+      "grad_norm": 4.222901377467788,
+      "learning_rate": 3.065589046163162e-06,
+      "loss": 0.5425,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4856165761416214,
+      "grad_norm": 3.783266465549031,
+      "learning_rate": 3.062168575202075e-06,
+      "loss": 0.542,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4860189096761215,
+      "grad_norm": 4.428497795054238,
+      "learning_rate": 3.0587469954893584e-06,
+      "loss": 0.6567,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4864212432106216,
+      "grad_norm": 3.691837740063112,
+      "learning_rate": 3.0553243137733125e-06,
+      "loss": 0.4558,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4868235767451217,
+      "grad_norm": 4.073296743584824,
+      "learning_rate": 3.051900536804411e-06,
+      "loss": 0.5664,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4872259102796218,
+      "grad_norm": 3.7192683213034234,
+      "learning_rate": 3.0484756713352858e-06,
+      "loss": 0.494,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4876282438141219,
+      "grad_norm": 3.6377595381026557,
+      "learning_rate": 3.045049724120718e-06,
+      "loss": 0.4885,
+      "step": 1212
+    },
+    {
+      "epoch": 0.488030577348622,
+      "grad_norm": 4.338683717827157,
+      "learning_rate": 3.0416227019176208e-06,
+      "loss": 0.6165,
+      "step": 1213
+    },
+    {
+      "epoch": 0.48843291088312213,
+      "grad_norm": 4.374868800699678,
+      "learning_rate": 3.0381946114850293e-06,
+      "loss": 0.5755,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4888352444176222,
+      "grad_norm": 4.54077234147531,
+      "learning_rate": 3.034765459584083e-06,
+      "loss": 0.5954,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4892375779521223,
+      "grad_norm": 4.166268412885333,
+      "learning_rate": 3.031335252978017e-06,
+      "loss": 0.6657,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4896399114866224,
+      "grad_norm": 3.675310094808539,
+      "learning_rate": 3.027903998432145e-06,
+      "loss": 0.4676,
+      "step": 1217
+    },
+    {
+      "epoch": 0.49004224502112254,
+      "grad_norm": 4.091571307318168,
+      "learning_rate": 3.0244717027138494e-06,
+      "loss": 0.579,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4904445785556226,
+      "grad_norm": 4.2620047470347595,
+      "learning_rate": 3.021038372592563e-06,
+      "loss": 0.577,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4908469120901227,
+      "grad_norm": 3.687909147915886,
+      "learning_rate": 3.017604014839762e-06,
+      "loss": 0.557,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4912492456246228,
+      "grad_norm": 4.003092692480048,
+      "learning_rate": 3.014168636228948e-06,
+      "loss": 0.5517,
+      "step": 1221
+    },
+    {
+      "epoch": 0.4916515791591229,
+      "grad_norm": 4.349298145062102,
+      "learning_rate": 3.010732243535636e-06,
+      "loss": 0.5991,
+      "step": 1222
+    },
+    {
+      "epoch": 0.49205391269362303,
+      "grad_norm": 3.9645619754575963,
+      "learning_rate": 3.0072948435373408e-06,
+      "loss": 0.5145,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4924562462281231,
+      "grad_norm": 4.216377733109406,
+      "learning_rate": 3.003856443013564e-06,
+      "loss": 0.6017,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4928585797626232,
+      "grad_norm": 3.9206543304610584,
+      "learning_rate": 3.0004170487457813e-06,
+      "loss": 0.5468,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4932609132971233,
+      "grad_norm": 4.196664240626691,
+      "learning_rate": 2.996976667517427e-06,
+      "loss": 0.5191,
+      "step": 1226
+    },
+    {
+      "epoch": 0.49366324683162344,
+      "grad_norm": 3.7775234111981493,
+      "learning_rate": 2.9935353061138833e-06,
+      "loss": 0.5094,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4940655803661235,
+      "grad_norm": 4.412452469732304,
+      "learning_rate": 2.990092971322464e-06,
+      "loss": 0.566,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4944679139006236,
+      "grad_norm": 3.967665512410053,
+      "learning_rate": 2.986649669932405e-06,
+      "loss": 0.5005,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4948702474351237,
+      "grad_norm": 3.8061725506913353,
+      "learning_rate": 2.983205408734846e-06,
+      "loss": 0.5214,
+      "step": 1230
+    },
+    {
+      "epoch": 0.49527258096962384,
+      "grad_norm": 3.615684737615807,
+      "learning_rate": 2.9797601945228214e-06,
+      "loss": 0.4647,
+      "step": 1231
+    },
+    {
+      "epoch": 0.49567491450412393,
+      "grad_norm": 4.27866020604673,
+      "learning_rate": 2.9763140340912463e-06,
+      "loss": 0.6409,
+      "step": 1232
+    },
+    {
+      "epoch": 0.496077248038624,
+      "grad_norm": 3.596076423790147,
+      "learning_rate": 2.972866934236898e-06,
+      "loss": 0.412,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4964795815731241,
+      "grad_norm": 4.0355620490752955,
+      "learning_rate": 2.9694189017584097e-06,
+      "loss": 0.5596,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4968819151076242,
+      "grad_norm": 4.148051125186017,
+      "learning_rate": 2.965969943456254e-06,
+      "loss": 0.5191,
+      "step": 1235
+    },
+    {
+      "epoch": 0.49728424864212434,
+      "grad_norm": 4.2692044767988,
+      "learning_rate": 2.962520066132728e-06,
+      "loss": 0.5066,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4976865821766244,
+      "grad_norm": 4.28381423937242,
+      "learning_rate": 2.9590692765919436e-06,
+      "loss": 0.598,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4980889157111245,
+      "grad_norm": 4.175358078657294,
+      "learning_rate": 2.955617581639809e-06,
+      "loss": 0.5953,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4984912492456246,
+      "grad_norm": 4.5191335184254955,
+      "learning_rate": 2.952164988084021e-06,
+      "loss": 0.6236,
+      "step": 1239
+    },
+    {
+      "epoch": 0.49889358278012474,
+      "grad_norm": 4.126996114146587,
+      "learning_rate": 2.9487115027340458e-06,
+      "loss": 0.5455,
+      "step": 1240
+    },
+    {
+      "epoch": 0.49929591631462483,
+      "grad_norm": 3.978032878939077,
+      "learning_rate": 2.9452571324011103e-06,
+      "loss": 0.5658,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4996982498491249,
+      "grad_norm": 4.494778767380183,
+      "learning_rate": 2.941801883898187e-06,
+      "loss": 0.6639,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5001005833836251,
+      "grad_norm": 3.7291179386624376,
+      "learning_rate": 2.9383457640399794e-06,
+      "loss": 0.4724,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5005029169181251,
+      "grad_norm": 4.0158904670389175,
+      "learning_rate": 2.9348887796429104e-06,
+      "loss": 0.5611,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5009052504526252,
+      "grad_norm": 3.93870398764487,
+      "learning_rate": 2.9314309375251065e-06,
+      "loss": 0.5596,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5013075839871253,
+      "grad_norm": 4.111802544399985,
+      "learning_rate": 2.9279722445063883e-06,
+      "loss": 0.5872,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5017099175216254,
+      "grad_norm": 3.948101448350297,
+      "learning_rate": 2.9245127074082535e-06,
+      "loss": 0.5131,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5021122510561256,
+      "grad_norm": 4.345129844652909,
+      "learning_rate": 2.9210523330538625e-06,
+      "loss": 0.6389,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5025145845906256,
+      "grad_norm": 3.6367326824064112,
+      "learning_rate": 2.9175911282680307e-06,
+      "loss": 0.4518,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5029169181251257,
+      "grad_norm": 4.027476673290246,
+      "learning_rate": 2.914129099877208e-06,
+      "loss": 0.5234,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5033192516596259,
+      "grad_norm": 3.450546244138055,
+      "learning_rate": 2.9106662547094724e-06,
+      "loss": 0.3807,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5037215851941259,
+      "grad_norm": 3.696528336339893,
+      "learning_rate": 2.907202599594508e-06,
+      "loss": 0.3884,
+      "step": 1252
+    },
+    {
+      "epoch": 0.504123918728626,
+      "grad_norm": 4.252344685184106,
+      "learning_rate": 2.9037381413636e-06,
+      "loss": 0.6935,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5045262522631261,
+      "grad_norm": 4.336362311318111,
+      "learning_rate": 2.900272886849616e-06,
+      "loss": 0.5331,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5049285857976262,
+      "grad_norm": 3.6842567382976688,
+      "learning_rate": 2.8968068428869955e-06,
+      "loss": 0.5545,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5053309193321264,
+      "grad_norm": 3.699409904804457,
+      "learning_rate": 2.8933400163117337e-06,
+      "loss": 0.3754,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5057332528666264,
+      "grad_norm": 4.205702275086939,
+      "learning_rate": 2.889872413961369e-06,
+      "loss": 0.6633,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5061355864011265,
+      "grad_norm": 3.6639853190462626,
+      "learning_rate": 2.886404042674972e-06,
+      "loss": 0.3892,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5065379199356266,
+      "grad_norm": 3.5640332071999765,
+      "learning_rate": 2.882934909293127e-06,
+      "loss": 0.4084,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5069402534701267,
+      "grad_norm": 4.0471975082058025,
+      "learning_rate": 2.8794650206579234e-06,
+      "loss": 0.5812,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5073425870046269,
+      "grad_norm": 4.0654643808596616,
+      "learning_rate": 2.875994383612939e-06,
+      "loss": 0.568,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5077449205391269,
+      "grad_norm": 4.330240433028303,
+      "learning_rate": 2.872523005003229e-06,
+      "loss": 0.5165,
+      "step": 1262
+    },
+    {
+      "epoch": 0.508147254073627,
+      "grad_norm": 4.015050113056623,
+      "learning_rate": 2.869050891675309e-06,
+      "loss": 0.5002,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5085495876081272,
+      "grad_norm": 4.393983632365189,
+      "learning_rate": 2.8655780504771476e-06,
+      "loss": 0.573,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5089519211426272,
+      "grad_norm": 3.193554396038774,
+      "learning_rate": 2.8621044882581433e-06,
+      "loss": 0.3709,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5093542546771274,
+      "grad_norm": 3.3841882285013813,
+      "learning_rate": 2.8586302118691223e-06,
+      "loss": 0.4387,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5097565882116274,
+      "grad_norm": 3.9970258932238742,
+      "learning_rate": 2.8551552281623157e-06,
+      "loss": 0.5225,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5101589217461275,
+      "grad_norm": 4.513606814044003,
+      "learning_rate": 2.85167954399135e-06,
+      "loss": 0.5528,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5105612552806277,
+      "grad_norm": 4.684727392976221,
+      "learning_rate": 2.848203166211235e-06,
+      "loss": 0.6793,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5109635888151277,
+      "grad_norm": 4.222583621631339,
+      "learning_rate": 2.8447261016783474e-06,
+      "loss": 0.6053,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5113659223496279,
+      "grad_norm": 3.8724976223748744,
+      "learning_rate": 2.841248357250418e-06,
+      "loss": 0.4797,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5117682558841279,
+      "grad_norm": 3.5925982833624652,
+      "learning_rate": 2.8377699397865184e-06,
+      "loss": 0.6524,
+      "step": 1272
+    },
+    {
+      "epoch": 0.512170589418628,
+      "grad_norm": 4.4527497576787285,
+      "learning_rate": 2.8342908561470496e-06,
+      "loss": 0.6426,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5125729229531282,
+      "grad_norm": 3.758694762638942,
+      "learning_rate": 2.8308111131937237e-06,
+      "loss": 0.4863,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5129752564876282,
+      "grad_norm": 3.432356742958401,
+      "learning_rate": 2.827330717789555e-06,
+      "loss": 0.5223,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5133775900221283,
+      "grad_norm": 4.33004996309828,
+      "learning_rate": 2.8238496767988433e-06,
+      "loss": 0.6004,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5137799235566285,
+      "grad_norm": 4.200608717420535,
+      "learning_rate": 2.820367997087164e-06,
+      "loss": 0.6228,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5141822570911285,
+      "grad_norm": 4.0453448815375035,
+      "learning_rate": 2.8168856855213493e-06,
+      "loss": 0.6242,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5145845906256287,
+      "grad_norm": 4.227997150733838,
+      "learning_rate": 2.81340274896948e-06,
+      "loss": 0.6175,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5149869241601287,
+      "grad_norm": 3.68997651052952,
+      "learning_rate": 2.8099191943008673e-06,
+      "loss": 0.3891,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5153892576946288,
+      "grad_norm": 3.983876769382806,
+      "learning_rate": 2.806435028386044e-06,
+      "loss": 0.4676,
+      "step": 1281
+    },
+    {
+      "epoch": 0.515791591229129,
+      "grad_norm": 3.7821657559918362,
+      "learning_rate": 2.8029502580967467e-06,
+      "loss": 0.5712,
+      "step": 1282
+    },
+    {
+      "epoch": 0.516193924763629,
+      "grad_norm": 4.613891160258615,
+      "learning_rate": 2.7994648903059034e-06,
+      "loss": 0.6892,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5165962582981292,
+      "grad_norm": 4.260048115855587,
+      "learning_rate": 2.7959789318876235e-06,
+      "loss": 0.587,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5169985918326292,
+      "grad_norm": 3.2651956443139905,
+      "learning_rate": 2.792492389717178e-06,
+      "loss": 0.4202,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5174009253671293,
+      "grad_norm": 3.8441567338972136,
+      "learning_rate": 2.7890052706709913e-06,
+      "loss": 0.5274,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5178032589016295,
+      "grad_norm": 4.244436141220556,
+      "learning_rate": 2.785517581626624e-06,
+      "loss": 0.5557,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5182055924361295,
+      "grad_norm": 3.777185766227343,
+      "learning_rate": 2.782029329462763e-06,
+      "loss": 0.5508,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5186079259706297,
+      "grad_norm": 4.024624316501224,
+      "learning_rate": 2.778540521059204e-06,
+      "loss": 0.4091,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5190102595051298,
+      "grad_norm": 3.429656579938711,
+      "learning_rate": 2.775051163296839e-06,
+      "loss": 0.373,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5194125930396298,
+      "grad_norm": 4.115516308278128,
+      "learning_rate": 2.771561263057647e-06,
+      "loss": 0.6296,
+      "step": 1291
+    },
+    {
+      "epoch": 0.51981492657413,
+      "grad_norm": 4.5340536303593,
+      "learning_rate": 2.7680708272246738e-06,
+      "loss": 0.5478,
+      "step": 1292
+    },
+    {
+      "epoch": 0.52021726010863,
+      "grad_norm": 4.1573055283514675,
+      "learning_rate": 2.7645798626820226e-06,
+      "loss": 0.5928,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5206195936431302,
+      "grad_norm": 3.5540381627057926,
+      "learning_rate": 2.7610883763148382e-06,
+      "loss": 0.4204,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5210219271776303,
+      "grad_norm": 3.7082383118644406,
+      "learning_rate": 2.7575963750092976e-06,
+      "loss": 0.5897,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5214242607121303,
+      "grad_norm": 3.670768517938044,
+      "learning_rate": 2.7541038656525906e-06,
+      "loss": 0.4524,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5218265942466305,
+      "grad_norm": 3.9599618494551336,
+      "learning_rate": 2.7506108551329084e-06,
+      "loss": 0.5334,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5222289277811305,
+      "grad_norm": 3.8332760152691425,
+      "learning_rate": 2.747117350339434e-06,
+      "loss": 0.5027,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5226312613156306,
+      "grad_norm": 3.5387447785248463,
+      "learning_rate": 2.743623358162322e-06,
+      "loss": 0.4453,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5230335948501308,
+      "grad_norm": 3.665534245478194,
+      "learning_rate": 2.74012888549269e-06,
+      "loss": 0.4541,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5234359283846308,
+      "grad_norm": 4.118363799618076,
+      "learning_rate": 2.736633939222604e-06,
+      "loss": 0.4863,
+      "step": 1301
+    },
+    {
+      "epoch": 0.523838261919131,
+      "grad_norm": 4.07159865432201,
+      "learning_rate": 2.733138526245061e-06,
+      "loss": 0.5836,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5242405954536311,
+      "grad_norm": 4.9413620392110955,
+      "learning_rate": 2.729642653453981e-06,
+      "loss": 0.6331,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5246429289881311,
+      "grad_norm": 4.349366316003809,
+      "learning_rate": 2.7261463277441913e-06,
+      "loss": 0.6584,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5250452625226313,
+      "grad_norm": 4.013818084372327,
+      "learning_rate": 2.7226495560114097e-06,
+      "loss": 0.6444,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5254475960571313,
+      "grad_norm": 3.9731448823583158,
+      "learning_rate": 2.719152345152237e-06,
+      "loss": 0.4717,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5258499295916315,
+      "grad_norm": 3.836192951376196,
+      "learning_rate": 2.7156547020641377e-06,
+      "loss": 0.4813,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5262522631261316,
+      "grad_norm": 4.526603353793837,
+      "learning_rate": 2.7121566336454295e-06,
+      "loss": 0.6223,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5266545966606316,
+      "grad_norm": 4.542236066581728,
+      "learning_rate": 2.70865814679527e-06,
+      "loss": 0.541,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5270569301951318,
+      "grad_norm": 4.477320709723044,
+      "learning_rate": 2.70515924841364e-06,
+      "loss": 0.4898,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5274592637296318,
+      "grad_norm": 4.188969544607867,
+      "learning_rate": 2.7016599454013324e-06,
+      "loss": 0.478,
+      "step": 1311
+    },
+    {
+      "epoch": 0.527861597264132,
+      "grad_norm": 3.57921084312417,
+      "learning_rate": 2.6981602446599408e-06,
+      "loss": 0.4797,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5282639307986321,
+      "grad_norm": 3.4828294543575784,
+      "learning_rate": 2.69466015309184e-06,
+      "loss": 0.4464,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5286662643331321,
+      "grad_norm": 3.6542821945407002,
+      "learning_rate": 2.691159677600177e-06,
+      "loss": 0.5241,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5290685978676323,
+      "grad_norm": 4.138582178147402,
+      "learning_rate": 2.6876588250888556e-06,
+      "loss": 0.4976,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5294709314021324,
+      "grad_norm": 4.156252344667509,
+      "learning_rate": 2.6841576024625247e-06,
+      "loss": 0.5346,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5298732649366324,
+      "grad_norm": 3.9783539180873118,
+      "learning_rate": 2.680656016626561e-06,
+      "loss": 0.5054,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5302755984711326,
+      "grad_norm": 4.167749289113631,
+      "learning_rate": 2.677154074487057e-06,
+      "loss": 0.5308,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5306779320056326,
+      "grad_norm": 3.6243864679157536,
+      "learning_rate": 2.6736517829508124e-06,
+      "loss": 0.4969,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5310802655401328,
+      "grad_norm": 4.080652153118834,
+      "learning_rate": 2.6701491489253107e-06,
+      "loss": 0.5166,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5314825990746329,
+      "grad_norm": 4.386897362383017,
+      "learning_rate": 2.6666461793187143e-06,
+      "loss": 0.4953,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5318849326091329,
+      "grad_norm": 4.286991486124671,
+      "learning_rate": 2.6631428810398453e-06,
+      "loss": 0.4804,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5322872661436331,
+      "grad_norm": 3.996180824655324,
+      "learning_rate": 2.659639260998176e-06,
+      "loss": 0.5182,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5326895996781331,
+      "grad_norm": 4.305112080018121,
+      "learning_rate": 2.6561353261038127e-06,
+      "loss": 0.6548,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5330919332126333,
+      "grad_norm": 3.6190226327616557,
+      "learning_rate": 2.6526310832674807e-06,
+      "loss": 0.5383,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5334942667471334,
+      "grad_norm": 4.316996697367821,
+      "learning_rate": 2.6491265394005157e-06,
+      "loss": 0.5516,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5338966002816334,
+      "grad_norm": 4.0648368660213325,
+      "learning_rate": 2.645621701414845e-06,
+      "loss": 0.6482,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5342989338161336,
+      "grad_norm": 3.7739876249793047,
+      "learning_rate": 2.6421165762229777e-06,
+      "loss": 0.4783,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5347012673506337,
+      "grad_norm": 3.789987275308021,
+      "learning_rate": 2.6386111707379873e-06,
+      "loss": 0.4525,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5351036008851338,
+      "grad_norm": 3.630575939542573,
+      "learning_rate": 2.635105491873502e-06,
+      "loss": 0.5329,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5355059344196339,
+      "grad_norm": 3.961557081061209,
+      "learning_rate": 2.6315995465436878e-06,
+      "loss": 0.5432,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5359082679541339,
+      "grad_norm": 3.8780055120857204,
+      "learning_rate": 2.628093341663236e-06,
+      "loss": 0.454,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5363106014886341,
+      "grad_norm": 4.031126634163913,
+      "learning_rate": 2.6245868841473525e-06,
+      "loss": 0.5452,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5367129350231342,
+      "grad_norm": 3.3788818500452997,
+      "learning_rate": 2.6210801809117377e-06,
+      "loss": 0.4364,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5371152685576343,
+      "grad_norm": 4.566291916332284,
+      "learning_rate": 2.61757323887258e-06,
+      "loss": 0.6427,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5375176020921344,
+      "grad_norm": 4.46318346856033,
+      "learning_rate": 2.614066064946536e-06,
+      "loss": 0.6142,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5379199356266344,
+      "grad_norm": 3.8724650951472377,
+      "learning_rate": 2.6105586660507214e-06,
+      "loss": 0.4865,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5383222691611346,
+      "grad_norm": 4.107266394665508,
+      "learning_rate": 2.607051049102696e-06,
+      "loss": 0.5618,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5387246026956347,
+      "grad_norm": 4.131578269950126,
+      "learning_rate": 2.6035432210204475e-06,
+      "loss": 0.6571,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5391269362301347,
+      "grad_norm": 4.135237490425349,
+      "learning_rate": 2.6000351887223817e-06,
+      "loss": 0.506,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5395292697646349,
+      "grad_norm": 4.2799950491642695,
+      "learning_rate": 2.5965269591273075e-06,
+      "loss": 0.5054,
+      "step": 1341
+    },
+    {
+      "epoch": 0.539931603299135,
+      "grad_norm": 4.111277509443795,
+      "learning_rate": 2.59301853915442e-06,
+      "loss": 0.5896,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5403339368336351,
+      "grad_norm": 3.8507472387773416,
+      "learning_rate": 2.5895099357232946e-06,
+      "loss": 0.4784,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5407362703681352,
+      "grad_norm": 4.553167071691355,
+      "learning_rate": 2.5860011557538644e-06,
+      "loss": 0.5889,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5411386039026352,
+      "grad_norm": 4.8377815361815175,
+      "learning_rate": 2.5824922061664125e-06,
+      "loss": 0.7153,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5415409374371354,
+      "grad_norm": 4.026163246451428,
+      "learning_rate": 2.578983093881557e-06,
+      "loss": 0.5578,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5419432709716355,
+      "grad_norm": 4.396870764657521,
+      "learning_rate": 2.5754738258202345e-06,
+      "loss": 0.7799,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5423456045061356,
+      "grad_norm": 4.349518153263542,
+      "learning_rate": 2.5719644089036916e-06,
+      "loss": 0.5919,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5427479380406357,
+      "grad_norm": 4.355976508524607,
+      "learning_rate": 2.568454850053467e-06,
+      "loss": 0.5903,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5431502715751357,
+      "grad_norm": 3.951612074252993,
+      "learning_rate": 2.5649451561913796e-06,
+      "loss": 0.5038,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5435526051096359,
+      "grad_norm": 4.030293462511148,
+      "learning_rate": 2.561435334239515e-06,
+      "loss": 0.5308,
+      "step": 1351
+    },
+    {
+      "epoch": 0.543954938644136,
+      "grad_norm": 4.530833691231785,
+      "learning_rate": 2.55792539112021e-06,
+      "loss": 0.5977,
+      "step": 1352
+    },
+    {
+      "epoch": 0.544357272178636,
+      "grad_norm": 4.1989735383766185,
+      "learning_rate": 2.5544153337560422e-06,
+      "loss": 0.5686,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5447596057131362,
+      "grad_norm": 4.32105387076282,
+      "learning_rate": 2.5509051690698137e-06,
+      "loss": 0.6257,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5451619392476363,
+      "grad_norm": 3.34919768311592,
+      "learning_rate": 2.5473949039845388e-06,
+      "loss": 0.4131,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5455642727821364,
+      "grad_norm": 4.128988896294745,
+      "learning_rate": 2.5438845454234284e-06,
+      "loss": 0.6007,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5459666063166365,
+      "grad_norm": 4.122788652813818,
+      "learning_rate": 2.5403741003098793e-06,
+      "loss": 0.5223,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5463689398511365,
+      "grad_norm": 4.420118909526267,
+      "learning_rate": 2.5368635755674586e-06,
+      "loss": 0.5506,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5467712733856367,
+      "grad_norm": 4.2195230085645345,
+      "learning_rate": 2.5333529781198895e-06,
+      "loss": 0.5989,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5471736069201368,
+      "grad_norm": 4.20795164503954,
+      "learning_rate": 2.5298423148910396e-06,
+      "loss": 0.4462,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5475759404546369,
+      "grad_norm": 4.010223353178494,
+      "learning_rate": 2.5263315928049063e-06,
+      "loss": 0.5192,
+      "step": 1361
+    },
+    {
+      "epoch": 0.547978273989137,
+      "grad_norm": 4.816701841105469,
+      "learning_rate": 2.5228208187856024e-06,
+      "loss": 0.6599,
+      "step": 1362
+    },
+    {
+      "epoch": 0.548380607523637,
+      "grad_norm": 4.023413458632839,
+      "learning_rate": 2.5193099997573437e-06,
+      "loss": 0.6185,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5487829410581372,
+      "grad_norm": 3.6516400727809115,
+      "learning_rate": 2.5157991426444343e-06,
+      "loss": 0.5859,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5491852745926373,
+      "grad_norm": 4.119751084462459,
+      "learning_rate": 2.5122882543712546e-06,
+      "loss": 0.5255,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5495876081271374,
+      "grad_norm": 3.957835467170672,
+      "learning_rate": 2.5087773418622447e-06,
+      "loss": 0.4659,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5499899416616375,
+      "grad_norm": 3.500258182406765,
+      "learning_rate": 2.505266412041893e-06,
+      "loss": 0.4385,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5503922751961376,
+      "grad_norm": 4.041408070624076,
+      "learning_rate": 2.5017554718347233e-06,
+      "loss": 0.5367,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5507946087306377,
+      "grad_norm": 4.091765389792879,
+      "learning_rate": 2.4982445281652775e-06,
+      "loss": 0.5744,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5511969422651378,
+      "grad_norm": 4.469717920616285,
+      "learning_rate": 2.4947335879581073e-06,
+      "loss": 0.5818,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5515992757996379,
+      "grad_norm": 3.9779088164073833,
+      "learning_rate": 2.4912226581377566e-06,
+      "loss": 0.5893,
+      "step": 1371
+    },
+    {
+      "epoch": 0.552001609334138,
+      "grad_norm": 3.627034776419984,
+      "learning_rate": 2.4877117456287463e-06,
+      "loss": 0.4902,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5524039428686381,
+      "grad_norm": 3.8562269767576405,
+      "learning_rate": 2.484200857355566e-06,
+      "loss": 0.6268,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5528062764031382,
+      "grad_norm": 4.192164362531014,
+      "learning_rate": 2.4806900002426567e-06,
+      "loss": 0.6121,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5532086099376383,
+      "grad_norm": 3.916801705190097,
+      "learning_rate": 2.477179181214398e-06,
+      "loss": 0.4926,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5536109434721384,
+      "grad_norm": 3.693356946438789,
+      "learning_rate": 2.473668407195095e-06,
+      "loss": 0.5189,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5540132770066385,
+      "grad_norm": 4.229576715855437,
+      "learning_rate": 2.4701576851089613e-06,
+      "loss": 0.6398,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5544156105411386,
+      "grad_norm": 3.6167470608574197,
+      "learning_rate": 2.4666470218801113e-06,
+      "loss": 0.5175,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5548179440756387,
+      "grad_norm": 5.070171032981664,
+      "learning_rate": 2.463136424432542e-06,
+      "loss": 0.4988,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5552202776101388,
+      "grad_norm": 4.058368062649386,
+      "learning_rate": 2.4596258996901215e-06,
+      "loss": 0.4313,
+      "step": 1380
+    },
+    {
+      "epoch": 0.555622611144639,
+      "grad_norm": 3.9184186874819464,
+      "learning_rate": 2.4561154545765724e-06,
+      "loss": 0.5338,
+      "step": 1381
+    },
+    {
+      "epoch": 0.556024944679139,
+      "grad_norm": 3.1244549306950784,
+      "learning_rate": 2.4526050960154616e-06,
+      "loss": 0.4289,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5564272782136391,
+      "grad_norm": 4.04400253855949,
+      "learning_rate": 2.4490948309301867e-06,
+      "loss": 0.5257,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5568296117481392,
+      "grad_norm": 3.9303715740181513,
+      "learning_rate": 2.445584666243959e-06,
+      "loss": 0.5149,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5572319452826393,
+      "grad_norm": 4.395655391522824,
+      "learning_rate": 2.442074608879791e-06,
+      "loss": 0.6249,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5576342788171395,
+      "grad_norm": 3.8268439749597576,
+      "learning_rate": 2.438564665760486e-06,
+      "loss": 0.5613,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5580366123516395,
+      "grad_norm": 4.255793020118025,
+      "learning_rate": 2.4350548438086212e-06,
+      "loss": 0.6155,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5584389458861396,
+      "grad_norm": 3.979739013550483,
+      "learning_rate": 2.4315451499465336e-06,
+      "loss": 0.5394,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5588412794206397,
+      "grad_norm": 3.781411432288367,
+      "learning_rate": 2.4280355910963097e-06,
+      "loss": 0.4507,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5592436129551398,
+      "grad_norm": 4.371781717934846,
+      "learning_rate": 2.4245261741797663e-06,
+      "loss": 0.5305,
+      "step": 1390
+    },
+    {
+      "epoch": 0.55964594648964,
+      "grad_norm": 3.52895283713103,
+      "learning_rate": 2.421016906118444e-06,
+      "loss": 0.5194,
+      "step": 1391
+    },
+    {
+      "epoch": 0.56004828002414,
+      "grad_norm": 3.872042926144624,
+      "learning_rate": 2.417507793833587e-06,
+      "loss": 0.4998,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5604506135586401,
+      "grad_norm": 3.391234256928363,
+      "learning_rate": 2.413998844246136e-06,
+      "loss": 0.4093,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5608529470931403,
+      "grad_norm": 3.9654243758806174,
+      "learning_rate": 2.4104900642767066e-06,
+      "loss": 0.4515,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5612552806276403,
+      "grad_norm": 3.5879075344069378,
+      "learning_rate": 2.4069814608455804e-06,
+      "loss": 0.5207,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5616576141621404,
+      "grad_norm": 4.104094820028554,
+      "learning_rate": 2.4034730408726938e-06,
+      "loss": 0.5625,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5620599476966405,
+      "grad_norm": 3.772298699595217,
+      "learning_rate": 2.3999648112776183e-06,
+      "loss": 0.5068,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5624622812311406,
+      "grad_norm": 3.688052093132043,
+      "learning_rate": 2.3964567789795533e-06,
+      "loss": 0.5042,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5628646147656408,
+      "grad_norm": 3.8931716258044498,
+      "learning_rate": 2.3929489508973047e-06,
+      "loss": 0.5485,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5632669483001408,
+      "grad_norm": 4.224728616738161,
+      "learning_rate": 2.389441333949279e-06,
+      "loss": 0.6582,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5636692818346409,
+      "grad_norm": 3.8494878180062897,
+      "learning_rate": 2.3859339350534646e-06,
+      "loss": 0.5379,
+      "step": 1401
+    },
+    {
+      "epoch": 0.564071615369141,
+      "grad_norm": 4.200325419085908,
+      "learning_rate": 2.382426761127421e-06,
+      "loss": 0.5997,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5644739489036411,
+      "grad_norm": 4.143252972394019,
+      "learning_rate": 2.378919819088263e-06,
+      "loss": 0.6195,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5648762824381413,
+      "grad_norm": 3.7174533902571145,
+      "learning_rate": 2.3754131158526483e-06,
+      "loss": 0.5317,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5652786159726413,
+      "grad_norm": 4.0056931921159356,
+      "learning_rate": 2.371906658336764e-06,
+      "loss": 0.5232,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5656809495071414,
+      "grad_norm": 3.830247393182765,
+      "learning_rate": 2.368400453456313e-06,
+      "loss": 0.4898,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5660832830416416,
+      "grad_norm": 3.7820185066794787,
+      "learning_rate": 2.364894508126499e-06,
+      "loss": 0.5438,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5664856165761416,
+      "grad_norm": 4.4898443665627,
+      "learning_rate": 2.3613888292620135e-06,
+      "loss": 0.6322,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5668879501106417,
+      "grad_norm": 4.157236186991902,
+      "learning_rate": 2.3578834237770227e-06,
+      "loss": 0.5978,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5672902836451418,
+      "grad_norm": 4.576188931404494,
+      "learning_rate": 2.354378298585155e-06,
+      "loss": 0.5774,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5676926171796419,
+      "grad_norm": 3.714487248907507,
+      "learning_rate": 2.3508734605994855e-06,
+      "loss": 0.4305,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5680949507141421,
+      "grad_norm": 3.9270791753173313,
+      "learning_rate": 2.3473689167325205e-06,
+      "loss": 0.4912,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5684972842486421,
+      "grad_norm": 3.655781362606588,
+      "learning_rate": 2.3438646738961886e-06,
+      "loss": 0.5125,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5688996177831422,
+      "grad_norm": 4.2097153394323845,
+      "learning_rate": 2.3403607390018246e-06,
+      "loss": 0.4273,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5693019513176423,
+      "grad_norm": 3.7418997500983,
+      "learning_rate": 2.3368571189601556e-06,
+      "loss": 0.4534,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5697042848521424,
+      "grad_norm": 3.7385152001909154,
+      "learning_rate": 2.3333538206812874e-06,
+      "loss": 0.4461,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5701066183866426,
+      "grad_norm": 4.243485694858217,
+      "learning_rate": 2.32985085107469e-06,
+      "loss": 0.4943,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5705089519211426,
+      "grad_norm": 4.064181274521891,
+      "learning_rate": 2.3263482170491884e-06,
+      "loss": 0.6696,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5709112854556427,
+      "grad_norm": 4.168819688893492,
+      "learning_rate": 2.3228459255129425e-06,
+      "loss": 0.5914,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5713136189901429,
+      "grad_norm": 3.299064883781313,
+      "learning_rate": 2.3193439833734403e-06,
+      "loss": 0.3987,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5717159525246429,
+      "grad_norm": 4.887635853989916,
+      "learning_rate": 2.3158423975374757e-06,
+      "loss": 0.4842,
+      "step": 1421
+    },
+    {
+      "epoch": 0.572118286059143,
+      "grad_norm": 3.4644648967520806,
+      "learning_rate": 2.312341174911145e-06,
+      "loss": 0.4643,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5725206195936431,
+      "grad_norm": 3.75941250297541,
+      "learning_rate": 2.3088403223998235e-06,
+      "loss": 0.4772,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5729229531281432,
+      "grad_norm": 3.4615485090255578,
+      "learning_rate": 2.3053398469081613e-06,
+      "loss": 0.5075,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5733252866626434,
+      "grad_norm": 4.2227564973481995,
+      "learning_rate": 2.30183975534006e-06,
+      "loss": 0.5125,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5737276201971434,
+      "grad_norm": 4.090888673139273,
+      "learning_rate": 2.298340054598668e-06,
+      "loss": 0.5183,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5741299537316435,
+      "grad_norm": 4.548845256935699,
+      "learning_rate": 2.294840751586361e-06,
+      "loss": 0.639,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5745322872661436,
+      "grad_norm": 3.901263423192965,
+      "learning_rate": 2.291341853204731e-06,
+      "loss": 0.623,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5749346208006437,
+      "grad_norm": 3.347357658304439,
+      "learning_rate": 2.287843366354571e-06,
+      "loss": 0.4009,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5753369543351439,
+      "grad_norm": 4.512229941763247,
+      "learning_rate": 2.284345297935863e-06,
+      "loss": 0.6336,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5757392878696439,
+      "grad_norm": 4.132485335550154,
+      "learning_rate": 2.2808476548477636e-06,
+      "loss": 0.5015,
+      "step": 1431
+    },
+    {
+      "epoch": 0.576141621404144,
+      "grad_norm": 4.226000440762799,
+      "learning_rate": 2.2773504439885903e-06,
+      "loss": 0.5063,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5765439549386442,
+      "grad_norm": 4.010604789353783,
+      "learning_rate": 2.27385367225581e-06,
+      "loss": 0.5085,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5769462884731442,
+      "grad_norm": 3.71838877160847,
+      "learning_rate": 2.2703573465460194e-06,
+      "loss": 0.483,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5773486220076444,
+      "grad_norm": 3.715720870136391,
+      "learning_rate": 2.2668614737549395e-06,
+      "loss": 0.5292,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5777509555421444,
+      "grad_norm": 3.7875519229722756,
+      "learning_rate": 2.263366060777397e-06,
+      "loss": 0.4504,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5781532890766445,
+      "grad_norm": 4.235567100677654,
+      "learning_rate": 2.25987111450731e-06,
+      "loss": 0.6075,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5785556226111447,
+      "grad_norm": 4.23015897289991,
+      "learning_rate": 2.256376641837679e-06,
+      "loss": 0.6036,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5789579561456447,
+      "grad_norm": 4.397376552533781,
+      "learning_rate": 2.2528826496605672e-06,
+      "loss": 0.6584,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5793602896801449,
+      "grad_norm": 3.987430088200103,
+      "learning_rate": 2.249389144867092e-06,
+      "loss": 0.5312,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5797626232146449,
+      "grad_norm": 4.349402058381357,
+      "learning_rate": 2.24589613434741e-06,
+      "loss": 0.4806,
+      "step": 1441
+    },
+    {
+      "epoch": 0.580164956749145,
+      "grad_norm": 4.389627692568862,
+      "learning_rate": 2.2424036249907032e-06,
+      "loss": 0.649,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5805672902836452,
+      "grad_norm": 3.889422512107434,
+      "learning_rate": 2.238911623685162e-06,
+      "loss": 0.5553,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5809696238181452,
+      "grad_norm": 4.1471977764513674,
+      "learning_rate": 2.235420137317978e-06,
+      "loss": 0.5527,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5813719573526454,
+      "grad_norm": 3.642246289302764,
+      "learning_rate": 2.2319291727753266e-06,
+      "loss": 0.4669,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5817742908871455,
+      "grad_norm": 3.7186046973608406,
+      "learning_rate": 2.228438736942354e-06,
+      "loss": 0.3984,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5821766244216455,
+      "grad_norm": 4.208948818822306,
+      "learning_rate": 2.2249488367031612e-06,
+      "loss": 0.574,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5825789579561457,
+      "grad_norm": 3.4944920810310633,
+      "learning_rate": 2.221459478940797e-06,
+      "loss": 0.4565,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5829812914906457,
+      "grad_norm": 3.8209272903130325,
+      "learning_rate": 2.2179706705372377e-06,
+      "loss": 0.4116,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5833836250251458,
+      "grad_norm": 4.562727204828236,
+      "learning_rate": 2.214482418373376e-06,
+      "loss": 0.6272,
+      "step": 1450
+    },
+    {
+      "epoch": 0.583785958559646,
+      "grad_norm": 3.986021082183279,
+      "learning_rate": 2.21099472932901e-06,
+      "loss": 0.4946,
+      "step": 1451
+    },
+    {
+      "epoch": 0.584188292094146,
+      "grad_norm": 3.619941584357637,
+      "learning_rate": 2.207507610282823e-06,
+      "loss": 0.5552,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5845906256286462,
+      "grad_norm": 4.265275522318885,
+      "learning_rate": 2.2040210681123773e-06,
+      "loss": 0.5387,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5849929591631462,
+      "grad_norm": 3.592671313761321,
+      "learning_rate": 2.2005351096940966e-06,
+      "loss": 0.4221,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5853952926976463,
+      "grad_norm": 4.0275300106962755,
+      "learning_rate": 2.1970497419032546e-06,
+      "loss": 0.5781,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5857976262321465,
+      "grad_norm": 3.8884576976533425,
+      "learning_rate": 2.1935649716139567e-06,
+      "loss": 0.5121,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5861999597666465,
+      "grad_norm": 4.481785017500937,
+      "learning_rate": 2.1900808056991335e-06,
+      "loss": 0.6321,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5866022933011467,
+      "grad_norm": 3.74252328127819,
+      "learning_rate": 2.186597251030521e-06,
+      "loss": 0.4935,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5870046268356468,
+      "grad_norm": 3.9659714356438442,
+      "learning_rate": 2.183114314478651e-06,
+      "loss": 0.5503,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5874069603701468,
+      "grad_norm": 3.435746642131932,
+      "learning_rate": 2.179632002912837e-06,
+      "loss": 0.3508,
+      "step": 1460
+    },
+    {
+      "epoch": 0.587809293904647,
+      "grad_norm": 4.32896011694038,
+      "learning_rate": 2.176150323201157e-06,
+      "loss": 0.5728,
+      "step": 1461
+    },
+    {
+      "epoch": 0.588211627439147,
+      "grad_norm": 4.1710482850436215,
+      "learning_rate": 2.1726692822104455e-06,
+      "loss": 0.5115,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5886139609736472,
+      "grad_norm": 4.1244090573644145,
+      "learning_rate": 2.169188886806277e-06,
+      "loss": 0.3078,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5890162945081473,
+      "grad_norm": 3.462366030557245,
+      "learning_rate": 2.1657091438529517e-06,
+      "loss": 0.472,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5894186280426473,
+      "grad_norm": 3.920420130653373,
+      "learning_rate": 2.1622300602134824e-06,
+      "loss": 0.5559,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5898209615771475,
+      "grad_norm": 4.574556709525141,
+      "learning_rate": 2.158751642749583e-06,
+      "loss": 0.5525,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5902232951116475,
+      "grad_norm": 3.8284931733353513,
+      "learning_rate": 2.155273898321653e-06,
+      "loss": 0.5681,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5906256286461476,
+      "grad_norm": 3.3784102201483677,
+      "learning_rate": 2.151796833788766e-06,
+      "loss": 0.4727,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5910279621806478,
+      "grad_norm": 4.047681657794816,
+      "learning_rate": 2.1483204560086507e-06,
+      "loss": 0.5665,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5914302957151478,
+      "grad_norm": 4.308357381109115,
+      "learning_rate": 2.144844771837685e-06,
+      "loss": 0.6112,
+      "step": 1470
+    },
+    {
+      "epoch": 0.591832629249648,
+      "grad_norm": 3.733873078989285,
+      "learning_rate": 2.1413697881308785e-06,
+      "loss": 0.494,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5922349627841481,
+      "grad_norm": 4.130337043839704,
+      "learning_rate": 2.1378955117418567e-06,
+      "loss": 0.445,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5926372963186481,
+      "grad_norm": 4.189878829812914,
+      "learning_rate": 2.1344219495228537e-06,
+      "loss": 0.569,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5930396298531483,
+      "grad_norm": 3.926746164109752,
+      "learning_rate": 2.1309491083246915e-06,
+      "loss": 0.5696,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5934419633876483,
+      "grad_norm": 3.688932516880435,
+      "learning_rate": 2.127476994996772e-06,
+      "loss": 0.5001,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5938442969221485,
+      "grad_norm": 4.084824375606503,
+      "learning_rate": 2.1240056163870614e-06,
+      "loss": 0.613,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5942466304566486,
+      "grad_norm": 4.614880824241214,
+      "learning_rate": 2.120534979342078e-06,
+      "loss": 0.5252,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5946489639911486,
+      "grad_norm": 3.0888633468235014,
+      "learning_rate": 2.1170650907068742e-06,
+      "loss": 0.4531,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5950512975256488,
+      "grad_norm": 4.213120019047452,
+      "learning_rate": 2.113595957325029e-06,
+      "loss": 0.6458,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5954536310601488,
+      "grad_norm": 3.610258712936895,
+      "learning_rate": 2.1101275860386313e-06,
+      "loss": 0.5652,
+      "step": 1480
+    },
+    {
+      "epoch": 0.595855964594649,
+      "grad_norm": 3.901115632452379,
+      "learning_rate": 2.1066599836882667e-06,
+      "loss": 0.4731,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5962582981291491,
+      "grad_norm": 4.576321666746087,
+      "learning_rate": 2.1031931571130053e-06,
+      "loss": 0.6799,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5966606316636491,
+      "grad_norm": 4.188708168775149,
+      "learning_rate": 2.0997271131503847e-06,
+      "loss": 0.5172,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5970629651981493,
+      "grad_norm": 3.993460741506799,
+      "learning_rate": 2.0962618586364005e-06,
+      "loss": 0.4828,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5974652987326494,
+      "grad_norm": 3.977543353419007,
+      "learning_rate": 2.092797400405493e-06,
+      "loss": 0.5368,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5978676322671495,
+      "grad_norm": 4.079669745022429,
+      "learning_rate": 2.0893337452905293e-06,
+      "loss": 0.4879,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5982699658016496,
+      "grad_norm": 4.000645952876263,
+      "learning_rate": 2.0858709001227924e-06,
+      "loss": 0.5338,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5986722993361496,
+      "grad_norm": 4.483257438703314,
+      "learning_rate": 2.0824088717319697e-06,
+      "loss": 0.5771,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5990746328706498,
+      "grad_norm": 3.493611740579638,
+      "learning_rate": 2.0789476669461375e-06,
+      "loss": 0.4758,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5994769664051499,
+      "grad_norm": 3.9869817394957385,
+      "learning_rate": 2.0754872925917465e-06,
+      "loss": 0.5482,
+      "step": 1490
+    },
+    {
+      "epoch": 0.59987929993965,
+      "grad_norm": 3.655771235240259,
+      "learning_rate": 2.072027755493612e-06,
+      "loss": 0.4924,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6002816334741501,
+      "grad_norm": 3.959985484700295,
+      "learning_rate": 2.0685690624748943e-06,
+      "loss": 0.5653,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6006839670086501,
+      "grad_norm": 3.4092513778358047,
+      "learning_rate": 2.0651112203570904e-06,
+      "loss": 0.4375,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6010863005431503,
+      "grad_norm": 4.050380522574125,
+      "learning_rate": 2.061654235960021e-06,
+      "loss": 0.5627,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6014886340776504,
+      "grad_norm": 4.329223167800351,
+      "learning_rate": 2.058198116101814e-06,
+      "loss": 0.5642,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6018909676121504,
+      "grad_norm": 3.475966176009881,
+      "learning_rate": 2.0547428675988906e-06,
+      "loss": 0.4417,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6022933011466506,
+      "grad_norm": 3.7228401404060443,
+      "learning_rate": 2.051288497265955e-06,
+      "loss": 0.4509,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6026956346811507,
+      "grad_norm": 3.9298800993895284,
+      "learning_rate": 2.04783501191598e-06,
+      "loss": 0.4373,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6030979682156508,
+      "grad_norm": 4.3431668642603105,
+      "learning_rate": 2.0443824183601917e-06,
+      "loss": 0.6361,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6035003017501509,
+      "grad_norm": 3.759176356037184,
+      "learning_rate": 2.0409307234080577e-06,
+      "loss": 0.4214,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6039026352846509,
+      "grad_norm": 3.881662137011941,
+      "learning_rate": 2.0374799338672723e-06,
+      "loss": 0.4868,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6043049688191511,
+      "grad_norm": 4.786052455382573,
+      "learning_rate": 2.034030056543747e-06,
+      "loss": 0.6772,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6047073023536512,
+      "grad_norm": 4.137240403825967,
+      "learning_rate": 2.0305810982415907e-06,
+      "loss": 0.4981,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6051096358881513,
+      "grad_norm": 3.9245214475911947,
+      "learning_rate": 2.0271330657631034e-06,
+      "loss": 0.461,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6055119694226514,
+      "grad_norm": 4.043701387123113,
+      "learning_rate": 2.023685965908755e-06,
+      "loss": 0.4861,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6059143029571514,
+      "grad_norm": 4.578623188458431,
+      "learning_rate": 2.020239805477179e-06,
+      "loss": 0.5516,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6063166364916516,
+      "grad_norm": 3.939503557149013,
+      "learning_rate": 2.0167945912651542e-06,
+      "loss": 0.5348,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6067189700261517,
+      "grad_norm": 4.274904255142657,
+      "learning_rate": 2.0133503300675963e-06,
+      "loss": 0.5992,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6071213035606517,
+      "grad_norm": 3.732143003199723,
+      "learning_rate": 2.0099070286775367e-06,
+      "loss": 0.4891,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6075236370951519,
+      "grad_norm": 4.093400863247009,
+      "learning_rate": 2.0064646938861175e-06,
+      "loss": 0.4785,
+      "step": 1510
+    },
+    {
+      "epoch": 0.607925970629652,
+      "grad_norm": 3.850637017187709,
+      "learning_rate": 2.0030233324825734e-06,
+      "loss": 0.5371,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6083283041641521,
+      "grad_norm": 3.726527595640385,
+      "learning_rate": 1.999582951254219e-06,
+      "loss": 0.4105,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6087306376986522,
+      "grad_norm": 4.0679286236770045,
+      "learning_rate": 1.9961435569864366e-06,
+      "loss": 0.524,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6091329712331522,
+      "grad_norm": 4.191017273712236,
+      "learning_rate": 1.99270515646266e-06,
+      "loss": 0.422,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6095353047676524,
+      "grad_norm": 4.549567455680612,
+      "learning_rate": 1.9892677564643643e-06,
+      "loss": 0.5753,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6099376383021525,
+      "grad_norm": 4.239380717200701,
+      "learning_rate": 1.985831363771052e-06,
+      "loss": 0.5251,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6103399718366526,
+      "grad_norm": 3.8382650524932433,
+      "learning_rate": 1.9823959851602387e-06,
+      "loss": 0.6749,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6107423053711527,
+      "grad_norm": 4.058628419689791,
+      "learning_rate": 1.9789616274074375e-06,
+      "loss": 0.5202,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6111446389056527,
+      "grad_norm": 4.1466696163681815,
+      "learning_rate": 1.9755282972861515e-06,
+      "loss": 0.4595,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6115469724401529,
+      "grad_norm": 4.193830162309673,
+      "learning_rate": 1.9720960015678555e-06,
+      "loss": 0.485,
+      "step": 1520
+    },
+    {
+      "epoch": 0.611949305974653,
+      "grad_norm": 4.239153001501294,
+      "learning_rate": 1.9686647470219835e-06,
+      "loss": 0.6782,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6123516395091531,
+      "grad_norm": 4.138990574689833,
+      "learning_rate": 1.965234540415918e-06,
+      "loss": 0.6284,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6127539730436532,
+      "grad_norm": 4.079634170381014,
+      "learning_rate": 1.9618053885149715e-06,
+      "loss": 0.4656,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6131563065781533,
+      "grad_norm": 3.9398678794833204,
+      "learning_rate": 1.9583772980823797e-06,
+      "loss": 0.4345,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6135586401126534,
+      "grad_norm": 3.8999859116475237,
+      "learning_rate": 1.9549502758792825e-06,
+      "loss": 0.4463,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6139609736471535,
+      "grad_norm": 3.8236374179780652,
+      "learning_rate": 1.9515243286647155e-06,
+      "loss": 0.5032,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6143633071816536,
+      "grad_norm": 3.9105069856639787,
+      "learning_rate": 1.9480994631955904e-06,
+      "loss": 0.5249,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6147656407161537,
+      "grad_norm": 4.090697174474606,
+      "learning_rate": 1.944675686226688e-06,
+      "loss": 0.5986,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6151679742506538,
+      "grad_norm": 4.026970204591501,
+      "learning_rate": 1.941253004510642e-06,
+      "loss": 0.5695,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6155703077851539,
+      "grad_norm": 3.649099842703245,
+      "learning_rate": 1.937831424797926e-06,
+      "loss": 0.436,
+      "step": 1530
+    },
+    {
+      "epoch": 0.615972641319654,
+      "grad_norm": 3.8376284628790773,
+      "learning_rate": 1.9344109538368388e-06,
+      "loss": 0.4986,
+      "step": 1531
+    },
+    {
+      "epoch": 0.616374974854154,
+      "grad_norm": 4.048438236221887,
+      "learning_rate": 1.9309915983734933e-06,
+      "loss": 0.5608,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6167773083886542,
+      "grad_norm": 4.036284880877361,
+      "learning_rate": 1.9275733651518038e-06,
+      "loss": 0.552,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6171796419231543,
+      "grad_norm": 4.098386930913388,
+      "learning_rate": 1.9241562609134694e-06,
+      "loss": 0.5564,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6175819754576544,
+      "grad_norm": 4.4349116098990065,
+      "learning_rate": 1.9207402923979633e-06,
+      "loss": 0.5937,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6179843089921545,
+      "grad_norm": 3.9955554180753556,
+      "learning_rate": 1.9173254663425165e-06,
+      "loss": 0.4773,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6183866425266547,
+      "grad_norm": 3.7895980004495553,
+      "learning_rate": 1.9139117894821107e-06,
+      "loss": 0.5117,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6187889760611547,
+      "grad_norm": 4.503044474467485,
+      "learning_rate": 1.910499268549459e-06,
+      "loss": 0.5728,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6191913095956548,
+      "grad_norm": 4.5048282169496865,
+      "learning_rate": 1.9070879102749944e-06,
+      "loss": 0.4952,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6195936431301549,
+      "grad_norm": 3.6152226921367516,
+      "learning_rate": 1.9036777213868567e-06,
+      "loss": 0.4901,
+      "step": 1540
+    },
+    {
+      "epoch": 0.619995976664655,
+      "grad_norm": 4.56665728701204,
+      "learning_rate": 1.9002687086108808e-06,
+      "loss": 0.6217,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6203983101991551,
+      "grad_norm": 3.8747261446432972,
+      "learning_rate": 1.8968608786705803e-06,
+      "loss": 0.4346,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6208006437336552,
+      "grad_norm": 3.553549231891504,
+      "learning_rate": 1.8934542382871365e-06,
+      "loss": 0.4086,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6212029772681553,
+      "grad_norm": 4.273007467466175,
+      "learning_rate": 1.8900487941793862e-06,
+      "loss": 0.5589,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6216053108026554,
+      "grad_norm": 4.033041498302765,
+      "learning_rate": 1.8866445530638034e-06,
+      "loss": 0.5258,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6220076443371555,
+      "grad_norm": 3.847764801262625,
+      "learning_rate": 1.883241521654492e-06,
+      "loss": 0.5286,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6224099778716556,
+      "grad_norm": 4.141005318996514,
+      "learning_rate": 1.87983970666317e-06,
+      "loss": 0.5428,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6228123114061557,
+      "grad_norm": 3.8545860722248286,
+      "learning_rate": 1.8764391147991556e-06,
+      "loss": 0.4504,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6232146449406558,
+      "grad_norm": 3.9064270159769543,
+      "learning_rate": 1.8730397527693544e-06,
+      "loss": 0.49,
+      "step": 1549
+    },
+    {
+      "epoch": 0.623616978475156,
+      "grad_norm": 3.5305364890500903,
+      "learning_rate": 1.8696416272782467e-06,
+      "loss": 0.4747,
+      "step": 1550
+    },
+    {
+      "epoch": 0.624019312009656,
+      "grad_norm": 4.367606237537799,
+      "learning_rate": 1.866244745027874e-06,
+      "loss": 0.6066,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6244216455441561,
+      "grad_norm": 3.778565429919078,
+      "learning_rate": 1.8628491127178272e-06,
+      "loss": 0.4401,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6248239790786562,
+      "grad_norm": 4.4484086477744675,
+      "learning_rate": 1.8594547370452288e-06,
+      "loss": 0.5702,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6252263126131563,
+      "grad_norm": 4.133384847499752,
+      "learning_rate": 1.8560616247047253e-06,
+      "loss": 0.4958,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6256286461476565,
+      "grad_norm": 4.481750156438252,
+      "learning_rate": 1.8526697823884721e-06,
+      "loss": 0.5982,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6260309796821565,
+      "grad_norm": 4.679844131200749,
+      "learning_rate": 1.849279216786117e-06,
+      "loss": 0.5699,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6264333132166566,
+      "grad_norm": 4.035894316525195,
+      "learning_rate": 1.8458899345847936e-06,
+      "loss": 0.6036,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6268356467511567,
+      "grad_norm": 3.3784184252716836,
+      "learning_rate": 1.8425019424690993e-06,
+      "loss": 0.3667,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6272379802856568,
+      "grad_norm": 3.7155353324243325,
+      "learning_rate": 1.839115247121091e-06,
+      "loss": 0.4744,
+      "step": 1559
+    },
+    {
+      "epoch": 0.627640313820157,
+      "grad_norm": 4.062605549457995,
+      "learning_rate": 1.8357298552202676e-06,
+      "loss": 0.4846,
+      "step": 1560
+    },
+    {
+      "epoch": 0.628042647354657,
+      "grad_norm": 3.968579655925073,
+      "learning_rate": 1.832345773443557e-06,
+      "loss": 0.4563,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6284449808891571,
+      "grad_norm": 3.859404877482336,
+      "learning_rate": 1.8289630084653009e-06,
+      "loss": 0.5339,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6288473144236573,
+      "grad_norm": 4.32493582991546,
+      "learning_rate": 1.825581566957247e-06,
+      "loss": 0.5206,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6292496479581573,
+      "grad_norm": 3.7620622738243474,
+      "learning_rate": 1.8222014555885303e-06,
+      "loss": 0.5187,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6296519814926574,
+      "grad_norm": 3.9245871423313616,
+      "learning_rate": 1.8188226810256647e-06,
+      "loss": 0.5302,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6300543150271575,
+      "grad_norm": 4.0730169054317305,
+      "learning_rate": 1.8154452499325273e-06,
+      "loss": 0.482,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6304566485616576,
+      "grad_norm": 4.0887356034823314,
+      "learning_rate": 1.8120691689703423e-06,
+      "loss": 0.6323,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6308589820961578,
+      "grad_norm": 4.134013360882183,
+      "learning_rate": 1.8086944447976745e-06,
+      "loss": 0.4215,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6312613156306578,
+      "grad_norm": 4.267968482401992,
+      "learning_rate": 1.805321084070412e-06,
+      "loss": 0.6421,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6316636491651579,
+      "grad_norm": 4.337432367486679,
+      "learning_rate": 1.8019490934417533e-06,
+      "loss": 0.614,
+      "step": 1570
+    },
+    {
+      "epoch": 0.632065982699658,
+      "grad_norm": 4.325604571245914,
+      "learning_rate": 1.7985784795621946e-06,
+      "loss": 0.6082,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6324683162341581,
+      "grad_norm": 4.7255506567581005,
+      "learning_rate": 1.7952092490795165e-06,
+      "loss": 0.5848,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6328706497686583,
+      "grad_norm": 3.9917358158112513,
+      "learning_rate": 1.7918414086387723e-06,
+      "loss": 0.51,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6332729833031583,
+      "grad_norm": 3.9759190309543806,
+      "learning_rate": 1.788474964882273e-06,
+      "loss": 0.4348,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6336753168376584,
+      "grad_norm": 3.9060781229407673,
+      "learning_rate": 1.7851099244495761e-06,
+      "loss": 0.5328,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6340776503721586,
+      "grad_norm": 3.794629121755811,
+      "learning_rate": 1.7817462939774683e-06,
+      "loss": 0.5406,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6344799839066586,
+      "grad_norm": 4.434204491241485,
+      "learning_rate": 1.778384080099959e-06,
+      "loss": 0.6955,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6348823174411587,
+      "grad_norm": 4.243047298973574,
+      "learning_rate": 1.775023289448261e-06,
+      "loss": 0.6021,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6352846509756588,
+      "grad_norm": 3.906621293593377,
+      "learning_rate": 1.7716639286507834e-06,
+      "loss": 0.5017,
+      "step": 1579
+    },
+    {
+      "epoch": 0.6356869845101589,
+      "grad_norm": 4.433075604901836,
+      "learning_rate": 1.7683060043331102e-06,
+      "loss": 0.6304,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6360893180446591,
+      "grad_norm": 4.0324680233039425,
+      "learning_rate": 1.764949523117997e-06,
+      "loss": 0.4593,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6364916515791591,
+      "grad_norm": 4.861200946299266,
+      "learning_rate": 1.7615944916253511e-06,
+      "loss": 0.5755,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6368939851136592,
+      "grad_norm": 3.4859779578870462,
+      "learning_rate": 1.7582409164722219e-06,
+      "loss": 0.4494,
+      "step": 1583
+    },
+    {
+      "epoch": 0.6372963186481593,
+      "grad_norm": 4.394364956739334,
+      "learning_rate": 1.7548888042727835e-06,
+      "loss": 0.5756,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6376986521826594,
+      "grad_norm": 4.474860851761582,
+      "learning_rate": 1.751538161638327e-06,
+      "loss": 0.5217,
+      "step": 1585
+    },
+    {
+      "epoch": 0.6381009857171596,
+      "grad_norm": 4.665357328027418,
+      "learning_rate": 1.748188995177245e-06,
+      "loss": 0.5599,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6385033192516596,
+      "grad_norm": 3.186769209343777,
+      "learning_rate": 1.744841311495019e-06,
+      "loss": 0.3409,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6389056527861597,
+      "grad_norm": 3.3506177658392744,
+      "learning_rate": 1.7414951171942057e-06,
+      "loss": 0.5176,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6393079863206599,
+      "grad_norm": 3.4136910306135935,
+      "learning_rate": 1.7381504188744218e-06,
+      "loss": 0.4464,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6397103198551599,
+      "grad_norm": 3.4917018490110467,
+      "learning_rate": 1.7348072231323382e-06,
+      "loss": 0.438,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6401126533896601,
+      "grad_norm": 3.421803275946972,
+      "learning_rate": 1.7314655365616584e-06,
+      "loss": 0.4189,
+      "step": 1591
+    },
+    {
+      "epoch": 0.6405149869241601,
+      "grad_norm": 4.433987296079437,
+      "learning_rate": 1.7281253657531123e-06,
+      "loss": 0.606,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6409173204586602,
+      "grad_norm": 3.6737703307949983,
+      "learning_rate": 1.7247867172944367e-06,
+      "loss": 0.4349,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6413196539931604,
+      "grad_norm": 3.8221357123925914,
+      "learning_rate": 1.721449597770369e-06,
+      "loss": 0.484,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6417219875276604,
+      "grad_norm": 4.069693321778052,
+      "learning_rate": 1.718114013762631e-06,
+      "loss": 0.6507,
+      "step": 1595
+    },
+    {
+      "epoch": 0.6421243210621606,
+      "grad_norm": 4.268934323506783,
+      "learning_rate": 1.714779971849915e-06,
+      "loss": 0.498,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6425266545966606,
+      "grad_norm": 4.426082622301671,
+      "learning_rate": 1.7114474786078716e-06,
+      "loss": 0.6755,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6429289881311607,
+      "grad_norm": 3.6655062613985288,
+      "learning_rate": 1.708116540609098e-06,
+      "loss": 0.5078,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6433313216656609,
+      "grad_norm": 4.096742196755438,
+      "learning_rate": 1.7047871644231226e-06,
+      "loss": 0.5085,
+      "step": 1599
+    },
+    {
+      "epoch": 0.6437336552001609,
+      "grad_norm": 4.341802218643631,
+      "learning_rate": 1.7014593566163949e-06,
+      "loss": 0.6022,
+      "step": 1600
+    },
+    {
+      "epoch": 0.644135988734661,
+      "grad_norm": 4.167303639683251,
+      "learning_rate": 1.6981331237522716e-06,
+      "loss": 0.595,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6445383222691611,
+      "grad_norm": 3.860753980307849,
+      "learning_rate": 1.6948084723910002e-06,
+      "loss": 0.4189,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6449406558036612,
+      "grad_norm": 3.801435117916446,
+      "learning_rate": 1.6914854090897124e-06,
+      "loss": 0.4691,
+      "step": 1603
+    },
+    {
+      "epoch": 0.6453429893381614,
+      "grad_norm": 4.309844638484278,
+      "learning_rate": 1.6881639404024063e-06,
+      "loss": 0.4802,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6457453228726614,
+      "grad_norm": 3.8218645060402285,
+      "learning_rate": 1.6848440728799346e-06,
+      "loss": 0.6087,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6461476564071615,
+      "grad_norm": 4.318515117319728,
+      "learning_rate": 1.6815258130699938e-06,
+      "loss": 0.6644,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6465499899416617,
+      "grad_norm": 3.8706029461346114,
+      "learning_rate": 1.6782091675171064e-06,
+      "loss": 0.5314,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6469523234761617,
+      "grad_norm": 4.588786654868146,
+      "learning_rate": 1.6748941427626142e-06,
+      "loss": 0.5713,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6473546570106619,
+      "grad_norm": 4.206226833574013,
+      "learning_rate": 1.671580745344661e-06,
+      "loss": 0.5207,
+      "step": 1609
+    },
+    {
+      "epoch": 0.6477569905451619,
+      "grad_norm": 4.1241193689590565,
+      "learning_rate": 1.6682689817981827e-06,
+      "loss": 0.5325,
+      "step": 1610
+    },
+    {
+      "epoch": 0.648159324079662,
+      "grad_norm": 4.013331656155903,
+      "learning_rate": 1.6649588586548888e-06,
+      "loss": 0.5593,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6485616576141622,
+      "grad_norm": 4.701503684635364,
+      "learning_rate": 1.6616503824432573e-06,
+      "loss": 0.6878,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6489639911486622,
+      "grad_norm": 4.047494584818948,
+      "learning_rate": 1.658343559688516e-06,
+      "loss": 0.5131,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6493663246831624,
+      "grad_norm": 4.0317139401208575,
+      "learning_rate": 1.6550383969126341e-06,
+      "loss": 0.512,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6497686582176624,
+      "grad_norm": 3.9682317404566034,
+      "learning_rate": 1.6517349006343025e-06,
+      "loss": 0.5921,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6501709917521625,
+      "grad_norm": 3.824331386002604,
+      "learning_rate": 1.6484330773689292e-06,
+      "loss": 0.4712,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6505733252866627,
+      "grad_norm": 3.7019715035054204,
+      "learning_rate": 1.645132933628621e-06,
+      "loss": 0.4426,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6509756588211627,
+      "grad_norm": 4.0459494986124245,
+      "learning_rate": 1.6418344759221734e-06,
+      "loss": 0.6728,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6513779923556628,
+      "grad_norm": 4.154812270453969,
+      "learning_rate": 1.6385377107550543e-06,
+      "loss": 0.559,
+      "step": 1619
+    },
+    {
+      "epoch": 0.651780325890163,
+      "grad_norm": 4.703387446268752,
+      "learning_rate": 1.6352426446293952e-06,
+      "loss": 0.5548,
+      "step": 1620
+    },
+    {
+      "epoch": 0.652182659424663,
+      "grad_norm": 3.894555312354832,
+      "learning_rate": 1.631949284043976e-06,
+      "loss": 0.4987,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6525849929591632,
+      "grad_norm": 4.281016820029128,
+      "learning_rate": 1.628657635494213e-06,
+      "loss": 0.6007,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6529873264936632,
+      "grad_norm": 3.708021853167261,
+      "learning_rate": 1.625367705472147e-06,
+      "loss": 0.5041,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6533896600281633,
+      "grad_norm": 4.184667272519736,
+      "learning_rate": 1.6220795004664264e-06,
+      "loss": 0.563,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6537919935626635,
+      "grad_norm": 4.386544791097414,
+      "learning_rate": 1.6187930269623001e-06,
+      "loss": 0.6591,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6541943270971635,
+      "grad_norm": 4.095997806253382,
+      "learning_rate": 1.6155082914416018e-06,
+      "loss": 0.5122,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6545966606316637,
+      "grad_norm": 3.97582656513418,
+      "learning_rate": 1.6122253003827353e-06,
+      "loss": 0.5805,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6549989941661637,
+      "grad_norm": 3.990901166079429,
+      "learning_rate": 1.6089440602606675e-06,
+      "loss": 0.569,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6554013277006638,
+      "grad_norm": 3.8624368116988403,
+      "learning_rate": 1.6056645775469077e-06,
+      "loss": 0.4729,
+      "step": 1629
+    },
+    {
+      "epoch": 0.655803661235164,
+      "grad_norm": 4.529362920268054,
+      "learning_rate": 1.6023868587095016e-06,
+      "loss": 0.5482,
+      "step": 1630
+    },
+    {
+      "epoch": 0.656205994769664,
+      "grad_norm": 4.017565552354934,
+      "learning_rate": 1.599110910213016e-06,
+      "loss": 0.4916,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6566083283041642,
+      "grad_norm": 4.4182156313212575,
+      "learning_rate": 1.595836738518527e-06,
+      "loss": 0.5792,
+      "step": 1632
+    },
+    {
+      "epoch": 0.6570106618386643,
+      "grad_norm": 3.5677962985553524,
+      "learning_rate": 1.592564350083603e-06,
+      "loss": 0.4697,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6574129953731643,
+      "grad_norm": 3.909711067921425,
+      "learning_rate": 1.5892937513622977e-06,
+      "loss": 0.4383,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6578153289076645,
+      "grad_norm": 4.098825788465001,
+      "learning_rate": 1.5860249488051355e-06,
+      "loss": 0.4745,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6582176624421645,
+      "grad_norm": 3.2004894026603616,
+      "learning_rate": 1.5827579488590977e-06,
+      "loss": 0.4775,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6586199959766647,
+      "grad_norm": 3.970341598297501,
+      "learning_rate": 1.5794927579676089e-06,
+      "loss": 0.5471,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6590223295111648,
+      "grad_norm": 4.241718162619275,
+      "learning_rate": 1.5762293825705276e-06,
+      "loss": 0.6544,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6594246630456648,
+      "grad_norm": 3.924819583311759,
+      "learning_rate": 1.5729678291041318e-06,
+      "loss": 0.5368,
+      "step": 1639
+    },
+    {
+      "epoch": 0.659826996580165,
+      "grad_norm": 3.8382551527997397,
+      "learning_rate": 1.5697081040011043e-06,
+      "loss": 0.5121,
+      "step": 1640
+    },
+    {
+      "epoch": 0.660229330114665,
+      "grad_norm": 3.6682266397110337,
+      "learning_rate": 1.5664502136905236e-06,
+      "loss": 0.4353,
+      "step": 1641
+    },
+    {
+      "epoch": 0.6606316636491651,
+      "grad_norm": 3.9060845453518143,
+      "learning_rate": 1.563194164597848e-06,
+      "loss": 0.4814,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6610339971836653,
+      "grad_norm": 3.985541657912861,
+      "learning_rate": 1.5599399631449052e-06,
+      "loss": 0.4643,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6614363307181653,
+      "grad_norm": 4.246558106503365,
+      "learning_rate": 1.5566876157498797e-06,
+      "loss": 0.5457,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6618386642526655,
+      "grad_norm": 4.538356504223212,
+      "learning_rate": 1.5534371288272992e-06,
+      "loss": 0.5586,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6622409977871656,
+      "grad_norm": 3.512817123486096,
+      "learning_rate": 1.5501885087880187e-06,
+      "loss": 0.3777,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6626433313216656,
+      "grad_norm": 3.611315735282059,
+      "learning_rate": 1.546941762039216e-06,
+      "loss": 0.4559,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6630456648561658,
+      "grad_norm": 3.804900233889627,
+      "learning_rate": 1.5436968949843705e-06,
+      "loss": 0.5593,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6634479983906658,
+      "grad_norm": 4.290281020900778,
+      "learning_rate": 1.5404539140232572e-06,
+      "loss": 0.5433,
+      "step": 1649
+    },
+    {
+      "epoch": 0.663850331925166,
+      "grad_norm": 4.261399800624298,
+      "learning_rate": 1.5372128255519298e-06,
+      "loss": 0.5669,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6642526654596661,
+      "grad_norm": 4.8206169191275094,
+      "learning_rate": 1.533973635962708e-06,
+      "loss": 0.535,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6646549989941661,
+      "grad_norm": 3.5705866846183882,
+      "learning_rate": 1.5307363516441693e-06,
+      "loss": 0.4436,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6650573325286663,
+      "grad_norm": 4.015284104793611,
+      "learning_rate": 1.5275009789811317e-06,
+      "loss": 0.4606,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6654596660631663,
+      "grad_norm": 3.746888831379427,
+      "learning_rate": 1.5242675243546435e-06,
+      "loss": 0.5344,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6658619995976665,
+      "grad_norm": 5.350374969998737,
+      "learning_rate": 1.5210359941419698e-06,
+      "loss": 0.7187,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6662643331321666,
+      "grad_norm": 4.992414873463243,
+      "learning_rate": 1.5178063947165795e-06,
+      "loss": 0.5597,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 3.9536138723501413,
+      "learning_rate": 1.514578732448135e-06,
+      "loss": 0.5345,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6670690002011668,
+      "grad_norm": 4.023274067891729,
+      "learning_rate": 1.5113530137024774e-06,
+      "loss": 0.4893,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6674713337356669,
+      "grad_norm": 4.108018429855622,
+      "learning_rate": 1.5081292448416158e-06,
+      "loss": 0.645,
+      "step": 1659
+    },
+    {
+      "epoch": 0.667873667270167,
+      "grad_norm": 3.817031046521698,
+      "learning_rate": 1.5049074322237097e-06,
+      "loss": 0.4838,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6682760008046671,
+      "grad_norm": 4.10734795999481,
+      "learning_rate": 1.5016875822030651e-06,
+      "loss": 0.5853,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6686783343391671,
+      "grad_norm": 4.104736058188134,
+      "learning_rate": 1.4984697011301142e-06,
+      "loss": 0.5032,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6690806678736673,
+      "grad_norm": 4.213736489738209,
+      "learning_rate": 1.4952537953514078e-06,
+      "loss": 0.4234,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6694830014081674,
+      "grad_norm": 4.592368631770386,
+      "learning_rate": 1.4920398712095985e-06,
+      "loss": 0.5323,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6698853349426674,
+      "grad_norm": 4.593839464020305,
+      "learning_rate": 1.4888279350434325e-06,
+      "loss": 0.5821,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6702876684771676,
+      "grad_norm": 3.881133737659202,
+      "learning_rate": 1.485617993187735e-06,
+      "loss": 0.5224,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6706900020116676,
+      "grad_norm": 4.088068946795046,
+      "learning_rate": 1.4824100519733985e-06,
+      "loss": 0.556,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6710923355461678,
+      "grad_norm": 4.328495324161391,
+      "learning_rate": 1.4792041177273668e-06,
+      "loss": 0.6097,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6714946690806679,
+      "grad_norm": 4.220316709489982,
+      "learning_rate": 1.4760001967726283e-06,
+      "loss": 0.6015,
+      "step": 1669
+    },
+    {
+      "epoch": 0.6718970026151679,
+      "grad_norm": 3.788396547796328,
+      "learning_rate": 1.472798295428199e-06,
+      "loss": 0.5863,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6722993361496681,
+      "grad_norm": 3.8086553750558494,
+      "learning_rate": 1.4695984200091129e-06,
+      "loss": 0.5155,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6727016696841682,
+      "grad_norm": 3.792467737324882,
+      "learning_rate": 1.4664005768264084e-06,
+      "loss": 0.4517,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6731040032186683,
+      "grad_norm": 4.528886592192105,
+      "learning_rate": 1.4632047721871134e-06,
+      "loss": 0.5514,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6735063367531684,
+      "grad_norm": 3.687134247150264,
+      "learning_rate": 1.4600110123942376e-06,
+      "loss": 0.5765,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6739086702876684,
+      "grad_norm": 3.651752055706122,
+      "learning_rate": 1.4568193037467574e-06,
+      "loss": 0.4434,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6743110038221686,
+      "grad_norm": 4.196555677667497,
+      "learning_rate": 1.453629652539604e-06,
+      "loss": 0.5832,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6747133373566687,
+      "grad_norm": 3.3768494467370047,
+      "learning_rate": 1.4504420650636485e-06,
+      "loss": 0.4094,
+      "step": 1677
+    },
+    {
+      "epoch": 0.6751156708911688,
+      "grad_norm": 3.9350013310511582,
+      "learning_rate": 1.4472565476056952e-06,
+      "loss": 0.4875,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6755180044256689,
+      "grad_norm": 3.9364895166754383,
+      "learning_rate": 1.444073106448462e-06,
+      "loss": 0.626,
+      "step": 1679
+    },
+    {
+      "epoch": 0.6759203379601689,
+      "grad_norm": 4.218774138421791,
+      "learning_rate": 1.4408917478705748e-06,
+      "loss": 0.5318,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6763226714946691,
+      "grad_norm": 3.557852927029882,
+      "learning_rate": 1.4377124781465517e-06,
+      "loss": 0.4564,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6767250050291692,
+      "grad_norm": 4.131292301591179,
+      "learning_rate": 1.4345353035467888e-06,
+      "loss": 0.5458,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6771273385636692,
+      "grad_norm": 4.261120314757306,
+      "learning_rate": 1.4313602303375523e-06,
+      "loss": 0.6402,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6775296720981694,
+      "grad_norm": 4.027439558487913,
+      "learning_rate": 1.4281872647809628e-06,
+      "loss": 0.5762,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6779320056326695,
+      "grad_norm": 3.9491401859678605,
+      "learning_rate": 1.4250164131349858e-06,
+      "loss": 0.4807,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6783343391671696,
+      "grad_norm": 3.60940986670999,
+      "learning_rate": 1.421847681653414e-06,
+      "loss": 0.3659,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6787366727016697,
+      "grad_norm": 3.5822407819650515,
+      "learning_rate": 1.4186810765858616e-06,
+      "loss": 0.5165,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6791390062361697,
+      "grad_norm": 4.161182758555724,
+      "learning_rate": 1.4155166041777483e-06,
+      "loss": 0.495,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6795413397706699,
+      "grad_norm": 4.076431248542256,
+      "learning_rate": 1.4123542706702868e-06,
+      "loss": 0.4569,
+      "step": 1689
+    },
+    {
+      "epoch": 0.67994367330517,
+      "grad_norm": 3.401944784975411,
+      "learning_rate": 1.409194082300473e-06,
+      "loss": 0.4661,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6803460068396701,
+      "grad_norm": 4.053896920545759,
+      "learning_rate": 1.4060360453010691e-06,
+      "loss": 0.6945,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6807483403741702,
+      "grad_norm": 3.6315876209619726,
+      "learning_rate": 1.4028801659005975e-06,
+      "loss": 0.5088,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6811506739086702,
+      "grad_norm": 3.9033263695876617,
+      "learning_rate": 1.3997264503233222e-06,
+      "loss": 0.5535,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6815530074431704,
+      "grad_norm": 3.6520508454132568,
+      "learning_rate": 1.3965749047892421e-06,
+      "loss": 0.4627,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6819553409776705,
+      "grad_norm": 4.253145100591827,
+      "learning_rate": 1.3934255355140735e-06,
+      "loss": 0.6059,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6823576745121706,
+      "grad_norm": 3.6442609114551168,
+      "learning_rate": 1.3902783487092436e-06,
+      "loss": 0.437,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6827600080466707,
+      "grad_norm": 3.6843000132081083,
+      "learning_rate": 1.387133350581873e-06,
+      "loss": 0.4741,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6831623415811708,
+      "grad_norm": 4.082924701084028,
+      "learning_rate": 1.383990547334767e-06,
+      "loss": 0.5523,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6835646751156709,
+      "grad_norm": 4.525193517476986,
+      "learning_rate": 1.3808499451663998e-06,
+      "loss": 0.5733,
+      "step": 1699
+    },
+    {
+      "epoch": 0.683967008650171,
+      "grad_norm": 3.8434172838289142,
+      "learning_rate": 1.377711550270907e-06,
+      "loss": 0.4042,
+      "step": 1700
+    },
+    {
+      "epoch": 0.684369342184671,
+      "grad_norm": 3.475844585621674,
+      "learning_rate": 1.3745753688380692e-06,
+      "loss": 0.4492,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6847716757191712,
+      "grad_norm": 4.106466165584271,
+      "learning_rate": 1.3714414070533022e-06,
+      "loss": 0.4795,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6851740092536713,
+      "grad_norm": 3.9408796988863477,
+      "learning_rate": 1.368309671097645e-06,
+      "loss": 0.4586,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6855763427881714,
+      "grad_norm": 3.892878635048642,
+      "learning_rate": 1.3651801671477435e-06,
+      "loss": 0.5256,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6859786763226715,
+      "grad_norm": 3.602589673252341,
+      "learning_rate": 1.3620529013758444e-06,
+      "loss": 0.4474,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6863810098571715,
+      "grad_norm": 4.29050518808401,
+      "learning_rate": 1.3589278799497802e-06,
+      "loss": 0.4799,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6867833433916717,
+      "grad_norm": 4.530188596691838,
+      "learning_rate": 1.3558051090329536e-06,
+      "loss": 0.6044,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6871856769261718,
+      "grad_norm": 3.8842771157381932,
+      "learning_rate": 1.352684594784333e-06,
+      "loss": 0.4759,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6875880104606719,
+      "grad_norm": 4.206785571596244,
+      "learning_rate": 1.3495663433584327e-06,
+      "loss": 0.58,
+      "step": 1709
+    },
+    {
+      "epoch": 0.687990343995172,
+      "grad_norm": 4.648975138578508,
+      "learning_rate": 1.3464503609053053e-06,
+      "loss": 0.5508,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6883926775296721,
+      "grad_norm": 4.513492173355037,
+      "learning_rate": 1.3433366535705292e-06,
+      "loss": 0.6606,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6887950110641722,
+      "grad_norm": 4.149085872177257,
+      "learning_rate": 1.3402252274951954e-06,
+      "loss": 0.5438,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6891973445986723,
+      "grad_norm": 4.3080582119504935,
+      "learning_rate": 1.3371160888158935e-06,
+      "loss": 0.7277,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6895996781331724,
+      "grad_norm": 4.195944255893043,
+      "learning_rate": 1.3340092436647045e-06,
+      "loss": 0.5682,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6900020116676725,
+      "grad_norm": 4.22717459587227,
+      "learning_rate": 1.330904698169184e-06,
+      "loss": 0.5394,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6904043452021726,
+      "grad_norm": 3.8328998433470334,
+      "learning_rate": 1.3278024584523542e-06,
+      "loss": 0.519,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6908066787366727,
+      "grad_norm": 3.809809815071977,
+      "learning_rate": 1.324702530632686e-06,
+      "loss": 0.4609,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6912090122711728,
+      "grad_norm": 4.408644772842878,
+      "learning_rate": 1.3216049208240945e-06,
+      "loss": 0.5259,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6916113458056729,
+      "grad_norm": 4.351040955206057,
+      "learning_rate": 1.3185096351359211e-06,
+      "loss": 0.531,
+      "step": 1719
+    },
+    {
+      "epoch": 0.692013679340173,
+      "grad_norm": 3.8862458710646135,
+      "learning_rate": 1.3154166796729245e-06,
+      "loss": 0.5521,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6924160128746731,
+      "grad_norm": 3.8732946833809283,
+      "learning_rate": 1.3123260605352647e-06,
+      "loss": 0.4054,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6928183464091732,
+      "grad_norm": 4.014263742357025,
+      "learning_rate": 1.3092377838184984e-06,
+      "loss": 0.5096,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6932206799436733,
+      "grad_norm": 3.8457014699738306,
+      "learning_rate": 1.3061518556135572e-06,
+      "loss": 0.5361,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6936230134781735,
+      "grad_norm": 4.700764158732767,
+      "learning_rate": 1.3030682820067447e-06,
+      "loss": 0.6363,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6940253470126735,
+      "grad_norm": 4.056861341126563,
+      "learning_rate": 1.2999870690797207e-06,
+      "loss": 0.5499,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6944276805471736,
+      "grad_norm": 3.824223111444158,
+      "learning_rate": 1.2969082229094853e-06,
+      "loss": 0.5049,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6948300140816737,
+      "grad_norm": 3.8066989110426332,
+      "learning_rate": 1.2938317495683744e-06,
+      "loss": 0.487,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6952323476161738,
+      "grad_norm": 4.382851529690001,
+      "learning_rate": 1.2907576551240425e-06,
+      "loss": 0.5016,
+      "step": 1728
+    },
+    {
+      "epoch": 0.695634681150674,
+      "grad_norm": 4.2709658357684415,
+      "learning_rate": 1.2876859456394537e-06,
+      "loss": 0.5287,
+      "step": 1729
+    },
+    {
+      "epoch": 0.696037014685174,
+      "grad_norm": 4.10428141385855,
+      "learning_rate": 1.284616627172865e-06,
+      "loss": 0.5048,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6964393482196741,
+      "grad_norm": 3.9240197132726133,
+      "learning_rate": 1.281549705777821e-06,
+      "loss": 0.5834,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6968416817541742,
+      "grad_norm": 3.917832710676123,
+      "learning_rate": 1.278485187503137e-06,
+      "loss": 0.4978,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6972440152886743,
+      "grad_norm": 3.9886241574028123,
+      "learning_rate": 1.2754230783928907e-06,
+      "loss": 0.5046,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6976463488231744,
+      "grad_norm": 3.8252585006207744,
+      "learning_rate": 1.2723633844864036e-06,
+      "loss": 0.4337,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6980486823576745,
+      "grad_norm": 4.327909530843214,
+      "learning_rate": 1.2693061118182399e-06,
+      "loss": 0.5663,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6984510158921746,
+      "grad_norm": 3.9729308104472225,
+      "learning_rate": 1.2662512664181826e-06,
+      "loss": 0.4753,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6988533494266748,
+      "grad_norm": 3.4193536839822434,
+      "learning_rate": 1.2631988543112314e-06,
+      "loss": 0.3984,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6992556829611748,
+      "grad_norm": 3.199663843957577,
+      "learning_rate": 1.2601488815175867e-06,
+      "loss": 0.3615,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6996580164956749,
+      "grad_norm": 4.275095481044542,
+      "learning_rate": 1.257101354052635e-06,
+      "loss": 0.635,
+      "step": 1739
+    },
+    {
+      "epoch": 0.700060350030175,
+      "grad_norm": 4.154894699295693,
+      "learning_rate": 1.2540562779269421e-06,
+      "loss": 0.468,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7004626835646751,
+      "grad_norm": 4.21271106717443,
+      "learning_rate": 1.2510136591462399e-06,
+      "loss": 0.557,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7008650170991753,
+      "grad_norm": 3.9488032178250276,
+      "learning_rate": 1.2479735037114118e-06,
+      "loss": 0.4743,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7012673506336753,
+      "grad_norm": 3.6437407466020453,
+      "learning_rate": 1.2449358176184848e-06,
+      "loss": 0.4958,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7016696841681754,
+      "grad_norm": 4.011291060246906,
+      "learning_rate": 1.2419006068586125e-06,
+      "loss": 0.5109,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7020720177026755,
+      "grad_norm": 3.7545390991024026,
+      "learning_rate": 1.2388678774180698e-06,
+      "loss": 0.5269,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7024743512371756,
+      "grad_norm": 3.821504931037996,
+      "learning_rate": 1.2358376352782358e-06,
+      "loss": 0.4905,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7028766847716758,
+      "grad_norm": 4.3580981207011655,
+      "learning_rate": 1.232809886415586e-06,
+      "loss": 0.4904,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7032790183061758,
+      "grad_norm": 3.539055820090707,
+      "learning_rate": 1.2297846368016748e-06,
+      "loss": 0.4124,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7036813518406759,
+      "grad_norm": 3.9844890147713254,
+      "learning_rate": 1.2267618924031317e-06,
+      "loss": 0.5321,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7040836853751761,
+      "grad_norm": 4.178112023828336,
+      "learning_rate": 1.2237416591816412e-06,
+      "loss": 0.6362,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7044860189096761,
+      "grad_norm": 4.0472275399599695,
+      "learning_rate": 1.220723943093939e-06,
+      "loss": 0.5101,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7048883524441762,
+      "grad_norm": 4.312523214514084,
+      "learning_rate": 1.2177087500917925e-06,
+      "loss": 0.5277,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7052906859786763,
+      "grad_norm": 4.133768938615848,
+      "learning_rate": 1.2146960861219954e-06,
+      "loss": 0.5271,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7056930195131764,
+      "grad_norm": 4.430552507921185,
+      "learning_rate": 1.211685957126353e-06,
+      "loss": 0.711,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7060953530476766,
+      "grad_norm": 4.119451270646705,
+      "learning_rate": 1.2086783690416704e-06,
+      "loss": 0.4358,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7064976865821766,
+      "grad_norm": 4.295693476571627,
+      "learning_rate": 1.205673327799743e-06,
+      "loss": 0.5466,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7069000201166767,
+      "grad_norm": 4.312446284251359,
+      "learning_rate": 1.2026708393273392e-06,
+      "loss": 0.5861,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7073023536511768,
+      "grad_norm": 3.9109516304734733,
+      "learning_rate": 1.1996709095461967e-06,
+      "loss": 0.4782,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7077046871856769,
+      "grad_norm": 4.425022693894618,
+      "learning_rate": 1.1966735443730052e-06,
+      "loss": 0.5627,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7081070207201771,
+      "grad_norm": 4.259706435608531,
+      "learning_rate": 1.193678749719397e-06,
+      "loss": 0.5274,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7085093542546771,
+      "grad_norm": 4.219902699982978,
+      "learning_rate": 1.190686531491932e-06,
+      "loss": 0.5397,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7089116877891772,
+      "grad_norm": 4.105043115699033,
+      "learning_rate": 1.1876968955920916e-06,
+      "loss": 0.4845,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7093140213236774,
+      "grad_norm": 3.905684008250733,
+      "learning_rate": 1.1847098479162644e-06,
+      "loss": 0.5769,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7097163548581774,
+      "grad_norm": 4.0434341189775616,
+      "learning_rate": 1.1817253943557308e-06,
+      "loss": 0.4338,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7101186883926776,
+      "grad_norm": 4.331648953621162,
+      "learning_rate": 1.1787435407966593e-06,
+      "loss": 0.6231,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7105210219271776,
+      "grad_norm": 4.225800580764582,
+      "learning_rate": 1.1757642931200864e-06,
+      "loss": 0.4946,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7109233554616777,
+      "grad_norm": 3.832934508222693,
+      "learning_rate": 1.1727876572019117e-06,
+      "loss": 0.5319,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7113256889961779,
+      "grad_norm": 3.994089631167334,
+      "learning_rate": 1.1698136389128833e-06,
+      "loss": 0.5005,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7117280225306779,
+      "grad_norm": 3.8399445750919425,
+      "learning_rate": 1.166842244118587e-06,
+      "loss": 0.4726,
+      "step": 1769
+    },
+    {
+      "epoch": 0.712130356065178,
+      "grad_norm": 4.043691249667768,
+      "learning_rate": 1.163873478679432e-06,
+      "loss": 0.4885,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7125326895996781,
+      "grad_norm": 4.26586065507147,
+      "learning_rate": 1.1609073484506441e-06,
+      "loss": 0.5955,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7129350231341782,
+      "grad_norm": 3.2723853825243006,
+      "learning_rate": 1.157943859282251e-06,
+      "loss": 0.393,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7133373566686784,
+      "grad_norm": 4.228399556429077,
+      "learning_rate": 1.1549830170190714e-06,
+      "loss": 0.6138,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7137396902031784,
+      "grad_norm": 3.703790105757895,
+      "learning_rate": 1.152024827500705e-06,
+      "loss": 0.4952,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7141420237376785,
+      "grad_norm": 4.037603384099717,
+      "learning_rate": 1.149069296561516e-06,
+      "loss": 0.5791,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7145443572721787,
+      "grad_norm": 3.958021889837016,
+      "learning_rate": 1.1461164300306292e-06,
+      "loss": 0.495,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7149466908066787,
+      "grad_norm": 4.142430978673458,
+      "learning_rate": 1.1431662337319107e-06,
+      "loss": 0.6534,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7153490243411789,
+      "grad_norm": 4.484413621549814,
+      "learning_rate": 1.1402187134839643e-06,
+      "loss": 0.4885,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7157513578756789,
+      "grad_norm": 3.6695545389300395,
+      "learning_rate": 1.1372738751001111e-06,
+      "loss": 0.4446,
+      "step": 1779
+    },
+    {
+      "epoch": 0.716153691410179,
+      "grad_norm": 3.492829736247777,
+      "learning_rate": 1.134331724388387e-06,
+      "loss": 0.5245,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7165560249446792,
+      "grad_norm": 3.8848835038152107,
+      "learning_rate": 1.1313922671515252e-06,
+      "loss": 0.5042,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7169583584791792,
+      "grad_norm": 4.4510516838349705,
+      "learning_rate": 1.128455509186948e-06,
+      "loss": 0.6898,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7173606920136794,
+      "grad_norm": 4.381017376532981,
+      "learning_rate": 1.1255214562867503e-06,
+      "loss": 0.5289,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7177630255481794,
+      "grad_norm": 3.5079718266262416,
+      "learning_rate": 1.122590114237696e-06,
+      "loss": 0.4724,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7181653590826795,
+      "grad_norm": 4.292226353007704,
+      "learning_rate": 1.1196614888212007e-06,
+      "loss": 0.5796,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7185676926171797,
+      "grad_norm": 4.052087753788445,
+      "learning_rate": 1.1167355858133226e-06,
+      "loss": 0.4938,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7189700261516797,
+      "grad_norm": 5.157980356295375,
+      "learning_rate": 1.1138124109847503e-06,
+      "loss": 0.5949,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7193723596861799,
+      "grad_norm": 3.285940613285488,
+      "learning_rate": 1.1108919701007907e-06,
+      "loss": 0.3908,
+      "step": 1788
+    },
+    {
+      "epoch": 0.71977469322068,
+      "grad_norm": 3.5335717637010937,
+      "learning_rate": 1.1079742689213593e-06,
+      "loss": 0.5013,
+      "step": 1789
+    },
+    {
+      "epoch": 0.72017702675518,
+      "grad_norm": 4.450180240936036,
+      "learning_rate": 1.1050593132009703e-06,
+      "loss": 0.6329,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7205793602896802,
+      "grad_norm": 4.404576812133769,
+      "learning_rate": 1.1021471086887187e-06,
+      "loss": 0.5781,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7209816938241802,
+      "grad_norm": 3.3602741770117177,
+      "learning_rate": 1.0992376611282776e-06,
+      "loss": 0.3621,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7213840273586803,
+      "grad_norm": 3.381272077924684,
+      "learning_rate": 1.0963309762578795e-06,
+      "loss": 0.384,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7217863608931805,
+      "grad_norm": 3.5780173251972025,
+      "learning_rate": 1.0934270598103094e-06,
+      "loss": 0.5372,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7221886944276805,
+      "grad_norm": 3.7738716419210814,
+      "learning_rate": 1.0905259175128931e-06,
+      "loss": 0.4336,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7225910279621807,
+      "grad_norm": 3.820342940812247,
+      "learning_rate": 1.0876275550874846e-06,
+      "loss": 0.5335,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7229933614966807,
+      "grad_norm": 4.176570559423123,
+      "learning_rate": 1.0847319782504532e-06,
+      "loss": 0.5038,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7233956950311808,
+      "grad_norm": 3.67077010562448,
+      "learning_rate": 1.0818391927126764e-06,
+      "loss": 0.4342,
+      "step": 1798
+    },
+    {
+      "epoch": 0.723798028565681,
+      "grad_norm": 3.9880069700930423,
+      "learning_rate": 1.078949204179526e-06,
+      "loss": 0.4953,
+      "step": 1799
+    },
+    {
+      "epoch": 0.724200362100181,
+      "grad_norm": 4.161334696726195,
+      "learning_rate": 1.0760620183508582e-06,
+      "loss": 0.5799,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7246026956346812,
+      "grad_norm": 4.126464363948578,
+      "learning_rate": 1.073177640920999e-06,
+      "loss": 0.5138,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7250050291691813,
+      "grad_norm": 4.1895057334300505,
+      "learning_rate": 1.0702960775787374e-06,
+      "loss": 0.5121,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7254073627036813,
+      "grad_norm": 3.884515999748438,
+      "learning_rate": 1.0674173340073121e-06,
+      "loss": 0.4891,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7258096962381815,
+      "grad_norm": 4.286867137892461,
+      "learning_rate": 1.0645414158844012e-06,
+      "loss": 0.6118,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7262120297726815,
+      "grad_norm": 3.8959234040057806,
+      "learning_rate": 1.0616683288821067e-06,
+      "loss": 0.5234,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7266143633071817,
+      "grad_norm": 4.239978882514847,
+      "learning_rate": 1.058798078666952e-06,
+      "loss": 0.5183,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7270166968416818,
+      "grad_norm": 3.7122065670739235,
+      "learning_rate": 1.0559306708998607e-06,
+      "loss": 0.501,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7274190303761818,
+      "grad_norm": 4.556831231079228,
+      "learning_rate": 1.0530661112361533e-06,
+      "loss": 0.5257,
+      "step": 1808
+    },
+    {
+      "epoch": 0.727821363910682,
+      "grad_norm": 4.308685532929913,
+      "learning_rate": 1.050204405325534e-06,
+      "loss": 0.5534,
+      "step": 1809
+    },
+    {
+      "epoch": 0.728223697445182,
+      "grad_norm": 3.9297951909048368,
+      "learning_rate": 1.0473455588120738e-06,
+      "loss": 0.4681,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7286260309796821,
+      "grad_norm": 3.98667448491781,
+      "learning_rate": 1.0444895773342092e-06,
+      "loss": 0.4785,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7290283645141823,
+      "grad_norm": 4.2934600586771845,
+      "learning_rate": 1.0416364665247236e-06,
+      "loss": 0.5408,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7294306980486823,
+      "grad_norm": 3.9347493779534,
+      "learning_rate": 1.0387862320107403e-06,
+      "loss": 0.5895,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7298330315831825,
+      "grad_norm": 4.338941092246621,
+      "learning_rate": 1.0359388794137068e-06,
+      "loss": 0.5648,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7302353651176826,
+      "grad_norm": 3.792647800273631,
+      "learning_rate": 1.0330944143493892e-06,
+      "loss": 0.5449,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7306376986521826,
+      "grad_norm": 3.821766624740813,
+      "learning_rate": 1.0302528424278585e-06,
+      "loss": 0.4544,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7310400321866828,
+      "grad_norm": 4.176558757203708,
+      "learning_rate": 1.0274141692534784e-06,
+      "loss": 0.4802,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7314423657211828,
+      "grad_norm": 3.5626608894002803,
+      "learning_rate": 1.0245784004248968e-06,
+      "loss": 0.4273,
+      "step": 1818
+    },
+    {
+      "epoch": 0.731844699255683,
+      "grad_norm": 3.8719939712343936,
+      "learning_rate": 1.0217455415350316e-06,
+      "loss": 0.5409,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7322470327901831,
+      "grad_norm": 4.537877117506395,
+      "learning_rate": 1.0189155981710642e-06,
+      "loss": 0.5705,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7326493663246831,
+      "grad_norm": 3.932184586794533,
+      "learning_rate": 1.0160885759144217e-06,
+      "loss": 0.4904,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7330516998591833,
+      "grad_norm": 4.266661778185627,
+      "learning_rate": 1.0132644803407753e-06,
+      "loss": 0.4769,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7334540333936833,
+      "grad_norm": 3.7589693532265827,
+      "learning_rate": 1.0104433170200188e-06,
+      "loss": 0.4891,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7338563669281835,
+      "grad_norm": 3.8562553041075263,
+      "learning_rate": 1.007625091516266e-06,
+      "loss": 0.417,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7342587004626836,
+      "grad_norm": 4.130360177598847,
+      "learning_rate": 1.0048098093878352e-06,
+      "loss": 0.5207,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7346610339971836,
+      "grad_norm": 4.16513812007776,
+      "learning_rate": 1.001997476187241e-06,
+      "loss": 0.5213,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7350633675316838,
+      "grad_norm": 4.4208377164885055,
+      "learning_rate": 9.991880974611809e-07,
+      "loss": 0.6269,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7354657010661839,
+      "grad_norm": 4.867032713419264,
+      "learning_rate": 9.96381678750524e-07,
+      "loss": 0.4808,
+      "step": 1828
+    },
+    {
+      "epoch": 0.735868034600684,
+      "grad_norm": 3.982545780740484,
+      "learning_rate": 9.935782255903034e-07,
+      "loss": 0.5066,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7362703681351841,
+      "grad_norm": 4.286351965336647,
+      "learning_rate": 9.907777435097029e-07,
+      "loss": 0.5151,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7366727016696841,
+      "grad_norm": 4.362229906545009,
+      "learning_rate": 9.879802380320473e-07,
+      "loss": 0.5936,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7370750352041843,
+      "grad_norm": 4.660993080328005,
+      "learning_rate": 9.851857146747873e-07,
+      "loss": 0.6829,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7374773687386844,
+      "grad_norm": 4.222997821431192,
+      "learning_rate": 9.82394178949497e-07,
+      "loss": 0.514,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7378797022731844,
+      "grad_norm": 4.369792251301722,
+      "learning_rate": 9.796056363618533e-07,
+      "loss": 0.5742,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7382820358076846,
+      "grad_norm": 4.422672878513299,
+      "learning_rate": 9.768200924116338e-07,
+      "loss": 0.568,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7386843693421846,
+      "grad_norm": 4.507594407110887,
+      "learning_rate": 9.740375525926988e-07,
+      "loss": 0.6017,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7390867028766848,
+      "grad_norm": 3.8125967279749835,
+      "learning_rate": 9.71258022392985e-07,
+      "loss": 0.5959,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7394890364111849,
+      "grad_norm": 3.7273585439154586,
+      "learning_rate": 9.684815072944946e-07,
+      "loss": 0.3825,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7398913699456849,
+      "grad_norm": 3.738805120372397,
+      "learning_rate": 9.657080127732807e-07,
+      "loss": 0.4931,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7402937034801851,
+      "grad_norm": 3.9828190911127526,
+      "learning_rate": 9.629375442994418e-07,
+      "loss": 0.5182,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7406960370146852,
+      "grad_norm": 3.9798516908695882,
+      "learning_rate": 9.601701073371045e-07,
+      "loss": 0.548,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7410983705491853,
+      "grad_norm": 3.7199861649210715,
+      "learning_rate": 9.574057073444195e-07,
+      "loss": 0.468,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7415007040836854,
+      "grad_norm": 4.0680484984054335,
+      "learning_rate": 9.546443497735467e-07,
+      "loss": 0.5782,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7419030376181854,
+      "grad_norm": 3.4109362486846115,
+      "learning_rate": 9.518860400706465e-07,
+      "loss": 0.4245,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7423053711526856,
+      "grad_norm": 3.931799967484987,
+      "learning_rate": 9.491307836758651e-07,
+      "loss": 0.4863,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7427077046871857,
+      "grad_norm": 4.211928044347116,
+      "learning_rate": 9.463785860233301e-07,
+      "loss": 0.5459,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7431100382216858,
+      "grad_norm": 4.0902800868050715,
+      "learning_rate": 9.436294525411357e-07,
+      "loss": 0.6375,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7435123717561859,
+      "grad_norm": 3.524293359481542,
+      "learning_rate": 9.408833886513303e-07,
+      "loss": 0.444,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7439147052906859,
+      "grad_norm": 3.6061388537955827,
+      "learning_rate": 9.381403997699117e-07,
+      "loss": 0.5217,
+      "step": 1849
+    },
+    {
+      "epoch": 0.7443170388251861,
+      "grad_norm": 3.6547603411472838,
+      "learning_rate": 9.354004913068096e-07,
+      "loss": 0.5229,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7447193723596862,
+      "grad_norm": 4.340731245011316,
+      "learning_rate": 9.326636686658808e-07,
+      "loss": 0.5924,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7451217058941862,
+      "grad_norm": 3.4273270658083166,
+      "learning_rate": 9.299299372448953e-07,
+      "loss": 0.4941,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7455240394286864,
+      "grad_norm": 3.9330296000490668,
+      "learning_rate": 9.271993024355263e-07,
+      "loss": 0.5988,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7459263729631865,
+      "grad_norm": 4.79047981984662,
+      "learning_rate": 9.244717696233388e-07,
+      "loss": 0.6193,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7463287064976866,
+      "grad_norm": 3.6277035412847924,
+      "learning_rate": 9.217473441877805e-07,
+      "loss": 0.4905,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7467310400321867,
+      "grad_norm": 4.652255454626382,
+      "learning_rate": 9.19026031502171e-07,
+      "loss": 0.6231,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7471333735666867,
+      "grad_norm": 4.139619345656057,
+      "learning_rate": 9.1630783693369e-07,
+      "loss": 0.4981,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7475357071011869,
+      "grad_norm": 5.059383612132635,
+      "learning_rate": 9.135927658433685e-07,
+      "loss": 0.7292,
+      "step": 1858
+    },
+    {
+      "epoch": 0.747938040635687,
+      "grad_norm": 4.664959181536374,
+      "learning_rate": 9.108808235860744e-07,
+      "loss": 0.6118,
+      "step": 1859
+    },
+    {
+      "epoch": 0.7483403741701871,
+      "grad_norm": 4.243108848129753,
+      "learning_rate": 9.081720155105076e-07,
+      "loss": 0.5616,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7487427077046872,
+      "grad_norm": 3.4693576707426548,
+      "learning_rate": 9.054663469591862e-07,
+      "loss": 0.4451,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7491450412391872,
+      "grad_norm": 4.037761942838199,
+      "learning_rate": 9.027638232684339e-07,
+      "loss": 0.5354,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7495473747736874,
+      "grad_norm": 4.037140323875402,
+      "learning_rate": 9.00064449768375e-07,
+      "loss": 0.3946,
+      "step": 1863
+    },
+    {
+      "epoch": 0.7499497083081875,
+      "grad_norm": 4.420498117292749,
+      "learning_rate": 8.97368231782918e-07,
+      "loss": 0.6091,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7503520418426876,
+      "grad_norm": 4.098542987137017,
+      "learning_rate": 8.946751746297494e-07,
+      "loss": 0.6347,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7507543753771877,
+      "grad_norm": 3.718783314907544,
+      "learning_rate": 8.919852836203224e-07,
+      "loss": 0.5523,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7511567089116878,
+      "grad_norm": 4.238567081312867,
+      "learning_rate": 8.892985640598434e-07,
+      "loss": 0.5716,
+      "step": 1867
+    },
+    {
+      "epoch": 0.7515590424461879,
+      "grad_norm": 3.878960051531824,
+      "learning_rate": 8.866150212472657e-07,
+      "loss": 0.5552,
+      "step": 1868
+    },
+    {
+      "epoch": 0.751961375980688,
+      "grad_norm": 4.0405823701369155,
+      "learning_rate": 8.839346604752763e-07,
+      "loss": 0.5286,
+      "step": 1869
+    },
+    {
+      "epoch": 0.752363709515188,
+      "grad_norm": 4.251404967808422,
+      "learning_rate": 8.81257487030287e-07,
+      "loss": 0.4872,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7527660430496882,
+      "grad_norm": 3.6336762556932163,
+      "learning_rate": 8.785835061924234e-07,
+      "loss": 0.4301,
+      "step": 1871
+    },
+    {
+      "epoch": 0.7531683765841883,
+      "grad_norm": 4.324873818573403,
+      "learning_rate": 8.759127232355128e-07,
+      "loss": 0.5405,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7535707101186884,
+      "grad_norm": 4.42447965555661,
+      "learning_rate": 8.732451434270767e-07,
+      "loss": 0.5209,
+      "step": 1873
+    },
+    {
+      "epoch": 0.7539730436531885,
+      "grad_norm": 4.251290292687751,
+      "learning_rate": 8.705807720283199e-07,
+      "loss": 0.5951,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7543753771876885,
+      "grad_norm": 4.171584555966765,
+      "learning_rate": 8.679196142941173e-07,
+      "loss": 0.6086,
+      "step": 1875
+    },
+    {
+      "epoch": 0.7547777107221887,
+      "grad_norm": 4.0958958819445614,
+      "learning_rate": 8.652616754730075e-07,
+      "loss": 0.5693,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7551800442566888,
+      "grad_norm": 3.683062028222347,
+      "learning_rate": 8.626069608071785e-07,
+      "loss": 0.4584,
+      "step": 1877
+    },
+    {
+      "epoch": 0.7555823777911889,
+      "grad_norm": 3.9012710111822924,
+      "learning_rate": 8.599554755324613e-07,
+      "loss": 0.4611,
+      "step": 1878
+    },
+    {
+      "epoch": 0.755984711325689,
+      "grad_norm": 3.5489182194802704,
+      "learning_rate": 8.573072248783165e-07,
+      "loss": 0.5083,
+      "step": 1879
+    },
+    {
+      "epoch": 0.7563870448601892,
+      "grad_norm": 3.9884722155489225,
+      "learning_rate": 8.546622140678265e-07,
+      "loss": 0.5094,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7567893783946892,
+      "grad_norm": 4.328049459143733,
+      "learning_rate": 8.520204483176814e-07,
+      "loss": 0.3656,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7571917119291893,
+      "grad_norm": 3.9339309591347087,
+      "learning_rate": 8.493819328381731e-07,
+      "loss": 0.5317,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7575940454636894,
+      "grad_norm": 4.2270154261756385,
+      "learning_rate": 8.467466728331828e-07,
+      "loss": 0.4588,
+      "step": 1883
+    },
+    {
+      "epoch": 0.7579963789981895,
+      "grad_norm": 3.303945186981914,
+      "learning_rate": 8.441146735001715e-07,
+      "loss": 0.4649,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7583987125326896,
+      "grad_norm": 4.054188026080457,
+      "learning_rate": 8.414859400301667e-07,
+      "loss": 0.5645,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7588010460671897,
+      "grad_norm": 4.17992148808867,
+      "learning_rate": 8.388604776077575e-07,
+      "loss": 0.5204,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7592033796016898,
+      "grad_norm": 4.428173356861454,
+      "learning_rate": 8.362382914110804e-07,
+      "loss": 0.6007,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7596057131361899,
+      "grad_norm": 4.255468940795954,
+      "learning_rate": 8.336193866118117e-07,
+      "loss": 0.5153,
+      "step": 1888
+    },
+    {
+      "epoch": 0.76000804667069,
+      "grad_norm": 4.025791541260584,
+      "learning_rate": 8.310037683751527e-07,
+      "loss": 0.4859,
+      "step": 1889
+    },
+    {
+      "epoch": 0.7604103802051901,
+      "grad_norm": 3.8573782279099533,
+      "learning_rate": 8.283914418598263e-07,
+      "loss": 0.5058,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7608127137396902,
+      "grad_norm": 3.332235385366874,
+      "learning_rate": 8.257824122180602e-07,
+      "loss": 0.3737,
+      "step": 1891
+    },
+    {
+      "epoch": 0.7612150472741903,
+      "grad_norm": 3.394561752123056,
+      "learning_rate": 8.231766845955818e-07,
+      "loss": 0.4197,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7616173808086905,
+      "grad_norm": 3.3171785759316443,
+      "learning_rate": 8.205742641316064e-07,
+      "loss": 0.417,
+      "step": 1893
+    },
+    {
+      "epoch": 0.7620197143431905,
+      "grad_norm": 4.487532016958629,
+      "learning_rate": 8.179751559588234e-07,
+      "loss": 0.5663,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7624220478776906,
+      "grad_norm": 4.546271967177341,
+      "learning_rate": 8.153793652033931e-07,
+      "loss": 0.5597,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7628243814121907,
+      "grad_norm": 3.914164846084329,
+      "learning_rate": 8.12786896984931e-07,
+      "loss": 0.4677,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7632267149466908,
+      "grad_norm": 3.852819616304247,
+      "learning_rate": 8.101977564165011e-07,
+      "loss": 0.503,
+      "step": 1897
+    },
+    {
+      "epoch": 0.763629048481191,
+      "grad_norm": 4.067929177534588,
+      "learning_rate": 8.076119486046019e-07,
+      "loss": 0.5396,
+      "step": 1898
+    },
+    {
+      "epoch": 0.764031382015691,
+      "grad_norm": 4.010571344025329,
+      "learning_rate": 8.050294786491611e-07,
+      "loss": 0.4829,
+      "step": 1899
+    },
+    {
+      "epoch": 0.7644337155501911,
+      "grad_norm": 4.199207618578668,
+      "learning_rate": 8.024503516435222e-07,
+      "loss": 0.6123,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7648360490846912,
+      "grad_norm": 4.313823267969129,
+      "learning_rate": 7.998745726744358e-07,
+      "loss": 0.5597,
+      "step": 1901
+    },
+    {
+      "epoch": 0.7652383826191913,
+      "grad_norm": 3.7661528466618956,
+      "learning_rate": 7.973021468220502e-07,
+      "loss": 0.5036,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7656407161536914,
+      "grad_norm": 4.017310363498917,
+      "learning_rate": 7.947330791598973e-07,
+      "loss": 0.5435,
+      "step": 1903
+    },
+    {
+      "epoch": 0.7660430496881915,
+      "grad_norm": 4.447099011053462,
+      "learning_rate": 7.921673747548895e-07,
+      "loss": 0.4892,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7664453832226916,
+      "grad_norm": 4.1491367299305555,
+      "learning_rate": 7.896050386673026e-07,
+      "loss": 0.5234,
+      "step": 1905
+    },
+    {
+      "epoch": 0.7668477167571918,
+      "grad_norm": 4.640959967611087,
+      "learning_rate": 7.870460759507722e-07,
+      "loss": 0.5619,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7672500502916918,
+      "grad_norm": 4.207681163367058,
+      "learning_rate": 7.844904916522784e-07,
+      "loss": 0.6313,
+      "step": 1907
+    },
+    {
+      "epoch": 0.7676523838261919,
+      "grad_norm": 4.624725241802207,
+      "learning_rate": 7.819382908121386e-07,
+      "loss": 0.6169,
+      "step": 1908
+    },
+    {
+      "epoch": 0.768054717360692,
+      "grad_norm": 3.891844454436063,
+      "learning_rate": 7.793894784639986e-07,
+      "loss": 0.4932,
+      "step": 1909
+    },
+    {
+      "epoch": 0.7684570508951921,
+      "grad_norm": 3.6005349724480573,
+      "learning_rate": 7.768440596348192e-07,
+      "loss": 0.5835,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7688593844296923,
+      "grad_norm": 4.073804566700674,
+      "learning_rate": 7.743020393448702e-07,
+      "loss": 0.4882,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7692617179641923,
+      "grad_norm": 3.959969867245135,
+      "learning_rate": 7.717634226077156e-07,
+      "loss": 0.4945,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7696640514986924,
+      "grad_norm": 3.599166244946097,
+      "learning_rate": 7.692282144302094e-07,
+      "loss": 0.3889,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7700663850331925,
+      "grad_norm": 3.539375774753211,
+      "learning_rate": 7.666964198124819e-07,
+      "loss": 0.523,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7704687185676926,
+      "grad_norm": 3.912187073942566,
+      "learning_rate": 7.64168043747932e-07,
+      "loss": 0.5567,
+      "step": 1915
+    },
+    {
+      "epoch": 0.7708710521021928,
+      "grad_norm": 4.034008046236846,
+      "learning_rate": 7.616430912232137e-07,
+      "loss": 0.5219,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7712733856366928,
+      "grad_norm": 4.074005253663992,
+      "learning_rate": 7.59121567218232e-07,
+      "loss": 0.5302,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7716757191711929,
+      "grad_norm": 3.4776790939949347,
+      "learning_rate": 7.566034767061265e-07,
+      "loss": 0.2992,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7720780527056931,
+      "grad_norm": 3.755741404325393,
+      "learning_rate": 7.540888246532688e-07,
+      "loss": 0.4337,
+      "step": 1919
+    },
+    {
+      "epoch": 0.7724803862401931,
+      "grad_norm": 3.854464881913325,
+      "learning_rate": 7.515776160192454e-07,
+      "loss": 0.5286,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7728827197746932,
+      "grad_norm": 3.74441109291879,
+      "learning_rate": 7.490698557568535e-07,
+      "loss": 0.5449,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7732850533091933,
+      "grad_norm": 4.570844263947877,
+      "learning_rate": 7.465655488120891e-07,
+      "loss": 0.6153,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7736873868436934,
+      "grad_norm": 3.7650085288496613,
+      "learning_rate": 7.440647001241363e-07,
+      "loss": 0.5166,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7740897203781936,
+      "grad_norm": 3.6719483984817236,
+      "learning_rate": 7.415673146253604e-07,
+      "loss": 0.4842,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7744920539126936,
+      "grad_norm": 4.057471839423865,
+      "learning_rate": 7.390733972412933e-07,
+      "loss": 0.6075,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7748943874471937,
+      "grad_norm": 4.4626847789976924,
+      "learning_rate": 7.365829528906293e-07,
+      "loss": 0.5292,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7752967209816938,
+      "grad_norm": 3.76677011259477,
+      "learning_rate": 7.340959864852124e-07,
+      "loss": 0.405,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7756990545161939,
+      "grad_norm": 3.8469846124186082,
+      "learning_rate": 7.316125029300275e-07,
+      "loss": 0.4827,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7761013880506941,
+      "grad_norm": 3.8371445392089827,
+      "learning_rate": 7.291325071231883e-07,
+      "loss": 0.5134,
+      "step": 1929
+    },
+    {
+      "epoch": 0.7765037215851941,
+      "grad_norm": 3.8383451450589887,
+      "learning_rate": 7.26656003955932e-07,
+      "loss": 0.4061,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7769060551196942,
+      "grad_norm": 4.037666657124557,
+      "learning_rate": 7.241829983126075e-07,
+      "loss": 0.4976,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7773083886541944,
+      "grad_norm": 3.980302342869356,
+      "learning_rate": 7.217134950706625e-07,
+      "loss": 0.5149,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7777107221886944,
+      "grad_norm": 3.7463100803001277,
+      "learning_rate": 7.192474991006416e-07,
+      "loss": 0.459,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7781130557231946,
+      "grad_norm": 3.9126255555142597,
+      "learning_rate": 7.167850152661673e-07,
+      "loss": 0.5537,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7785153892576946,
+      "grad_norm": 4.306777769767276,
+      "learning_rate": 7.143260484239386e-07,
+      "loss": 0.4997,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7789177227921947,
+      "grad_norm": 4.472035583740821,
+      "learning_rate": 7.118706034237172e-07,
+      "loss": 0.6373,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7793200563266949,
+      "grad_norm": 3.8082808763272356,
+      "learning_rate": 7.094186851083187e-07,
+      "loss": 0.5713,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7797223898611949,
+      "grad_norm": 3.907355998411772,
+      "learning_rate": 7.069702983136023e-07,
+      "loss": 0.4916,
+      "step": 1938
+    },
+    {
+      "epoch": 0.780124723395695,
+      "grad_norm": 3.4404115391335623,
+      "learning_rate": 7.045254478684626e-07,
+      "loss": 0.4552,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7805270569301951,
+      "grad_norm": 4.072479874224982,
+      "learning_rate": 7.020841385948199e-07,
+      "loss": 0.5061,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7809293904646952,
+      "grad_norm": 5.014496338999608,
+      "learning_rate": 6.996463753076102e-07,
+      "loss": 0.5401,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7813317239991954,
+      "grad_norm": 4.5982221422710134,
+      "learning_rate": 6.972121628147766e-07,
+      "loss": 0.6385,
+      "step": 1942
+    },
+    {
+      "epoch": 0.7817340575336954,
+      "grad_norm": 3.866309336941733,
+      "learning_rate": 6.94781505917256e-07,
+      "loss": 0.5007,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7821363910681955,
+      "grad_norm": 4.000641087223532,
+      "learning_rate": 6.923544094089762e-07,
+      "loss": 0.4097,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7825387246026957,
+      "grad_norm": 4.183627371809607,
+      "learning_rate": 6.899308780768423e-07,
+      "loss": 0.4986,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7829410581371957,
+      "grad_norm": 3.834784873841407,
+      "learning_rate": 6.875109167007255e-07,
+      "loss": 0.4719,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7833433916716959,
+      "grad_norm": 4.624045837237235,
+      "learning_rate": 6.85094530053459e-07,
+      "loss": 0.5742,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7837457252061959,
+      "grad_norm": 4.968557042198568,
+      "learning_rate": 6.826817229008232e-07,
+      "loss": 0.6943,
+      "step": 1948
+    },
+    {
+      "epoch": 0.784148058740696,
+      "grad_norm": 3.798467897259186,
+      "learning_rate": 6.802725000015406e-07,
+      "loss": 0.5297,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7845503922751962,
+      "grad_norm": 3.974644659610978,
+      "learning_rate": 6.778668661072643e-07,
+      "loss": 0.4803,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7849527258096962,
+      "grad_norm": 4.048333500653427,
+      "learning_rate": 6.754648259625673e-07,
+      "loss": 0.5941,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7853550593441964,
+      "grad_norm": 4.3035958811807875,
+      "learning_rate": 6.730663843049367e-07,
+      "loss": 0.5955,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7857573928786964,
+      "grad_norm": 3.6544635394961764,
+      "learning_rate": 6.706715458647615e-07,
+      "loss": 0.4851,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7861597264131965,
+      "grad_norm": 4.218917868283341,
+      "learning_rate": 6.682803153653241e-07,
+      "loss": 0.6081,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7865620599476967,
+      "grad_norm": 3.617134711107577,
+      "learning_rate": 6.658926975227923e-07,
+      "loss": 0.4414,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7869643934821967,
+      "grad_norm": 4.041925903109421,
+      "learning_rate": 6.635086970462057e-07,
+      "loss": 0.4589,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7873667270166969,
+      "grad_norm": 3.9272206421591944,
+      "learning_rate": 6.611283186374723e-07,
+      "loss": 0.4697,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7877690605511969,
+      "grad_norm": 4.242979124153208,
+      "learning_rate": 6.587515669913558e-07,
+      "loss": 0.4185,
+      "step": 1958
+    },
+    {
+      "epoch": 0.788171394085697,
+      "grad_norm": 3.38359591341345,
+      "learning_rate": 6.563784467954668e-07,
+      "loss": 0.3828,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7885737276201972,
+      "grad_norm": 4.348954859955942,
+      "learning_rate": 6.540089627302523e-07,
+      "loss": 0.5046,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7889760611546972,
+      "grad_norm": 4.120739882383607,
+      "learning_rate": 6.516431194689907e-07,
+      "loss": 0.5227,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7893783946891973,
+      "grad_norm": 3.6740741354898305,
+      "learning_rate": 6.492809216777762e-07,
+      "loss": 0.4134,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7897807282236975,
+      "grad_norm": 3.690318810926642,
+      "learning_rate": 6.469223740155159e-07,
+      "loss": 0.4901,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7901830617581975,
+      "grad_norm": 4.2381781760809485,
+      "learning_rate": 6.445674811339175e-07,
+      "loss": 0.4225,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7905853952926977,
+      "grad_norm": 4.106368765680081,
+      "learning_rate": 6.422162476774788e-07,
+      "loss": 0.5645,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7909877288271977,
+      "grad_norm": 3.706517698008605,
+      "learning_rate": 6.398686782834815e-07,
+      "loss": 0.4995,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7913900623616978,
+      "grad_norm": 4.432403407342872,
+      "learning_rate": 6.375247775819809e-07,
+      "loss": 0.5127,
+      "step": 1967
+    },
+    {
+      "epoch": 0.791792395896198,
+      "grad_norm": 4.022769337353194,
+      "learning_rate": 6.351845501957971e-07,
+      "loss": 0.4525,
+      "step": 1968
+    },
+    {
+      "epoch": 0.792194729430698,
+      "grad_norm": 4.43806071523494,
+      "learning_rate": 6.328480007405027e-07,
+      "loss": 0.5067,
+      "step": 1969
+    },
+    {
+      "epoch": 0.7925970629651982,
+      "grad_norm": 3.7265320273813027,
+      "learning_rate": 6.305151338244192e-07,
+      "loss": 0.4277,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7929993964996982,
+      "grad_norm": 4.06221248413846,
+      "learning_rate": 6.281859540486043e-07,
+      "loss": 0.6297,
+      "step": 1971
+    },
+    {
+      "epoch": 0.7934017300341983,
+      "grad_norm": 4.391919257148749,
+      "learning_rate": 6.258604660068443e-07,
+      "loss": 0.4994,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7938040635686985,
+      "grad_norm": 4.252838693305146,
+      "learning_rate": 6.23538674285642e-07,
+      "loss": 0.5949,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7942063971031985,
+      "grad_norm": 4.325947652020019,
+      "learning_rate": 6.212205834642132e-07,
+      "loss": 0.5428,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7946087306376987,
+      "grad_norm": 4.185176650642697,
+      "learning_rate": 6.189061981144715e-07,
+      "loss": 0.5732,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7950110641721988,
+      "grad_norm": 3.2994931408989627,
+      "learning_rate": 6.165955228010242e-07,
+      "loss": 0.37,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7954133977066988,
+      "grad_norm": 3.9581400842405277,
+      "learning_rate": 6.14288562081162e-07,
+      "loss": 0.6003,
+      "step": 1977
+    },
+    {
+      "epoch": 0.795815731241199,
+      "grad_norm": 3.855839195313672,
+      "learning_rate": 6.11985320504847e-07,
+      "loss": 0.5755,
+      "step": 1978
+    },
+    {
+      "epoch": 0.796218064775699,
+      "grad_norm": 4.28483670710571,
+      "learning_rate": 6.096858026147079e-07,
+      "loss": 0.5674,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7966203983101992,
+      "grad_norm": 3.946092310322622,
+      "learning_rate": 6.073900129460286e-07,
+      "loss": 0.4429,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7970227318446993,
+      "grad_norm": 3.9442626812628725,
+      "learning_rate": 6.050979560267414e-07,
+      "loss": 0.4953,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7974250653791993,
+      "grad_norm": 4.023203847345498,
+      "learning_rate": 6.028096363774136e-07,
+      "loss": 0.5437,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7978273989136995,
+      "grad_norm": 4.5536613551126335,
+      "learning_rate": 6.005250585112437e-07,
+      "loss": 0.6177,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7982297324481995,
+      "grad_norm": 3.8312030174786313,
+      "learning_rate": 5.982442269340505e-07,
+      "loss": 0.4801,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7986320659826996,
+      "grad_norm": 3.8614656531277896,
+      "learning_rate": 5.959671461442634e-07,
+      "loss": 0.4395,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7990343995171998,
+      "grad_norm": 4.191209006767766,
+      "learning_rate": 5.936938206329148e-07,
+      "loss": 0.4913,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7994367330516998,
+      "grad_norm": 4.201209209604027,
+      "learning_rate": 5.914242548836291e-07,
+      "loss": 0.567,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7998390665862,
+      "grad_norm": 4.711191490796992,
+      "learning_rate": 5.891584533726177e-07,
+      "loss": 0.6445,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8002414001207001,
+      "grad_norm": 4.050282830317415,
+      "learning_rate": 5.868964205686653e-07,
+      "loss": 0.4854,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8006437336552001,
+      "grad_norm": 3.4477078683288838,
+      "learning_rate": 5.84638160933127e-07,
+      "loss": 0.4578,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8010460671897003,
+      "grad_norm": 3.977635083967608,
+      "learning_rate": 5.82383678919912e-07,
+      "loss": 0.4966,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8014484007242003,
+      "grad_norm": 4.254984216161458,
+      "learning_rate": 5.801329789754826e-07,
+      "loss": 0.5101,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8018507342587005,
+      "grad_norm": 4.1472621686996005,
+      "learning_rate": 5.778860655388399e-07,
+      "loss": 0.6366,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8022530677932006,
+      "grad_norm": 4.0712688803226715,
+      "learning_rate": 5.75642943041518e-07,
+      "loss": 0.4908,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8026554013277006,
+      "grad_norm": 4.69998065888643,
+      "learning_rate": 5.734036159075743e-07,
+      "loss": 0.6769,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8030577348622008,
+      "grad_norm": 3.5455182945935144,
+      "learning_rate": 5.711680885535784e-07,
+      "loss": 0.4177,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8034600683967008,
+      "grad_norm": 3.853566513480482,
+      "learning_rate": 5.689363653886085e-07,
+      "loss": 0.4434,
+      "step": 1997
+    },
+    {
+      "epoch": 0.803862401931201,
+      "grad_norm": 4.050333441948282,
+      "learning_rate": 5.667084508142386e-07,
+      "loss": 0.4822,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8042647354657011,
+      "grad_norm": 4.442980318720399,
+      "learning_rate": 5.644843492245322e-07,
+      "loss": 0.4872,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8046670690002011,
+      "grad_norm": 4.457653531733048,
+      "learning_rate": 5.622640650060301e-07,
+      "loss": 0.6444,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8050694025347013,
+      "grad_norm": 3.921778521793536,
+      "learning_rate": 5.600476025377463e-07,
+      "loss": 0.5391,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8054717360692014,
+      "grad_norm": 3.8081912666035755,
+      "learning_rate": 5.578349661911578e-07,
+      "loss": 0.4802,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8058740696037014,
+      "grad_norm": 4.001868301091059,
+      "learning_rate": 5.556261603301924e-07,
+      "loss": 0.5459,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8062764031382016,
+      "grad_norm": 3.847308434546458,
+      "learning_rate": 5.534211893112268e-07,
+      "loss": 0.4606,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8066787366727016,
+      "grad_norm": 4.15027221954597,
+      "learning_rate": 5.512200574830712e-07,
+      "loss": 0.5705,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8070810702072018,
+      "grad_norm": 3.8098780638264027,
+      "learning_rate": 5.490227691869656e-07,
+      "loss": 0.491,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8074834037417019,
+      "grad_norm": 3.9362205164209616,
+      "learning_rate": 5.468293287565696e-07,
+      "loss": 0.4342,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8078857372762019,
+      "grad_norm": 3.621695424225381,
+      "learning_rate": 5.446397405179534e-07,
+      "loss": 0.4943,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8082880708107021,
+      "grad_norm": 4.894139218776488,
+      "learning_rate": 5.424540087895888e-07,
+      "loss": 0.7108,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8086904043452021,
+      "grad_norm": 4.43306730596569,
+      "learning_rate": 5.402721378823422e-07,
+      "loss": 0.5759,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8090927378797023,
+      "grad_norm": 4.303012456014276,
+      "learning_rate": 5.380941320994659e-07,
+      "loss": 0.575,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8094950714142024,
+      "grad_norm": 4.313613214512251,
+      "learning_rate": 5.35919995736589e-07,
+      "loss": 0.4761,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8098974049487024,
+      "grad_norm": 3.8521996380430394,
+      "learning_rate": 5.337497330817074e-07,
+      "loss": 0.5352,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8102997384832026,
+      "grad_norm": 4.528703112399277,
+      "learning_rate": 5.315833484151786e-07,
+      "loss": 0.5959,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8107020720177027,
+      "grad_norm": 3.878303004429545,
+      "learning_rate": 5.294208460097125e-07,
+      "loss": 0.4815,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8111044055522028,
+      "grad_norm": 4.490283907824032,
+      "learning_rate": 5.272622301303587e-07,
+      "loss": 0.6368,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8115067390867029,
+      "grad_norm": 4.279104111581364,
+      "learning_rate": 5.251075050345056e-07,
+      "loss": 0.5216,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8119090726212029,
+      "grad_norm": 3.6608651815923086,
+      "learning_rate": 5.229566749718645e-07,
+      "loss": 0.4809,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8123114061557031,
+      "grad_norm": 4.666782295826111,
+      "learning_rate": 5.208097441844668e-07,
+      "loss": 0.504,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8127137396902032,
+      "grad_norm": 3.84925143665575,
+      "learning_rate": 5.186667169066523e-07,
+      "loss": 0.5423,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8131160732247033,
+      "grad_norm": 3.578156563304788,
+      "learning_rate": 5.165275973650633e-07,
+      "loss": 0.3936,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8135184067592034,
+      "grad_norm": 4.074315611637044,
+      "learning_rate": 5.143923897786329e-07,
+      "loss": 0.6302,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8139207402937034,
+      "grad_norm": 4.085845386699454,
+      "learning_rate": 5.122610983585808e-07,
+      "loss": 0.5789,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8143230738282036,
+      "grad_norm": 4.000394354287652,
+      "learning_rate": 5.101337273084017e-07,
+      "loss": 0.4229,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8147254073627037,
+      "grad_norm": 3.835079915243062,
+      "learning_rate": 5.080102808238585e-07,
+      "loss": 0.5005,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8151277408972037,
+      "grad_norm": 3.590556700281588,
+      "learning_rate": 5.058907630929749e-07,
+      "loss": 0.4452,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8155300744317039,
+      "grad_norm": 4.254640813185372,
+      "learning_rate": 5.037751782960234e-07,
+      "loss": 0.6328,
+      "step": 2027
+    },
+    {
+      "epoch": 0.815932407966204,
+      "grad_norm": 4.290085466017056,
+      "learning_rate": 5.016635306055223e-07,
+      "loss": 0.5785,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8163347415007041,
+      "grad_norm": 4.159416054122212,
+      "learning_rate": 4.995558241862242e-07,
+      "loss": 0.5129,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8167370750352042,
+      "grad_norm": 3.7061504996754153,
+      "learning_rate": 4.974520631951069e-07,
+      "loss": 0.5305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8171394085697042,
+      "grad_norm": 4.178397491082031,
+      "learning_rate": 4.953522517813689e-07,
+      "loss": 0.5855,
+      "step": 2031
+    },
+    {
+      "epoch": 0.8175417421042044,
+      "grad_norm": 4.154495627484645,
+      "learning_rate": 4.932563940864168e-07,
+      "loss": 0.4164,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8179440756387045,
+      "grad_norm": 4.283562955009001,
+      "learning_rate": 4.911644942438614e-07,
+      "loss": 0.4942,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8183464091732046,
+      "grad_norm": 3.307871102947408,
+      "learning_rate": 4.890765563795072e-07,
+      "loss": 0.4552,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8187487427077047,
+      "grad_norm": 4.15735416688211,
+      "learning_rate": 4.86992584611343e-07,
+      "loss": 0.5797,
+      "step": 2035
+    },
+    {
+      "epoch": 0.8191510762422047,
+      "grad_norm": 4.111111195260256,
+      "learning_rate": 4.849125830495366e-07,
+      "loss": 0.5776,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8195534097767049,
+      "grad_norm": 3.534887324822181,
+      "learning_rate": 4.828365557964257e-07,
+      "loss": 0.4913,
+      "step": 2037
+    },
+    {
+      "epoch": 0.819955743311205,
+      "grad_norm": 3.8702017938683606,
+      "learning_rate": 4.807645069465095e-07,
+      "loss": 0.4638,
+      "step": 2038
+    },
+    {
+      "epoch": 0.820358076845705,
+      "grad_norm": 4.105810885704208,
+      "learning_rate": 4.786964405864403e-07,
+      "loss": 0.51,
+      "step": 2039
+    },
+    {
+      "epoch": 0.8207604103802052,
+      "grad_norm": 3.7721413153025183,
+      "learning_rate": 4.766323607950149e-07,
+      "loss": 0.5327,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8211627439147053,
+      "grad_norm": 4.604383899796593,
+      "learning_rate": 4.7457227164316914e-07,
+      "loss": 0.6425,
+      "step": 2041
+    },
+    {
+      "epoch": 0.8215650774492054,
+      "grad_norm": 4.7452724983070595,
+      "learning_rate": 4.725161771939679e-07,
+      "loss": 0.5968,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8219674109837055,
+      "grad_norm": 4.13939509952924,
+      "learning_rate": 4.704640815025971e-07,
+      "loss": 0.5454,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8223697445182055,
+      "grad_norm": 4.07811911040908,
+      "learning_rate": 4.684159886163553e-07,
+      "loss": 0.4846,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8227720780527057,
+      "grad_norm": 4.223857527886817,
+      "learning_rate": 4.663719025746477e-07,
+      "loss": 0.548,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8231744115872058,
+      "grad_norm": 4.055398221449843,
+      "learning_rate": 4.643318274089756e-07,
+      "loss": 0.4762,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8235767451217059,
+      "grad_norm": 3.93292252326897,
+      "learning_rate": 4.6229576714293067e-07,
+      "loss": 0.5506,
+      "step": 2047
+    },
+    {
+      "epoch": 0.823979078656206,
+      "grad_norm": 3.5755501548250357,
+      "learning_rate": 4.6026372579218686e-07,
+      "loss": 0.5277,
+      "step": 2048
+    },
+    {
+      "epoch": 0.824381412190706,
+      "grad_norm": 3.900846576287911,
+      "learning_rate": 4.5823570736448913e-07,
+      "loss": 0.4289,
+      "step": 2049
+    },
+    {
+      "epoch": 0.8247837457252062,
+      "grad_norm": 4.422345897561269,
+      "learning_rate": 4.5621171585965046e-07,
+      "loss": 0.5439,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8251860792597063,
+      "grad_norm": 3.86188238287213,
+      "learning_rate": 4.5419175526954074e-07,
+      "loss": 0.4971,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8255884127942064,
+      "grad_norm": 4.149840183197456,
+      "learning_rate": 4.5217582957808064e-07,
+      "loss": 0.4913,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8259907463287065,
+      "grad_norm": 4.14442670188834,
+      "learning_rate": 4.501639427612309e-07,
+      "loss": 0.593,
+      "step": 2053
+    },
+    {
+      "epoch": 0.8263930798632066,
+      "grad_norm": 4.160435006991155,
+      "learning_rate": 4.4815609878698847e-07,
+      "loss": 0.5702,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8267954133977067,
+      "grad_norm": 4.699994325668033,
+      "learning_rate": 4.46152301615376e-07,
+      "loss": 0.6052,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8271977469322068,
+      "grad_norm": 4.158216057441291,
+      "learning_rate": 4.441525551984349e-07,
+      "loss": 0.6375,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8276000804667069,
+      "grad_norm": 5.052807884367178,
+      "learning_rate": 4.421568634802165e-07,
+      "loss": 0.687,
+      "step": 2057
+    },
+    {
+      "epoch": 0.828002414001207,
+      "grad_norm": 3.9353061661541133,
+      "learning_rate": 4.40165230396776e-07,
+      "loss": 0.4774,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8284047475357071,
+      "grad_norm": 4.01193456780117,
+      "learning_rate": 4.3817765987616474e-07,
+      "loss": 0.4932,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8288070810702072,
+      "grad_norm": 4.976729164334829,
+      "learning_rate": 4.361941558384189e-07,
+      "loss": 0.6807,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8292094146047073,
+      "grad_norm": 4.339682582742787,
+      "learning_rate": 4.3421472219555725e-07,
+      "loss": 0.4727,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8296117481392074,
+      "grad_norm": 4.503522087170257,
+      "learning_rate": 4.3223936285156836e-07,
+      "loss": 0.6022,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8300140816737075,
+      "grad_norm": 4.009166626990359,
+      "learning_rate": 4.3026808170240686e-07,
+      "loss": 0.6318,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8304164152082076,
+      "grad_norm": 4.433472277211533,
+      "learning_rate": 4.283008826359833e-07,
+      "loss": 0.6059,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8308187487427077,
+      "grad_norm": 4.239654566860423,
+      "learning_rate": 4.263377695321577e-07,
+      "loss": 0.5319,
+      "step": 2065
+    },
+    {
+      "epoch": 0.8312210822772078,
+      "grad_norm": 4.33872236149242,
+      "learning_rate": 4.2437874626273024e-07,
+      "loss": 0.5192,
+      "step": 2066
+    },
+    {
+      "epoch": 0.831623415811708,
+      "grad_norm": 4.396517305178651,
+      "learning_rate": 4.2242381669143603e-07,
+      "loss": 0.6266,
+      "step": 2067
+    },
+    {
+      "epoch": 0.832025749346208,
+      "grad_norm": 4.213477790744039,
+      "learning_rate": 4.204729846739358e-07,
+      "loss": 0.6738,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8324280828807081,
+      "grad_norm": 3.312219624247552,
+      "learning_rate": 4.1852625405780914e-07,
+      "loss": 0.4833,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8328304164152082,
+      "grad_norm": 3.9392383731854257,
+      "learning_rate": 4.165836286825464e-07,
+      "loss": 0.4513,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8332327499497083,
+      "grad_norm": 3.8224909723307734,
+      "learning_rate": 4.1464511237954056e-07,
+      "loss": 0.5628,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8336350834842085,
+      "grad_norm": 4.5465506777870885,
+      "learning_rate": 4.1271070897208166e-07,
+      "loss": 0.5761,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8340374170187085,
+      "grad_norm": 3.87942801566727,
+      "learning_rate": 4.107804222753464e-07,
+      "loss": 0.4364,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8344397505532086,
+      "grad_norm": 3.901824805945848,
+      "learning_rate": 4.088542560963937e-07,
+      "loss": 0.4897,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8348420840877087,
+      "grad_norm": 4.2369387169530155,
+      "learning_rate": 4.0693221423415447e-07,
+      "loss": 0.6186,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8352444176222088,
+      "grad_norm": 4.1329987533853005,
+      "learning_rate": 4.0501430047942603e-07,
+      "loss": 0.5358,
+      "step": 2076
+    },
+    {
+      "epoch": 0.835646751156709,
+      "grad_norm": 4.533598883351018,
+      "learning_rate": 4.03100518614864e-07,
+      "loss": 0.5522,
+      "step": 2077
+    },
+    {
+      "epoch": 0.836049084691209,
+      "grad_norm": 3.998009974659786,
+      "learning_rate": 4.011908724149746e-07,
+      "loss": 0.3743,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8364514182257091,
+      "grad_norm": 3.417865078970055,
+      "learning_rate": 3.9928536564610767e-07,
+      "loss": 0.4035,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8368537517602093,
+      "grad_norm": 3.9749987307006345,
+      "learning_rate": 3.9738400206644766e-07,
+      "loss": 0.6477,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8372560852947093,
+      "grad_norm": 4.037697255647064,
+      "learning_rate": 3.9548678542600886e-07,
+      "loss": 0.5661,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8376584188292094,
+      "grad_norm": 4.431095797135675,
+      "learning_rate": 3.935937194666267e-07,
+      "loss": 0.4818,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8380607523637095,
+      "grad_norm": 4.260045700942229,
+      "learning_rate": 3.9170480792194985e-07,
+      "loss": 0.4263,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8384630858982096,
+      "grad_norm": 4.051159338313962,
+      "learning_rate": 3.8982005451743243e-07,
+      "loss": 0.487,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8388654194327098,
+      "grad_norm": 3.505632578812268,
+      "learning_rate": 3.8793946297032926e-07,
+      "loss": 0.5173,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8392677529672098,
+      "grad_norm": 3.4621570841739917,
+      "learning_rate": 3.8606303698968643e-07,
+      "loss": 0.3609,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8396700865017099,
+      "grad_norm": 3.708654324006204,
+      "learning_rate": 3.841907802763331e-07,
+      "loss": 0.4437,
+      "step": 2087
+    },
+    {
+      "epoch": 0.84007242003621,
+      "grad_norm": 3.3369961693311447,
+      "learning_rate": 3.8232269652287713e-07,
+      "loss": 0.4295,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8404747535707101,
+      "grad_norm": 3.8475631558462884,
+      "learning_rate": 3.8045878941369434e-07,
+      "loss": 0.5628,
+      "step": 2089
+    },
+    {
+      "epoch": 0.8408770871052103,
+      "grad_norm": 3.827602289260064,
+      "learning_rate": 3.7859906262492503e-07,
+      "loss": 0.5158,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8412794206397103,
+      "grad_norm": 4.093693428637088,
+      "learning_rate": 3.7674351982446347e-07,
+      "loss": 0.4742,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8416817541742104,
+      "grad_norm": 4.310046930338328,
+      "learning_rate": 3.7489216467195304e-07,
+      "loss": 0.6097,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8420840877087106,
+      "grad_norm": 4.017527345482499,
+      "learning_rate": 3.730450008187761e-07,
+      "loss": 0.5075,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8424864212432106,
+      "grad_norm": 3.946920016020563,
+      "learning_rate": 3.7120203190805034e-07,
+      "loss": 0.4275,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8428887547777107,
+      "grad_norm": 4.145255268119006,
+      "learning_rate": 3.6936326157461925e-07,
+      "loss": 0.5914,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8432910883122108,
+      "grad_norm": 4.2460232246744525,
+      "learning_rate": 3.6752869344504586e-07,
+      "loss": 0.5152,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8436934218467109,
+      "grad_norm": 3.95033426326261,
+      "learning_rate": 3.656983311376042e-07,
+      "loss": 0.4448,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8440957553812111,
+      "grad_norm": 3.8931938435117286,
+      "learning_rate": 3.638721782622745e-07,
+      "loss": 0.499,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8444980889157111,
+      "grad_norm": 4.3118732957405985,
+      "learning_rate": 3.620502384207342e-07,
+      "loss": 0.49,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8449004224502112,
+      "grad_norm": 4.005112629756995,
+      "learning_rate": 3.602325152063524e-07,
+      "loss": 0.4125,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8453027559847113,
+      "grad_norm": 3.9812849673226114,
+      "learning_rate": 3.5841901220418e-07,
+      "loss": 0.5451,
+      "step": 2101
+    },
+    {
+      "epoch": 0.8457050895192114,
+      "grad_norm": 3.948178072792975,
+      "learning_rate": 3.566097329909468e-07,
+      "loss": 0.6427,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8461074230537116,
+      "grad_norm": 3.7188722401214016,
+      "learning_rate": 3.5480468113504946e-07,
+      "loss": 0.405,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8465097565882116,
+      "grad_norm": 3.913596983806647,
+      "learning_rate": 3.530038601965491e-07,
+      "loss": 0.4635,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8469120901227117,
+      "grad_norm": 4.411578338116122,
+      "learning_rate": 3.5120727372716295e-07,
+      "loss": 0.5511,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8473144236572119,
+      "grad_norm": 3.282083239679109,
+      "learning_rate": 3.494149252702536e-07,
+      "loss": 0.4441,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8477167571917119,
+      "grad_norm": 4.011319326831284,
+      "learning_rate": 3.476268183608281e-07,
+      "loss": 0.5018,
+      "step": 2107
+    },
+    {
+      "epoch": 0.848119090726212,
+      "grad_norm": 4.284870919327636,
+      "learning_rate": 3.4584295652552667e-07,
+      "loss": 0.4846,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8485214242607121,
+      "grad_norm": 4.276416896965254,
+      "learning_rate": 3.4406334328261765e-07,
+      "loss": 0.5376,
+      "step": 2109
+    },
+    {
+      "epoch": 0.8489237577952122,
+      "grad_norm": 3.834125175179826,
+      "learning_rate": 3.4228798214198974e-07,
+      "loss": 0.5651,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8493260913297124,
+      "grad_norm": 3.9930841906813517,
+      "learning_rate": 3.405168766051445e-07,
+      "loss": 0.3851,
+      "step": 2111
+    },
+    {
+      "epoch": 0.8497284248642124,
+      "grad_norm": 4.2112351410317626,
+      "learning_rate": 3.3875003016519137e-07,
+      "loss": 0.5431,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8501307583987125,
+      "grad_norm": 4.52788654812861,
+      "learning_rate": 3.369874463068398e-07,
+      "loss": 0.5849,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8505330919332126,
+      "grad_norm": 3.8509825855331163,
+      "learning_rate": 3.3522912850639106e-07,
+      "loss": 0.5748,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8509354254677127,
+      "grad_norm": 3.510509250668585,
+      "learning_rate": 3.334750802317338e-07,
+      "loss": 0.4397,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8513377590022129,
+      "grad_norm": 3.8349389092945114,
+      "learning_rate": 3.3172530494233467e-07,
+      "loss": 0.4769,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8517400925367129,
+      "grad_norm": 3.8288751826393197,
+      "learning_rate": 3.2997980608923426e-07,
+      "loss": 0.5434,
+      "step": 2117
+    },
+    {
+      "epoch": 0.852142426071213,
+      "grad_norm": 3.9001403159636374,
+      "learning_rate": 3.282385871150387e-07,
+      "loss": 0.4523,
+      "step": 2118
+    },
+    {
+      "epoch": 0.8525447596057132,
+      "grad_norm": 3.6933669568597978,
+      "learning_rate": 3.26501651453911e-07,
+      "loss": 0.3905,
+      "step": 2119
+    },
+    {
+      "epoch": 0.8529470931402132,
+      "grad_norm": 3.811615026103731,
+      "learning_rate": 3.247690025315689e-07,
+      "loss": 0.4305,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8533494266747134,
+      "grad_norm": 3.6210800698038925,
+      "learning_rate": 3.2304064376527417e-07,
+      "loss": 0.5332,
+      "step": 2121
+    },
+    {
+      "epoch": 0.8537517602092134,
+      "grad_norm": 4.231885438858511,
+      "learning_rate": 3.2131657856382753e-07,
+      "loss": 0.5911,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8541540937437135,
+      "grad_norm": 4.464935656077254,
+      "learning_rate": 3.195968103275618e-07,
+      "loss": 0.5702,
+      "step": 2123
+    },
+    {
+      "epoch": 0.8545564272782137,
+      "grad_norm": 4.3078684853827465,
+      "learning_rate": 3.17881342448334e-07,
+      "loss": 0.5691,
+      "step": 2124
+    },
+    {
+      "epoch": 0.8549587608127137,
+      "grad_norm": 3.8241264163503255,
+      "learning_rate": 3.1617017830952084e-07,
+      "loss": 0.5375,
+      "step": 2125
+    },
+    {
+      "epoch": 0.8553610943472139,
+      "grad_norm": 3.519525197941784,
+      "learning_rate": 3.1446332128601015e-07,
+      "loss": 0.4416,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8557634278817139,
+      "grad_norm": 3.4697912989841524,
+      "learning_rate": 3.12760774744196e-07,
+      "loss": 0.4459,
+      "step": 2127
+    },
+    {
+      "epoch": 0.856165761416214,
+      "grad_norm": 3.7851796176468455,
+      "learning_rate": 3.1106254204196883e-07,
+      "loss": 0.4682,
+      "step": 2128
+    },
+    {
+      "epoch": 0.8565680949507142,
+      "grad_norm": 3.936606771811409,
+      "learning_rate": 3.093686265287138e-07,
+      "loss": 0.5729,
+      "step": 2129
+    },
+    {
+      "epoch": 0.8569704284852142,
+      "grad_norm": 4.190455136191216,
+      "learning_rate": 3.076790315452988e-07,
+      "loss": 0.5277,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8573727620197144,
+      "grad_norm": 4.108892268475147,
+      "learning_rate": 3.05993760424072e-07,
+      "loss": 0.4329,
+      "step": 2131
+    },
+    {
+      "epoch": 0.8577750955542145,
+      "grad_norm": 4.390564938079403,
+      "learning_rate": 3.043128164888537e-07,
+      "loss": 0.5011,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8581774290887145,
+      "grad_norm": 4.367475352759402,
+      "learning_rate": 3.0263620305492877e-07,
+      "loss": 0.4797,
+      "step": 2133
+    },
+    {
+      "epoch": 0.8585797626232147,
+      "grad_norm": 3.8444843612579924,
+      "learning_rate": 3.009639234290418e-07,
+      "loss": 0.5327,
+      "step": 2134
+    },
+    {
+      "epoch": 0.8589820961577147,
+      "grad_norm": 3.6090606638950318,
+      "learning_rate": 2.992959809093901e-07,
+      "loss": 0.4413,
+      "step": 2135
+    },
+    {
+      "epoch": 0.8593844296922148,
+      "grad_norm": 4.388230282282369,
+      "learning_rate": 2.976323787856175e-07,
+      "loss": 0.5717,
+      "step": 2136
+    },
+    {
+      "epoch": 0.859786763226715,
+      "grad_norm": 4.333473286328121,
+      "learning_rate": 2.9597312033880557e-07,
+      "loss": 0.6314,
+      "step": 2137
+    },
+    {
+      "epoch": 0.860189096761215,
+      "grad_norm": 3.890510795148687,
+      "learning_rate": 2.943182088414701e-07,
+      "loss": 0.433,
+      "step": 2138
+    },
+    {
+      "epoch": 0.8605914302957152,
+      "grad_norm": 4.234653691147445,
+      "learning_rate": 2.9266764755755406e-07,
+      "loss": 0.54,
+      "step": 2139
+    },
+    {
+      "epoch": 0.8609937638302152,
+      "grad_norm": 4.183758834087082,
+      "learning_rate": 2.910214397424202e-07,
+      "loss": 0.4845,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8613960973647153,
+      "grad_norm": 3.799668423286151,
+      "learning_rate": 2.8937958864284353e-07,
+      "loss": 0.4333,
+      "step": 2141
+    },
+    {
+      "epoch": 0.8617984308992155,
+      "grad_norm": 4.133997388279442,
+      "learning_rate": 2.877420974970088e-07,
+      "loss": 0.4396,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8622007644337155,
+      "grad_norm": 4.0525547445688614,
+      "learning_rate": 2.861089695345007e-07,
+      "loss": 0.4857,
+      "step": 2143
+    },
+    {
+      "epoch": 0.8626030979682157,
+      "grad_norm": 3.7880628956269904,
+      "learning_rate": 2.8448020797629735e-07,
+      "loss": 0.5046,
+      "step": 2144
+    },
+    {
+      "epoch": 0.8630054315027158,
+      "grad_norm": 4.143267859487349,
+      "learning_rate": 2.8285581603476747e-07,
+      "loss": 0.5291,
+      "step": 2145
+    },
+    {
+      "epoch": 0.8634077650372158,
+      "grad_norm": 3.8047765913147993,
+      "learning_rate": 2.8123579691365923e-07,
+      "loss": 0.5513,
+      "step": 2146
+    },
+    {
+      "epoch": 0.863810098571716,
+      "grad_norm": 4.041711864076377,
+      "learning_rate": 2.796201538080981e-07,
+      "loss": 0.448,
+      "step": 2147
+    },
+    {
+      "epoch": 0.864212432106216,
+      "grad_norm": 4.193683351365294,
+      "learning_rate": 2.7800888990457855e-07,
+      "loss": 0.5598,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8646147656407162,
+      "grad_norm": 3.6613935569874583,
+      "learning_rate": 2.7640200838095793e-07,
+      "loss": 0.4067,
+      "step": 2149
+    },
+    {
+      "epoch": 0.8650170991752163,
+      "grad_norm": 4.104620267298875,
+      "learning_rate": 2.7479951240644953e-07,
+      "loss": 0.5648,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8654194327097163,
+      "grad_norm": 3.7600215255952256,
+      "learning_rate": 2.7320140514161846e-07,
+      "loss": 0.3946,
+      "step": 2151
+    },
+    {
+      "epoch": 0.8658217662442165,
+      "grad_norm": 3.9525358112645965,
+      "learning_rate": 2.7160768973837293e-07,
+      "loss": 0.53,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8662240997787165,
+      "grad_norm": 3.5450533427675226,
+      "learning_rate": 2.7001836933995995e-07,
+      "loss": 0.5298,
+      "step": 2153
+    },
+    {
+      "epoch": 0.8666264333132166,
+      "grad_norm": 3.6668666878628815,
+      "learning_rate": 2.684334470809588e-07,
+      "loss": 0.5056,
+      "step": 2154
+    },
+    {
+      "epoch": 0.8670287668477168,
+      "grad_norm": 4.224237505147509,
+      "learning_rate": 2.6685292608727236e-07,
+      "loss": 0.5218,
+      "step": 2155
+    },
+    {
+      "epoch": 0.8674311003822168,
+      "grad_norm": 3.69557439406065,
+      "learning_rate": 2.652768094761257e-07,
+      "loss": 0.3756,
+      "step": 2156
+    },
+    {
+      "epoch": 0.867833433916717,
+      "grad_norm": 4.281090167368473,
+      "learning_rate": 2.637051003560548e-07,
+      "loss": 0.6402,
+      "step": 2157
+    },
+    {
+      "epoch": 0.8682357674512171,
+      "grad_norm": 3.9633212522861054,
+      "learning_rate": 2.621378018269047e-07,
+      "loss": 0.5075,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8686381009857171,
+      "grad_norm": 3.8583945548771825,
+      "learning_rate": 2.6057491697982e-07,
+      "loss": 0.6033,
+      "step": 2159
+    },
+    {
+      "epoch": 0.8690404345202173,
+      "grad_norm": 3.3558856738663487,
+      "learning_rate": 2.5901644889724186e-07,
+      "loss": 0.4015,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8694427680547173,
+      "grad_norm": 3.5749358804434483,
+      "learning_rate": 2.574624006528992e-07,
+      "loss": 0.4944,
+      "step": 2161
+    },
+    {
+      "epoch": 0.8698451015892175,
+      "grad_norm": 3.86587826976141,
+      "learning_rate": 2.55912775311804e-07,
+      "loss": 0.4891,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8702474351237176,
+      "grad_norm": 3.9917581972004785,
+      "learning_rate": 2.543675759302464e-07,
+      "loss": 0.5079,
+      "step": 2163
+    },
+    {
+      "epoch": 0.8706497686582176,
+      "grad_norm": 4.393976999836958,
+      "learning_rate": 2.5282680555578477e-07,
+      "loss": 0.6145,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8710521021927178,
+      "grad_norm": 4.04228191167184,
+      "learning_rate": 2.51290467227244e-07,
+      "loss": 0.5429,
+      "step": 2165
+    },
+    {
+      "epoch": 0.8714544357272178,
+      "grad_norm": 4.4282484060475324,
+      "learning_rate": 2.497585639747077e-07,
+      "loss": 0.6325,
+      "step": 2166
+    },
+    {
+      "epoch": 0.871856769261718,
+      "grad_norm": 3.816275160612967,
+      "learning_rate": 2.482310988195125e-07,
+      "loss": 0.5543,
+      "step": 2167
+    },
+    {
+      "epoch": 0.8722591027962181,
+      "grad_norm": 4.1145602142866675,
+      "learning_rate": 2.4670807477424007e-07,
+      "loss": 0.5527,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8726614363307181,
+      "grad_norm": 4.231321602541556,
+      "learning_rate": 2.4518949484271525e-07,
+      "loss": 0.6496,
+      "step": 2169
+    },
+    {
+      "epoch": 0.8730637698652183,
+      "grad_norm": 4.220523454555819,
+      "learning_rate": 2.4367536201999695e-07,
+      "loss": 0.5652,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8734661033997184,
+      "grad_norm": 3.938889256449136,
+      "learning_rate": 2.421656792923724e-07,
+      "loss": 0.456,
+      "step": 2171
+    },
+    {
+      "epoch": 0.8738684369342185,
+      "grad_norm": 4.5217908297671485,
+      "learning_rate": 2.406604496373535e-07,
+      "loss": 0.6334,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8742707704687186,
+      "grad_norm": 3.7526929266944773,
+      "learning_rate": 2.391596760236681e-07,
+      "loss": 0.4743,
+      "step": 2173
+    },
+    {
+      "epoch": 0.8746731040032186,
+      "grad_norm": 3.710705344410006,
+      "learning_rate": 2.376633614112561e-07,
+      "loss": 0.5661,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8750754375377188,
+      "grad_norm": 3.2609967694053243,
+      "learning_rate": 2.361715087512631e-07,
+      "loss": 0.3589,
+      "step": 2175
+    },
+    {
+      "epoch": 0.8754777710722189,
+      "grad_norm": 4.108101136746485,
+      "learning_rate": 2.346841209860351e-07,
+      "loss": 0.4368,
+      "step": 2176
+    },
+    {
+      "epoch": 0.875880104606719,
+      "grad_norm": 3.763267262463342,
+      "learning_rate": 2.3320120104911053e-07,
+      "loss": 0.4953,
+      "step": 2177
+    },
+    {
+      "epoch": 0.8762824381412191,
+      "grad_norm": 5.025294007587999,
+      "learning_rate": 2.317227518652168e-07,
+      "loss": 0.6524,
+      "step": 2178
+    },
+    {
+      "epoch": 0.8766847716757191,
+      "grad_norm": 3.9367012271057584,
+      "learning_rate": 2.302487763502645e-07,
+      "loss": 0.4149,
+      "step": 2179
+    },
+    {
+      "epoch": 0.8770871052102193,
+      "grad_norm": 4.236893564539156,
+      "learning_rate": 2.2877927741134036e-07,
+      "loss": 0.5532,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8774894387447194,
+      "grad_norm": 3.5424218920413373,
+      "learning_rate": 2.2731425794670135e-07,
+      "loss": 0.4691,
+      "step": 2181
+    },
+    {
+      "epoch": 0.8778917722792194,
+      "grad_norm": 4.04407474745495,
+      "learning_rate": 2.2585372084577033e-07,
+      "loss": 0.5125,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8782941058137196,
+      "grad_norm": 3.8612028242647147,
+      "learning_rate": 2.2439766898913e-07,
+      "loss": 0.5256,
+      "step": 2183
+    },
+    {
+      "epoch": 0.8786964393482197,
+      "grad_norm": 4.271299959657279,
+      "learning_rate": 2.2294610524851685e-07,
+      "loss": 0.4944,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8790987728827198,
+      "grad_norm": 4.465135814281888,
+      "learning_rate": 2.2149903248681464e-07,
+      "loss": 0.6834,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8795011064172199,
+      "grad_norm": 3.9117028793633173,
+      "learning_rate": 2.2005645355805127e-07,
+      "loss": 0.4605,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8799034399517199,
+      "grad_norm": 3.8402220541360292,
+      "learning_rate": 2.1861837130738945e-07,
+      "loss": 0.4467,
+      "step": 2187
+    },
+    {
+      "epoch": 0.8803057734862201,
+      "grad_norm": 4.223182944719613,
+      "learning_rate": 2.1718478857112506e-07,
+      "loss": 0.5347,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8807081070207202,
+      "grad_norm": 3.8811613151194213,
+      "learning_rate": 2.1575570817667952e-07,
+      "loss": 0.5265,
+      "step": 2189
+    },
+    {
+      "epoch": 0.8811104405552203,
+      "grad_norm": 3.8705546986459956,
+      "learning_rate": 2.143311329425926e-07,
+      "loss": 0.4427,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8815127740897204,
+      "grad_norm": 4.5414937373063955,
+      "learning_rate": 2.1291106567852105e-07,
+      "loss": 0.6442,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8819151076242204,
+      "grad_norm": 3.588299387450043,
+      "learning_rate": 2.1149550918522887e-07,
+      "loss": 0.4173,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8823174411587206,
+      "grad_norm": 4.165225372551161,
+      "learning_rate": 2.100844662545848e-07,
+      "loss": 0.5496,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8827197746932207,
+      "grad_norm": 4.217297273907552,
+      "learning_rate": 2.086779396695554e-07,
+      "loss": 0.5329,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8831221082277207,
+      "grad_norm": 3.980212414160413,
+      "learning_rate": 2.0727593220419812e-07,
+      "loss": 0.4682,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8835244417622209,
+      "grad_norm": 4.015069610064507,
+      "learning_rate": 2.0587844662365963e-07,
+      "loss": 0.4975,
+      "step": 2196
+    },
+    {
+      "epoch": 0.883926775296721,
+      "grad_norm": 3.675033120744641,
+      "learning_rate": 2.0448548568416715e-07,
+      "loss": 0.4932,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8843291088312211,
+      "grad_norm": 3.7917456823069178,
+      "learning_rate": 2.030970521330247e-07,
+      "loss": 0.4523,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8847314423657212,
+      "grad_norm": 4.2084887531102195,
+      "learning_rate": 2.0171314870860548e-07,
+      "loss": 0.5797,
+      "step": 2199
+    },
+    {
+      "epoch": 0.8851337759002212,
+      "grad_norm": 4.1920678247066245,
+      "learning_rate": 2.0033377814034999e-07,
+      "loss": 0.5458,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8855361094347214,
+      "grad_norm": 4.341033030777012,
+      "learning_rate": 1.9895894314875707e-07,
+      "loss": 0.5069,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8859384429692215,
+      "grad_norm": 4.251489033634488,
+      "learning_rate": 1.9758864644538188e-07,
+      "loss": 0.5658,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8863407765037216,
+      "grad_norm": 4.118913583140747,
+      "learning_rate": 1.9622289073282675e-07,
+      "loss": 0.5186,
+      "step": 2203
+    },
+    {
+      "epoch": 0.8867431100382217,
+      "grad_norm": 4.17198274381597,
+      "learning_rate": 1.9486167870473915e-07,
+      "loss": 0.5394,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8871454435727217,
+      "grad_norm": 4.1140466186765865,
+      "learning_rate": 1.9350501304580577e-07,
+      "loss": 0.6302,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8875477771072219,
+      "grad_norm": 4.252952260852059,
+      "learning_rate": 1.9215289643174485e-07,
+      "loss": 0.5715,
+      "step": 2206
+    },
+    {
+      "epoch": 0.887950110641722,
+      "grad_norm": 3.9053915680472673,
+      "learning_rate": 1.9080533152930486e-07,
+      "loss": 0.502,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8883524441762221,
+      "grad_norm": 3.702225135993101,
+      "learning_rate": 1.8946232099625482e-07,
+      "loss": 0.3917,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8887547777107222,
+      "grad_norm": 3.5875962058050863,
+      "learning_rate": 1.8812386748138234e-07,
+      "loss": 0.3413,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8891571112452223,
+      "grad_norm": 3.6894845944247723,
+      "learning_rate": 1.8678997362448743e-07,
+      "loss": 0.4334,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8895594447797224,
+      "grad_norm": 3.977873097022255,
+      "learning_rate": 1.8546064205637758e-07,
+      "loss": 0.4746,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8899617783142225,
+      "grad_norm": 3.4865669801340147,
+      "learning_rate": 1.8413587539886064e-07,
+      "loss": 0.4173,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8903641118487226,
+      "grad_norm": 4.35862373407535,
+      "learning_rate": 1.828156762647429e-07,
+      "loss": 0.5653,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8907664453832227,
+      "grad_norm": 4.2652495550914775,
+      "learning_rate": 1.81500047257821e-07,
+      "loss": 0.5071,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8911687789177228,
+      "grad_norm": 3.285190042692873,
+      "learning_rate": 1.8018899097287906e-07,
+      "loss": 0.3731,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8915711124522229,
+      "grad_norm": 3.782205241606253,
+      "learning_rate": 1.78882509995682e-07,
+      "loss": 0.4907,
+      "step": 2216
+    },
+    {
+      "epoch": 0.891973445986723,
+      "grad_norm": 4.391188417961942,
+      "learning_rate": 1.7758060690297035e-07,
+      "loss": 0.3594,
+      "step": 2217
+    },
+    {
+      "epoch": 0.892375779521223,
+      "grad_norm": 3.172977998916911,
+      "learning_rate": 1.7628328426245677e-07,
+      "loss": 0.3866,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8927781130557232,
+      "grad_norm": 4.109857114119739,
+      "learning_rate": 1.7499054463281985e-07,
+      "loss": 0.5489,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8931804465902233,
+      "grad_norm": 3.6387926247562214,
+      "learning_rate": 1.7370239056369946e-07,
+      "loss": 0.3892,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8935827801247234,
+      "grad_norm": 3.897940749417438,
+      "learning_rate": 1.7241882459569008e-07,
+      "loss": 0.6287,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8939851136592235,
+      "grad_norm": 3.968476012728493,
+      "learning_rate": 1.7113984926033832e-07,
+      "loss": 0.459,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8943874471937237,
+      "grad_norm": 4.209938898766111,
+      "learning_rate": 1.6986546708013722e-07,
+      "loss": 0.5426,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8947897807282237,
+      "grad_norm": 3.791580194250541,
+      "learning_rate": 1.6859568056852016e-07,
+      "loss": 0.498,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8951921142627238,
+      "grad_norm": 3.8639160637014665,
+      "learning_rate": 1.673304922298563e-07,
+      "loss": 0.4049,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8955944477972239,
+      "grad_norm": 4.147228062957371,
+      "learning_rate": 1.6606990455944634e-07,
+      "loss": 0.5258,
+      "step": 2226
+    },
+    {
+      "epoch": 0.895996781331724,
+      "grad_norm": 4.247308433616599,
+      "learning_rate": 1.648139200435178e-07,
+      "loss": 0.5252,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8963991148662241,
+      "grad_norm": 4.057558052673904,
+      "learning_rate": 1.635625411592179e-07,
+      "loss": 0.4851,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8968014484007242,
+      "grad_norm": 4.529371570037193,
+      "learning_rate": 1.6231577037461228e-07,
+      "loss": 0.6091,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8972037819352243,
+      "grad_norm": 4.092889891063284,
+      "learning_rate": 1.6107361014867622e-07,
+      "loss": 0.5051,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8976061154697244,
+      "grad_norm": 4.057510211081129,
+      "learning_rate": 1.598360629312931e-07,
+      "loss": 0.5379,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8980084490042245,
+      "grad_norm": 4.08275289824886,
+      "learning_rate": 1.586031311632477e-07,
+      "loss": 0.5542,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8984107825387246,
+      "grad_norm": 5.7116466876850875,
+      "learning_rate": 1.5737481727622212e-07,
+      "loss": 0.5021,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8988131160732247,
+      "grad_norm": 3.507507523132951,
+      "learning_rate": 1.5615112369278957e-07,
+      "loss": 0.391,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8992154496077248,
+      "grad_norm": 3.7113726990024194,
+      "learning_rate": 1.5493205282641228e-07,
+      "loss": 0.4515,
+      "step": 2235
+    },
+    {
+      "epoch": 0.899617783142225,
+      "grad_norm": 3.512898207115907,
+      "learning_rate": 1.5371760708143447e-07,
+      "loss": 0.4935,
+      "step": 2236
+    },
+    {
+      "epoch": 0.900020116676725,
+      "grad_norm": 4.213491522925021,
+      "learning_rate": 1.525077888530782e-07,
+      "loss": 0.538,
+      "step": 2237
+    },
+    {
+      "epoch": 0.9004224502112251,
+      "grad_norm": 4.482285849278304,
+      "learning_rate": 1.5130260052743955e-07,
+      "loss": 0.6545,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9008247837457252,
+      "grad_norm": 3.7492718480617384,
+      "learning_rate": 1.5010204448148214e-07,
+      "loss": 0.4442,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9012271172802253,
+      "grad_norm": 3.9838020107249195,
+      "learning_rate": 1.489061230830338e-07,
+      "loss": 0.4953,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9016294508147255,
+      "grad_norm": 3.9003308620272796,
+      "learning_rate": 1.4771483869078228e-07,
+      "loss": 0.5208,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9020317843492255,
+      "grad_norm": 3.788631062888711,
+      "learning_rate": 1.4652819365426867e-07,
+      "loss": 0.5613,
+      "step": 2242
+    },
+    {
+      "epoch": 0.9024341178837256,
+      "grad_norm": 4.676276663949487,
+      "learning_rate": 1.453461903138853e-07,
+      "loss": 0.5352,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9028364514182257,
+      "grad_norm": 4.525346271987768,
+      "learning_rate": 1.441688310008682e-07,
+      "loss": 0.5099,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9032387849527258,
+      "grad_norm": 3.625199458865597,
+      "learning_rate": 1.4299611803729547e-07,
+      "loss": 0.4766,
+      "step": 2245
+    },
+    {
+      "epoch": 0.903641118487226,
+      "grad_norm": 3.8753089109297787,
+      "learning_rate": 1.4182805373608055e-07,
+      "loss": 0.5326,
+      "step": 2246
+    },
+    {
+      "epoch": 0.904043452021726,
+      "grad_norm": 4.021228834417677,
+      "learning_rate": 1.4066464040096956e-07,
+      "loss": 0.5691,
+      "step": 2247
+    },
+    {
+      "epoch": 0.9044457855562261,
+      "grad_norm": 4.6752863808538105,
+      "learning_rate": 1.395058803265334e-07,
+      "loss": 0.7598,
+      "step": 2248
+    },
+    {
+      "epoch": 0.9048481190907263,
+      "grad_norm": 3.7179471187703514,
+      "learning_rate": 1.3835177579816783e-07,
+      "loss": 0.3955,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9052504526252263,
+      "grad_norm": 5.00701237752423,
+      "learning_rate": 1.3720232909208541e-07,
+      "loss": 0.6319,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9056527861597264,
+      "grad_norm": 3.7834417036951895,
+      "learning_rate": 1.3605754247531245e-07,
+      "loss": 0.5174,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9060551196942265,
+      "grad_norm": 3.680557056193272,
+      "learning_rate": 1.3491741820568366e-07,
+      "loss": 0.4133,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9064574532287266,
+      "grad_norm": 4.046295696144512,
+      "learning_rate": 1.337819585318395e-07,
+      "loss": 0.6443,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9068597867632268,
+      "grad_norm": 4.363328152667618,
+      "learning_rate": 1.3265116569322e-07,
+      "loss": 0.5437,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9072621202977268,
+      "grad_norm": 4.437499912355556,
+      "learning_rate": 1.3152504192006005e-07,
+      "loss": 0.5439,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9076644538322269,
+      "grad_norm": 3.904533797833089,
+      "learning_rate": 1.3040358943338742e-07,
+      "loss": 0.4058,
+      "step": 2256
+    },
+    {
+      "epoch": 0.908066787366727,
+      "grad_norm": 3.990363526582056,
+      "learning_rate": 1.2928681044501513e-07,
+      "loss": 0.4204,
+      "step": 2257
+    },
+    {
+      "epoch": 0.9084691209012271,
+      "grad_norm": 4.0760801231394606,
+      "learning_rate": 1.281747071575404e-07,
+      "loss": 0.6104,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9088714544357273,
+      "grad_norm": 4.256580027345346,
+      "learning_rate": 1.270672817643376e-07,
+      "loss": 0.6544,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9092737879702273,
+      "grad_norm": 4.71579332313011,
+      "learning_rate": 1.2596453644955597e-07,
+      "loss": 0.6273,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9096761215047274,
+      "grad_norm": 3.483958223270458,
+      "learning_rate": 1.2486647338811248e-07,
+      "loss": 0.4546,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9100784550392276,
+      "grad_norm": 3.9093619147070946,
+      "learning_rate": 1.2377309474569165e-07,
+      "loss": 0.4985,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9104807885737276,
+      "grad_norm": 4.057320495855987,
+      "learning_rate": 1.226844026787377e-07,
+      "loss": 0.5448,
+      "step": 2263
+    },
+    {
+      "epoch": 0.9108831221082277,
+      "grad_norm": 3.5360903799800503,
+      "learning_rate": 1.2160039933445222e-07,
+      "loss": 0.4164,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9112854556427278,
+      "grad_norm": 4.087420290794404,
+      "learning_rate": 1.2052108685078834e-07,
+      "loss": 0.594,
+      "step": 2265
+    },
+    {
+      "epoch": 0.9116877891772279,
+      "grad_norm": 3.9832710853164763,
+      "learning_rate": 1.1944646735644882e-07,
+      "loss": 0.452,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9120901227117281,
+      "grad_norm": 4.282566514995135,
+      "learning_rate": 1.1837654297087997e-07,
+      "loss": 0.5473,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9124924562462281,
+      "grad_norm": 3.9435273915235656,
+      "learning_rate": 1.1731131580426796e-07,
+      "loss": 0.6224,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9128947897807282,
+      "grad_norm": 4.3103292605724155,
+      "learning_rate": 1.1625078795753476e-07,
+      "loss": 0.5467,
+      "step": 2269
+    },
+    {
+      "epoch": 0.9132971233152283,
+      "grad_norm": 4.22573396631058,
+      "learning_rate": 1.1519496152233439e-07,
+      "loss": 0.5037,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9136994568497284,
+      "grad_norm": 3.7886525775152338,
+      "learning_rate": 1.1414383858104756e-07,
+      "loss": 0.4498,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9141017903842286,
+      "grad_norm": 3.2224197154073617,
+      "learning_rate": 1.1309742120677897e-07,
+      "loss": 0.4794,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9145041239187286,
+      "grad_norm": 4.705013349470272,
+      "learning_rate": 1.1205571146335303e-07,
+      "loss": 0.6041,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9149064574532287,
+      "grad_norm": 4.6403427413148535,
+      "learning_rate": 1.1101871140530851e-07,
+      "loss": 0.6308,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9153087909877289,
+      "grad_norm": 3.819381857552772,
+      "learning_rate": 1.0998642307789576e-07,
+      "loss": 0.5461,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9157111245222289,
+      "grad_norm": 3.997200472540085,
+      "learning_rate": 1.0895884851707255e-07,
+      "loss": 0.4926,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9161134580567291,
+      "grad_norm": 4.162425768576248,
+      "learning_rate": 1.079359897494997e-07,
+      "loss": 0.489,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9165157915912291,
+      "grad_norm": 3.8422971328140454,
+      "learning_rate": 1.0691784879253736e-07,
+      "loss": 0.5291,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9169181251257292,
+      "grad_norm": 4.112928250644688,
+      "learning_rate": 1.0590442765423953e-07,
+      "loss": 0.4926,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9173204586602294,
+      "grad_norm": 3.9378155981778664,
+      "learning_rate": 1.0489572833335349e-07,
+      "loss": 0.4722,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9177227921947294,
+      "grad_norm": 4.631927118939465,
+      "learning_rate": 1.038917528193123e-07,
+      "loss": 0.5274,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9181251257292296,
+      "grad_norm": 4.035377760065258,
+      "learning_rate": 1.0289250309223314e-07,
+      "loss": 0.5788,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9185274592637296,
+      "grad_norm": 3.868825647605808,
+      "learning_rate": 1.0189798112291177e-07,
+      "loss": 0.5046,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9189297927982297,
+      "grad_norm": 4.194215283322997,
+      "learning_rate": 1.0090818887282112e-07,
+      "loss": 0.6029,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9193321263327299,
+      "grad_norm": 4.215786203028234,
+      "learning_rate": 9.99231282941035e-08,
+      "loss": 0.4987,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9197344598672299,
+      "grad_norm": 4.911243421308879,
+      "learning_rate": 9.894280132957124e-08,
+      "loss": 0.6475,
+      "step": 2286
+    },
+    {
+      "epoch": 0.92013679340173,
+      "grad_norm": 4.551350447477857,
+      "learning_rate": 9.79672099126991e-08,
+      "loss": 0.6439,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9205391269362302,
+      "grad_norm": 4.0194710398308615,
+      "learning_rate": 9.69963559676232e-08,
+      "loss": 0.5376,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9209414604707302,
+      "grad_norm": 3.83021269408099,
+      "learning_rate": 9.60302414091352e-08,
+      "loss": 0.5093,
+      "step": 2289
+    },
+    {
+      "epoch": 0.9213437940052304,
+      "grad_norm": 4.1940419822947534,
+      "learning_rate": 9.506886814268007e-08,
+      "loss": 0.5702,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9217461275397304,
+      "grad_norm": 4.341103613511488,
+      "learning_rate": 9.41122380643511e-08,
+      "loss": 0.5271,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9221484610742305,
+      "grad_norm": 4.421018842084841,
+      "learning_rate": 9.316035306088684e-08,
+      "loss": 0.5923,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9225507946087307,
+      "grad_norm": 4.192150302264493,
+      "learning_rate": 9.221321500966746e-08,
+      "loss": 0.5475,
+      "step": 2293
+    },
+    {
+      "epoch": 0.9229531281432307,
+      "grad_norm": 4.56023548088152,
+      "learning_rate": 9.127082577871038e-08,
+      "loss": 0.6055,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9233554616777309,
+      "grad_norm": 3.727375694339219,
+      "learning_rate": 9.033318722666801e-08,
+      "loss": 0.3706,
+      "step": 2295
+    },
+    {
+      "epoch": 0.9237577952122309,
+      "grad_norm": 3.7651015398148897,
+      "learning_rate": 8.940030120282162e-08,
+      "loss": 0.5118,
+      "step": 2296
+    },
+    {
+      "epoch": 0.924160128746731,
+      "grad_norm": 4.305229814421829,
+      "learning_rate": 8.847216954708055e-08,
+      "loss": 0.5453,
+      "step": 2297
+    },
+    {
+      "epoch": 0.9245624622812312,
+      "grad_norm": 4.1747564223004066,
+      "learning_rate": 8.754879408997636e-08,
+      "loss": 0.5156,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9249647958157312,
+      "grad_norm": 3.649399704340268,
+      "learning_rate": 8.663017665266032e-08,
+      "loss": 0.3701,
+      "step": 2299
+    },
+    {
+      "epoch": 0.9253671293502314,
+      "grad_norm": 3.7596577501273734,
+      "learning_rate": 8.571631904690014e-08,
+      "loss": 0.4399,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9257694628847315,
+      "grad_norm": 4.230082435346815,
+      "learning_rate": 8.480722307507461e-08,
+      "loss": 0.4874,
+      "step": 2301
+    },
+    {
+      "epoch": 0.9261717964192315,
+      "grad_norm": 3.9612627542935197,
+      "learning_rate": 8.390289053017226e-08,
+      "loss": 0.5485,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9265741299537317,
+      "grad_norm": 3.5438909670467003,
+      "learning_rate": 8.300332319578664e-08,
+      "loss": 0.3951,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9269764634882317,
+      "grad_norm": 3.2833308636054217,
+      "learning_rate": 8.210852284611299e-08,
+      "loss": 0.407,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9273787970227318,
+      "grad_norm": 3.4444311785022865,
+      "learning_rate": 8.121849124594405e-08,
+      "loss": 0.3925,
+      "step": 2305
+    },
+    {
+      "epoch": 0.927781130557232,
+      "grad_norm": 4.548198093339736,
+      "learning_rate": 8.033323015066818e-08,
+      "loss": 0.719,
+      "step": 2306
+    },
+    {
+      "epoch": 0.928183464091732,
+      "grad_norm": 3.6123982321182284,
+      "learning_rate": 7.945274130626484e-08,
+      "loss": 0.3828,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9285857976262322,
+      "grad_norm": 3.177723392355348,
+      "learning_rate": 7.857702644930133e-08,
+      "loss": 0.353,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9289881311607322,
+      "grad_norm": 4.133100828896777,
+      "learning_rate": 7.770608730692852e-08,
+      "loss": 0.4977,
+      "step": 2309
+    },
+    {
+      "epoch": 0.9293904646952323,
+      "grad_norm": 3.562342202197354,
+      "learning_rate": 7.683992559687881e-08,
+      "loss": 0.3892,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9297927982297325,
+      "grad_norm": 3.898366854418511,
+      "learning_rate": 7.597854302746288e-08,
+      "loss": 0.4583,
+      "step": 2311
+    },
+    {
+      "epoch": 0.9301951317642325,
+      "grad_norm": 4.612414420760638,
+      "learning_rate": 7.512194129756428e-08,
+      "loss": 0.5581,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9305974652987327,
+      "grad_norm": 3.988029769560431,
+      "learning_rate": 7.42701220966388e-08,
+      "loss": 0.4744,
+      "step": 2313
+    },
+    {
+      "epoch": 0.9309997988332327,
+      "grad_norm": 4.481476524243926,
+      "learning_rate": 7.342308710470814e-08,
+      "loss": 0.59,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9314021323677328,
+      "grad_norm": 4.562506876878105,
+      "learning_rate": 7.258083799235982e-08,
+      "loss": 0.6742,
+      "step": 2315
+    },
+    {
+      "epoch": 0.931804465902233,
+      "grad_norm": 4.628598132831451,
+      "learning_rate": 7.174337642074147e-08,
+      "loss": 0.6008,
+      "step": 2316
+    },
+    {
+      "epoch": 0.932206799436733,
+      "grad_norm": 4.283379928862786,
+      "learning_rate": 7.091070404155854e-08,
+      "loss": 0.5053,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9326091329712332,
+      "grad_norm": 3.9776126177559155,
+      "learning_rate": 7.008282249707065e-08,
+      "loss": 0.4839,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9330114665057333,
+      "grad_norm": 4.069433607383545,
+      "learning_rate": 6.925973342008897e-08,
+      "loss": 0.453,
+      "step": 2319
+    },
+    {
+      "epoch": 0.9334138000402333,
+      "grad_norm": 4.037972743715134,
+      "learning_rate": 6.844143843397238e-08,
+      "loss": 0.5427,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9338161335747335,
+      "grad_norm": 3.547048708842676,
+      "learning_rate": 6.76279391526241e-08,
+      "loss": 0.4561,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9342184671092335,
+      "grad_norm": 3.7831575419487256,
+      "learning_rate": 6.681923718049016e-08,
+      "loss": 0.5184,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9346208006437337,
+      "grad_norm": 3.7960605831677428,
+      "learning_rate": 6.601533411255328e-08,
+      "loss": 0.428,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9350231341782338,
+      "grad_norm": 3.699019611927255,
+      "learning_rate": 6.52162315343327e-08,
+      "loss": 0.4711,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9354254677127338,
+      "grad_norm": 3.7499929742301465,
+      "learning_rate": 6.44219310218791e-08,
+      "loss": 0.494,
+      "step": 2325
+    },
+    {
+      "epoch": 0.935827801247234,
+      "grad_norm": 3.982088440613016,
+      "learning_rate": 6.363243414177239e-08,
+      "loss": 0.5013,
+      "step": 2326
+    },
+    {
+      "epoch": 0.936230134781734,
+      "grad_norm": 3.5405567172339034,
+      "learning_rate": 6.284774245111868e-08,
+      "loss": 0.4151,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9366324683162341,
+      "grad_norm": 3.9069783277866894,
+      "learning_rate": 6.206785749754613e-08,
+      "loss": 0.5257,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9370348018507343,
+      "grad_norm": 4.560908800971963,
+      "learning_rate": 6.12927808192032e-08,
+      "loss": 0.5827,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9374371353852343,
+      "grad_norm": 3.6067893534884967,
+      "learning_rate": 6.052251394475544e-08,
+      "loss": 0.4533,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9378394689197345,
+      "grad_norm": 4.228506402107548,
+      "learning_rate": 5.975705839338208e-08,
+      "loss": 0.5367,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9382418024542346,
+      "grad_norm": 4.377928096595712,
+      "learning_rate": 5.89964156747716e-08,
+      "loss": 0.487,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9386441359887346,
+      "grad_norm": 4.001917632135577,
+      "learning_rate": 5.8240587289122576e-08,
+      "loss": 0.5991,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9390464695232348,
+      "grad_norm": 3.89102054624204,
+      "learning_rate": 5.7489574727136734e-08,
+      "loss": 0.5289,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9394488030577348,
+      "grad_norm": 3.9529925607216314,
+      "learning_rate": 5.67433794700184e-08,
+      "loss": 0.5516,
+      "step": 2335
+    },
+    {
+      "epoch": 0.939851136592235,
+      "grad_norm": 3.9965015035833114,
+      "learning_rate": 5.600200298947034e-08,
+      "loss": 0.5758,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9402534701267351,
+      "grad_norm": 4.354462501990094,
+      "learning_rate": 5.52654467476918e-08,
+      "loss": 0.5793,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9406558036612351,
+      "grad_norm": 4.199766571714081,
+      "learning_rate": 5.453371219737491e-08,
+      "loss": 0.5688,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9410581371957353,
+      "grad_norm": 3.3551684650390454,
+      "learning_rate": 5.380680078170275e-08,
+      "loss": 0.3524,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9414604707302353,
+      "grad_norm": 4.408645627917106,
+      "learning_rate": 5.3084713934344613e-08,
+      "loss": 0.5442,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9418628042647355,
+      "grad_norm": 4.4356672674608895,
+      "learning_rate": 5.236745307945545e-08,
+      "loss": 0.5028,
+      "step": 2341
+    },
+    {
+      "epoch": 0.9422651377992356,
+      "grad_norm": 3.46236072019421,
+      "learning_rate": 5.165501963167174e-08,
+      "loss": 0.3921,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9426674713337356,
+      "grad_norm": 4.4269078668736075,
+      "learning_rate": 5.094741499610839e-08,
+      "loss": 0.5983,
+      "step": 2343
+    },
+    {
+      "epoch": 0.9430698048682358,
+      "grad_norm": 4.503787561410546,
+      "learning_rate": 5.02446405683582e-08,
+      "loss": 0.585,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9434721384027359,
+      "grad_norm": 3.658547311389223,
+      "learning_rate": 4.9546697734485224e-08,
+      "loss": 0.4762,
+      "step": 2345
+    },
+    {
+      "epoch": 0.943874471937236,
+      "grad_norm": 4.353601396348682,
+      "learning_rate": 4.885358787102584e-08,
+      "loss": 0.6102,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9442768054717361,
+      "grad_norm": 4.143732790296102,
+      "learning_rate": 4.816531234498406e-08,
+      "loss": 0.6124,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9446791390062361,
+      "grad_norm": 4.3975559021632575,
+      "learning_rate": 4.7481872513829575e-08,
+      "loss": 0.5633,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9450814725407363,
+      "grad_norm": 4.066972538920918,
+      "learning_rate": 4.680326972549387e-08,
+      "loss": 0.5118,
+      "step": 2349
+    },
+    {
+      "epoch": 0.9454838060752364,
+      "grad_norm": 4.03878765773838,
+      "learning_rate": 4.6129505318369127e-08,
+      "loss": 0.4778,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9458861396097364,
+      "grad_norm": 4.201321170733065,
+      "learning_rate": 4.546058062130487e-08,
+      "loss": 0.5128,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9462884731442366,
+      "grad_norm": 3.8094183727880395,
+      "learning_rate": 4.4796496953605493e-08,
+      "loss": 0.5441,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9466908066787366,
+      "grad_norm": 3.5299699606250834,
+      "learning_rate": 4.413725562502691e-08,
+      "loss": 0.4678,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9470931402132368,
+      "grad_norm": 3.9087226991955717,
+      "learning_rate": 4.348285793577517e-08,
+      "loss": 0.5329,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9474954737477369,
+      "grad_norm": 4.0121109992844755,
+      "learning_rate": 4.2833305176503136e-08,
+      "loss": 0.5837,
+      "step": 2355
+    },
+    {
+      "epoch": 0.9478978072822369,
+      "grad_norm": 4.544145628732853,
+      "learning_rate": 4.218859862830826e-08,
+      "loss": 0.3829,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9483001408167371,
+      "grad_norm": 3.89076415048523,
+      "learning_rate": 4.154873956272981e-08,
+      "loss": 0.4691,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9487024743512372,
+      "grad_norm": 4.1725175451999,
+      "learning_rate": 4.091372924174636e-08,
+      "loss": 0.4885,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9491048078857373,
+      "grad_norm": 4.3450167781713285,
+      "learning_rate": 4.028356891777385e-08,
+      "loss": 0.5124,
+      "step": 2359
+    },
+    {
+      "epoch": 0.9495071414202374,
+      "grad_norm": 4.4322372412295605,
+      "learning_rate": 3.9658259833662284e-08,
+      "loss": 0.5234,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9499094749547374,
+      "grad_norm": 3.707524532117256,
+      "learning_rate": 3.903780322269374e-08,
+      "loss": 0.4149,
+      "step": 2361
+    },
+    {
+      "epoch": 0.9503118084892376,
+      "grad_norm": 5.065679790929374,
+      "learning_rate": 3.842220030858018e-08,
+      "loss": 0.688,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9507141420237377,
+      "grad_norm": 3.786912616101777,
+      "learning_rate": 3.781145230546013e-08,
+      "loss": 0.43,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9511164755582378,
+      "grad_norm": 3.6227665097152224,
+      "learning_rate": 3.720556041789752e-08,
+      "loss": 0.4382,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9515188090927379,
+      "grad_norm": 3.999125768684484,
+      "learning_rate": 3.660452584087815e-08,
+      "loss": 0.5716,
+      "step": 2365
+    },
+    {
+      "epoch": 0.9519211426272379,
+      "grad_norm": 3.409457081881056,
+      "learning_rate": 3.600834975980877e-08,
+      "loss": 0.5407,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9523234761617381,
+      "grad_norm": 3.892658686650397,
+      "learning_rate": 3.5417033350512444e-08,
+      "loss": 0.4589,
+      "step": 2367
+    },
+    {
+      "epoch": 0.9527258096962382,
+      "grad_norm": 4.000437062384583,
+      "learning_rate": 3.483057777922905e-08,
+      "loss": 0.4875,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9531281432307382,
+      "grad_norm": 4.046666707315621,
+      "learning_rate": 3.42489842026103e-08,
+      "loss": 0.5162,
+      "step": 2369
+    },
+    {
+      "epoch": 0.9535304767652384,
+      "grad_norm": 4.290796216817974,
+      "learning_rate": 3.3672253767719485e-08,
+      "loss": 0.6233,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9539328102997385,
+      "grad_norm": 4.023014608954607,
+      "learning_rate": 3.310038761202839e-08,
+      "loss": 0.5062,
+      "step": 2371
+    },
+    {
+      "epoch": 0.9543351438342386,
+      "grad_norm": 4.43210170452369,
+      "learning_rate": 3.2533386863414784e-08,
+      "loss": 0.5594,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9547374773687387,
+      "grad_norm": 3.8979482975499624,
+      "learning_rate": 3.1971252640160545e-08,
+      "loss": 0.5128,
+      "step": 2373
+    },
+    {
+      "epoch": 0.9551398109032387,
+      "grad_norm": 4.162386221402383,
+      "learning_rate": 3.141398605094992e-08,
+      "loss": 0.5848,
+      "step": 2374
+    },
+    {
+      "epoch": 0.9555421444377389,
+      "grad_norm": 4.022603602017269,
+      "learning_rate": 3.086158819486651e-08,
+      "loss": 0.5771,
+      "step": 2375
+    },
+    {
+      "epoch": 0.955944477972239,
+      "grad_norm": 3.8920410241916863,
+      "learning_rate": 3.031406016139077e-08,
+      "loss": 0.5498,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9563468115067391,
+      "grad_norm": 4.084788242660989,
+      "learning_rate": 2.9771403030399713e-08,
+      "loss": 0.598,
+      "step": 2377
+    },
+    {
+      "epoch": 0.9567491450412392,
+      "grad_norm": 4.048190861072185,
+      "learning_rate": 2.923361787216278e-08,
+      "loss": 0.4405,
+      "step": 2378
+    },
+    {
+      "epoch": 0.9571514785757392,
+      "grad_norm": 4.502700834332798,
+      "learning_rate": 2.8700705747340963e-08,
+      "loss": 0.5169,
+      "step": 2379
+    },
+    {
+      "epoch": 0.9575538121102394,
+      "grad_norm": 4.054509620797073,
+      "learning_rate": 2.8172667706983792e-08,
+      "loss": 0.53,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9579561456447395,
+      "grad_norm": 3.939484097293469,
+      "learning_rate": 2.7649504792528204e-08,
+      "loss": 0.5566,
+      "step": 2381
+    },
+    {
+      "epoch": 0.9583584791792396,
+      "grad_norm": 4.299563624727568,
+      "learning_rate": 2.7131218035796324e-08,
+      "loss": 0.5603,
+      "step": 2382
+    },
+    {
+      "epoch": 0.9587608127137397,
+      "grad_norm": 3.723520114625019,
+      "learning_rate": 2.6617808458991867e-08,
+      "loss": 0.4346,
+      "step": 2383
+    },
+    {
+      "epoch": 0.9591631462482398,
+      "grad_norm": 3.667012045051573,
+      "learning_rate": 2.6109277074700955e-08,
+      "loss": 0.3638,
+      "step": 2384
+    },
+    {
+      "epoch": 0.9595654797827399,
+      "grad_norm": 3.725825844080702,
+      "learning_rate": 2.5605624885887414e-08,
+      "loss": 0.4541,
+      "step": 2385
+    },
+    {
+      "epoch": 0.95996781331724,
+      "grad_norm": 3.7094680890029776,
+      "learning_rate": 2.5106852885892764e-08,
+      "loss": 0.4693,
+      "step": 2386
+    },
+    {
+      "epoch": 0.96037014685174,
+      "grad_norm": 3.7229547246513937,
+      "learning_rate": 2.461296205843261e-08,
+      "loss": 0.5144,
+      "step": 2387
+    },
+    {
+      "epoch": 0.9607724803862402,
+      "grad_norm": 4.159612295768075,
+      "learning_rate": 2.412395337759693e-08,
+      "loss": 0.522,
+      "step": 2388
+    },
+    {
+      "epoch": 0.9611748139207403,
+      "grad_norm": 3.8986140359372286,
+      "learning_rate": 2.3639827807844796e-08,
+      "loss": 0.5237,
+      "step": 2389
+    },
+    {
+      "epoch": 0.9615771474552404,
+      "grad_norm": 4.121244015419092,
+      "learning_rate": 2.3160586304006026e-08,
+      "loss": 0.538,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9619794809897405,
+      "grad_norm": 4.166559231330322,
+      "learning_rate": 2.268622981127705e-08,
+      "loss": 0.5822,
+      "step": 2391
+    },
+    {
+      "epoch": 0.9623818145242405,
+      "grad_norm": 3.9211071919172813,
+      "learning_rate": 2.2216759265220044e-08,
+      "loss": 0.4897,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9627841480587407,
+      "grad_norm": 4.349097825226008,
+      "learning_rate": 2.1752175591760184e-08,
+      "loss": 0.5211,
+      "step": 2393
+    },
+    {
+      "epoch": 0.9631864815932408,
+      "grad_norm": 3.9282047251137677,
+      "learning_rate": 2.1292479707184787e-08,
+      "loss": 0.5741,
+      "step": 2394
+    },
+    {
+      "epoch": 0.9635888151277409,
+      "grad_norm": 4.159989263744697,
+      "learning_rate": 2.0837672518141395e-08,
+      "loss": 0.5238,
+      "step": 2395
+    },
+    {
+      "epoch": 0.963991148662241,
+      "grad_norm": 3.8198728228381293,
+      "learning_rate": 2.0387754921634972e-08,
+      "loss": 0.5144,
+      "step": 2396
+    },
+    {
+      "epoch": 0.9643934821967411,
+      "grad_norm": 3.346255200741844,
+      "learning_rate": 1.9942727805027373e-08,
+      "loss": 0.4384,
+      "step": 2397
+    },
+    {
+      "epoch": 0.9647958157312412,
+      "grad_norm": 3.8469019196516845,
+      "learning_rate": 1.9502592046035107e-08,
+      "loss": 0.486,
+      "step": 2398
+    },
+    {
+      "epoch": 0.9651981492657413,
+      "grad_norm": 3.5776773638982893,
+      "learning_rate": 1.9067348512726847e-08,
+      "loss": 0.4414,
+      "step": 2399
+    },
+    {
+      "epoch": 0.9656004828002414,
+      "grad_norm": 4.031486537261451,
+      "learning_rate": 1.8636998063523705e-08,
+      "loss": 0.5328,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9660028163347415,
+      "grad_norm": 4.260664576753229,
+      "learning_rate": 1.8211541547195067e-08,
+      "loss": 0.5136,
+      "step": 2401
+    },
+    {
+      "epoch": 0.9664051498692416,
+      "grad_norm": 4.152186034058617,
+      "learning_rate": 1.7790979802858876e-08,
+      "loss": 0.6316,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9668074834037417,
+      "grad_norm": 3.995357070807909,
+      "learning_rate": 1.7375313659978855e-08,
+      "loss": 0.5377,
+      "step": 2403
+    },
+    {
+      "epoch": 0.9672098169382418,
+      "grad_norm": 3.4651763373212474,
+      "learning_rate": 1.6964543938363388e-08,
+      "loss": 0.4157,
+      "step": 2404
+    },
+    {
+      "epoch": 0.9676121504727418,
+      "grad_norm": 4.360198459376172,
+      "learning_rate": 1.6558671448163866e-08,
+      "loss": 0.6273,
+      "step": 2405
+    },
+    {
+      "epoch": 0.968014484007242,
+      "grad_norm": 4.006498747343065,
+      "learning_rate": 1.6157696989873017e-08,
+      "loss": 0.4914,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9684168175417421,
+      "grad_norm": 3.7398733702546822,
+      "learning_rate": 1.5761621354322965e-08,
+      "loss": 0.4022,
+      "step": 2407
+    },
+    {
+      "epoch": 0.9688191510762422,
+      "grad_norm": 4.1110217996882294,
+      "learning_rate": 1.5370445322684392e-08,
+      "loss": 0.4698,
+      "step": 2408
+    },
+    {
+      "epoch": 0.9692214846107423,
+      "grad_norm": 4.362254470529691,
+      "learning_rate": 1.498416966646432e-08,
+      "loss": 0.5598,
+      "step": 2409
+    },
+    {
+      "epoch": 0.9696238181452425,
+      "grad_norm": 3.8880330136972328,
+      "learning_rate": 1.4602795147504734e-08,
+      "loss": 0.473,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9700261516797425,
+      "grad_norm": 4.110874787245324,
+      "learning_rate": 1.4226322517981728e-08,
+      "loss": 0.4901,
+      "step": 2411
+    },
+    {
+      "epoch": 0.9704284852142426,
+      "grad_norm": 4.138814644855194,
+      "learning_rate": 1.3854752520403303e-08,
+      "loss": 0.5618,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9708308187487427,
+      "grad_norm": 4.481371755440245,
+      "learning_rate": 1.3488085887607972e-08,
+      "loss": 0.6152,
+      "step": 2413
+    },
+    {
+      "epoch": 0.9712331522832428,
+      "grad_norm": 3.6050373310270176,
+      "learning_rate": 1.3126323342763648e-08,
+      "loss": 0.5525,
+      "step": 2414
+    },
+    {
+      "epoch": 0.971635485817743,
+      "grad_norm": 4.061084112826385,
+      "learning_rate": 1.2769465599365982e-08,
+      "loss": 0.4619,
+      "step": 2415
+    },
+    {
+      "epoch": 0.972037819352243,
+      "grad_norm": 3.8920028399978097,
+      "learning_rate": 1.24175133612367e-08,
+      "loss": 0.4833,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9724401528867431,
+      "grad_norm": 3.909035344776143,
+      "learning_rate": 1.2070467322522762e-08,
+      "loss": 0.5188,
+      "step": 2417
+    },
+    {
+      "epoch": 0.9728424864212432,
+      "grad_norm": 4.225663392112197,
+      "learning_rate": 1.1728328167695258e-08,
+      "loss": 0.6696,
+      "step": 2418
+    },
+    {
+      "epoch": 0.9732448199557433,
+      "grad_norm": 4.6741890030042725,
+      "learning_rate": 1.1391096571546633e-08,
+      "loss": 0.5815,
+      "step": 2419
+    },
+    {
+      "epoch": 0.9736471534902434,
+      "grad_norm": 4.51608876386972,
+      "learning_rate": 1.1058773199190965e-08,
+      "loss": 0.5173,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9740494870247435,
+      "grad_norm": 4.0654982873308665,
+      "learning_rate": 1.0731358706061456e-08,
+      "loss": 0.5749,
+      "step": 2421
+    },
+    {
+      "epoch": 0.9744518205592436,
+      "grad_norm": 4.2304161552160835,
+      "learning_rate": 1.0408853737909896e-08,
+      "loss": 0.5813,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9748541540937438,
+      "grad_norm": 3.4742799346124493,
+      "learning_rate": 1.0091258930805259e-08,
+      "loss": 0.4705,
+      "step": 2423
+    },
+    {
+      "epoch": 0.9752564876282438,
+      "grad_norm": 4.08672880848899,
+      "learning_rate": 9.778574911132044e-09,
+      "loss": 0.516,
+      "step": 2424
+    },
+    {
+      "epoch": 0.9756588211627439,
+      "grad_norm": 3.669930620914309,
+      "learning_rate": 9.470802295589442e-09,
+      "loss": 0.4138,
+      "step": 2425
+    },
+    {
+      "epoch": 0.976061154697244,
+      "grad_norm": 3.5860479619434673,
+      "learning_rate": 9.16794169119023e-09,
+      "loss": 0.4224,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9764634882317441,
+      "grad_norm": 4.642054584241187,
+      "learning_rate": 8.869993695258816e-09,
+      "loss": 0.611,
+      "step": 2427
+    },
+    {
+      "epoch": 0.9768658217662443,
+      "grad_norm": 3.977155407637389,
+      "learning_rate": 8.576958895431531e-09,
+      "loss": 0.5681,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9772681553007443,
+      "grad_norm": 3.6884328747774857,
+      "learning_rate": 8.28883786965301e-09,
+      "loss": 0.4945,
+      "step": 2429
+    },
+    {
+      "epoch": 0.9776704888352444,
+      "grad_norm": 4.023502331465119,
+      "learning_rate": 8.005631186177864e-09,
+      "loss": 0.4201,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9780728223697445,
+      "grad_norm": 3.901800973146699,
+      "learning_rate": 7.727339403568179e-09,
+      "loss": 0.4977,
+      "step": 2431
+    },
+    {
+      "epoch": 0.9784751559042446,
+      "grad_norm": 3.9967038572140856,
+      "learning_rate": 7.45396307069185e-09,
+      "loss": 0.6418,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9788774894387448,
+      "grad_norm": 4.291073567899429,
+      "learning_rate": 7.185502726722305e-09,
+      "loss": 0.6601,
+      "step": 2433
+    },
+    {
+      "epoch": 0.9792798229732448,
+      "grad_norm": 3.974019553234897,
+      "learning_rate": 6.921958901137948e-09,
+      "loss": 0.4781,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9796821565077449,
+      "grad_norm": 3.833029352112665,
+      "learning_rate": 6.66333211371939e-09,
+      "loss": 0.4373,
+      "step": 2435
+    },
+    {
+      "epoch": 0.9800844900422451,
+      "grad_norm": 4.398810707509668,
+      "learning_rate": 6.409622874550547e-09,
+      "loss": 0.5795,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9804868235767451,
+      "grad_norm": 3.5697882170733557,
+      "learning_rate": 6.160831684015878e-09,
+      "loss": 0.398,
+      "step": 2437
+    },
+    {
+      "epoch": 0.9808891571112452,
+      "grad_norm": 3.904433602202901,
+      "learning_rate": 5.9169590328003735e-09,
+      "loss": 0.5029,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9812914906457453,
+      "grad_norm": 3.7353800882065715,
+      "learning_rate": 5.678005401888176e-09,
+      "loss": 0.4911,
+      "step": 2439
+    },
+    {
+      "epoch": 0.9816938241802454,
+      "grad_norm": 3.795949925181833,
+      "learning_rate": 5.443971262561465e-09,
+      "loss": 0.4405,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9820961577147456,
+      "grad_norm": 3.617609885039579,
+      "learning_rate": 5.214857076400736e-09,
+      "loss": 0.4329,
+      "step": 2441
+    },
+    {
+      "epoch": 0.9824984912492456,
+      "grad_norm": 3.5249358490195637,
+      "learning_rate": 4.990663295282305e-09,
+      "loss": 0.4285,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9829008247837457,
+      "grad_norm": 4.142775577116706,
+      "learning_rate": 4.771390361377748e-09,
+      "loss": 0.4777,
+      "step": 2443
+    },
+    {
+      "epoch": 0.9833031583182458,
+      "grad_norm": 4.312344250176257,
+      "learning_rate": 4.5570387071544595e-09,
+      "loss": 0.6627,
+      "step": 2444
+    },
+    {
+      "epoch": 0.9837054918527459,
+      "grad_norm": 3.67386973137855,
+      "learning_rate": 4.347608755372601e-09,
+      "loss": 0.4431,
+      "step": 2445
+    },
+    {
+      "epoch": 0.9841078253872461,
+      "grad_norm": 4.191733748165821,
+      "learning_rate": 4.143100919086207e-09,
+      "loss": 0.4884,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9845101589217461,
+      "grad_norm": 4.38016174030762,
+      "learning_rate": 3.943515601640968e-09,
+      "loss": 0.5822,
+      "step": 2447
+    },
+    {
+      "epoch": 0.9849124924562462,
+      "grad_norm": 4.227081529941771,
+      "learning_rate": 3.748853196673952e-09,
+      "loss": 0.49,
+      "step": 2448
+    },
+    {
+      "epoch": 0.9853148259907464,
+      "grad_norm": 3.6169704926991124,
+      "learning_rate": 3.5591140881138797e-09,
+      "loss": 0.5642,
+      "step": 2449
+    },
+    {
+      "epoch": 0.9857171595252464,
+      "grad_norm": 3.228283450712886,
+      "learning_rate": 3.374298650178076e-09,
+      "loss": 0.3926,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9861194930597466,
+      "grad_norm": 3.996805720599477,
+      "learning_rate": 3.194407247373854e-09,
+      "loss": 0.5175,
+      "step": 2451
+    },
+    {
+      "epoch": 0.9865218265942466,
+      "grad_norm": 3.786750548322366,
+      "learning_rate": 3.0194402344965733e-09,
+      "loss": 0.4496,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9869241601287467,
+      "grad_norm": 3.730366038034417,
+      "learning_rate": 2.8493979566296403e-09,
+      "loss": 0.4701,
+      "step": 2453
+    },
+    {
+      "epoch": 0.9873264936632469,
+      "grad_norm": 4.616677722561612,
+      "learning_rate": 2.6842807491433974e-09,
+      "loss": 0.6207,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9877288271977469,
+      "grad_norm": 3.9983804243694645,
+      "learning_rate": 2.524088937694291e-09,
+      "loss": 0.4734,
+      "step": 2455
+    },
+    {
+      "epoch": 0.988131160732247,
+      "grad_norm": 4.364062652259154,
+      "learning_rate": 2.3688228382251487e-09,
+      "loss": 0.6254,
+      "step": 2456
+    },
+    {
+      "epoch": 0.9885334942667471,
+      "grad_norm": 3.97492888774164,
+      "learning_rate": 2.218482756963236e-09,
+      "loss": 0.5645,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9889358278012472,
+      "grad_norm": 3.9671569540217493,
+      "learning_rate": 2.07306899042109e-09,
+      "loss": 0.4821,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9893381613357474,
+      "grad_norm": 3.808293451273474,
+      "learning_rate": 1.9325818253945748e-09,
+      "loss": 0.4502,
+      "step": 2459
+    },
+    {
+      "epoch": 0.9897404948702474,
+      "grad_norm": 4.467378355634954,
+      "learning_rate": 1.7970215389628842e-09,
+      "loss": 0.5963,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9901428284047475,
+      "grad_norm": 4.625199727103312,
+      "learning_rate": 1.6663883984888163e-09,
+      "loss": 0.6419,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9905451619392477,
+      "grad_norm": 3.9326948253065184,
+      "learning_rate": 1.5406826616168324e-09,
+      "loss": 0.5575,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9909474954737477,
+      "grad_norm": 4.4568921518738485,
+      "learning_rate": 1.4199045762730569e-09,
+      "loss": 0.6483,
+      "step": 2463
+    },
+    {
+      "epoch": 0.9913498290082479,
+      "grad_norm": 4.351427492104435,
+      "learning_rate": 1.3040543806658313e-09,
+      "loss": 0.5948,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9917521625427479,
+      "grad_norm": 3.4818450271582857,
+      "learning_rate": 1.1931323032832177e-09,
+      "loss": 0.3535,
+      "step": 2465
+    },
+    {
+      "epoch": 0.992154496077248,
+      "grad_norm": 4.11227816177031,
+      "learning_rate": 1.0871385628941077e-09,
+      "loss": 0.5317,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9925568296117482,
+      "grad_norm": 4.418916055074932,
+      "learning_rate": 9.860733685479463e-10,
+      "loss": 0.6433,
+      "step": 2467
+    },
+    {
+      "epoch": 0.9929591631462482,
+      "grad_norm": 3.6128250845259724,
+      "learning_rate": 8.899369195727869e-10,
+      "loss": 0.5447,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9933614966807484,
+      "grad_norm": 4.364090118207612,
+      "learning_rate": 7.987294055761264e-10,
+      "loss": 0.6099,
+      "step": 2469
+    },
+    {
+      "epoch": 0.9937638302152484,
+      "grad_norm": 3.739315819157321,
+      "learning_rate": 7.124510064446255e-10,
+      "loss": 0.4998,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9941661637497485,
+      "grad_norm": 4.035011038094296,
+      "learning_rate": 6.311018923432776e-10,
+      "loss": 0.4558,
+      "step": 2471
+    },
+    {
+      "epoch": 0.9945684972842487,
+      "grad_norm": 3.577550936704116,
+      "learning_rate": 5.546822237145755e-10,
+      "loss": 0.4408,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9949708308187487,
+      "grad_norm": 3.7748701163123584,
+      "learning_rate": 4.831921512796211e-10,
+      "loss": 0.4403,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9953731643532489,
+      "grad_norm": 3.7403701320267686,
+      "learning_rate": 4.166318160361837e-10,
+      "loss": 0.4989,
+      "step": 2474
+    },
+    {
+      "epoch": 0.995775497887749,
+      "grad_norm": 3.8787870872673857,
+      "learning_rate": 3.5500134925980925e-10,
+      "loss": 0.5066,
+      "step": 2475
+    },
+    {
+      "epoch": 0.996177831422249,
+      "grad_norm": 4.321954937291901,
+      "learning_rate": 2.983008725029879e-10,
+      "loss": 0.498,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9965801649567492,
+      "grad_norm": 3.574339539948393,
+      "learning_rate": 2.4653049759432167e-10,
+      "loss": 0.4164,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9969824984912492,
+      "grad_norm": 3.7965194640316478,
+      "learning_rate": 1.9969032663935683e-10,
+      "loss": 0.4769,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9973848320257493,
+      "grad_norm": 4.23022008583803,
+      "learning_rate": 1.5778045202002878e-10,
+      "loss": 0.4481,
+      "step": 2479
+    },
+    {
+      "epoch": 0.9977871655602495,
+      "grad_norm": 4.071535851310751,
+      "learning_rate": 1.2080095639410704e-10,
+      "loss": 0.535,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9981894990947495,
+      "grad_norm": 3.9007108087469646,
+      "learning_rate": 8.875191269547279e-11,
+      "loss": 0.542,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9985918326292497,
+      "grad_norm": 4.05416067842318,
+      "learning_rate": 6.163338413328612e-11,
+      "loss": 0.5738,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9989941661637497,
+      "grad_norm": 3.9020981788420386,
+      "learning_rate": 3.944542419337394e-11,
+      "loss": 0.4897,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9993964996982498,
+      "grad_norm": 3.6120331011996085,
+      "learning_rate": 2.2188076636009415e-11,
+      "loss": 0.5195,
+      "step": 2484
+    },
+    {
+      "epoch": 0.99979883323275,
+      "grad_norm": 3.4028120769775385,
+      "learning_rate": 9.861375497577375e-12,
+      "loss": 0.3763,
+      "step": 2485
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4028120769775385,
+      "learning_rate": 2.4653450900191666e-12,
+      "loss": 0.4064,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0,
+      "step": 2486,
+      "total_flos": 882515131957248.0,
+      "train_loss": 0.5852410430066162,
+      "train_runtime": 194010.5784,
+      "train_samples_per_second": 1.64,
+      "train_steps_per_second": 0.013
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2486,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 882515131957248.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}