diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -10,7002 +10,7002 @@
   "log_history": [
     {
       "epoch": 0.00044483985765124553,
-      "grad_norm": 0.5906195640563965,
+      "grad_norm": 0.5906099081039429,
       "learning_rate": 4e-05,
       "loss": 1.504,
       "step": 1
     },
     {
       "epoch": 0.0008896797153024911,
-      "grad_norm": 0.8538110852241516,
+      "grad_norm": 0.8538464903831482,
       "learning_rate": 8e-05,
       "loss": 1.4735,
       "step": 2
     },
     {
       "epoch": 0.0013345195729537367,
-      "grad_norm": 0.4623548686504364,
+      "grad_norm": 0.46255260705947876,
       "learning_rate": 0.00012,
       "loss": 1.3887,
       "step": 3
     },
     {
       "epoch": 0.0017793594306049821,
-      "grad_norm": 0.40841907262802124,
+      "grad_norm": 0.40842387080192566,
       "learning_rate": 0.00016,
       "loss": 1.2993,
       "step": 4
     },
     {
       "epoch": 0.002224199288256228,
-      "grad_norm": 0.4153430759906769,
+      "grad_norm": 0.41691651940345764,
       "learning_rate": 0.0002,
       "loss": 1.3105,
       "step": 5
     },
     {
       "epoch": 0.0026690391459074734,
-      "grad_norm": 0.4014444947242737,
+      "grad_norm": 0.40152719616889954,
       "learning_rate": 0.00019979899497487438,
       "loss": 1.1942,
       "step": 6
     },
     {
       "epoch": 0.003113879003558719,
-      "grad_norm": 0.39270123839378357,
+      "grad_norm": 0.39211127161979675,
       "learning_rate": 0.00019959798994974876,
       "loss": 1.1688,
       "step": 7
     },
     {
       "epoch": 0.0035587188612099642,
-      "grad_norm": 0.4414786696434021,
+      "grad_norm": 0.43954452872276306,
       "learning_rate": 0.00019939698492462313,
-      "loss": 1.121,
+      "loss": 1.1211,
       "step": 8
     },
     {
       "epoch": 0.00400355871886121,
-      "grad_norm": 0.3911682069301605,
+      "grad_norm": 0.38755497336387634,
       "learning_rate": 0.0001991959798994975,
       "loss": 1.1039,
       "step": 9
     },
     {
       "epoch": 0.004448398576512456,
-      "grad_norm": 0.3546655476093292,
+      "grad_norm": 0.3547092378139496,
       "learning_rate": 0.00019899497487437187,
-      "loss": 1.1259,
+      "loss": 1.126,
       "step": 10
     },
     {
       "epoch": 0.004893238434163701,
-      "grad_norm": 0.37283238768577576,
+      "grad_norm": 0.37269020080566406,
       "learning_rate": 0.00019879396984924622,
       "loss": 1.1599,
       "step": 11
     },
     {
       "epoch": 0.005338078291814947,
-      "grad_norm": 0.35209548473358154,
+      "grad_norm": 0.3519953191280365,
       "learning_rate": 0.00019859296482412062,
       "loss": 1.2022,
       "step": 12
     },
     {
       "epoch": 0.005782918149466192,
-      "grad_norm": 0.3735697865486145,
+      "grad_norm": 0.3725995421409607,
       "learning_rate": 0.000198391959798995,
-      "loss": 1.2544,
+      "loss": 1.2543,
       "step": 13
     },
     {
       "epoch": 0.006227758007117438,
-      "grad_norm": 0.324276328086853,
+      "grad_norm": 0.32365697622299194,
       "learning_rate": 0.00019819095477386937,
-      "loss": 1.1451,
+      "loss": 1.145,
       "step": 14
     },
     {
       "epoch": 0.0066725978647686835,
-      "grad_norm": 0.44405269622802734,
+      "grad_norm": 0.44411811232566833,
       "learning_rate": 0.0001979899497487437,
-      "loss": 1.2581,
+      "loss": 1.2582,
       "step": 15
     },
     {
       "epoch": 0.0071174377224199285,
-      "grad_norm": 0.40987834334373474,
+      "grad_norm": 0.4044501781463623,
       "learning_rate": 0.0001977889447236181,
-      "loss": 1.1422,
+      "loss": 1.1424,
       "step": 16
     },
     {
       "epoch": 0.007562277580071174,
-      "grad_norm": 0.31967031955718994,
+      "grad_norm": 0.3195618689060211,
       "learning_rate": 0.00019758793969849249,
-      "loss": 1.1521,
+      "loss": 1.1522,
       "step": 17
     },
     {
       "epoch": 0.00800711743772242,
-      "grad_norm": 0.3634157180786133,
+      "grad_norm": 0.35177746415138245,
       "learning_rate": 0.00019738693467336683,
       "loss": 1.0172,
       "step": 18
     },
     {
       "epoch": 0.008451957295373666,
-      "grad_norm": 0.37751761078834534,
+      "grad_norm": 0.37688982486724854,
       "learning_rate": 0.0001971859296482412,
       "loss": 1.1343,
       "step": 19
     },
     {
       "epoch": 0.008896797153024912,
-      "grad_norm": 0.3501513600349426,
+      "grad_norm": 0.3498818278312683,
       "learning_rate": 0.0001969849246231156,
-      "loss": 1.1137,
+      "loss": 1.1138,
       "step": 20
     },
     {
       "epoch": 0.009341637010676156,
-      "grad_norm": 0.41018760204315186,
+      "grad_norm": 0.4116188585758209,
       "learning_rate": 0.00019678391959798995,
-      "loss": 1.0601,
+      "loss": 1.0602,
       "step": 21
     },
     {
       "epoch": 0.009786476868327402,
-      "grad_norm": 0.5334712266921997,
+      "grad_norm": 0.5222630500793457,
       "learning_rate": 0.00019658291457286432,
-      "loss": 1.2332,
+      "loss": 1.2328,
       "step": 22
     },
     {
       "epoch": 0.010231316725978648,
-      "grad_norm": 0.5504945516586304,
+      "grad_norm": 0.5468437075614929,
       "learning_rate": 0.0001963819095477387,
-      "loss": 1.1928,
+      "loss": 1.1929,
       "step": 23
     },
     {
       "epoch": 0.010676156583629894,
-      "grad_norm": 0.39751070737838745,
+      "grad_norm": 0.39658740162849426,
       "learning_rate": 0.0001961809045226131,
-      "loss": 0.9774,
+      "loss": 0.9775,
       "step": 24
     },
     {
       "epoch": 0.01112099644128114,
-      "grad_norm": 0.31996771693229675,
+      "grad_norm": 0.3198809027671814,
       "learning_rate": 0.00019597989949748744,
       "loss": 1.042,
       "step": 25
     },
     {
       "epoch": 0.011565836298932384,
-      "grad_norm": 0.49899256229400635,
+      "grad_norm": 0.49676308035850525,
       "learning_rate": 0.00019577889447236181,
-      "loss": 1.0384,
+      "loss": 1.0378,
       "step": 26
     },
     {
       "epoch": 0.01201067615658363,
-      "grad_norm": 0.37340229749679565,
+      "grad_norm": 0.3704735338687897,
       "learning_rate": 0.0001955778894472362,
-      "loss": 1.0307,
+      "loss": 1.0303,
       "step": 27
     },
     {
       "epoch": 0.012455516014234875,
-      "grad_norm": 0.33216455578804016,
+      "grad_norm": 0.33084699511528015,
       "learning_rate": 0.00019537688442211056,
-      "loss": 1.0253,
+      "loss": 1.0252,
       "step": 28
     },
     {
       "epoch": 0.012900355871886121,
-      "grad_norm": 0.3582470715045929,
+      "grad_norm": 0.35819771885871887,
       "learning_rate": 0.00019517587939698493,
-      "loss": 1.0751,
+      "loss": 1.0752,
       "step": 29
     },
     {
       "epoch": 0.013345195729537367,
-      "grad_norm": 0.3682143986225128,
+      "grad_norm": 0.36553651094436646,
       "learning_rate": 0.0001949748743718593,
-      "loss": 1.0633,
+      "loss": 1.0637,
       "step": 30
     },
     {
       "epoch": 0.013790035587188613,
-      "grad_norm": 0.4105437695980072,
+      "grad_norm": 0.41030144691467285,
       "learning_rate": 0.00019477386934673368,
-      "loss": 1.1865,
+      "loss": 1.1858,
       "step": 31
     },
     {
       "epoch": 0.014234875444839857,
-      "grad_norm": 0.3709512948989868,
+      "grad_norm": 0.3703097105026245,
       "learning_rate": 0.00019457286432160805,
-      "loss": 1.0198,
+      "loss": 1.0193,
       "step": 32
     },
     {
       "epoch": 0.014679715302491103,
-      "grad_norm": 0.3693946599960327,
+      "grad_norm": 0.37004703283309937,
       "learning_rate": 0.00019437185929648243,
       "loss": 1.0405,
       "step": 33
     },
     {
       "epoch": 0.015124555160142349,
-      "grad_norm": 0.335426390171051,
+      "grad_norm": 0.33804813027381897,
       "learning_rate": 0.0001941708542713568,
-      "loss": 0.9301,
+      "loss": 0.9303,
       "step": 34
     },
     {
       "epoch": 0.015569395017793594,
-      "grad_norm": 0.5277281403541565,
+      "grad_norm": 0.5308049917221069,
       "learning_rate": 0.00019396984924623117,
-      "loss": 0.9615,
+      "loss": 0.961,
       "step": 35
     },
     {
       "epoch": 0.01601423487544484,
-      "grad_norm": 0.40925148129463196,
+      "grad_norm": 0.4099597632884979,
       "learning_rate": 0.00019376884422110552,
-      "loss": 1.1093,
+      "loss": 1.1084,
       "step": 36
     },
     {
       "epoch": 0.016459074733096084,
-      "grad_norm": 0.40884852409362793,
+      "grad_norm": 0.4100741446018219,
       "learning_rate": 0.00019356783919597992,
       "loss": 1.063,
       "step": 37
     },
     {
       "epoch": 0.016903914590747332,
-      "grad_norm": 0.3492753207683563,
+      "grad_norm": 0.3497965335845947,
       "learning_rate": 0.0001933668341708543,
-      "loss": 1.0865,
+      "loss": 1.0864,
       "step": 38
     },
     {
       "epoch": 0.017348754448398576,
-      "grad_norm": 0.3660659193992615,
+      "grad_norm": 0.3670850694179535,
       "learning_rate": 0.00019316582914572864,
       "loss": 1.071,
       "step": 39
     },
     {
       "epoch": 0.017793594306049824,
-      "grad_norm": 0.3863303065299988,
+      "grad_norm": 0.38884881138801575,
       "learning_rate": 0.000192964824120603,
-      "loss": 1.0096,
+      "loss": 1.0091,
       "step": 40
     },
     {
       "epoch": 0.018238434163701068,
-      "grad_norm": 0.34570327401161194,
+      "grad_norm": 0.3449464738368988,
       "learning_rate": 0.0001927638190954774,
       "loss": 1.0753,
       "step": 41
     },
     {
       "epoch": 0.018683274021352312,
-      "grad_norm": 0.35912126302719116,
+      "grad_norm": 0.35679587721824646,
       "learning_rate": 0.00019256281407035178,
-      "loss": 1.0997,
+      "loss": 1.0992,
       "step": 42
     },
     {
       "epoch": 0.01912811387900356,
-      "grad_norm": 0.36907467246055603,
+      "grad_norm": 0.3696941137313843,
       "learning_rate": 0.00019236180904522613,
-      "loss": 1.0419,
+      "loss": 1.0413,
       "step": 43
     },
     {
       "epoch": 0.019572953736654804,
-      "grad_norm": 0.3746166229248047,
+      "grad_norm": 0.37430745363235474,
       "learning_rate": 0.0001921608040201005,
       "loss": 1.0424,
       "step": 44
     },
     {
       "epoch": 0.02001779359430605,
-      "grad_norm": 0.33465832471847534,
+      "grad_norm": 0.33420661091804504,
       "learning_rate": 0.0001919597989949749,
-      "loss": 1.0335,
+      "loss": 1.0339,
       "step": 45
     },
     {
       "epoch": 0.020462633451957295,
-      "grad_norm": 0.3497348427772522,
+      "grad_norm": 0.35058847069740295,
       "learning_rate": 0.00019175879396984925,
-      "loss": 1.0282,
+      "loss": 1.0291,
       "step": 46
     },
     {
       "epoch": 0.02090747330960854,
-      "grad_norm": 0.34102576971054077,
+      "grad_norm": 0.3396761119365692,
       "learning_rate": 0.00019155778894472362,
-      "loss": 1.1197,
+      "loss": 1.1198,
       "step": 47
     },
     {
       "epoch": 0.021352313167259787,
-      "grad_norm": 0.31394055485725403,
+      "grad_norm": 0.3136732280254364,
       "learning_rate": 0.000191356783919598,
-      "loss": 1.0298,
+      "loss": 1.0288,
       "step": 48
     },
     {
       "epoch": 0.02179715302491103,
-      "grad_norm": 0.505631148815155,
+      "grad_norm": 0.5155500173568726,
       "learning_rate": 0.0001911557788944724,
-      "loss": 1.0349,
+      "loss": 1.0353,
       "step": 49
     },
     {
       "epoch": 0.02224199288256228,
-      "grad_norm": 0.3988608121871948,
+      "grad_norm": 0.39861586689949036,
       "learning_rate": 0.00019095477386934674,
-      "loss": 0.9868,
+      "loss": 0.9865,
       "step": 50
     },
     {
       "epoch": 0.022686832740213523,
-      "grad_norm": 0.3646712899208069,
+      "grad_norm": 0.36214709281921387,
       "learning_rate": 0.0001907537688442211,
-      "loss": 1.0884,
+      "loss": 1.0886,
       "step": 51
     },
     {
       "epoch": 0.023131672597864767,
-      "grad_norm": 0.35094475746154785,
+      "grad_norm": 0.35173431038856506,
       "learning_rate": 0.00019055276381909548,
-      "loss": 1.1207,
+      "loss": 1.1213,
       "step": 52
     },
     {
       "epoch": 0.023576512455516015,
-      "grad_norm": 0.34856799244880676,
+      "grad_norm": 0.34853485226631165,
       "learning_rate": 0.00019035175879396986,
-      "loss": 1.0121,
+      "loss": 1.0123,
       "step": 53
     },
     {
       "epoch": 0.02402135231316726,
-      "grad_norm": 0.37425950169563293,
+      "grad_norm": 0.3735390305519104,
       "learning_rate": 0.00019015075376884423,
-      "loss": 1.1073,
+      "loss": 1.1078,
       "step": 54
     },
     {
       "epoch": 0.024466192170818506,
-      "grad_norm": 0.3625994026660919,
+      "grad_norm": 0.3629351854324341,
       "learning_rate": 0.0001899497487437186,
-      "loss": 1.0863,
+      "loss": 1.0864,
       "step": 55
     },
     {
       "epoch": 0.02491103202846975,
-      "grad_norm": 0.295337051153183,
+      "grad_norm": 0.2966906428337097,
       "learning_rate": 0.00018974874371859298,
-      "loss": 1.0561,
+      "loss": 1.057,
       "step": 56
     },
     {
       "epoch": 0.025355871886120998,
-      "grad_norm": 0.34300458431243896,
+      "grad_norm": 0.34373167157173157,
       "learning_rate": 0.00018954773869346732,
-      "loss": 0.9955,
+      "loss": 0.9951,
       "step": 57
     },
     {
       "epoch": 0.025800711743772242,
-      "grad_norm": 0.45535075664520264,
+      "grad_norm": 0.454569548368454,
       "learning_rate": 0.00018934673366834172,
-      "loss": 1.0623,
+      "loss": 1.0618,
       "step": 58
     },
     {
       "epoch": 0.026245551601423486,
-      "grad_norm": 0.3560766577720642,
+      "grad_norm": 0.358252614736557,
       "learning_rate": 0.0001891457286432161,
-      "loss": 1.0259,
+      "loss": 1.0242,
       "step": 59
     },
     {
       "epoch": 0.026690391459074734,
-      "grad_norm": 0.33135348558425903,
+      "grad_norm": 0.33208489418029785,
       "learning_rate": 0.00018894472361809047,
-      "loss": 1.0823,
+      "loss": 1.0812,
       "step": 60
     },
     {
       "epoch": 0.027135231316725978,
-      "grad_norm": 0.33519864082336426,
+      "grad_norm": 0.3399137556552887,
       "learning_rate": 0.00018874371859296481,
-      "loss": 0.8893,
+      "loss": 0.8889,
       "step": 61
     },
     {
       "epoch": 0.027580071174377226,
-      "grad_norm": 0.4156452417373657,
+      "grad_norm": 0.41684690117836,
       "learning_rate": 0.00018854271356783921,
-      "loss": 0.888,
+      "loss": 0.8872,
       "step": 62
     },
     {
       "epoch": 0.02802491103202847,
-      "grad_norm": 0.5566922426223755,
+      "grad_norm": 0.5529223680496216,
       "learning_rate": 0.0001883417085427136,
-      "loss": 0.9465,
+      "loss": 0.95,
       "step": 63
     },
     {
       "epoch": 0.028469750889679714,
-      "grad_norm": 0.3567597270011902,
+      "grad_norm": 0.35778260231018066,
       "learning_rate": 0.00018814070351758793,
-      "loss": 1.044,
+      "loss": 1.0439,
       "step": 64
     },
     {
       "epoch": 0.02891459074733096,
-      "grad_norm": 0.3937987685203552,
+      "grad_norm": 0.387169748544693,
       "learning_rate": 0.0001879396984924623,
-      "loss": 1.1491,
+      "loss": 1.1486,
       "step": 65
     },
     {
       "epoch": 0.029359430604982206,
-      "grad_norm": 0.3297279477119446,
+      "grad_norm": 0.32700011134147644,
       "learning_rate": 0.0001877386934673367,
-      "loss": 1.1387,
+      "loss": 1.1394,
       "step": 66
     },
     {
       "epoch": 0.029804270462633453,
-      "grad_norm": 0.4475502371788025,
+      "grad_norm": 0.45792707800865173,
       "learning_rate": 0.00018753768844221108,
-      "loss": 1.1832,
+      "loss": 1.1815,
       "step": 67
     },
     {
       "epoch": 0.030249110320284697,
-      "grad_norm": 0.4012911021709442,
+      "grad_norm": 0.4012312591075897,
       "learning_rate": 0.00018733668341708543,
-      "loss": 1.1017,
+      "loss": 1.1024,
       "step": 68
     },
     {
       "epoch": 0.03069395017793594,
-      "grad_norm": 0.36798158288002014,
+      "grad_norm": 0.3783579468727112,
       "learning_rate": 0.0001871356783919598,
-      "loss": 0.9889,
+      "loss": 0.9901,
       "step": 69
     },
     {
       "epoch": 0.03113879003558719,
-      "grad_norm": 0.3330882787704468,
+      "grad_norm": 0.33383217453956604,
       "learning_rate": 0.0001869346733668342,
-      "loss": 1.0825,
+      "loss": 1.0832,
       "step": 70
     },
     {
       "epoch": 0.03158362989323844,
-      "grad_norm": 0.41344648599624634,
+      "grad_norm": 0.41429656744003296,
       "learning_rate": 0.00018673366834170854,
-      "loss": 1.0599,
+      "loss": 1.0593,
       "step": 71
     },
     {
       "epoch": 0.03202846975088968,
-      "grad_norm": 0.5002074837684631,
+      "grad_norm": 0.5010104179382324,
       "learning_rate": 0.00018653266331658292,
-      "loss": 1.0586,
+      "loss": 1.0602,
       "step": 72
     },
     {
       "epoch": 0.032473309608540925,
-      "grad_norm": 0.3306974470615387,
+      "grad_norm": 0.33280321955680847,
       "learning_rate": 0.0001863316582914573,
       "loss": 1.0004,
       "step": 73
     },
     {
       "epoch": 0.03291814946619217,
-      "grad_norm": 0.394704133272171,
+      "grad_norm": 0.39433741569519043,
       "learning_rate": 0.0001861306532663317,
-      "loss": 1.0083,
+      "loss": 1.0071,
       "step": 74
     },
     {
       "epoch": 0.03336298932384341,
-      "grad_norm": 0.3663751184940338,
+      "grad_norm": 0.3676820397377014,
       "learning_rate": 0.00018592964824120604,
-      "loss": 0.8553,
+      "loss": 0.8539,
       "step": 75
     },
     {
       "epoch": 0.033807829181494664,
-      "grad_norm": 0.37072721123695374,
+      "grad_norm": 0.3701139986515045,
       "learning_rate": 0.0001857286432160804,
-      "loss": 1.1199,
+      "loss": 1.1197,
       "step": 76
     },
     {
       "epoch": 0.03425266903914591,
-      "grad_norm": 0.37048038840293884,
+      "grad_norm": 0.3730039596557617,
       "learning_rate": 0.00018552763819095478,
-      "loss": 1.0716,
+      "loss": 1.0727,
       "step": 77
     },
     {
       "epoch": 0.03469750889679715,
-      "grad_norm": 0.34426698088645935,
+      "grad_norm": 0.3400503098964691,
       "learning_rate": 0.00018532663316582915,
-      "loss": 1.0378,
+      "loss": 1.0385,
       "step": 78
     },
     {
       "epoch": 0.035142348754448396,
-      "grad_norm": 0.3649255037307739,
+      "grad_norm": 0.36764055490493774,
       "learning_rate": 0.00018512562814070353,
-      "loss": 1.0843,
+      "loss": 1.0818,
       "step": 79
     },
     {
       "epoch": 0.03558718861209965,
-      "grad_norm": 0.37399303913116455,
+      "grad_norm": 0.37534597516059875,
       "learning_rate": 0.0001849246231155779,
-      "loss": 0.9741,
+      "loss": 0.9744,
       "step": 80
     },
     {
       "epoch": 0.03603202846975089,
-      "grad_norm": 0.3313739001750946,
+      "grad_norm": 0.33084404468536377,
       "learning_rate": 0.00018472361809045227,
-      "loss": 0.9883,
+      "loss": 0.9885,
       "step": 81
     },
     {
       "epoch": 0.036476868327402136,
-      "grad_norm": 0.36365818977355957,
+      "grad_norm": 0.36626842617988586,
       "learning_rate": 0.00018452261306532662,
-      "loss": 1.0434,
+      "loss": 1.0442,
       "step": 82
     },
     {
       "epoch": 0.03692170818505338,
-      "grad_norm": 0.33248934149742126,
+      "grad_norm": 0.5151258111000061,
       "learning_rate": 0.00018432160804020102,
-      "loss": 1.0635,
+      "loss": 1.0693,
       "step": 83
     },
     {
       "epoch": 0.037366548042704624,
-      "grad_norm": 0.361134797334671,
+      "grad_norm": 0.36287805438041687,
       "learning_rate": 0.0001841206030150754,
-      "loss": 1.0248,
+      "loss": 1.0257,
       "step": 84
     },
     {
       "epoch": 0.037811387900355875,
-      "grad_norm": 0.3499568700790405,
+      "grad_norm": 0.35327619314193726,
       "learning_rate": 0.00018391959798994977,
-      "loss": 0.9905,
+      "loss": 0.9902,
       "step": 85
     },
     {
       "epoch": 0.03825622775800712,
-      "grad_norm": 0.3246215879917145,
+      "grad_norm": 0.32582369446754456,
       "learning_rate": 0.0001837185929648241,
-      "loss": 1.0431,
+      "loss": 1.0425,
       "step": 86
     },
     {
       "epoch": 0.03870106761565836,
-      "grad_norm": 0.39000576734542847,
+      "grad_norm": 0.38856419920921326,
       "learning_rate": 0.0001835175879396985,
-      "loss": 0.995,
+      "loss": 0.9939,
       "step": 87
     },
     {
       "epoch": 0.03914590747330961,
-      "grad_norm": 0.32663071155548096,
+      "grad_norm": 0.3371952772140503,
       "learning_rate": 0.00018331658291457288,
-      "loss": 1.0976,
+      "loss": 1.0968,
       "step": 88
     },
     {
       "epoch": 0.03959074733096085,
-      "grad_norm": 0.35704660415649414,
+      "grad_norm": 0.36074796319007874,
       "learning_rate": 0.00018311557788944723,
-      "loss": 1.088,
+      "loss": 1.0878,
       "step": 89
     },
     {
       "epoch": 0.0400355871886121,
-      "grad_norm": 0.3753162622451782,
+      "grad_norm": 0.39816179871559143,
       "learning_rate": 0.0001829145728643216,
-      "loss": 1.0509,
+      "loss": 1.0504,
       "step": 90
     },
     {
       "epoch": 0.04048042704626335,
-      "grad_norm": 0.45394080877304077,
+      "grad_norm": 0.4431018829345703,
       "learning_rate": 0.000182713567839196,
-      "loss": 0.8953,
+      "loss": 0.8954,
       "step": 91
     },
     {
       "epoch": 0.04092526690391459,
-      "grad_norm": 0.35750091075897217,
+      "grad_norm": 0.3593955338001251,
       "learning_rate": 0.00018251256281407038,
       "loss": 1.0031,
       "step": 92
     },
     {
       "epoch": 0.041370106761565835,
-      "grad_norm": 0.2915020287036896,
+      "grad_norm": 0.2924874722957611,
       "learning_rate": 0.00018231155778894472,
-      "loss": 0.9292,
+      "loss": 0.9301,
       "step": 93
     },
     {
       "epoch": 0.04181494661921708,
-      "grad_norm": 0.29004380106925964,
+      "grad_norm": 0.29153579473495483,
       "learning_rate": 0.0001821105527638191,
-      "loss": 0.9423,
+      "loss": 0.9428,
       "step": 94
     },
     {
       "epoch": 0.04225978647686833,
-      "grad_norm": 0.3510340750217438,
+      "grad_norm": 0.3542582392692566,
       "learning_rate": 0.0001819095477386935,
-      "loss": 0.9933,
+      "loss": 0.9941,
       "step": 95
     },
     {
       "epoch": 0.042704626334519574,
-      "grad_norm": 0.3953409194946289,
+      "grad_norm": 0.3997493088245392,
       "learning_rate": 0.00018170854271356784,
-      "loss": 1.0174,
+      "loss": 1.0187,
       "step": 96
     },
     {
       "epoch": 0.04314946619217082,
-      "grad_norm": 0.3693845272064209,
+      "grad_norm": 0.3728165626525879,
       "learning_rate": 0.00018150753768844221,
-      "loss": 1.1309,
+      "loss": 1.1312,
       "step": 97
     },
     {
       "epoch": 0.04359430604982206,
-      "grad_norm": 0.33780723810195923,
+      "grad_norm": 0.35325804352760315,
       "learning_rate": 0.0001813065326633166,
-      "loss": 0.995,
+      "loss": 0.9955,
       "step": 98
     },
     {
       "epoch": 0.04403914590747331,
-      "grad_norm": 0.39444178342819214,
+      "grad_norm": 0.39074355363845825,
       "learning_rate": 0.00018110552763819096,
-      "loss": 1.0567,
+      "loss": 1.0561,
       "step": 99
     },
     {
       "epoch": 0.04448398576512456,
-      "grad_norm": 0.3430976867675781,
+      "grad_norm": 0.346187949180603,
       "learning_rate": 0.00018090452261306533,
-      "loss": 0.956,
+      "loss": 0.9568,
       "step": 100
     },
     {
       "epoch": 0.0449288256227758,
-      "grad_norm": 0.31412309408187866,
+      "grad_norm": 0.314586877822876,
       "learning_rate": 0.0001807035175879397,
-      "loss": 1.1141,
+      "loss": 1.1153,
       "step": 101
     },
     {
       "epoch": 0.045373665480427046,
-      "grad_norm": 0.3321763575077057,
+      "grad_norm": 0.3357396423816681,
       "learning_rate": 0.00018050251256281408,
-      "loss": 1.0245,
+      "loss": 1.0259,
       "step": 102
     },
     {
       "epoch": 0.04581850533807829,
-      "grad_norm": 0.3730502128601074,
+      "grad_norm": 0.3750225007534027,
       "learning_rate": 0.00018030150753768845,
-      "loss": 1.0961,
+      "loss": 1.096,
       "step": 103
     },
     {
       "epoch": 0.046263345195729534,
-      "grad_norm": 0.33242374658584595,
+      "grad_norm": 0.3401976227760315,
       "learning_rate": 0.00018010050251256282,
-      "loss": 0.9899,
+      "loss": 0.9911,
       "step": 104
     },
     {
       "epoch": 0.046708185053380785,
-      "grad_norm": 0.3853437900543213,
+      "grad_norm": 0.3866574764251709,
       "learning_rate": 0.0001798994974874372,
-      "loss": 0.9559,
+      "loss": 0.9558,
       "step": 105
     },
     {
       "epoch": 0.04715302491103203,
-      "grad_norm": 0.33745744824409485,
+      "grad_norm": 0.3394455313682556,
       "learning_rate": 0.00017969849246231157,
-      "loss": 1.0458,
+      "loss": 1.0463,
       "step": 106
     },
     {
       "epoch": 0.04759786476868327,
-      "grad_norm": 0.3601361811161041,
+      "grad_norm": 0.36298561096191406,
       "learning_rate": 0.00017949748743718592,
-      "loss": 1.0801,
+      "loss": 1.0807,
       "step": 107
     },
     {
       "epoch": 0.04804270462633452,
-      "grad_norm": 0.4157007932662964,
+      "grad_norm": 0.41382816433906555,
       "learning_rate": 0.00017929648241206032,
-      "loss": 0.9852,
+      "loss": 0.9859,
       "step": 108
     },
     {
       "epoch": 0.04848754448398576,
-      "grad_norm": 0.2746104598045349,
+      "grad_norm": 0.2769756019115448,
       "learning_rate": 0.0001790954773869347,
-      "loss": 1.0247,
+      "loss": 1.0251,
       "step": 109
     },
     {
       "epoch": 0.04893238434163701,
-      "grad_norm": 0.3099765181541443,
+      "grad_norm": 0.31425032019615173,
       "learning_rate": 0.00017889447236180906,
       "loss": 0.9642,
       "step": 110
     },
     {
       "epoch": 0.04937722419928826,
-      "grad_norm": 0.3376833498477936,
+      "grad_norm": 0.3441152274608612,
       "learning_rate": 0.0001786934673366834,
-      "loss": 1.0203,
+      "loss": 1.0211,
       "step": 111
     },
     {
       "epoch": 0.0498220640569395,
-      "grad_norm": 0.3929193913936615,
+      "grad_norm": 0.40452349185943604,
       "learning_rate": 0.0001784924623115578,
-      "loss": 1.1084,
+      "loss": 1.109,
       "step": 112
     },
     {
       "epoch": 0.050266903914590745,
-      "grad_norm": 0.3145126402378082,
+      "grad_norm": 0.32000288367271423,
       "learning_rate": 0.00017829145728643218,
-      "loss": 1.124,
+      "loss": 1.1237,
       "step": 113
     },
     {
       "epoch": 0.050711743772241996,
-      "grad_norm": 0.3065921664237976,
+      "grad_norm": 0.31428951025009155,
       "learning_rate": 0.00017809045226130653,
-      "loss": 1.0453,
+      "loss": 1.0443,
       "step": 114
     },
     {
       "epoch": 0.05115658362989324,
-      "grad_norm": 0.3715677857398987,
+      "grad_norm": 0.37118658423423767,
       "learning_rate": 0.0001778894472361809,
-      "loss": 0.9841,
+      "loss": 0.9842,
       "step": 115
     },
     {
       "epoch": 0.051601423487544484,
-      "grad_norm": 0.36604928970336914,
+      "grad_norm": 0.3640018403530121,
       "learning_rate": 0.0001776884422110553,
-      "loss": 0.9419,
+      "loss": 0.942,
       "step": 116
     },
     {
       "epoch": 0.05204626334519573,
-      "grad_norm": 0.37098243832588196,
+      "grad_norm": 0.3735273778438568,
       "learning_rate": 0.00017748743718592967,
-      "loss": 1.0133,
+      "loss": 1.0128,
       "step": 117
     },
     {
       "epoch": 0.05249110320284697,
-      "grad_norm": 0.3711595833301544,
+      "grad_norm": 0.37252792716026306,
       "learning_rate": 0.00017728643216080402,
-      "loss": 1.0324,
+      "loss": 1.0308,
       "step": 118
     },
     {
       "epoch": 0.052935943060498224,
-      "grad_norm": 0.3096163868904114,
+      "grad_norm": 0.3111330568790436,
       "learning_rate": 0.0001770854271356784,
-      "loss": 1.0086,
+      "loss": 1.0082,
       "step": 119
     },
     {
       "epoch": 0.05338078291814947,
-      "grad_norm": 0.4001842737197876,
+      "grad_norm": 0.40468236804008484,
       "learning_rate": 0.0001768844221105528,
-      "loss": 1.026,
+      "loss": 1.0273,
       "step": 120
     },
     {
       "epoch": 0.05382562277580071,
-      "grad_norm": 0.29020652174949646,
+      "grad_norm": 0.3018098473548889,
       "learning_rate": 0.00017668341708542714,
-      "loss": 0.9578,
+      "loss": 0.9599,
       "step": 121
     },
     {
       "epoch": 0.054270462633451956,
-      "grad_norm": 0.35916441679000854,
+      "grad_norm": 0.3560699224472046,
       "learning_rate": 0.0001764824120603015,
-      "loss": 1.0502,
+      "loss": 1.0508,
       "step": 122
     },
     {
       "epoch": 0.0547153024911032,
-      "grad_norm": 0.36931201815605164,
+      "grad_norm": 0.3692304193973541,
       "learning_rate": 0.00017628140703517588,
-      "loss": 0.9456,
+      "loss": 0.9454,
       "step": 123
     },
     {
       "epoch": 0.05516014234875445,
-      "grad_norm": 0.37013575434684753,
+      "grad_norm": 0.38090547919273376,
       "learning_rate": 0.00017608040201005026,
-      "loss": 1.0472,
+      "loss": 1.0469,
       "step": 124
     },
     {
       "epoch": 0.055604982206405695,
-      "grad_norm": 0.3738909959793091,
+      "grad_norm": 0.37586429715156555,
       "learning_rate": 0.00017587939698492463,
-      "loss": 1.0194,
+      "loss": 1.0205,
       "step": 125
     },
     {
       "epoch": 0.05604982206405694,
-      "grad_norm": 0.4676179885864258,
+      "grad_norm": 0.4673503041267395,
       "learning_rate": 0.000175678391959799,
-      "loss": 1.1556,
+      "loss": 1.1555,
       "step": 126
     },
     {
       "epoch": 0.056494661921708184,
-      "grad_norm": 0.430128812789917,
+      "grad_norm": 0.42561087012290955,
       "learning_rate": 0.00017547738693467338,
-      "loss": 1.1472,
+      "loss": 1.1458,
       "step": 127
     },
     {
       "epoch": 0.05693950177935943,
-      "grad_norm": 0.3177819848060608,
+      "grad_norm": 0.3206027150154114,
       "learning_rate": 0.00017527638190954775,
-      "loss": 1.0159,
+      "loss": 1.0165,
       "step": 128
     },
     {
       "epoch": 0.05738434163701068,
-      "grad_norm": 0.35535189509391785,
+      "grad_norm": 0.3541994094848633,
       "learning_rate": 0.00017507537688442212,
-      "loss": 1.0388,
+      "loss": 1.0377,
       "step": 129
     },
     {
       "epoch": 0.05782918149466192,
-      "grad_norm": 0.41556811332702637,
+      "grad_norm": 0.4175245761871338,
       "learning_rate": 0.0001748743718592965,
-      "loss": 1.0366,
+      "loss": 1.0371,
       "step": 130
     },
     {
       "epoch": 0.05827402135231317,
-      "grad_norm": 0.28733861446380615,
+      "grad_norm": 0.2889978587627411,
       "learning_rate": 0.00017467336683417087,
-      "loss": 1.041,
+      "loss": 1.0413,
       "step": 131
     },
     {
       "epoch": 0.05871886120996441,
-      "grad_norm": 0.36875680088996887,
+      "grad_norm": 0.36901023983955383,
       "learning_rate": 0.00017447236180904521,
-      "loss": 1.0408,
+      "loss": 1.042,
       "step": 132
     },
     {
       "epoch": 0.059163701067615655,
-      "grad_norm": 0.30114784836769104,
+      "grad_norm": 0.3048287332057953,
       "learning_rate": 0.00017427135678391961,
-      "loss": 1.0582,
+      "loss": 1.0587,
       "step": 133
     },
     {
       "epoch": 0.059608540925266906,
-      "grad_norm": 0.397324800491333,
+      "grad_norm": 0.40319743752479553,
       "learning_rate": 0.000174070351758794,
-      "loss": 1.0715,
+      "loss": 1.0727,
       "step": 134
     },
     {
       "epoch": 0.06005338078291815,
-      "grad_norm": 0.30825501680374146,
+      "grad_norm": 0.30953213572502136,
       "learning_rate": 0.00017386934673366836,
-      "loss": 0.8977,
+      "loss": 0.8989,
       "step": 135
     },
     {
       "epoch": 0.060498220640569395,
-      "grad_norm": 0.34875619411468506,
+      "grad_norm": 0.3466584384441376,
       "learning_rate": 0.0001736683417085427,
-      "loss": 1.0427,
+      "loss": 1.0438,
       "step": 136
     },
     {
       "epoch": 0.06094306049822064,
-      "grad_norm": 0.44976702332496643,
+      "grad_norm": 0.4506623148918152,
       "learning_rate": 0.0001734673366834171,
-      "loss": 1.0594,
+      "loss": 1.0601,
       "step": 137
     },
     {
       "epoch": 0.06138790035587188,
-      "grad_norm": 0.4009503722190857,
+      "grad_norm": 0.4022829830646515,
       "learning_rate": 0.00017326633165829148,
-      "loss": 1.0156,
+      "loss": 1.0149,
       "step": 138
     },
     {
       "epoch": 0.061832740213523134,
-      "grad_norm": 0.32680168747901917,
+      "grad_norm": 0.32555732131004333,
       "learning_rate": 0.00017306532663316582,
-      "loss": 1.0518,
+      "loss": 1.0512,
       "step": 139
     },
     {
       "epoch": 0.06227758007117438,
-      "grad_norm": 0.3297230005264282,
+      "grad_norm": 0.33276137709617615,
       "learning_rate": 0.0001728643216080402,
-      "loss": 0.982,
+      "loss": 0.9819,
       "step": 140
     },
     {
       "epoch": 0.06272241992882563,
-      "grad_norm": 0.3223113417625427,
+      "grad_norm": 0.32384124398231506,
       "learning_rate": 0.0001726633165829146,
-      "loss": 0.9451,
+      "loss": 0.9455,
       "step": 141
     },
     {
       "epoch": 0.06316725978647687,
-      "grad_norm": 0.3400196135044098,
+      "grad_norm": 0.3403018116950989,
       "learning_rate": 0.00017246231155778897,
       "loss": 0.9247,
       "step": 142
     },
     {
       "epoch": 0.06361209964412812,
-      "grad_norm": 0.36805739998817444,
+      "grad_norm": 0.3674178421497345,
       "learning_rate": 0.00017226130653266332,
-      "loss": 1.0732,
+      "loss": 1.074,
       "step": 143
     },
     {
       "epoch": 0.06405693950177936,
-      "grad_norm": 0.40036970376968384,
+      "grad_norm": 0.39759331941604614,
       "learning_rate": 0.0001720603015075377,
-      "loss": 1.0966,
+      "loss": 1.0983,
       "step": 144
     },
     {
       "epoch": 0.0645017793594306,
-      "grad_norm": 0.3755742311477661,
+      "grad_norm": 0.3763502240180969,
       "learning_rate": 0.00017185929648241206,
-      "loss": 0.9863,
+      "loss": 0.9859,
       "step": 145
     },
     {
       "epoch": 0.06494661921708185,
-      "grad_norm": 0.4415525496006012,
+      "grad_norm": 0.44169896841049194,
       "learning_rate": 0.00017165829145728644,
-      "loss": 1.0933,
+      "loss": 1.0924,
       "step": 146
     },
     {
       "epoch": 0.0653914590747331,
-      "grad_norm": 0.3682146966457367,
+      "grad_norm": 0.3722588121891022,
       "learning_rate": 0.0001714572864321608,
-      "loss": 0.9468,
+      "loss": 0.9457,
       "step": 147
     },
     {
       "epoch": 0.06583629893238434,
-      "grad_norm": 0.3670349717140198,
+      "grad_norm": 0.3731984496116638,
       "learning_rate": 0.00017125628140703518,
-      "loss": 1.1367,
+      "loss": 1.1355,
       "step": 148
     },
     {
       "epoch": 0.06628113879003558,
-      "grad_norm": 0.30022603273391724,
+      "grad_norm": 0.29889795184135437,
       "learning_rate": 0.00017105527638190955,
-      "loss": 0.9387,
+      "loss": 0.9392,
       "step": 149
     },
     {
       "epoch": 0.06672597864768683,
-      "grad_norm": 0.36275291442871094,
+      "grad_norm": 0.36229395866394043,
       "learning_rate": 0.00017085427135678393,
-      "loss": 0.8882,
+      "loss": 0.8904,
       "step": 150
     },
     {
       "epoch": 0.06717081850533808,
-      "grad_norm": 0.4646683633327484,
+      "grad_norm": 0.454025536775589,
       "learning_rate": 0.0001706532663316583,
-      "loss": 0.8556,
+      "loss": 0.857,
       "step": 151
     },
     {
       "epoch": 0.06761565836298933,
-      "grad_norm": 0.34833550453186035,
+      "grad_norm": 0.34767961502075195,
       "learning_rate": 0.00017045226130653267,
-      "loss": 1.0124,
+      "loss": 1.0121,
       "step": 152
     },
     {
       "epoch": 0.06806049822064057,
-      "grad_norm": 0.34811916947364807,
+      "grad_norm": 0.3490912616252899,
       "learning_rate": 0.00017025125628140705,
-      "loss": 0.8668,
+      "loss": 0.8651,
       "step": 153
     },
     {
       "epoch": 0.06850533807829182,
-      "grad_norm": 0.42011362314224243,
+      "grad_norm": 0.41010549664497375,
       "learning_rate": 0.00017005025125628142,
-      "loss": 0.9818,
+      "loss": 0.9816,
       "step": 154
     },
     {
       "epoch": 0.06895017793594306,
-      "grad_norm": 0.34665143489837646,
+      "grad_norm": 0.348345011472702,
       "learning_rate": 0.0001698492462311558,
-      "loss": 0.9958,
+      "loss": 0.9971,
       "step": 155
     },
     {
       "epoch": 0.0693950177935943,
-      "grad_norm": 0.3390451967716217,
+      "grad_norm": 0.3389308452606201,
       "learning_rate": 0.00016964824120603016,
-      "loss": 0.9835,
+      "loss": 0.9833,
       "step": 156
     },
     {
       "epoch": 0.06983985765124555,
-      "grad_norm": 0.3538704812526703,
+      "grad_norm": 0.3532242178916931,
       "learning_rate": 0.0001694472361809045,
-      "loss": 1.102,
+      "loss": 1.1019,
       "step": 157
     },
     {
       "epoch": 0.07028469750889679,
-      "grad_norm": 0.40229830145835876,
+      "grad_norm": 0.3977639675140381,
       "learning_rate": 0.0001692462311557789,
-      "loss": 1.1512,
+      "loss": 1.1519,
       "step": 158
     },
     {
       "epoch": 0.07072953736654804,
-      "grad_norm": 0.37556755542755127,
+      "grad_norm": 0.3684947192668915,
       "learning_rate": 0.00016904522613065328,
-      "loss": 1.0165,
+      "loss": 1.0164,
       "step": 159
     },
     {
       "epoch": 0.0711743772241993,
-      "grad_norm": 0.33789652585983276,
+      "grad_norm": 0.3370361328125,
       "learning_rate": 0.00016884422110552766,
-      "loss": 0.9962,
+      "loss": 0.9961,
       "step": 160
     },
     {
       "epoch": 0.07161921708185054,
-      "grad_norm": 0.39385226368904114,
+      "grad_norm": 0.3965550661087036,
       "learning_rate": 0.000168643216080402,
-      "loss": 1.0538,
+      "loss": 1.0542,
       "step": 161
     },
     {
       "epoch": 0.07206405693950178,
-      "grad_norm": 0.39948439598083496,
+      "grad_norm": 0.39642298221588135,
       "learning_rate": 0.0001684422110552764,
-      "loss": 1.0472,
+      "loss": 1.0452,
       "step": 162
     },
     {
       "epoch": 0.07250889679715303,
-      "grad_norm": 0.3786976635456085,
+      "grad_norm": 0.3771675229072571,
       "learning_rate": 0.00016824120603015078,
-      "loss": 1.0288,
+      "loss": 1.0284,
       "step": 163
     },
     {
       "epoch": 0.07295373665480427,
-      "grad_norm": 0.3729378879070282,
+      "grad_norm": 0.3752223253250122,
       "learning_rate": 0.00016804020100502512,
-      "loss": 0.9856,
+      "loss": 0.985,
       "step": 164
     },
     {
       "epoch": 0.07339857651245552,
-      "grad_norm": 0.36763879656791687,
+      "grad_norm": 0.3507731258869171,
       "learning_rate": 0.0001678391959798995,
-      "loss": 0.9711,
+      "loss": 0.9716,
       "step": 165
     },
     {
       "epoch": 0.07384341637010676,
-      "grad_norm": 0.38680917024612427,
+      "grad_norm": 0.3909834027290344,
       "learning_rate": 0.0001676381909547739,
-      "loss": 0.8915,
+      "loss": 0.8909,
       "step": 166
     },
     {
       "epoch": 0.074288256227758,
-      "grad_norm": 0.4449096620082855,
+      "grad_norm": 0.5423752665519714,
       "learning_rate": 0.00016743718592964827,
-      "loss": 0.9931,
+      "loss": 0.9922,
       "step": 167
     },
     {
       "epoch": 0.07473309608540925,
-      "grad_norm": 0.39082011580467224,
+      "grad_norm": 0.39162302017211914,
       "learning_rate": 0.0001672361809045226,
-      "loss": 1.0084,
+      "loss": 1.0075,
       "step": 168
     },
     {
       "epoch": 0.07517793594306049,
-      "grad_norm": 0.37978312373161316,
+      "grad_norm": 0.3779022693634033,
       "learning_rate": 0.00016703517587939699,
-      "loss": 0.9449,
+      "loss": 0.9451,
       "step": 169
     },
     {
       "epoch": 0.07562277580071175,
-      "grad_norm": 0.32479336857795715,
+      "grad_norm": 0.3254469633102417,
       "learning_rate": 0.00016683417085427136,
-      "loss": 0.8941,
+      "loss": 0.8934,
       "step": 170
     },
     {
       "epoch": 0.076067615658363,
-      "grad_norm": 0.35645660758018494,
+      "grad_norm": 0.3539400100708008,
       "learning_rate": 0.00016663316582914573,
-      "loss": 0.9952,
+      "loss": 0.9944,
       "step": 171
     },
     {
       "epoch": 0.07651245551601424,
-      "grad_norm": 0.40985000133514404,
+      "grad_norm": 0.41099923849105835,
       "learning_rate": 0.0001664321608040201,
-      "loss": 1.0187,
+      "loss": 1.018,
       "step": 172
     },
     {
       "epoch": 0.07695729537366548,
-      "grad_norm": 0.4032020568847656,
+      "grad_norm": 0.40103086829185486,
       "learning_rate": 0.00016623115577889448,
-      "loss": 0.9356,
+      "loss": 0.9354,
       "step": 173
     },
     {
       "epoch": 0.07740213523131673,
-      "grad_norm": 0.3428996801376343,
+      "grad_norm": 0.3429498076438904,
       "learning_rate": 0.00016603015075376885,
-      "loss": 0.9326,
+      "loss": 0.9328,
       "step": 174
     },
     {
       "epoch": 0.07784697508896797,
-      "grad_norm": 0.2806510925292969,
+      "grad_norm": 0.2685195505619049,
       "learning_rate": 0.00016582914572864322,
-      "loss": 0.6118,
+      "loss": 0.6091,
       "step": 175
     },
     {
       "epoch": 0.07829181494661921,
-      "grad_norm": 0.3010331392288208,
+      "grad_norm": 0.2993873655796051,
       "learning_rate": 0.0001656281407035176,
-      "loss": 1.0065,
+      "loss": 1.0054,
       "step": 176
     },
     {
       "epoch": 0.07873665480427046,
-      "grad_norm": 0.2989744246006012,
+      "grad_norm": 0.29764705896377563,
       "learning_rate": 0.00016542713567839197,
-      "loss": 0.948,
+      "loss": 0.9484,
       "step": 177
     },
     {
       "epoch": 0.0791814946619217,
-      "grad_norm": 0.3374849259853363,
+      "grad_norm": 0.33408114314079285,
       "learning_rate": 0.00016522613065326634,
-      "loss": 1.0899,
+      "loss": 1.0898,
       "step": 178
     },
     {
       "epoch": 0.07962633451957295,
-      "grad_norm": 0.38859498500823975,
+      "grad_norm": 0.3865182399749756,
       "learning_rate": 0.00016502512562814072,
-      "loss": 1.035,
+      "loss": 1.0371,
       "step": 179
     },
     {
       "epoch": 0.0800711743772242,
-      "grad_norm": 0.34474217891693115,
+      "grad_norm": 0.34564244747161865,
       "learning_rate": 0.0001648241206030151,
-      "loss": 0.9826,
+      "loss": 0.9825,
       "step": 180
     },
     {
       "epoch": 0.08051601423487545,
-      "grad_norm": 0.2957572937011719,
+      "grad_norm": 0.29182514548301697,
       "learning_rate": 0.00016462311557788946,
-      "loss": 0.8905,
+      "loss": 0.8901,
       "step": 181
     },
     {
       "epoch": 0.0809608540925267,
-      "grad_norm": 0.3576017916202545,
+      "grad_norm": 0.35713285207748413,
       "learning_rate": 0.0001644221105527638,
-      "loss": 1.0266,
+      "loss": 1.0267,
       "step": 182
     },
     {
       "epoch": 0.08140569395017794,
-      "grad_norm": 0.38848915696144104,
+      "grad_norm": 0.38320329785346985,
       "learning_rate": 0.0001642211055276382,
-      "loss": 0.9883,
+      "loss": 0.9875,
       "step": 183
     },
     {
       "epoch": 0.08185053380782918,
-      "grad_norm": 0.4087153375148773,
+      "grad_norm": 0.4025184214115143,
       "learning_rate": 0.00016402010050251258,
-      "loss": 1.1882,
+      "loss": 1.1871,
       "step": 184
     },
     {
       "epoch": 0.08229537366548043,
-      "grad_norm": 0.3468906879425049,
+      "grad_norm": 0.34486982226371765,
       "learning_rate": 0.00016381909547738695,
-      "loss": 0.9818,
+      "loss": 0.982,
       "step": 185
     },
     {
       "epoch": 0.08274021352313167,
-      "grad_norm": 0.4617744982242584,
+      "grad_norm": 0.45098572969436646,
       "learning_rate": 0.0001636180904522613,
-      "loss": 0.9065,
+      "loss": 0.9052,
       "step": 186
     },
     {
       "epoch": 0.08318505338078291,
-      "grad_norm": 0.4207713305950165,
+      "grad_norm": 0.4179227352142334,
       "learning_rate": 0.0001634170854271357,
-      "loss": 0.9988,
+      "loss": 0.9985,
       "step": 187
     },
     {
       "epoch": 0.08362989323843416,
-      "grad_norm": 0.3632107675075531,
+      "grad_norm": 0.36076629161834717,
       "learning_rate": 0.00016321608040201007,
-      "loss": 0.9049,
+      "loss": 0.9035,
       "step": 188
     },
     {
       "epoch": 0.08407473309608542,
-      "grad_norm": 0.4303954839706421,
+      "grad_norm": 0.4315630793571472,
       "learning_rate": 0.00016301507537688442,
-      "loss": 0.9801,
+      "loss": 0.9815,
       "step": 189
     },
     {
       "epoch": 0.08451957295373666,
-      "grad_norm": 0.3233638107776642,
+      "grad_norm": 0.31583619117736816,
       "learning_rate": 0.0001628140703517588,
-      "loss": 1.1457,
+      "loss": 1.1461,
       "step": 190
     },
     {
       "epoch": 0.0849644128113879,
-      "grad_norm": 0.36270812153816223,
+      "grad_norm": 0.3578287661075592,
       "learning_rate": 0.00016261306532663316,
-      "loss": 0.9804,
+      "loss": 0.9802,
       "step": 191
     },
     {
       "epoch": 0.08540925266903915,
-      "grad_norm": 0.3483213484287262,
+      "grad_norm": 0.35003262758255005,
       "learning_rate": 0.00016241206030150756,
-      "loss": 0.9412,
+      "loss": 0.9406,
       "step": 192
     },
     {
       "epoch": 0.08585409252669039,
-      "grad_norm": 0.33415859937667847,
+      "grad_norm": 0.33165210485458374,
       "learning_rate": 0.0001622110552763819,
-      "loss": 1.0479,
+      "loss": 1.0484,
       "step": 193
     },
     {
       "epoch": 0.08629893238434164,
-      "grad_norm": 0.3390806317329407,
+      "grad_norm": 0.33711302280426025,
       "learning_rate": 0.00016201005025125628,
-      "loss": 1.0266,
+      "loss": 1.0255,
       "step": 194
     },
     {
       "epoch": 0.08674377224199288,
-      "grad_norm": 0.3439917266368866,
+      "grad_norm": 0.3443413972854614,
       "learning_rate": 0.00016180904522613066,
-      "loss": 0.9495,
+      "loss": 0.9512,
       "step": 195
     },
     {
       "epoch": 0.08718861209964412,
-      "grad_norm": 0.5404065847396851,
+      "grad_norm": 0.5403580069541931,
       "learning_rate": 0.00016160804020100503,
-      "loss": 1.0176,
+      "loss": 1.0158,
       "step": 196
     },
     {
       "epoch": 0.08763345195729537,
-      "grad_norm": 0.40834537148475647,
+      "grad_norm": 0.40549829602241516,
       "learning_rate": 0.0001614070351758794,
-      "loss": 1.0119,
+      "loss": 1.0102,
       "step": 197
     },
     {
       "epoch": 0.08807829181494661,
-      "grad_norm": 0.41839244961738586,
+      "grad_norm": 0.43879571557044983,
       "learning_rate": 0.00016120603015075378,
-      "loss": 1.0006,
+      "loss": 1.0015,
       "step": 198
     },
     {
       "epoch": 0.08852313167259787,
-      "grad_norm": 0.36179983615875244,
+      "grad_norm": 0.3620396852493286,
       "learning_rate": 0.00016100502512562815,
-      "loss": 1.0247,
+      "loss": 1.0248,
       "step": 199
     },
     {
       "epoch": 0.08896797153024912,
-      "grad_norm": 0.34288227558135986,
+      "grad_norm": 0.339257150888443,
       "learning_rate": 0.00016080402010050252,
-      "loss": 0.9545,
+      "loss": 0.9548,
       "step": 200
     },
     {
       "epoch": 0.08941281138790036,
-      "grad_norm": 0.37949255108833313,
+      "grad_norm": 0.3820551037788391,
       "learning_rate": 0.0001606030150753769,
-      "loss": 1.0024,
+      "loss": 1.0026,
       "step": 201
     },
     {
       "epoch": 0.0898576512455516,
-      "grad_norm": 0.4100090563297272,
+      "grad_norm": 0.4135708808898926,
       "learning_rate": 0.00016040201005025127,
-      "loss": 1.0465,
+      "loss": 1.0476,
       "step": 202
     },
     {
       "epoch": 0.09030249110320285,
-      "grad_norm": 0.3741397559642792,
+      "grad_norm": 0.3772624731063843,
       "learning_rate": 0.00016020100502512564,
-      "loss": 0.8969,
+      "loss": 0.8981,
       "step": 203
     },
     {
       "epoch": 0.09074733096085409,
-      "grad_norm": 0.4092821180820465,
+      "grad_norm": 0.42693498730659485,
       "learning_rate": 0.00016,
-      "loss": 1.3054,
+      "loss": 1.3064,
       "step": 204
     },
     {
       "epoch": 0.09119217081850534,
-      "grad_norm": 0.29491907358169556,
+      "grad_norm": 0.2962593138217926,
       "learning_rate": 0.00015979899497487439,
-      "loss": 1.0039,
+      "loss": 1.0059,
       "step": 205
     },
     {
       "epoch": 0.09163701067615658,
-      "grad_norm": 0.38052767515182495,
+      "grad_norm": 0.3942723274230957,
       "learning_rate": 0.00015959798994974876,
-      "loss": 0.9034,
+      "loss": 0.9041,
       "step": 206
     },
     {
       "epoch": 0.09208185053380782,
-      "grad_norm": 0.34436798095703125,
+      "grad_norm": 0.34500858187675476,
       "learning_rate": 0.0001593969849246231,
-      "loss": 0.9621,
+      "loss": 0.9624,
       "step": 207
     },
     {
       "epoch": 0.09252669039145907,
-      "grad_norm": 0.41748183965682983,
+      "grad_norm": 0.41078099608421326,
       "learning_rate": 0.0001591959798994975,
-      "loss": 1.0112,
+      "loss": 1.0105,
       "step": 208
     },
     {
       "epoch": 0.09297153024911033,
-      "grad_norm": 0.35025399923324585,
+      "grad_norm": 0.3484095335006714,
       "learning_rate": 0.00015899497487437188,
-      "loss": 1.0594,
+      "loss": 1.0591,
       "step": 209
     },
     {
       "epoch": 0.09341637010676157,
-      "grad_norm": 0.3578777611255646,
+      "grad_norm": 0.3550896644592285,
       "learning_rate": 0.00015879396984924625,
-      "loss": 1.0191,
+      "loss": 1.0196,
       "step": 210
     },
     {
       "epoch": 0.09386120996441281,
-      "grad_norm": 0.3903699517250061,
+      "grad_norm": 0.3832697570323944,
       "learning_rate": 0.0001585929648241206,
-      "loss": 1.0492,
+      "loss": 1.048,
       "step": 211
     },
     {
       "epoch": 0.09430604982206406,
-      "grad_norm": 0.34571701288223267,
+      "grad_norm": 0.34110692143440247,
       "learning_rate": 0.000158391959798995,
-      "loss": 1.07,
+      "loss": 1.0706,
       "step": 212
     },
     {
       "epoch": 0.0947508896797153,
-      "grad_norm": 0.3766401410102844,
+      "grad_norm": 0.3769712746143341,
       "learning_rate": 0.00015819095477386937,
-      "loss": 0.9508,
+      "loss": 0.9514,
       "step": 213
     },
     {
       "epoch": 0.09519572953736655,
-      "grad_norm": 0.3765754699707031,
+      "grad_norm": 0.3767399489879608,
       "learning_rate": 0.00015798994974874372,
-      "loss": 0.9604,
+      "loss": 0.9611,
       "step": 214
     },
     {
       "epoch": 0.09564056939501779,
-      "grad_norm": 0.39237186312675476,
+      "grad_norm": 0.3955567479133606,
       "learning_rate": 0.0001577889447236181,
-      "loss": 0.9678,
+      "loss": 0.9683,
       "step": 215
     },
     {
       "epoch": 0.09608540925266904,
-      "grad_norm": 0.43379107117652893,
+      "grad_norm": 0.4317300021648407,
       "learning_rate": 0.00015758793969849246,
-      "loss": 0.9569,
+      "loss": 0.957,
       "step": 216
     },
     {
       "epoch": 0.09653024911032028,
-      "grad_norm": 0.44235309958457947,
+      "grad_norm": 0.4412591755390167,
       "learning_rate": 0.00015738693467336686,
-      "loss": 1.047,
+      "loss": 1.0473,
       "step": 217
     },
     {
       "epoch": 0.09697508896797152,
-      "grad_norm": 0.37492284178733826,
+      "grad_norm": 0.37174472212791443,
       "learning_rate": 0.0001571859296482412,
-      "loss": 0.9247,
+      "loss": 0.9258,
       "step": 218
     },
     {
       "epoch": 0.09741992882562278,
-      "grad_norm": 0.44187766313552856,
+      "grad_norm": 0.4425841271877289,
       "learning_rate": 0.00015698492462311558,
-      "loss": 1.0136,
+      "loss": 1.0147,
       "step": 219
     },
     {
       "epoch": 0.09786476868327403,
-      "grad_norm": 0.41393113136291504,
+      "grad_norm": 0.40715131163597107,
       "learning_rate": 0.00015678391959798995,
-      "loss": 1.0051,
+      "loss": 1.0028,
       "step": 220
     },
     {
       "epoch": 0.09830960854092527,
-      "grad_norm": 0.3643686771392822,
+      "grad_norm": 0.3638611435890198,
       "learning_rate": 0.00015658291457286433,
-      "loss": 1.1171,
+      "loss": 1.1178,
       "step": 221
     },
     {
       "epoch": 0.09875444839857651,
-      "grad_norm": 0.351253479719162,
+      "grad_norm": 0.3498910963535309,
       "learning_rate": 0.0001563819095477387,
-      "loss": 1.0466,
+      "loss": 1.0467,
       "step": 222
     },
     {
       "epoch": 0.09919928825622776,
-      "grad_norm": 0.35708218812942505,
+      "grad_norm": 0.36899253726005554,
       "learning_rate": 0.00015618090452261307,
-      "loss": 1.0688,
+      "loss": 1.0702,
       "step": 223
     },
     {
       "epoch": 0.099644128113879,
-      "grad_norm": 0.3984560966491699,
+      "grad_norm": 0.398636132478714,
       "learning_rate": 0.00015597989949748745,
-      "loss": 0.9606,
+      "loss": 0.9641,
       "step": 224
     },
     {
       "epoch": 0.10008896797153025,
-      "grad_norm": 0.3412669897079468,
+      "grad_norm": 0.3399724066257477,
       "learning_rate": 0.00015577889447236182,
-      "loss": 0.9781,
+      "loss": 0.9786,
       "step": 225
     },
     {
       "epoch": 0.10053380782918149,
-      "grad_norm": 0.36613014340400696,
+      "grad_norm": 0.3582400679588318,
       "learning_rate": 0.0001555778894472362,
-      "loss": 0.9449,
+      "loss": 0.9441,
       "step": 226
     },
     {
       "epoch": 0.10097864768683273,
-      "grad_norm": 0.3568494915962219,
+      "grad_norm": 0.365993469953537,
       "learning_rate": 0.00015537688442211056,
-      "loss": 0.9679,
+      "loss": 0.9669,
       "step": 227
     },
     {
       "epoch": 0.10142348754448399,
-      "grad_norm": 0.4687019884586334,
+      "grad_norm": 0.46609994769096375,
       "learning_rate": 0.00015517587939698494,
-      "loss": 1.0386,
+      "loss": 1.0387,
       "step": 228
     },
     {
       "epoch": 0.10186832740213524,
-      "grad_norm": 0.40447989106178284,
+      "grad_norm": 0.4048779010772705,
       "learning_rate": 0.0001549748743718593,
-      "loss": 0.9812,
+      "loss": 0.9808,
       "step": 229
     },
     {
       "epoch": 0.10231316725978648,
-      "grad_norm": 0.39272576570510864,
+      "grad_norm": 0.3965264856815338,
       "learning_rate": 0.00015477386934673368,
-      "loss": 1.0165,
+      "loss": 1.0191,
       "step": 230
     },
     {
       "epoch": 0.10275800711743772,
-      "grad_norm": 0.3780848979949951,
+      "grad_norm": 0.379210501909256,
       "learning_rate": 0.00015457286432160806,
-      "loss": 0.9606,
+      "loss": 0.9608,
       "step": 231
     },
     {
       "epoch": 0.10320284697508897,
-      "grad_norm": 0.46794167160987854,
+      "grad_norm": 0.4589402675628662,
       "learning_rate": 0.0001543718592964824,
-      "loss": 0.9305,
+      "loss": 0.9267,
       "step": 232
     },
     {
       "epoch": 0.10364768683274021,
-      "grad_norm": 0.39729002118110657,
+      "grad_norm": 0.3889247179031372,
       "learning_rate": 0.0001541708542713568,
-      "loss": 1.0113,
+      "loss": 1.0115,
       "step": 233
     },
     {
       "epoch": 0.10409252669039146,
-      "grad_norm": 0.4045659005641937,
+      "grad_norm": 0.3973037004470825,
       "learning_rate": 0.00015396984924623117,
-      "loss": 0.9961,
+      "loss": 0.9967,
       "step": 234
     },
     {
       "epoch": 0.1045373665480427,
-      "grad_norm": 0.35122302174568176,
+      "grad_norm": 0.35045430064201355,
       "learning_rate": 0.00015376884422110555,
-      "loss": 0.9891,
+      "loss": 0.9899,
       "step": 235
     },
     {
       "epoch": 0.10498220640569395,
-      "grad_norm": 0.37050357460975647,
+      "grad_norm": 0.36741137504577637,
       "learning_rate": 0.0001535678391959799,
-      "loss": 1.0024,
+      "loss": 1.0016,
       "step": 236
     },
     {
       "epoch": 0.10542704626334519,
-      "grad_norm": 0.40677326917648315,
+      "grad_norm": 0.4001176953315735,
       "learning_rate": 0.00015336683417085427,
-      "loss": 1.0948,
+      "loss": 1.0941,
       "step": 237
     },
     {
       "epoch": 0.10587188612099645,
-      "grad_norm": 0.35067057609558105,
+      "grad_norm": 0.3526718020439148,
       "learning_rate": 0.00015316582914572867,
-      "loss": 0.9897,
+      "loss": 0.9905,
       "step": 238
     },
     {
       "epoch": 0.10631672597864769,
-      "grad_norm": 0.3920201063156128,
+      "grad_norm": 0.39122429490089417,
       "learning_rate": 0.000152964824120603,
-      "loss": 1.0114,
+      "loss": 1.0116,
       "step": 239
     },
     {
       "epoch": 0.10676156583629894,
-      "grad_norm": 0.3613733649253845,
+      "grad_norm": 0.36022770404815674,
       "learning_rate": 0.00015276381909547739,
-      "loss": 0.9508,
+      "loss": 0.9501,
       "step": 240
     },
     {
       "epoch": 0.10720640569395018,
-      "grad_norm": 0.35030075907707214,
+      "grad_norm": 0.3478608727455139,
       "learning_rate": 0.00015256281407035176,
-      "loss": 1.0019,
+      "loss": 1.0017,
       "step": 241
     },
     {
       "epoch": 0.10765124555160142,
-      "grad_norm": 0.4036657512187958,
+      "grad_norm": 0.36006009578704834,
       "learning_rate": 0.00015236180904522613,
-      "loss": 0.9347,
+      "loss": 0.935,
       "step": 242
     },
     {
       "epoch": 0.10809608540925267,
-      "grad_norm": 0.3680381178855896,
+      "grad_norm": 0.3560234606266022,
       "learning_rate": 0.0001521608040201005,
-      "loss": 0.9541,
+      "loss": 0.9552,
       "step": 243
     },
     {
       "epoch": 0.10854092526690391,
-      "grad_norm": 0.7577487826347351,
+      "grad_norm": 0.6785101890563965,
       "learning_rate": 0.00015195979899497488,
-      "loss": 1.0019,
+      "loss": 0.9972,
       "step": 244
     },
     {
       "epoch": 0.10898576512455516,
-      "grad_norm": 0.32293733954429626,
+      "grad_norm": 0.32202988862991333,
       "learning_rate": 0.00015175879396984925,
-      "loss": 0.9768,
+      "loss": 0.9763,
       "step": 245
     },
     {
       "epoch": 0.1094306049822064,
-      "grad_norm": 0.32959771156311035,
+      "grad_norm": 0.33148205280303955,
       "learning_rate": 0.00015155778894472362,
-      "loss": 1.0606,
+      "loss": 1.0615,
       "step": 246
     },
     {
       "epoch": 0.10987544483985764,
-      "grad_norm": 0.3493881821632385,
+      "grad_norm": 0.35082313418388367,
       "learning_rate": 0.000151356783919598,
-      "loss": 1.0317,
+      "loss": 1.0332,
       "step": 247
     },
     {
       "epoch": 0.1103202846975089,
-      "grad_norm": 0.3482673168182373,
+      "grad_norm": 0.34773561358451843,
       "learning_rate": 0.00015115577889447237,
-      "loss": 0.9157,
+      "loss": 0.9172,
       "step": 248
     },
     {
       "epoch": 0.11076512455516015,
-      "grad_norm": 0.35461658239364624,
+      "grad_norm": 0.35180824995040894,
       "learning_rate": 0.00015095477386934674,
-      "loss": 1.0753,
+      "loss": 1.077,
       "step": 249
     },
     {
       "epoch": 0.11120996441281139,
-      "grad_norm": 0.3124016225337982,
+      "grad_norm": 0.33543917536735535,
       "learning_rate": 0.00015075376884422112,
-      "loss": 0.9352,
+      "loss": 0.9345,
       "step": 250
     },
     {
       "epoch": 0.11165480427046263,
-      "grad_norm": 0.3451753854751587,
+      "grad_norm": 0.3516584038734436,
       "learning_rate": 0.0001505527638190955,
-      "loss": 0.948,
+      "loss": 0.9481,
       "step": 251
     },
     {
       "epoch": 0.11209964412811388,
-      "grad_norm": 0.3436938524246216,
+      "grad_norm": 0.3360784947872162,
       "learning_rate": 0.00015035175879396986,
-      "loss": 0.9277,
+      "loss": 0.9285,
       "step": 252
     },
     {
       "epoch": 0.11254448398576512,
-      "grad_norm": 0.3544940650463104,
+      "grad_norm": 0.3512001633644104,
       "learning_rate": 0.00015015075376884423,
-      "loss": 0.976,
+      "loss": 0.9745,
       "step": 253
     },
     {
       "epoch": 0.11298932384341637,
-      "grad_norm": 0.34947720170021057,
+      "grad_norm": 0.34287190437316895,
       "learning_rate": 0.0001499497487437186,
-      "loss": 1.0602,
+      "loss": 1.0596,
       "step": 254
     },
     {
       "epoch": 0.11343416370106761,
-      "grad_norm": 0.38455042243003845,
+      "grad_norm": 0.3830087184906006,
       "learning_rate": 0.00014974874371859298,
-      "loss": 0.9524,
+      "loss": 0.9534,
       "step": 255
     },
     {
       "epoch": 0.11387900355871886,
-      "grad_norm": 0.4009455442428589,
+      "grad_norm": 0.3950470983982086,
       "learning_rate": 0.00014954773869346735,
-      "loss": 1.0465,
+      "loss": 1.045,
       "step": 256
     },
     {
       "epoch": 0.11432384341637011,
-      "grad_norm": 0.5961493849754333,
+      "grad_norm": 0.5723717212677002,
       "learning_rate": 0.0001493467336683417,
-      "loss": 0.8681,
+      "loss": 0.8671,
       "step": 257
     },
     {
       "epoch": 0.11476868327402136,
-      "grad_norm": 0.47233107686042786,
+      "grad_norm": 0.4654160737991333,
       "learning_rate": 0.0001491457286432161,
-      "loss": 0.9355,
+      "loss": 0.936,
       "step": 258
     },
     {
       "epoch": 0.1152135231316726,
-      "grad_norm": 0.41602540016174316,
+      "grad_norm": 0.4141252338886261,
       "learning_rate": 0.00014894472361809047,
-      "loss": 1.0256,
+      "loss": 1.0275,
       "step": 259
     },
     {
       "epoch": 0.11565836298932385,
-      "grad_norm": 0.3379324972629547,
+      "grad_norm": 0.3363032042980194,
       "learning_rate": 0.00014874371859296482,
-      "loss": 0.8982,
+      "loss": 0.899,
       "step": 260
     },
     {
       "epoch": 0.11610320284697509,
-      "grad_norm": 0.3502661883831024,
+      "grad_norm": 0.35972562432289124,
       "learning_rate": 0.0001485427135678392,
-      "loss": 0.8932,
+      "loss": 0.8928,
       "step": 261
     },
     {
       "epoch": 0.11654804270462633,
-      "grad_norm": 0.35349032282829285,
+      "grad_norm": 0.35416486859321594,
       "learning_rate": 0.00014834170854271356,
-      "loss": 0.9541,
+      "loss": 0.9557,
       "step": 262
     },
     {
       "epoch": 0.11699288256227758,
-      "grad_norm": 0.37270882725715637,
+      "grad_norm": 0.3734114170074463,
       "learning_rate": 0.00014814070351758796,
-      "loss": 0.9566,
+      "loss": 0.956,
       "step": 263
     },
     {
       "epoch": 0.11743772241992882,
-      "grad_norm": 0.3453938364982605,
+      "grad_norm": 0.3472467362880707,
       "learning_rate": 0.0001479396984924623,
-      "loss": 0.9901,
+      "loss": 0.9911,
       "step": 264
     },
     {
       "epoch": 0.11788256227758007,
-      "grad_norm": 0.4339182376861572,
+      "grad_norm": 0.433698832988739,
       "learning_rate": 0.00014773869346733668,
-      "loss": 0.9542,
+      "loss": 0.9547,
       "step": 265
     },
     {
       "epoch": 0.11832740213523131,
-      "grad_norm": 0.3721541166305542,
+      "grad_norm": 0.3658886253833771,
       "learning_rate": 0.00014753768844221106,
-      "loss": 0.9075,
+      "loss": 0.9044,
       "step": 266
     },
     {
       "epoch": 0.11877224199288257,
-      "grad_norm": 0.3452603220939636,
+      "grad_norm": 0.3511579632759094,
       "learning_rate": 0.00014733668341708543,
-      "loss": 0.9762,
+      "loss": 0.9766,
       "step": 267
     },
     {
       "epoch": 0.11921708185053381,
-      "grad_norm": 0.33102792501449585,
+      "grad_norm": 0.3257918357849121,
       "learning_rate": 0.0001471356783919598,
-      "loss": 0.9648,
+      "loss": 0.9641,
       "step": 268
     },
     {
       "epoch": 0.11966192170818506,
-      "grad_norm": 0.38129130005836487,
+      "grad_norm": 0.3833535611629486,
       "learning_rate": 0.00014693467336683417,
-      "loss": 1.0283,
+      "loss": 1.0307,
       "step": 269
     },
     {
       "epoch": 0.1201067615658363,
-      "grad_norm": 0.3465835452079773,
+      "grad_norm": 0.3672519326210022,
       "learning_rate": 0.00014673366834170855,
-      "loss": 0.9383,
+      "loss": 0.941,
       "step": 270
     },
     {
       "epoch": 0.12055160142348754,
-      "grad_norm": 0.31291693449020386,
+      "grad_norm": 0.3263191282749176,
       "learning_rate": 0.00014653266331658292,
-      "loss": 1.0326,
+      "loss": 1.0327,
       "step": 271
     },
     {
       "epoch": 0.12099644128113879,
-      "grad_norm": 0.40111976861953735,
+      "grad_norm": 0.3983098864555359,
       "learning_rate": 0.0001463316582914573,
-      "loss": 1.0133,
+      "loss": 1.011,
       "step": 272
     },
     {
       "epoch": 0.12144128113879003,
-      "grad_norm": 0.3763119578361511,
+      "grad_norm": 0.37602588534355164,
       "learning_rate": 0.00014613065326633167,
-      "loss": 1.0317,
+      "loss": 1.0328,
       "step": 273
     },
     {
       "epoch": 0.12188612099644128,
-      "grad_norm": 0.3577355742454529,
+      "grad_norm": 0.354548841714859,
       "learning_rate": 0.00014592964824120604,
-      "loss": 1.0477,
+      "loss": 1.0472,
       "step": 274
     },
     {
       "epoch": 0.12233096085409252,
-      "grad_norm": 0.3452180027961731,
+      "grad_norm": 0.343363493680954,
       "learning_rate": 0.0001457286432160804,
-      "loss": 1.0083,
+      "loss": 1.0053,
       "step": 275
     },
     {
       "epoch": 0.12277580071174377,
-      "grad_norm": 0.40135836601257324,
+      "grad_norm": 0.3958703577518463,
       "learning_rate": 0.00014552763819095479,
-      "loss": 0.9175,
+      "loss": 0.917,
       "step": 276
     },
     {
       "epoch": 0.12322064056939502,
-      "grad_norm": 0.45712020993232727,
+      "grad_norm": 0.4580914378166199,
       "learning_rate": 0.00014532663316582916,
-      "loss": 0.9563,
+      "loss": 0.961,
       "step": 277
     },
     {
       "epoch": 0.12366548042704627,
-      "grad_norm": 0.3986268937587738,
+      "grad_norm": 0.39822980761528015,
       "learning_rate": 0.00014512562814070353,
-      "loss": 0.926,
+      "loss": 0.9276,
       "step": 278
     },
     {
       "epoch": 0.12411032028469751,
-      "grad_norm": 0.3132750988006592,
+      "grad_norm": 0.3082531988620758,
       "learning_rate": 0.0001449246231155779,
-      "loss": 0.9503,
+      "loss": 0.9496,
       "step": 279
     },
     {
       "epoch": 0.12455516014234876,
-      "grad_norm": 0.34486275911331177,
+      "grad_norm": 0.34612587094306946,
       "learning_rate": 0.00014472361809045228,
-      "loss": 0.9375,
+      "loss": 0.9371,
       "step": 280
     },
     {
       "epoch": 0.125,
-      "grad_norm": 0.3309858739376068,
+      "grad_norm": 0.3305722177028656,
       "learning_rate": 0.00014452261306532665,
-      "loss": 0.9276,
+      "loss": 0.9265,
       "step": 281
     },
     {
       "epoch": 0.12544483985765126,
-      "grad_norm": 0.3567025363445282,
+      "grad_norm": 0.35687321424484253,
       "learning_rate": 0.000144321608040201,
-      "loss": 0.9924,
+      "loss": 0.9935,
       "step": 282
     },
     {
       "epoch": 0.1258896797153025,
-      "grad_norm": 0.35258856415748596,
+      "grad_norm": 0.3545040488243103,
       "learning_rate": 0.00014412060301507537,
-      "loss": 0.974,
+      "loss": 0.973,
       "step": 283
     },
     {
       "epoch": 0.12633451957295375,
-      "grad_norm": 0.33701589703559875,
+      "grad_norm": 0.3379552960395813,
       "learning_rate": 0.00014391959798994977,
-      "loss": 1.0021,
+      "loss": 1.0022,
       "step": 284
     },
     {
       "epoch": 0.12677935943060498,
-      "grad_norm": 0.3263789117336273,
+      "grad_norm": 0.32599905133247375,
       "learning_rate": 0.00014371859296482411,
-      "loss": 1.051,
+      "loss": 1.0497,
       "step": 285
     },
     {
       "epoch": 0.12722419928825623,
-      "grad_norm": 0.39523646235466003,
+      "grad_norm": 0.38973209261894226,
       "learning_rate": 0.0001435175879396985,
-      "loss": 0.9349,
+      "loss": 0.9373,
       "step": 286
     },
     {
       "epoch": 0.12766903914590746,
-      "grad_norm": 0.3561248183250427,
+      "grad_norm": 0.37509605288505554,
       "learning_rate": 0.00014331658291457286,
-      "loss": 0.869,
+      "loss": 0.8695,
       "step": 287
     },
     {
       "epoch": 0.12811387900355872,
-      "grad_norm": 0.43630316853523254,
+      "grad_norm": 0.33525556325912476,
       "learning_rate": 0.00014311557788944726,
-      "loss": 0.9773,
+      "loss": 0.9768,
       "step": 288
     },
     {
       "epoch": 0.12855871886120995,
-      "grad_norm": 0.3555675446987152,
+      "grad_norm": 0.3496398329734802,
       "learning_rate": 0.0001429145728643216,
-      "loss": 1.0048,
+      "loss": 1.0044,
       "step": 289
     },
     {
       "epoch": 0.1290035587188612,
-      "grad_norm": 0.40323683619499207,
+      "grad_norm": 0.39922475814819336,
       "learning_rate": 0.00014271356783919598,
-      "loss": 0.999,
+      "loss": 0.9998,
       "step": 290
     },
     {
       "epoch": 0.12944839857651247,
-      "grad_norm": 0.3344346284866333,
+      "grad_norm": 0.33336395025253296,
       "learning_rate": 0.00014251256281407035,
-      "loss": 0.9882,
+      "loss": 0.9871,
       "step": 291
     },
     {
       "epoch": 0.1298932384341637,
-      "grad_norm": 0.3018104135990143,
+      "grad_norm": 0.30318179726600647,
       "learning_rate": 0.00014231155778894473,
-      "loss": 0.9665,
+      "loss": 0.9652,
       "step": 292
     },
     {
       "epoch": 0.13033807829181496,
-      "grad_norm": 0.3336434066295624,
+      "grad_norm": 0.33315086364746094,
       "learning_rate": 0.0001421105527638191,
-      "loss": 0.9786,
+      "loss": 0.9793,
       "step": 293
     },
     {
       "epoch": 0.1307829181494662,
-      "grad_norm": 0.3364970088005066,
+      "grad_norm": 0.33998075127601624,
       "learning_rate": 0.00014190954773869347,
-      "loss": 0.9436,
+      "loss": 0.945,
       "step": 294
     },
     {
       "epoch": 0.13122775800711745,
-      "grad_norm": 0.3738161027431488,
+      "grad_norm": 0.3731788396835327,
       "learning_rate": 0.00014170854271356784,
-      "loss": 0.9884,
+      "loss": 0.9881,
       "step": 295
     },
     {
       "epoch": 0.13167259786476868,
-      "grad_norm": 0.34548112750053406,
+      "grad_norm": 0.34216803312301636,
       "learning_rate": 0.00014150753768844222,
-      "loss": 0.9578,
+      "loss": 0.9576,
       "step": 296
     },
     {
       "epoch": 0.13211743772241993,
-      "grad_norm": 0.4063248336315155,
+      "grad_norm": 0.4074036180973053,
       "learning_rate": 0.0001413065326633166,
-      "loss": 1.0649,
+      "loss": 1.0654,
       "step": 297
     },
     {
       "epoch": 0.13256227758007116,
-      "grad_norm": 0.4038452208042145,
+      "grad_norm": 0.401727557182312,
       "learning_rate": 0.00014110552763819096,
-      "loss": 0.9824,
+      "loss": 0.9813,
       "step": 298
     },
     {
       "epoch": 0.13300711743772242,
-      "grad_norm": 0.3010196089744568,
+      "grad_norm": 0.3025040626525879,
       "learning_rate": 0.00014090452261306534,
-      "loss": 0.9492,
+      "loss": 0.9491,
       "step": 299
     },
     {
       "epoch": 0.13345195729537365,
-      "grad_norm": 0.3451170325279236,
+      "grad_norm": 0.3456086814403534,
       "learning_rate": 0.0001407035175879397,
-      "loss": 1.1038,
+      "loss": 1.102,
       "step": 300
     },
     {
       "epoch": 0.1338967971530249,
-      "grad_norm": 0.3674055337905884,
+      "grad_norm": 0.36754533648490906,
       "learning_rate": 0.00014050251256281408,
-      "loss": 0.9764,
+      "loss": 0.9794,
       "step": 301
     },
     {
       "epoch": 0.13434163701067617,
-      "grad_norm": 0.28219959139823914,
+      "grad_norm": 0.2782064378261566,
       "learning_rate": 0.00014030150753768846,
-      "loss": 0.8152,
+      "loss": 0.8153,
       "step": 302
     },
     {
       "epoch": 0.1347864768683274,
-      "grad_norm": 0.3264709711074829,
+      "grad_norm": 0.35259565711021423,
       "learning_rate": 0.0001401005025125628,
-      "loss": 1.0336,
+      "loss": 1.032,
       "step": 303
     },
     {
       "epoch": 0.13523131672597866,
-      "grad_norm": 0.3442930579185486,
+      "grad_norm": 0.3422435224056244,
       "learning_rate": 0.0001398994974874372,
-      "loss": 0.9622,
+      "loss": 0.9601,
       "step": 304
     },
     {
       "epoch": 0.1356761565836299,
-      "grad_norm": 0.4365461468696594,
+      "grad_norm": 0.4378672242164612,
       "learning_rate": 0.00013969849246231157,
-      "loss": 0.9243,
+      "loss": 0.9241,
       "step": 305
     },
     {
       "epoch": 0.13612099644128114,
-      "grad_norm": 0.3317567706108093,
+      "grad_norm": 0.33044546842575073,
       "learning_rate": 0.00013949748743718595,
-      "loss": 0.9619,
+      "loss": 0.9607,
       "step": 306
     },
     {
       "epoch": 0.13656583629893237,
-      "grad_norm": 0.3600742816925049,
+      "grad_norm": 0.36180225014686584,
       "learning_rate": 0.0001392964824120603,
-      "loss": 1.0869,
+      "loss": 1.0878,
       "step": 307
     },
     {
       "epoch": 0.13701067615658363,
-      "grad_norm": 0.414174348115921,
+      "grad_norm": 0.4124760627746582,
       "learning_rate": 0.00013909547738693467,
-      "loss": 0.9696,
+      "loss": 0.9698,
       "step": 308
     },
     {
       "epoch": 0.13745551601423486,
-      "grad_norm": 0.3307357430458069,
+      "grad_norm": 0.33320683240890503,
       "learning_rate": 0.00013889447236180907,
-      "loss": 0.8673,
+      "loss": 0.869,
       "step": 309
     },
     {
       "epoch": 0.13790035587188612,
-      "grad_norm": 0.3574574291706085,
+      "grad_norm": 0.3595818877220154,
       "learning_rate": 0.0001386934673366834,
-      "loss": 1.0225,
+      "loss": 1.0227,
       "step": 310
     },
     {
       "epoch": 0.13834519572953738,
-      "grad_norm": 0.3453276455402374,
+      "grad_norm": 0.35221633315086365,
       "learning_rate": 0.00013849246231155778,
-      "loss": 0.956,
+      "loss": 0.9576,
       "step": 311
     },
     {
       "epoch": 0.1387900355871886,
-      "grad_norm": 0.3369787633419037,
+      "grad_norm": 0.3357265889644623,
       "learning_rate": 0.00013829145728643216,
-      "loss": 0.9878,
+      "loss": 0.9877,
       "step": 312
     },
     {
       "epoch": 0.13923487544483987,
-      "grad_norm": 0.36153116822242737,
+      "grad_norm": 0.3630138039588928,
       "learning_rate": 0.00013809045226130656,
-      "loss": 0.9906,
+      "loss": 0.9918,
       "step": 313
     },
     {
       "epoch": 0.1396797153024911,
-      "grad_norm": 0.3499455749988556,
+      "grad_norm": 0.3503361642360687,
       "learning_rate": 0.0001378894472361809,
-      "loss": 1.0581,
+      "loss": 1.0585,
       "step": 314
     },
     {
       "epoch": 0.14012455516014236,
-      "grad_norm": 0.3346167206764221,
+      "grad_norm": 0.33437854051589966,
       "learning_rate": 0.00013768844221105528,
-      "loss": 1.1294,
+      "loss": 1.1292,
       "step": 315
     },
     {
       "epoch": 0.14056939501779359,
-      "grad_norm": 0.3535563051700592,
+      "grad_norm": 0.3571662902832031,
       "learning_rate": 0.00013748743718592965,
-      "loss": 1.0081,
+      "loss": 1.0084,
       "step": 316
     },
     {
       "epoch": 0.14101423487544484,
-      "grad_norm": 0.3797767758369446,
+      "grad_norm": 0.3780195415019989,
       "learning_rate": 0.00013728643216080402,
-      "loss": 0.9821,
+      "loss": 0.98,
       "step": 317
     },
     {
       "epoch": 0.14145907473309607,
-      "grad_norm": 0.35535991191864014,
+      "grad_norm": 0.3566683232784271,
       "learning_rate": 0.0001370854271356784,
-      "loss": 0.9698,
+      "loss": 0.9694,
       "step": 318
     },
     {
       "epoch": 0.14190391459074733,
-      "grad_norm": 0.3628135323524475,
+      "grad_norm": 0.36311331391334534,
       "learning_rate": 0.00013688442211055277,
-      "loss": 0.8782,
+      "loss": 0.8768,
       "step": 319
     },
     {
       "epoch": 0.1423487544483986,
-      "grad_norm": 0.3347620666027069,
+      "grad_norm": 0.3234863877296448,
       "learning_rate": 0.00013668341708542714,
-      "loss": 0.9383,
+      "loss": 0.9384,
       "step": 320
     },
     {
       "epoch": 0.14279359430604982,
-      "grad_norm": 0.3387628197669983,
+      "grad_norm": 0.3360319137573242,
       "learning_rate": 0.00013648241206030151,
-      "loss": 0.9846,
+      "loss": 0.9831,
       "step": 321
     },
     {
       "epoch": 0.14323843416370108,
-      "grad_norm": 0.33139505982398987,
+      "grad_norm": 0.3335811197757721,
       "learning_rate": 0.0001362814070351759,
       "loss": 0.9541,
       "step": 322
     },
     {
       "epoch": 0.1436832740213523,
-      "grad_norm": 0.3996013104915619,
+      "grad_norm": 0.40138107538223267,
       "learning_rate": 0.00013608040201005026,
-      "loss": 0.9682,
+      "loss": 0.9647,
       "step": 323
     },
     {
       "epoch": 0.14412811387900357,
-      "grad_norm": 0.3609457314014435,
+      "grad_norm": 0.36284399032592773,
       "learning_rate": 0.00013587939698492463,
-      "loss": 0.9758,
+      "loss": 0.9763,
       "step": 324
     },
     {
       "epoch": 0.1445729537366548,
-      "grad_norm": 0.31257274746894836,
+      "grad_norm": 0.29980310797691345,
       "learning_rate": 0.000135678391959799,
-      "loss": 0.9793,
+      "loss": 0.9773,
       "step": 325
     },
     {
       "epoch": 0.14501779359430605,
-      "grad_norm": 0.3800847828388214,
+      "grad_norm": 0.37920883297920227,
       "learning_rate": 0.00013547738693467338,
-      "loss": 1.0033,
+      "loss": 1.0024,
       "step": 326
     },
     {
       "epoch": 0.14546263345195729,
-      "grad_norm": 0.32256045937538147,
+      "grad_norm": 0.32456889748573303,
       "learning_rate": 0.00013527638190954775,
-      "loss": 0.9944,
+      "loss": 0.995,
       "step": 327
     },
     {
       "epoch": 0.14590747330960854,
-      "grad_norm": 0.36587563157081604,
+      "grad_norm": 0.368257999420166,
       "learning_rate": 0.0001350753768844221,
-      "loss": 1.0223,
+      "loss": 1.0205,
       "step": 328
     },
     {
       "epoch": 0.14635231316725977,
-      "grad_norm": 0.37276628613471985,
+      "grad_norm": 0.37282678484916687,
       "learning_rate": 0.00013487437185929647,
-      "loss": 0.8559,
+      "loss": 0.8557,
       "step": 329
     },
     {
       "epoch": 0.14679715302491103,
-      "grad_norm": 0.3412066698074341,
+      "grad_norm": 0.34274178743362427,
       "learning_rate": 0.00013467336683417087,
-      "loss": 0.8826,
+      "loss": 0.8833,
       "step": 330
     },
     {
       "epoch": 0.1472419928825623,
-      "grad_norm": 0.33895769715309143,
+      "grad_norm": 0.34231486916542053,
       "learning_rate": 0.00013447236180904524,
-      "loss": 0.9517,
+      "loss": 0.9529,
       "step": 331
     },
     {
       "epoch": 0.14768683274021352,
-      "grad_norm": 0.3830057680606842,
+      "grad_norm": 0.38436999917030334,
       "learning_rate": 0.0001342713567839196,
-      "loss": 1.0307,
+      "loss": 1.029,
       "step": 332
     },
     {
       "epoch": 0.14813167259786478,
-      "grad_norm": 0.3927343487739563,
+      "grad_norm": 0.400931179523468,
       "learning_rate": 0.00013407035175879396,
-      "loss": 0.9916,
+      "loss": 0.9913,
       "step": 333
     },
     {
       "epoch": 0.148576512455516,
-      "grad_norm": 0.3220495283603668,
+      "grad_norm": 0.32551515102386475,
       "learning_rate": 0.00013386934673366836,
-      "loss": 0.9876,
+      "loss": 0.9879,
       "step": 334
     },
     {
       "epoch": 0.14902135231316727,
-      "grad_norm": 0.35559162497520447,
+      "grad_norm": 0.35669583082199097,
       "learning_rate": 0.0001336683417085427,
-      "loss": 1.0066,
+      "loss": 1.0058,
       "step": 335
     },
     {
       "epoch": 0.1494661921708185,
-      "grad_norm": 0.36672064661979675,
+      "grad_norm": 0.36875060200691223,
       "learning_rate": 0.00013346733668341708,
-      "loss": 0.9865,
+      "loss": 0.9858,
       "step": 336
     },
     {
       "epoch": 0.14991103202846975,
-      "grad_norm": 0.3622550964355469,
+      "grad_norm": 0.36278796195983887,
       "learning_rate": 0.00013326633165829146,
-      "loss": 1.0956,
+      "loss": 1.097,
       "step": 337
     },
     {
       "epoch": 0.15035587188612098,
-      "grad_norm": 0.3213232159614563,
+      "grad_norm": 0.31908801198005676,
       "learning_rate": 0.00013306532663316586,
-      "loss": 0.9552,
+      "loss": 0.955,
       "step": 338
     },
     {
       "epoch": 0.15080071174377224,
-      "grad_norm": 0.3889192044734955,
+      "grad_norm": 0.38954317569732666,
       "learning_rate": 0.0001328643216080402,
-      "loss": 0.9417,
+      "loss": 0.9402,
       "step": 339
     },
     {
       "epoch": 0.1512455516014235,
-      "grad_norm": 0.32667845487594604,
+      "grad_norm": 0.3285794258117676,
       "learning_rate": 0.00013266331658291457,
-      "loss": 0.9319,
+      "loss": 0.9327,
       "step": 340
     },
     {
       "epoch": 0.15169039145907473,
-      "grad_norm": 0.44496315717697144,
+      "grad_norm": 0.4542711675167084,
       "learning_rate": 0.00013246231155778895,
-      "loss": 0.7937,
+      "loss": 0.7939,
       "step": 341
     },
     {
       "epoch": 0.152135231316726,
-      "grad_norm": 0.371467649936676,
+      "grad_norm": 0.3730446696281433,
       "learning_rate": 0.00013226130653266332,
-      "loss": 0.9998,
+      "loss": 1.0014,
       "step": 342
     },
     {
       "epoch": 0.15258007117437722,
-      "grad_norm": 0.3509642779827118,
+      "grad_norm": 0.3543410301208496,
       "learning_rate": 0.0001320603015075377,
-      "loss": 0.9615,
+      "loss": 0.9617,
       "step": 343
     },
     {
       "epoch": 0.15302491103202848,
-      "grad_norm": 0.32483533024787903,
+      "grad_norm": 0.32625502347946167,
       "learning_rate": 0.00013185929648241207,
-      "loss": 0.9976,
+      "loss": 0.9975,
       "step": 344
     },
     {
       "epoch": 0.1534697508896797,
-      "grad_norm": 0.38299068808555603,
+      "grad_norm": 0.39967095851898193,
       "learning_rate": 0.00013165829145728644,
-      "loss": 0.8436,
+      "loss": 0.8421,
       "step": 345
     },
     {
       "epoch": 0.15391459074733096,
-      "grad_norm": 0.36412617564201355,
+      "grad_norm": 0.36405855417251587,
       "learning_rate": 0.0001314572864321608,
       "loss": 1.0338,
       "step": 346
     },
     {
       "epoch": 0.1543594306049822,
-      "grad_norm": 0.3295581638813019,
+      "grad_norm": 0.32427796721458435,
       "learning_rate": 0.00013125628140703518,
-      "loss": 0.9696,
+      "loss": 0.971,
       "step": 347
     },
     {
       "epoch": 0.15480427046263345,
-      "grad_norm": 0.34034737944602966,
+      "grad_norm": 0.3405332565307617,
       "learning_rate": 0.00013105527638190956,
-      "loss": 0.9977,
+      "loss": 0.9974,
       "step": 348
     },
     {
       "epoch": 0.1552491103202847,
-      "grad_norm": 0.3623081147670746,
+      "grad_norm": 0.3611313998699188,
       "learning_rate": 0.00013085427135678393,
-      "loss": 1.0032,
+      "loss": 1.0026,
       "step": 349
     },
     {
       "epoch": 0.15569395017793594,
-      "grad_norm": 0.3889707922935486,
+      "grad_norm": 0.3901672661304474,
       "learning_rate": 0.0001306532663316583,
-      "loss": 0.9245,
+      "loss": 0.924,
       "step": 350
     },
     {
       "epoch": 0.1561387900355872,
-      "grad_norm": 0.3887019455432892,
+      "grad_norm": 0.38940900564193726,
       "learning_rate": 0.00013045226130653268,
-      "loss": 1.0836,
+      "loss": 1.0827,
       "step": 351
     },
     {
       "epoch": 0.15658362989323843,
-      "grad_norm": 0.32099735736846924,
+      "grad_norm": 0.3226456046104431,
       "learning_rate": 0.00013025125628140705,
-      "loss": 0.907,
+      "loss": 0.9063,
       "step": 352
     },
     {
       "epoch": 0.1570284697508897,
-      "grad_norm": 0.29804012179374695,
+      "grad_norm": 0.2980092763900757,
       "learning_rate": 0.0001300502512562814,
-      "loss": 0.9691,
+      "loss": 0.9688,
       "step": 353
     },
     {
       "epoch": 0.15747330960854092,
-      "grad_norm": 0.3739759922027588,
+      "grad_norm": 0.3783577084541321,
       "learning_rate": 0.00012984924623115577,
-      "loss": 1.0826,
+      "loss": 1.0825,
       "step": 354
     },
     {
       "epoch": 0.15791814946619218,
-      "grad_norm": 0.3806516230106354,
+      "grad_norm": 0.3832230269908905,
       "learning_rate": 0.00012964824120603017,
-      "loss": 1.0638,
+      "loss": 1.0627,
       "step": 355
     },
     {
       "epoch": 0.1583629893238434,
-      "grad_norm": 0.3504475951194763,
+      "grad_norm": 0.34628695249557495,
       "learning_rate": 0.00012944723618090454,
-      "loss": 0.8802,
+      "loss": 0.8811,
       "step": 356
     },
     {
       "epoch": 0.15880782918149466,
-      "grad_norm": 0.4315776228904724,
+      "grad_norm": 0.43190130591392517,
       "learning_rate": 0.0001292462311557789,
-      "loss": 1.0159,
+      "loss": 1.0152,
       "step": 357
     },
     {
       "epoch": 0.1592526690391459,
-      "grad_norm": 0.3647462725639343,
+      "grad_norm": 0.36563345789909363,
       "learning_rate": 0.00012904522613065326,
-      "loss": 0.985,
+      "loss": 0.9848,
       "step": 358
     },
     {
       "epoch": 0.15969750889679715,
-      "grad_norm": 0.30111998319625854,
+      "grad_norm": 0.2992965281009674,
       "learning_rate": 0.00012884422110552766,
-      "loss": 0.951,
+      "loss": 0.9507,
       "step": 359
     },
     {
       "epoch": 0.1601423487544484,
-      "grad_norm": 0.3357127904891968,
+      "grad_norm": 0.3296932876110077,
       "learning_rate": 0.000128643216080402,
-      "loss": 1.1002,
+      "loss": 1.1,
       "step": 360
     },
     {
       "epoch": 0.16058718861209964,
-      "grad_norm": 0.4213799834251404,
+      "grad_norm": 0.41810277104377747,
       "learning_rate": 0.00012844221105527638,
-      "loss": 0.8264,
+      "loss": 0.8272,
       "step": 361
     },
     {
       "epoch": 0.1610320284697509,
-      "grad_norm": 0.4168066084384918,
+      "grad_norm": 0.41459906101226807,
       "learning_rate": 0.00012824120603015075,
-      "loss": 1.0548,
+      "loss": 1.0549,
       "step": 362
     },
     {
       "epoch": 0.16147686832740213,
-      "grad_norm": 0.36040782928466797,
+      "grad_norm": 0.35844656825065613,
       "learning_rate": 0.00012804020100502515,
-      "loss": 1.0988,
+      "loss": 1.0996,
       "step": 363
     },
     {
       "epoch": 0.1619217081850534,
-      "grad_norm": 0.315931499004364,
+      "grad_norm": 0.31563690304756165,
       "learning_rate": 0.0001278391959798995,
-      "loss": 0.8874,
+      "loss": 0.8873,
       "step": 364
     },
     {
       "epoch": 0.16236654804270462,
-      "grad_norm": 0.3526541292667389,
+      "grad_norm": 0.3521721065044403,
       "learning_rate": 0.00012763819095477387,
-      "loss": 0.8408,
+      "loss": 0.8421,
       "step": 365
     },
     {
       "epoch": 0.16281138790035588,
-      "grad_norm": 0.315824031829834,
+      "grad_norm": 0.3149266839027405,
       "learning_rate": 0.00012743718592964824,
-      "loss": 0.9537,
+      "loss": 0.9538,
       "step": 366
     },
     {
       "epoch": 0.1632562277580071,
-      "grad_norm": 0.36629655957221985,
+      "grad_norm": 0.36552533507347107,
       "learning_rate": 0.00012723618090452262,
-      "loss": 0.9732,
+      "loss": 0.9734,
       "step": 367
     },
     {
       "epoch": 0.16370106761565836,
-      "grad_norm": 0.3385011553764343,
+      "grad_norm": 0.33934077620506287,
       "learning_rate": 0.000127035175879397,
-      "loss": 1.0207,
+      "loss": 1.0199,
       "step": 368
     },
     {
       "epoch": 0.16414590747330962,
-      "grad_norm": 0.43948590755462646,
+      "grad_norm": 0.4329892098903656,
       "learning_rate": 0.00012683417085427136,
-      "loss": 0.8719,
+      "loss": 0.872,
       "step": 369
     },
     {
       "epoch": 0.16459074733096085,
-      "grad_norm": 0.4001463055610657,
+      "grad_norm": 0.4003223776817322,
       "learning_rate": 0.00012663316582914574,
-      "loss": 1.085,
+      "loss": 1.0864,
       "step": 370
     },
     {
       "epoch": 0.1650355871886121,
-      "grad_norm": 0.37441739439964294,
+      "grad_norm": 0.37618136405944824,
       "learning_rate": 0.0001264321608040201,
-      "loss": 0.9799,
+      "loss": 0.9797,
       "step": 371
     },
     {
       "epoch": 0.16548042704626334,
-      "grad_norm": 0.29792410135269165,
+      "grad_norm": 0.30325135588645935,
       "learning_rate": 0.00012623115577889448,
-      "loss": 0.8901,
+      "loss": 0.8896,
       "step": 372
     },
     {
       "epoch": 0.1659252669039146,
-      "grad_norm": 0.42688125371932983,
+      "grad_norm": 0.42738157510757446,
       "learning_rate": 0.00012603015075376885,
-      "loss": 1.085,
+      "loss": 1.0881,
       "step": 373
     },
     {
       "epoch": 0.16637010676156583,
-      "grad_norm": 0.3854696750640869,
+      "grad_norm": 0.38146787881851196,
       "learning_rate": 0.00012582914572864323,
-      "loss": 0.9271,
+      "loss": 0.9281,
       "step": 374
     },
     {
       "epoch": 0.16681494661921709,
-      "grad_norm": 0.40170320868492126,
+      "grad_norm": 0.40179941058158875,
       "learning_rate": 0.0001256281407035176,
-      "loss": 0.9355,
+      "loss": 0.9364,
       "step": 375
     },
     {
       "epoch": 0.16725978647686832,
-      "grad_norm": 0.3294379711151123,
+      "grad_norm": 0.33039966225624084,
       "learning_rate": 0.00012542713567839197,
-      "loss": 0.9221,
+      "loss": 0.9216,
       "step": 376
     },
     {
       "epoch": 0.16770462633451957,
-      "grad_norm": 0.3526048958301544,
+      "grad_norm": 0.3515557050704956,
       "learning_rate": 0.00012522613065326635,
-      "loss": 1.0483,
+      "loss": 1.0472,
       "step": 377
     },
     {
       "epoch": 0.16814946619217083,
-      "grad_norm": 0.3107386529445648,
+      "grad_norm": 0.3095014691352844,
       "learning_rate": 0.0001250251256281407,
-      "loss": 0.9166,
+      "loss": 0.9153,
       "step": 378
     },
     {
       "epoch": 0.16859430604982206,
-      "grad_norm": 0.4228864908218384,
+      "grad_norm": 0.42122480273246765,
       "learning_rate": 0.00012482412060301507,
-      "loss": 1.0207,
+      "loss": 1.0208,
       "step": 379
     },
     {
       "epoch": 0.16903914590747332,
-      "grad_norm": 0.3822677731513977,
+      "grad_norm": 0.3838384747505188,
       "learning_rate": 0.00012462311557788947,
-      "loss": 0.9935,
+      "loss": 0.9936,
       "step": 380
     },
     {
       "epoch": 0.16948398576512455,
-      "grad_norm": 0.3852001428604126,
+      "grad_norm": 0.38645121455192566,
       "learning_rate": 0.00012442211055276384,
-      "loss": 0.9831,
+      "loss": 0.983,
       "step": 381
     },
     {
       "epoch": 0.1699288256227758,
-      "grad_norm": 0.3484053313732147,
+      "grad_norm": 0.34789687395095825,
       "learning_rate": 0.00012422110552763818,
-      "loss": 0.9327,
+      "loss": 0.9341,
       "step": 382
     },
     {
       "epoch": 0.17037366548042704,
-      "grad_norm": 0.3499116897583008,
+      "grad_norm": 0.3516675531864166,
       "learning_rate": 0.00012402010050251256,
-      "loss": 1.0521,
+      "loss": 1.052,
       "step": 383
     },
     {
       "epoch": 0.1708185053380783,
-      "grad_norm": 0.362118661403656,
+      "grad_norm": 0.36350032687187195,
       "learning_rate": 0.00012381909547738696,
-      "loss": 1.0482,
+      "loss": 1.047,
       "step": 384
     },
     {
       "epoch": 0.17126334519572953,
-      "grad_norm": 0.3913699686527252,
+      "grad_norm": 0.3919387757778168,
       "learning_rate": 0.0001236180904522613,
-      "loss": 0.9908,
+      "loss": 0.9899,
       "step": 385
     },
     {
       "epoch": 0.17170818505338079,
-      "grad_norm": 0.3362598121166229,
+      "grad_norm": 0.3369230628013611,
       "learning_rate": 0.00012341708542713568,
-      "loss": 0.9814,
+      "loss": 0.981,
       "step": 386
     },
     {
       "epoch": 0.17215302491103202,
-      "grad_norm": 0.34024205803871155,
+      "grad_norm": 0.34190815687179565,
       "learning_rate": 0.00012321608040201005,
-      "loss": 0.9747,
+      "loss": 0.9727,
       "step": 387
     },
     {
       "epoch": 0.17259786476868327,
-      "grad_norm": 0.35168662667274475,
+      "grad_norm": 0.3579177260398865,
       "learning_rate": 0.00012301507537688445,
-      "loss": 1.0067,
+      "loss": 1.0071,
       "step": 388
     },
     {
       "epoch": 0.17304270462633453,
-      "grad_norm": 0.30861225724220276,
+      "grad_norm": 0.307522714138031,
       "learning_rate": 0.0001228140703517588,
-      "loss": 1.06,
+      "loss": 1.0614,
       "step": 389
     },
     {
       "epoch": 0.17348754448398576,
-      "grad_norm": 0.3759268522262573,
+      "grad_norm": 0.3733164966106415,
       "learning_rate": 0.00012261306532663317,
-      "loss": 1.1586,
+      "loss": 1.1578,
       "step": 390
     },
     {
       "epoch": 0.17393238434163702,
-      "grad_norm": 0.3894938826560974,
+      "grad_norm": 0.3878190517425537,
       "learning_rate": 0.00012241206030150754,
-      "loss": 0.9451,
+      "loss": 0.9454,
       "step": 391
     },
     {
       "epoch": 0.17437722419928825,
-      "grad_norm": 0.3796166777610779,
+      "grad_norm": 0.37619033455848694,
       "learning_rate": 0.00012221105527638191,
-      "loss": 0.9602,
+      "loss": 0.9608,
       "step": 392
     },
     {
       "epoch": 0.1748220640569395,
-      "grad_norm": 0.32060232758522034,
+      "grad_norm": 0.3204808533191681,
       "learning_rate": 0.00012201005025125629,
-      "loss": 1.0365,
+      "loss": 1.0344,
       "step": 393
     },
     {
       "epoch": 0.17526690391459074,
-      "grad_norm": 0.33367520570755005,
+      "grad_norm": 0.3382740616798401,
       "learning_rate": 0.00012180904522613066,
-      "loss": 0.9616,
+      "loss": 0.9624,
       "step": 394
     },
     {
       "epoch": 0.175711743772242,
-      "grad_norm": 0.3084149658679962,
+      "grad_norm": 0.3102628290653229,
       "learning_rate": 0.00012160804020100502,
-      "loss": 0.8514,
+      "loss": 0.8531,
       "step": 395
     },
     {
       "epoch": 0.17615658362989323,
-      "grad_norm": 0.32909664511680603,
+      "grad_norm": 0.32901301980018616,
       "learning_rate": 0.00012140703517587942,
       "loss": 0.9773,
       "step": 396
     },
     {
       "epoch": 0.17660142348754448,
-      "grad_norm": 0.3140466511249542,
+      "grad_norm": 0.3148498833179474,
       "learning_rate": 0.00012120603015075378,
-      "loss": 0.9808,
+      "loss": 0.9834,
       "step": 397
     },
     {
       "epoch": 0.17704626334519574,
-      "grad_norm": 0.34496647119522095,
+      "grad_norm": 0.34393250942230225,
       "learning_rate": 0.00012100502512562815,
-      "loss": 0.8233,
+      "loss": 0.8226,
       "step": 398
     },
     {
       "epoch": 0.17749110320284697,
-      "grad_norm": 0.4260810315608978,
+      "grad_norm": 0.415526807308197,
       "learning_rate": 0.00012080402010050251,
-      "loss": 0.9703,
+      "loss": 0.9692,
       "step": 399
     },
     {
       "epoch": 0.17793594306049823,
-      "grad_norm": 0.3465157747268677,
+      "grad_norm": 0.34324660897254944,
       "learning_rate": 0.00012060301507537688,
-      "loss": 1.0121,
+      "loss": 1.0135,
       "step": 400
     },
     {
       "epoch": 0.17838078291814946,
-      "grad_norm": 0.3023923635482788,
+      "grad_norm": 0.3012760579586029,
       "learning_rate": 0.00012040201005025127,
-      "loss": 0.9899,
+      "loss": 0.9889,
       "step": 401
     },
     {
       "epoch": 0.17882562277580072,
-      "grad_norm": 0.3486076593399048,
+      "grad_norm": 0.3476797342300415,
       "learning_rate": 0.00012020100502512563,
-      "loss": 0.9325,
+      "loss": 0.9328,
       "step": 402
     },
     {
       "epoch": 0.17927046263345195,
-      "grad_norm": 0.3666176199913025,
+      "grad_norm": 0.36807379126548767,
       "learning_rate": 0.00012,
-      "loss": 1.019,
+      "loss": 1.0193,
       "step": 403
     },
     {
       "epoch": 0.1797153024911032,
-      "grad_norm": 0.3321322798728943,
+      "grad_norm": 0.3321215808391571,
       "learning_rate": 0.00011979899497487436,
       "loss": 0.9197,
       "step": 404
     },
     {
       "epoch": 0.18016014234875444,
-      "grad_norm": 0.4836527407169342,
+      "grad_norm": 0.48590320348739624,
       "learning_rate": 0.00011959798994974876,
-      "loss": 0.8488,
+      "loss": 0.8509,
       "step": 405
     },
     {
       "epoch": 0.1806049822064057,
-      "grad_norm": 0.3428821861743927,
+      "grad_norm": 0.3456606864929199,
       "learning_rate": 0.00011939698492462312,
-      "loss": 1.0314,
+      "loss": 1.033,
       "step": 406
     },
     {
       "epoch": 0.18104982206405695,
-      "grad_norm": 0.38532841205596924,
+      "grad_norm": 0.3836311101913452,
       "learning_rate": 0.0001191959798994975,
-      "loss": 0.9393,
+      "loss": 0.9396,
       "step": 407
     },
     {
       "epoch": 0.18149466192170818,
-      "grad_norm": 0.3802700340747833,
+      "grad_norm": 0.3805595338344574,
       "learning_rate": 0.00011899497487437185,
-      "loss": 0.9136,
+      "loss": 0.9135,
       "step": 408
     },
     {
       "epoch": 0.18193950177935944,
-      "grad_norm": 0.3403851389884949,
+      "grad_norm": 0.34188127517700195,
       "learning_rate": 0.00011879396984924624,
-      "loss": 0.9552,
+      "loss": 0.9551,
       "step": 409
     },
     {
       "epoch": 0.18238434163701067,
-      "grad_norm": 0.3999098241329193,
+      "grad_norm": 0.40114229917526245,
       "learning_rate": 0.00011859296482412061,
-      "loss": 0.9292,
+      "loss": 0.93,
       "step": 410
     },
     {
       "epoch": 0.18282918149466193,
-      "grad_norm": 0.3740805685520172,
+      "grad_norm": 0.3679947555065155,
       "learning_rate": 0.00011839195979899497,
-      "loss": 1.0809,
+      "loss": 1.0823,
       "step": 411
     },
     {
       "epoch": 0.18327402135231316,
-      "grad_norm": 0.42133694887161255,
+      "grad_norm": 0.4190071225166321,
       "learning_rate": 0.00011819095477386935,
-      "loss": 0.8661,
+      "loss": 0.8648,
       "step": 412
     },
     {
       "epoch": 0.18371886120996442,
-      "grad_norm": 0.39346155524253845,
+      "grad_norm": 0.39405253529548645,
       "learning_rate": 0.00011798994974874373,
       "loss": 1.0682,
       "step": 413
     },
     {
       "epoch": 0.18416370106761565,
-      "grad_norm": 0.3395063281059265,
+      "grad_norm": 0.33679550886154175,
       "learning_rate": 0.0001177889447236181,
-      "loss": 0.9672,
+      "loss": 0.9674,
       "step": 414
     },
     {
       "epoch": 0.1846085409252669,
-      "grad_norm": 0.3930981755256653,
+      "grad_norm": 0.3880975842475891,
       "learning_rate": 0.00011758793969849247,
-      "loss": 0.9354,
+      "loss": 0.9367,
       "step": 415
     },
     {
       "epoch": 0.18505338078291814,
-      "grad_norm": 0.40588700771331787,
+      "grad_norm": 0.4066067337989807,
       "learning_rate": 0.00011738693467336684,
-      "loss": 0.9723,
+      "loss": 0.9712,
       "step": 416
     },
     {
       "epoch": 0.1854982206405694,
-      "grad_norm": 0.4842967689037323,
+      "grad_norm": 0.48199501633644104,
       "learning_rate": 0.00011718592964824122,
-      "loss": 0.9098,
+      "loss": 0.9133,
       "step": 417
     },
     {
       "epoch": 0.18594306049822065,
-      "grad_norm": 0.34736424684524536,
+      "grad_norm": 0.34463343024253845,
       "learning_rate": 0.00011698492462311558,
-      "loss": 0.9383,
+      "loss": 0.9384,
       "step": 418
     },
     {
       "epoch": 0.18638790035587188,
-      "grad_norm": 0.41721245646476746,
+      "grad_norm": 0.4196937680244446,
       "learning_rate": 0.00011678391959798996,
-      "loss": 0.9804,
+      "loss": 0.9803,
       "step": 419
     },
     {
       "epoch": 0.18683274021352314,
-      "grad_norm": 0.320420503616333,
+      "grad_norm": 0.3185271620750427,
       "learning_rate": 0.00011658291457286432,
-      "loss": 0.9266,
+      "loss": 0.9267,
       "step": 420
     },
     {
       "epoch": 0.18727758007117437,
-      "grad_norm": 0.3732924461364746,
+      "grad_norm": 0.36499395966529846,
       "learning_rate": 0.00011638190954773872,
-      "loss": 0.9703,
+      "loss": 0.9721,
       "step": 421
     },
     {
       "epoch": 0.18772241992882563,
-      "grad_norm": 0.3584558069705963,
+      "grad_norm": 0.3580315411090851,
       "learning_rate": 0.00011618090452261308,
-      "loss": 0.8832,
+      "loss": 0.8834,
       "step": 422
     },
     {
       "epoch": 0.18816725978647686,
-      "grad_norm": 0.3892074525356293,
+      "grad_norm": 0.38039615750312805,
       "learning_rate": 0.00011597989949748745,
-      "loss": 0.9617,
+      "loss": 0.9638,
       "step": 423
     },
     {
       "epoch": 0.18861209964412812,
-      "grad_norm": 0.3820159435272217,
+      "grad_norm": 0.38077226281166077,
       "learning_rate": 0.00011577889447236181,
-      "loss": 0.9092,
+      "loss": 0.9077,
       "step": 424
     },
     {
       "epoch": 0.18905693950177935,
-      "grad_norm": 0.33480194211006165,
+      "grad_norm": 0.33230000734329224,
       "learning_rate": 0.00011557788944723618,
-      "loss": 1.0104,
+      "loss": 1.0115,
       "step": 425
     },
     {
       "epoch": 0.1895017793594306,
-      "grad_norm": 0.41536465287208557,
+      "grad_norm": 0.41263043880462646,
       "learning_rate": 0.00011537688442211057,
-      "loss": 0.9157,
+      "loss": 0.9162,
       "step": 426
     },
     {
       "epoch": 0.18994661921708186,
-      "grad_norm": 0.3305935263633728,
+      "grad_norm": 0.32407188415527344,
       "learning_rate": 0.00011517587939698493,
-      "loss": 0.9622,
+      "loss": 0.9618,
       "step": 427
     },
     {
       "epoch": 0.1903914590747331,
-      "grad_norm": 0.4637777805328369,
+      "grad_norm": 0.4661211669445038,
       "learning_rate": 0.0001149748743718593,
-      "loss": 0.9317,
+      "loss": 0.931,
       "step": 428
     },
     {
       "epoch": 0.19083629893238435,
-      "grad_norm": 0.38072511553764343,
+      "grad_norm": 0.37407800555229187,
       "learning_rate": 0.00011477386934673366,
-      "loss": 1.0061,
+      "loss": 1.0046,
       "step": 429
     },
     {
       "epoch": 0.19128113879003558,
-      "grad_norm": 0.35834434628486633,
+      "grad_norm": 0.3580109477043152,
       "learning_rate": 0.00011457286432160806,
-      "loss": 1.0108,
+      "loss": 1.0114,
       "step": 430
     },
     {
       "epoch": 0.19172597864768684,
-      "grad_norm": 0.4265679717063904,
+      "grad_norm": 0.42048606276512146,
       "learning_rate": 0.00011437185929648242,
-      "loss": 0.9451,
+      "loss": 0.9468,
       "step": 431
     },
     {
       "epoch": 0.19217081850533807,
-      "grad_norm": 0.38988345861434937,
+      "grad_norm": 0.38913312554359436,
       "learning_rate": 0.00011417085427135679,
-      "loss": 1.1026,
+      "loss": 1.1037,
       "step": 432
     },
     {
       "epoch": 0.19261565836298933,
-      "grad_norm": 0.35931700468063354,
+      "grad_norm": 0.35675281286239624,
       "learning_rate": 0.00011396984924623115,
-      "loss": 0.9296,
+      "loss": 0.9292,
       "step": 433
     },
     {
       "epoch": 0.19306049822064056,
-      "grad_norm": 0.3506704568862915,
+      "grad_norm": 0.3579562306404114,
       "learning_rate": 0.00011376884422110554,
-      "loss": 0.942,
+      "loss": 0.9413,
       "step": 434
     },
     {
       "epoch": 0.19350533807829182,
-      "grad_norm": 0.38077566027641296,
+      "grad_norm": 0.3751574456691742,
       "learning_rate": 0.00011356783919597991,
-      "loss": 1.0012,
+      "loss": 1.0018,
       "step": 435
     },
     {
       "epoch": 0.19395017793594305,
-      "grad_norm": 0.3146056532859802,
+      "grad_norm": 0.3086400032043457,
       "learning_rate": 0.00011336683417085427,
-      "loss": 0.9361,
+      "loss": 0.9355,
       "step": 436
     },
     {
       "epoch": 0.1943950177935943,
-      "grad_norm": 0.42160287499427795,
+      "grad_norm": 0.41924887895584106,
       "learning_rate": 0.00011316582914572864,
-      "loss": 0.9426,
+      "loss": 0.942,
       "step": 437
     },
     {
       "epoch": 0.19483985765124556,
-      "grad_norm": 0.33514153957366943,
+      "grad_norm": 0.3349624276161194,
       "learning_rate": 0.00011296482412060303,
-      "loss": 1.0237,
+      "loss": 1.0232,
       "step": 438
     },
     {
       "epoch": 0.1952846975088968,
-      "grad_norm": 0.46194738149642944,
+      "grad_norm": 0.4605553448200226,
       "learning_rate": 0.0001127638190954774,
-      "loss": 1.095,
+      "loss": 1.0958,
       "step": 439
     },
     {
       "epoch": 0.19572953736654805,
-      "grad_norm": 0.40086501836776733,
+      "grad_norm": 0.3877868950366974,
       "learning_rate": 0.00011256281407035176,
-      "loss": 0.9239,
+      "loss": 0.9227,
       "step": 440
     },
     {
       "epoch": 0.19617437722419928,
-      "grad_norm": 0.34976860880851746,
+      "grad_norm": 0.3474135100841522,
       "learning_rate": 0.00011236180904522614,
-      "loss": 0.8445,
+      "loss": 0.8446,
       "step": 441
     },
     {
       "epoch": 0.19661921708185054,
-      "grad_norm": 0.35818928480148315,
+      "grad_norm": 0.3451891541481018,
       "learning_rate": 0.00011216080402010052,
-      "loss": 0.9626,
+      "loss": 0.9641,
       "step": 442
     },
     {
       "epoch": 0.19706405693950177,
-      "grad_norm": 0.35546717047691345,
+      "grad_norm": 0.3565372824668884,
       "learning_rate": 0.00011195979899497488,
-      "loss": 0.981,
+      "loss": 0.9807,
       "step": 443
     },
     {
       "epoch": 0.19750889679715303,
-      "grad_norm": 0.32887890934944153,
+      "grad_norm": 0.3283828794956207,
       "learning_rate": 0.00011175879396984925,
-      "loss": 0.9953,
+      "loss": 0.9949,
       "step": 444
     },
     {
       "epoch": 0.19795373665480426,
-      "grad_norm": 0.3613591492176056,
+      "grad_norm": 0.3562460243701935,
       "learning_rate": 0.00011155778894472361,
-      "loss": 0.9798,
+      "loss": 0.9807,
       "step": 445
     },
     {
       "epoch": 0.19839857651245552,
-      "grad_norm": 0.3338494896888733,
+      "grad_norm": 0.33841797709465027,
       "learning_rate": 0.00011135678391959799,
-      "loss": 1.0195,
+      "loss": 1.0213,
       "step": 446
     },
     {
       "epoch": 0.19884341637010677,
-      "grad_norm": 0.3248537480831146,
+      "grad_norm": 0.3265354633331299,
       "learning_rate": 0.00011115577889447237,
-      "loss": 0.9145,
+      "loss": 0.9138,
       "step": 447
     },
     {
       "epoch": 0.199288256227758,
-      "grad_norm": 0.35757869482040405,
+      "grad_norm": 0.35441333055496216,
       "learning_rate": 0.00011095477386934675,
-      "loss": 0.898,
+      "loss": 0.8982,
       "step": 448
     },
     {
       "epoch": 0.19973309608540926,
-      "grad_norm": 0.35583075881004333,
+      "grad_norm": 0.3564106822013855,
       "learning_rate": 0.0001107537688442211,
-      "loss": 1.0303,
+      "loss": 1.0298,
       "step": 449
     },
     {
       "epoch": 0.2001779359430605,
-      "grad_norm": 0.5167235136032104,
+      "grad_norm": 0.510447084903717,
       "learning_rate": 0.00011055276381909548,
-      "loss": 0.8168,
+      "loss": 0.818,
       "step": 450
     },
     {
       "epoch": 0.20062277580071175,
-      "grad_norm": 0.3596641719341278,
+      "grad_norm": 0.3568657338619232,
       "learning_rate": 0.00011035175879396986,
-      "loss": 0.9184,
+      "loss": 0.9187,
       "step": 451
     },
     {
       "epoch": 0.20106761565836298,
-      "grad_norm": 0.4046080410480499,
+      "grad_norm": 0.40208712220191956,
       "learning_rate": 0.00011015075376884422,
-      "loss": 1.0039,
+      "loss": 1.0041,
       "step": 452
     },
     {
       "epoch": 0.20151245551601424,
-      "grad_norm": 0.36985480785369873,
+      "grad_norm": 0.36585667729377747,
       "learning_rate": 0.0001099497487437186,
-      "loss": 0.9409,
+      "loss": 0.9413,
       "step": 453
     },
     {
       "epoch": 0.20195729537366547,
-      "grad_norm": 0.36848029494285583,
+      "grad_norm": 0.36290839314460754,
       "learning_rate": 0.00010974874371859296,
-      "loss": 1.0075,
+      "loss": 1.0084,
       "step": 454
     },
     {
       "epoch": 0.20240213523131673,
-      "grad_norm": 0.3421315550804138,
+      "grad_norm": 0.34766581654548645,
       "learning_rate": 0.00010954773869346736,
       "loss": 1.0761,
       "step": 455
     },
     {
       "epoch": 0.20284697508896798,
-      "grad_norm": 0.35291051864624023,
+      "grad_norm": 0.3324412703514099,
       "learning_rate": 0.00010934673366834172,
-      "loss": 0.9096,
+      "loss": 0.9078,
       "step": 456
     },
     {
       "epoch": 0.20329181494661921,
-      "grad_norm": 0.3597058355808258,
+      "grad_norm": 0.35148242115974426,
       "learning_rate": 0.00010914572864321609,
-      "loss": 0.9133,
+      "loss": 0.9148,
       "step": 457
     },
     {
       "epoch": 0.20373665480427047,
-      "grad_norm": 0.3638782501220703,
+      "grad_norm": 0.36268171668052673,
       "learning_rate": 0.00010894472361809045,
-      "loss": 0.9936,
+      "loss": 0.9915,
       "step": 458
     },
     {
       "epoch": 0.2041814946619217,
-      "grad_norm": 0.33646896481513977,
+      "grad_norm": 0.3343249261379242,
       "learning_rate": 0.00010874371859296483,
-      "loss": 0.9592,
+      "loss": 0.9596,
       "step": 459
     },
     {
       "epoch": 0.20462633451957296,
-      "grad_norm": 0.3791368901729584,
+      "grad_norm": 0.3835364878177643,
       "learning_rate": 0.00010854271356783921,
-      "loss": 0.9708,
+      "loss": 0.9718,
       "step": 460
     },
     {
       "epoch": 0.2050711743772242,
-      "grad_norm": 0.3495243489742279,
+      "grad_norm": 0.3455188274383545,
       "learning_rate": 0.00010834170854271357,
-      "loss": 0.9908,
+      "loss": 0.9918,
       "step": 461
     },
     {
       "epoch": 0.20551601423487545,
-      "grad_norm": 0.34329208731651306,
+      "grad_norm": 0.3390166759490967,
       "learning_rate": 0.00010814070351758794,
-      "loss": 0.9722,
+      "loss": 0.9718,
       "step": 462
     },
     {
       "epoch": 0.20596085409252668,
-      "grad_norm": 0.40529054403305054,
+      "grad_norm": 0.40748950839042664,
       "learning_rate": 0.00010793969849246233,
-      "loss": 0.9758,
+      "loss": 0.9766,
       "step": 463
     },
     {
       "epoch": 0.20640569395017794,
-      "grad_norm": 0.44128406047821045,
+      "grad_norm": 0.41326215863227844,
       "learning_rate": 0.0001077386934673367,
-      "loss": 0.9661,
+      "loss": 0.9639,
       "step": 464
     },
     {
       "epoch": 0.20685053380782917,
-      "grad_norm": 0.3696465790271759,
+      "grad_norm": 0.36672449111938477,
       "learning_rate": 0.00010753768844221106,
-      "loss": 1.0031,
+      "loss": 1.0027,
       "step": 465
     },
     {
       "epoch": 0.20729537366548043,
-      "grad_norm": 0.36873361468315125,
+      "grad_norm": 0.3668520152568817,
       "learning_rate": 0.00010733668341708543,
-      "loss": 0.9141,
+      "loss": 0.9147,
       "step": 466
     },
     {
       "epoch": 0.20774021352313168,
-      "grad_norm": 0.40861397981643677,
+      "grad_norm": 0.40608394145965576,
       "learning_rate": 0.00010713567839195982,
-      "loss": 0.9727,
+      "loss": 0.973,
       "step": 467
     },
     {
       "epoch": 0.20818505338078291,
-      "grad_norm": 0.3468811511993408,
+      "grad_norm": 0.3467264175415039,
       "learning_rate": 0.00010693467336683418,
       "loss": 1.0182,
       "step": 468
     },
     {
       "epoch": 0.20862989323843417,
-      "grad_norm": 0.37425243854522705,
+      "grad_norm": 0.3724631071090698,
       "learning_rate": 0.00010673366834170855,
-      "loss": 0.9268,
+      "loss": 0.9269,
       "step": 469
     },
     {
       "epoch": 0.2090747330960854,
-      "grad_norm": 0.4296034574508667,
+      "grad_norm": 0.4459652006626129,
       "learning_rate": 0.00010653266331658291,
-      "loss": 1.116,
+      "loss": 1.115,
       "step": 470
     },
     {
       "epoch": 0.20951957295373666,
-      "grad_norm": 0.3852652609348297,
+      "grad_norm": 0.3838179111480713,
       "learning_rate": 0.00010633165829145728,
-      "loss": 0.9378,
+      "loss": 0.9367,
       "step": 471
     },
     {
       "epoch": 0.2099644128113879,
-      "grad_norm": 0.34109607338905334,
+      "grad_norm": 0.3366687595844269,
       "learning_rate": 0.00010613065326633167,
-      "loss": 1.004,
+      "loss": 1.0043,
       "step": 472
     },
     {
       "epoch": 0.21040925266903915,
-      "grad_norm": 0.4025616943836212,
+      "grad_norm": 0.39886242151260376,
       "learning_rate": 0.00010592964824120604,
-      "loss": 0.96,
+      "loss": 0.9596,
       "step": 473
     },
     {
       "epoch": 0.21085409252669038,
-      "grad_norm": 0.3672776222229004,
+      "grad_norm": 0.36448466777801514,
       "learning_rate": 0.0001057286432160804,
-      "loss": 0.8744,
+      "loss": 0.874,
       "step": 474
     },
     {
       "epoch": 0.21129893238434164,
-      "grad_norm": 0.36405524611473083,
+      "grad_norm": 0.36318713426589966,
       "learning_rate": 0.00010552763819095478,
-      "loss": 0.8957,
+      "loss": 0.8955,
       "step": 475
     },
     {
       "epoch": 0.2117437722419929,
-      "grad_norm": 0.41747644543647766,
+      "grad_norm": 0.41962435841560364,
       "learning_rate": 0.00010532663316582916,
-      "loss": 1.0165,
+      "loss": 1.0139,
       "step": 476
     },
     {
       "epoch": 0.21218861209964412,
-      "grad_norm": 0.3478085994720459,
+      "grad_norm": 0.34656140208244324,
       "learning_rate": 0.00010512562814070352,
-      "loss": 0.9776,
+      "loss": 0.9768,
       "step": 477
     },
     {
       "epoch": 0.21263345195729538,
-      "grad_norm": 0.365633487701416,
+      "grad_norm": 0.36636775732040405,
       "learning_rate": 0.0001049246231155779,
       "loss": 0.9826,
       "step": 478
     },
     {
       "epoch": 0.2130782918149466,
-      "grad_norm": 0.3211476802825928,
+      "grad_norm": 0.3224871754646301,
       "learning_rate": 0.00010472361809045225,
-      "loss": 0.9228,
+      "loss": 0.9229,
       "step": 479
     },
     {
       "epoch": 0.21352313167259787,
-      "grad_norm": 0.3374654948711395,
+      "grad_norm": 0.33149847388267517,
       "learning_rate": 0.00010452261306532664,
-      "loss": 0.9379,
+      "loss": 0.9374,
       "step": 480
     },
     {
       "epoch": 0.2139679715302491,
-      "grad_norm": 0.34640708565711975,
+      "grad_norm": 0.356975793838501,
       "learning_rate": 0.00010432160804020101,
-      "loss": 0.8678,
+      "loss": 0.8682,
       "step": 481
     },
     {
       "epoch": 0.21441281138790036,
-      "grad_norm": 0.38176846504211426,
+      "grad_norm": 0.3753889799118042,
       "learning_rate": 0.00010412060301507539,
-      "loss": 0.9597,
+      "loss": 0.961,
       "step": 482
     },
     {
       "epoch": 0.2148576512455516,
-      "grad_norm": 0.34589311480522156,
+      "grad_norm": 0.3479616641998291,
       "learning_rate": 0.00010391959798994975,
       "loss": 0.9423,
       "step": 483
     },
     {
       "epoch": 0.21530249110320285,
-      "grad_norm": 0.4062221348285675,
+      "grad_norm": 0.4025513529777527,
       "learning_rate": 0.00010371859296482413,
-      "loss": 0.9926,
+      "loss": 0.9917,
       "step": 484
     },
     {
       "epoch": 0.2157473309608541,
-      "grad_norm": 0.3403069078922272,
+      "grad_norm": 0.3388558328151703,
       "learning_rate": 0.0001035175879396985,
-      "loss": 0.9953,
+      "loss": 0.994,
       "step": 485
     },
     {
       "epoch": 0.21619217081850534,
-      "grad_norm": 0.3341009318828583,
+      "grad_norm": 0.3347098231315613,
       "learning_rate": 0.00010331658291457286,
-      "loss": 0.9807,
+      "loss": 0.9796,
       "step": 486
     },
     {
       "epoch": 0.2166370106761566,
-      "grad_norm": 0.29718342423439026,
+      "grad_norm": 0.29469332098960876,
       "learning_rate": 0.00010311557788944724,
       "loss": 0.9296,
       "step": 487
     },
     {
       "epoch": 0.21708185053380782,
-      "grad_norm": 0.4219815135002136,
+      "grad_norm": 0.4186575412750244,
       "learning_rate": 0.00010291457286432162,
-      "loss": 1.0007,
+      "loss": 1.0021,
       "step": 488
     },
     {
       "epoch": 0.21752669039145908,
-      "grad_norm": 0.30951768159866333,
+      "grad_norm": 0.3087356388568878,
       "learning_rate": 0.00010271356783919598,
-      "loss": 0.9686,
+      "loss": 0.968,
       "step": 489
     },
     {
       "epoch": 0.2179715302491103,
-      "grad_norm": 0.38819047808647156,
+      "grad_norm": 0.3883945643901825,
       "learning_rate": 0.00010251256281407036,
-      "loss": 1.1859,
+      "loss": 1.1866,
       "step": 490
     },
     {
       "epoch": 0.21841637010676157,
-      "grad_norm": 0.3185841739177704,
+      "grad_norm": 0.31850650906562805,
       "learning_rate": 0.00010231155778894473,
-      "loss": 0.9523,
+      "loss": 0.9544,
       "step": 491
     },
     {
       "epoch": 0.2188612099644128,
-      "grad_norm": 0.3971594274044037,
+      "grad_norm": 0.40497350692749023,
       "learning_rate": 0.00010211055276381909,
-      "loss": 1.0133,
+      "loss": 1.0134,
       "step": 492
     },
     {
       "epoch": 0.21930604982206406,
-      "grad_norm": 0.31492552161216736,
+      "grad_norm": 0.31457439064979553,
       "learning_rate": 0.00010190954773869348,
-      "loss": 0.8269,
+      "loss": 0.8261,
       "step": 493
     },
     {
       "epoch": 0.2197508896797153,
-      "grad_norm": 0.3949122130870819,
+      "grad_norm": 0.398622065782547,
       "learning_rate": 0.00010170854271356785,
-      "loss": 1.0379,
+      "loss": 1.0384,
       "step": 494
     },
     {
       "epoch": 0.22019572953736655,
-      "grad_norm": 0.30094975233078003,
+      "grad_norm": 0.30925434827804565,
       "learning_rate": 0.00010150753768844221,
-      "loss": 0.9071,
+      "loss": 0.9075,
       "step": 495
     },
     {
       "epoch": 0.2206405693950178,
-      "grad_norm": 0.3560626208782196,
+      "grad_norm": 0.3536154329776764,
       "learning_rate": 0.00010130653266331658,
-      "loss": 0.9853,
+      "loss": 0.9856,
       "step": 496
     },
     {
       "epoch": 0.22108540925266904,
-      "grad_norm": 0.4022250771522522,
+      "grad_norm": 0.3990980386734009,
       "learning_rate": 0.00010110552763819097,
-      "loss": 0.9823,
+      "loss": 0.9822,
       "step": 497
     },
     {
       "epoch": 0.2215302491103203,
-      "grad_norm": 0.3849669098854065,
+      "grad_norm": 0.3845369517803192,
       "learning_rate": 0.00010090452261306533,
-      "loss": 0.9176,
+      "loss": 0.9191,
       "step": 498
     },
     {
       "epoch": 0.22197508896797152,
-      "grad_norm": 0.47436994314193726,
+      "grad_norm": 0.45938462018966675,
       "learning_rate": 0.0001007035175879397,
-      "loss": 0.9948,
+      "loss": 0.994,
       "step": 499
     },
     {
       "epoch": 0.22241992882562278,
-      "grad_norm": 0.32288622856140137,
+      "grad_norm": 0.32502633333206177,
       "learning_rate": 0.00010050251256281407,
-      "loss": 0.9409,
+      "loss": 0.9397,
       "step": 500
     },
     {
       "epoch": 0.222864768683274,
-      "grad_norm": 0.33923619985580444,
+      "grad_norm": 0.3375188410282135,
       "learning_rate": 0.00010030150753768846,
-      "loss": 0.9096,
+      "loss": 0.909,
       "step": 501
     },
     {
       "epoch": 0.22330960854092527,
-      "grad_norm": 0.3114798665046692,
+      "grad_norm": 0.31290072202682495,
       "learning_rate": 0.00010010050251256282,
-      "loss": 0.9112,
+      "loss": 0.9114,
       "step": 502
     },
     {
       "epoch": 0.2237544483985765,
-      "grad_norm": 0.37581518292427063,
+      "grad_norm": 0.37251392006874084,
       "learning_rate": 9.989949748743719e-05,
-      "loss": 0.9831,
+      "loss": 0.9812,
       "step": 503
     },
     {
       "epoch": 0.22419928825622776,
-      "grad_norm": 0.33060386776924133,
+      "grad_norm": 0.33114826679229736,
       "learning_rate": 9.969849246231156e-05,
-      "loss": 0.8942,
+      "loss": 0.893,
       "step": 504
     },
     {
       "epoch": 0.22464412811387902,
-      "grad_norm": 0.32445380091667175,
+      "grad_norm": 0.3244706392288208,
       "learning_rate": 9.949748743718594e-05,
-      "loss": 0.9665,
+      "loss": 0.9659,
       "step": 505
     },
     {
       "epoch": 0.22508896797153025,
-      "grad_norm": 0.3652149438858032,
+      "grad_norm": 0.3634791374206543,
       "learning_rate": 9.929648241206031e-05,
-      "loss": 1.0325,
+      "loss": 1.0315,
       "step": 506
     },
     {
       "epoch": 0.2255338078291815,
-      "grad_norm": 0.3789691925048828,
+      "grad_norm": 0.38301393389701843,
       "learning_rate": 9.909547738693468e-05,
-      "loss": 0.9488,
+      "loss": 0.9492,
       "step": 507
     },
     {
       "epoch": 0.22597864768683273,
-      "grad_norm": 0.35577788949012756,
+      "grad_norm": 0.3449389338493347,
       "learning_rate": 9.889447236180906e-05,
-      "loss": 0.9324,
+      "loss": 0.9327,
       "step": 508
     },
     {
       "epoch": 0.226423487544484,
-      "grad_norm": 0.37785062193870544,
+      "grad_norm": 0.3777461051940918,
       "learning_rate": 9.869346733668342e-05,
-      "loss": 0.9934,
+      "loss": 0.9938,
       "step": 509
     },
     {
       "epoch": 0.22686832740213522,
-      "grad_norm": 0.3577893078327179,
+      "grad_norm": 0.3581281006336212,
       "learning_rate": 9.84924623115578e-05,
-      "loss": 1.0183,
+      "loss": 1.0184,
       "step": 510
     },
     {
       "epoch": 0.22731316725978648,
-      "grad_norm": 0.3955710232257843,
+      "grad_norm": 0.3965352475643158,
       "learning_rate": 9.829145728643216e-05,
-      "loss": 0.9138,
+      "loss": 0.9131,
       "step": 511
     },
     {
       "epoch": 0.2277580071174377,
-      "grad_norm": 0.333099365234375,
+      "grad_norm": 0.33017244935035706,
       "learning_rate": 9.809045226130655e-05,
-      "loss": 0.9533,
+      "loss": 0.9547,
       "step": 512
     },
     {
       "epoch": 0.22820284697508897,
-      "grad_norm": 0.358026921749115,
+      "grad_norm": 0.35746607184410095,
       "learning_rate": 9.788944723618091e-05,
-      "loss": 0.9919,
+      "loss": 0.9917,
       "step": 513
     },
     {
       "epoch": 0.22864768683274023,
-      "grad_norm": 0.3610975742340088,
+      "grad_norm": 0.36392533779144287,
       "learning_rate": 9.768844221105528e-05,
-      "loss": 1.0059,
+      "loss": 1.0056,
       "step": 514
     },
     {
       "epoch": 0.22909252669039146,
-      "grad_norm": 0.32842594385147095,
+      "grad_norm": 0.3272344768047333,
       "learning_rate": 9.748743718592965e-05,
-      "loss": 0.9877,
+      "loss": 0.9895,
       "step": 515
     },
     {
       "epoch": 0.22953736654804271,
-      "grad_norm": 0.4672807455062866,
+      "grad_norm": 0.4714422821998596,
       "learning_rate": 9.728643216080403e-05,
-      "loss": 1.0148,
+      "loss": 1.0156,
       "step": 516
     },
     {
       "epoch": 0.22998220640569395,
-      "grad_norm": 0.3504094183444977,
+      "grad_norm": 0.3458651006221771,
       "learning_rate": 9.70854271356784e-05,
-      "loss": 0.8774,
+      "loss": 0.8773,
       "step": 517
     },
     {
       "epoch": 0.2304270462633452,
-      "grad_norm": 0.3299584984779358,
+      "grad_norm": 0.32996249198913574,
       "learning_rate": 9.688442211055276e-05,
-      "loss": 0.9018,
+      "loss": 0.9028,
       "step": 518
     },
     {
       "epoch": 0.23087188612099643,
-      "grad_norm": 0.3598071336746216,
+      "grad_norm": 0.3559573292732239,
       "learning_rate": 9.668341708542715e-05,
-      "loss": 0.9,
+      "loss": 0.8993,
       "step": 519
     },
     {
       "epoch": 0.2313167259786477,
-      "grad_norm": 0.366554856300354,
+      "grad_norm": 0.36793214082717896,
       "learning_rate": 9.64824120603015e-05,
-      "loss": 0.8612,
+      "loss": 0.8616,
       "step": 520
     },
     {
       "epoch": 0.23176156583629892,
-      "grad_norm": 0.3408771753311157,
+      "grad_norm": 0.3426240086555481,
       "learning_rate": 9.628140703517589e-05,
-      "loss": 0.9482,
+      "loss": 0.9483,
       "step": 521
     },
     {
       "epoch": 0.23220640569395018,
-      "grad_norm": 0.4127398133277893,
+      "grad_norm": 0.4089488685131073,
       "learning_rate": 9.608040201005025e-05,
-      "loss": 0.9427,
+      "loss": 0.9446,
       "step": 522
     },
     {
       "epoch": 0.2326512455516014,
-      "grad_norm": 0.4046645164489746,
+      "grad_norm": 0.40365123748779297,
       "learning_rate": 9.587939698492462e-05,
-      "loss": 1.0469,
+      "loss": 1.0485,
       "step": 523
     },
     {
       "epoch": 0.23309608540925267,
-      "grad_norm": 0.33988115191459656,
+      "grad_norm": 0.3387534022331238,
       "learning_rate": 9.5678391959799e-05,
-      "loss": 0.888,
+      "loss": 0.8881,
       "step": 524
     },
     {
       "epoch": 0.23354092526690393,
-      "grad_norm": 0.35615596175193787,
+      "grad_norm": 0.3568766117095947,
       "learning_rate": 9.547738693467337e-05,
-      "loss": 0.9522,
+      "loss": 0.9514,
       "step": 525
     },
     {
       "epoch": 0.23398576512455516,
-      "grad_norm": 0.41455528140068054,
+      "grad_norm": 0.42054617404937744,
       "learning_rate": 9.527638190954774e-05,
-      "loss": 0.9495,
+      "loss": 0.9481,
       "step": 526
     },
     {
       "epoch": 0.23443060498220641,
-      "grad_norm": 0.4435769021511078,
+      "grad_norm": 0.441377192735672,
       "learning_rate": 9.507537688442212e-05,
-      "loss": 0.9251,
+      "loss": 0.927,
       "step": 527
     },
     {
       "epoch": 0.23487544483985764,
-      "grad_norm": 0.34414413571357727,
+      "grad_norm": 0.3475041389465332,
       "learning_rate": 9.487437185929649e-05,
-      "loss": 0.9746,
+      "loss": 0.9735,
       "step": 528
     },
     {
       "epoch": 0.2353202846975089,
-      "grad_norm": 0.42099055647850037,
+      "grad_norm": 0.4227800667285919,
       "learning_rate": 9.467336683417086e-05,
-      "loss": 1.0079,
+      "loss": 1.0058,
       "step": 529
     },
     {
       "epoch": 0.23576512455516013,
-      "grad_norm": 0.3634389042854309,
+      "grad_norm": 0.3622898459434509,
       "learning_rate": 9.447236180904523e-05,
-      "loss": 0.8452,
+      "loss": 0.8462,
       "step": 530
     },
     {
       "epoch": 0.2362099644128114,
-      "grad_norm": 0.45584288239479065,
+      "grad_norm": 0.4561034142971039,
       "learning_rate": 9.427135678391961e-05,
-      "loss": 0.9618,
+      "loss": 0.9615,
       "step": 531
     },
     {
       "epoch": 0.23665480427046262,
-      "grad_norm": 0.3558484613895416,
+      "grad_norm": 0.35412004590034485,
       "learning_rate": 9.407035175879397e-05,
-      "loss": 1.007,
+      "loss": 1.0062,
       "step": 532
     },
     {
       "epoch": 0.23709964412811388,
-      "grad_norm": 0.41819700598716736,
+      "grad_norm": 0.4144454598426819,
       "learning_rate": 9.386934673366835e-05,
-      "loss": 0.8806,
+      "loss": 0.8798,
       "step": 533
     },
     {
       "epoch": 0.23754448398576514,
-      "grad_norm": 0.39965230226516724,
+      "grad_norm": 0.40296846628189087,
       "learning_rate": 9.366834170854271e-05,
-      "loss": 1.0749,
+      "loss": 1.0744,
       "step": 534
     },
     {
       "epoch": 0.23798932384341637,
-      "grad_norm": 0.35675838589668274,
+      "grad_norm": 0.35528820753097534,
       "learning_rate": 9.34673366834171e-05,
-      "loss": 1.0074,
+      "loss": 1.0083,
       "step": 535
     },
     {
       "epoch": 0.23843416370106763,
-      "grad_norm": 0.37170112133026123,
+      "grad_norm": 0.37445664405822754,
       "learning_rate": 9.326633165829146e-05,
-      "loss": 0.9623,
+      "loss": 0.9631,
       "step": 536
     },
     {
       "epoch": 0.23887900355871886,
-      "grad_norm": 0.35851290822029114,
+      "grad_norm": 0.360051691532135,
       "learning_rate": 9.306532663316585e-05,
-      "loss": 0.9361,
+      "loss": 0.936,
       "step": 537
     },
     {
       "epoch": 0.2393238434163701,
-      "grad_norm": 0.4395543038845062,
+      "grad_norm": 0.440403550863266,
       "learning_rate": 9.28643216080402e-05,
-      "loss": 0.871,
+      "loss": 0.8711,
       "step": 538
     },
     {
       "epoch": 0.23976868327402134,
-      "grad_norm": 0.2868823707103729,
+      "grad_norm": 0.2862887680530548,
       "learning_rate": 9.266331658291458e-05,
-      "loss": 0.9066,
+      "loss": 0.9069,
       "step": 539
     },
     {
       "epoch": 0.2402135231316726,
-      "grad_norm": 0.3684181869029999,
+      "grad_norm": 0.3707970976829529,
       "learning_rate": 9.246231155778895e-05,
-      "loss": 1.0657,
+      "loss": 1.0669,
       "step": 540
     },
     {
       "epoch": 0.24065836298932383,
-      "grad_norm": 0.347599059343338,
+      "grad_norm": 0.3454754948616028,
       "learning_rate": 9.226130653266331e-05,
-      "loss": 1.0081,
+      "loss": 1.0074,
       "step": 541
     },
     {
       "epoch": 0.2411032028469751,
-      "grad_norm": 0.32958024740219116,
+      "grad_norm": 0.3313436806201935,
       "learning_rate": 9.20603015075377e-05,
-      "loss": 0.9427,
+      "loss": 0.9425,
       "step": 542
     },
     {
       "epoch": 0.24154804270462635,
-      "grad_norm": 0.3672914505004883,
+      "grad_norm": 0.36515554785728455,
       "learning_rate": 9.185929648241206e-05,
-      "loss": 0.994,
+      "loss": 0.9946,
       "step": 543
     },
     {
       "epoch": 0.24199288256227758,
-      "grad_norm": 0.3725748658180237,
+      "grad_norm": 0.3730347454547882,
       "learning_rate": 9.165829145728644e-05,
-      "loss": 0.9835,
+      "loss": 0.9838,
       "step": 544
     },
     {
       "epoch": 0.24243772241992884,
-      "grad_norm": 0.39337027072906494,
+      "grad_norm": 0.39922618865966797,
       "learning_rate": 9.14572864321608e-05,
-      "loss": 0.8137,
+      "loss": 0.815,
       "step": 545
     },
     {
       "epoch": 0.24288256227758007,
-      "grad_norm": 0.32661673426628113,
+      "grad_norm": 0.31671297550201416,
       "learning_rate": 9.125628140703519e-05,
-      "loss": 1.0025,
+      "loss": 1.0014,
       "step": 546
     },
     {
       "epoch": 0.24332740213523132,
-      "grad_norm": 0.3912467956542969,
+      "grad_norm": 0.39499327540397644,
       "learning_rate": 9.105527638190955e-05,
-      "loss": 0.9087,
+      "loss": 0.9104,
       "step": 547
     },
     {
       "epoch": 0.24377224199288255,
-      "grad_norm": 0.37240296602249146,
+      "grad_norm": 0.3678281009197235,
       "learning_rate": 9.085427135678392e-05,
-      "loss": 0.9275,
+      "loss": 0.927,
       "step": 548
     },
     {
       "epoch": 0.2442170818505338,
-      "grad_norm": 0.3758089244365692,
+      "grad_norm": 0.372120201587677,
       "learning_rate": 9.06532663316583e-05,
-      "loss": 0.8224,
+      "loss": 0.8212,
       "step": 549
     },
     {
       "epoch": 0.24466192170818504,
-      "grad_norm": 0.3426155745983124,
+      "grad_norm": 0.3431313931941986,
       "learning_rate": 9.045226130653267e-05,
-      "loss": 0.8894,
+      "loss": 0.8883,
       "step": 550
     },
     {
       "epoch": 0.2451067615658363,
-      "grad_norm": 0.3958059549331665,
+      "grad_norm": 0.391932874917984,
       "learning_rate": 9.025125628140704e-05,
-      "loss": 0.9429,
+      "loss": 0.9425,
       "step": 551
     },
     {
       "epoch": 0.24555160142348753,
-      "grad_norm": 0.4533610939979553,
+      "grad_norm": 0.4534127116203308,
       "learning_rate": 9.005025125628141e-05,
-      "loss": 1.2454,
+      "loss": 1.2452,
       "step": 552
     },
     {
       "epoch": 0.2459964412811388,
-      "grad_norm": 0.36734533309936523,
+      "grad_norm": 0.37136268615722656,
       "learning_rate": 8.984924623115579e-05,
-      "loss": 1.0402,
+      "loss": 1.0424,
       "step": 553
     },
     {
       "epoch": 0.24644128113879005,
-      "grad_norm": 0.42336615920066833,
+      "grad_norm": 0.41206422448158264,
       "learning_rate": 8.964824120603016e-05,
-      "loss": 1.1425,
+      "loss": 1.1418,
       "step": 554
     },
     {
       "epoch": 0.24688612099644128,
-      "grad_norm": 0.44834333658218384,
+      "grad_norm": 0.440790057182312,
       "learning_rate": 8.944723618090453e-05,
-      "loss": 0.8959,
+      "loss": 0.8978,
       "step": 555
     },
     {
       "epoch": 0.24733096085409254,
-      "grad_norm": 0.3234824240207672,
+      "grad_norm": 0.32154756784439087,
       "learning_rate": 8.92462311557789e-05,
-      "loss": 0.9743,
+      "loss": 0.974,
       "step": 556
     },
     {
       "epoch": 0.24777580071174377,
-      "grad_norm": 0.3193085491657257,
+      "grad_norm": 0.3203801214694977,
       "learning_rate": 8.904522613065326e-05,
-      "loss": 0.9685,
+      "loss": 0.9684,
       "step": 557
     },
     {
       "epoch": 0.24822064056939502,
-      "grad_norm": 0.34897491335868835,
+      "grad_norm": 0.3494798243045807,
       "learning_rate": 8.884422110552765e-05,
-      "loss": 0.9132,
+      "loss": 0.9127,
       "step": 558
     },
     {
       "epoch": 0.24866548042704625,
-      "grad_norm": 0.3368080258369446,
+      "grad_norm": 0.3334081470966339,
       "learning_rate": 8.864321608040201e-05,
-      "loss": 0.9678,
+      "loss": 0.9685,
       "step": 559
     },
     {
       "epoch": 0.2491103202846975,
-      "grad_norm": 0.3948473036289215,
+      "grad_norm": 0.3990677297115326,
       "learning_rate": 8.84422110552764e-05,
-      "loss": 0.993,
+      "loss": 0.9923,
       "step": 560
     },
     {
       "epoch": 0.24955516014234874,
-      "grad_norm": 0.40010324120521545,
+      "grad_norm": 0.39742422103881836,
       "learning_rate": 8.824120603015076e-05,
-      "loss": 0.9779,
+      "loss": 0.9786,
       "step": 561
     },
     {
       "epoch": 0.25,
-      "grad_norm": 0.40043747425079346,
+      "grad_norm": 0.396680623292923,
       "learning_rate": 8.804020100502513e-05,
-      "loss": 0.9633,
+      "loss": 0.9636,
       "step": 562
     },
     {
       "epoch": 0.25044483985765126,
-      "grad_norm": 0.3239743709564209,
+      "grad_norm": 0.32014915347099304,
       "learning_rate": 8.78391959798995e-05,
-      "loss": 0.939,
+      "loss": 0.9392,
       "step": 563
     },
     {
       "epoch": 0.2508896797153025,
-      "grad_norm": 0.3126469552516937,
+      "grad_norm": 0.31222036480903625,
       "learning_rate": 8.763819095477387e-05,
-      "loss": 0.8804,
+      "loss": 0.8806,
       "step": 564
     },
     {
       "epoch": 0.2513345195729537,
-      "grad_norm": 0.32982298731803894,
+      "grad_norm": 0.3277677595615387,
       "learning_rate": 8.743718592964825e-05,
-      "loss": 0.8698,
+      "loss": 0.8699,
       "step": 565
     },
     {
       "epoch": 0.251779359430605,
-      "grad_norm": 0.4134942293167114,
+      "grad_norm": 0.42860186100006104,
       "learning_rate": 8.723618090452261e-05,
-      "loss": 0.9995,
+      "loss": 0.9997,
       "step": 566
     },
     {
       "epoch": 0.25222419928825623,
-      "grad_norm": 0.34916219115257263,
+      "grad_norm": 0.35114365816116333,
       "learning_rate": 8.7035175879397e-05,
-      "loss": 0.9463,
+      "loss": 0.947,
       "step": 567
     },
     {
       "epoch": 0.2526690391459075,
-      "grad_norm": 0.3170456886291504,
+      "grad_norm": 0.3189033269882202,
       "learning_rate": 8.683417085427135e-05,
-      "loss": 0.9028,
+      "loss": 0.904,
       "step": 568
     },
     {
       "epoch": 0.2531138790035587,
-      "grad_norm": 0.377886027097702,
+      "grad_norm": 0.37736451625823975,
       "learning_rate": 8.663316582914574e-05,
-      "loss": 0.9503,
+      "loss": 0.9518,
       "step": 569
     },
     {
       "epoch": 0.25355871886120995,
-      "grad_norm": 0.3546507656574249,
+      "grad_norm": 0.357546329498291,
       "learning_rate": 8.64321608040201e-05,
-      "loss": 0.9509,
+      "loss": 0.9499,
       "step": 570
     },
     {
       "epoch": 0.2540035587188612,
-      "grad_norm": 0.38051891326904297,
+      "grad_norm": 0.37832340598106384,
       "learning_rate": 8.623115577889449e-05,
-      "loss": 0.8949,
+      "loss": 0.895,
       "step": 571
     },
     {
       "epoch": 0.25444839857651247,
-      "grad_norm": 0.2983281910419464,
+      "grad_norm": 0.3017910122871399,
       "learning_rate": 8.603015075376884e-05,
-      "loss": 1.0411,
+      "loss": 1.041,
       "step": 572
     },
     {
       "epoch": 0.2548932384341637,
-      "grad_norm": 0.3371431529521942,
+      "grad_norm": 0.3350992798805237,
       "learning_rate": 8.582914572864322e-05,
-      "loss": 0.93,
+      "loss": 0.9303,
       "step": 573
     },
     {
       "epoch": 0.25533807829181493,
-      "grad_norm": 0.39484888315200806,
+      "grad_norm": 0.3973081707954407,
       "learning_rate": 8.562814070351759e-05,
-      "loss": 0.9633,
+      "loss": 0.9663,
       "step": 574
     },
     {
       "epoch": 0.2557829181494662,
-      "grad_norm": 0.34015023708343506,
+      "grad_norm": 0.3381018340587616,
       "learning_rate": 8.542713567839196e-05,
-      "loss": 0.944,
+      "loss": 0.9427,
       "step": 575
     },
     {
       "epoch": 0.25622775800711745,
-      "grad_norm": 0.32014408707618713,
+      "grad_norm": 0.3191261887550354,
       "learning_rate": 8.522613065326634e-05,
-      "loss": 0.829,
+      "loss": 0.8298,
       "step": 576
     },
     {
       "epoch": 0.2566725978647687,
-      "grad_norm": 0.42777976393699646,
+      "grad_norm": 0.4260123372077942,
       "learning_rate": 8.502512562814071e-05,
-      "loss": 1.0107,
+      "loss": 1.0098,
       "step": 577
     },
     {
       "epoch": 0.2571174377224199,
-      "grad_norm": 0.35483554005622864,
+      "grad_norm": 0.34333735704421997,
       "learning_rate": 8.482412060301508e-05,
-      "loss": 0.9974,
+      "loss": 1.0,
       "step": 578
     },
     {
       "epoch": 0.25756227758007116,
-      "grad_norm": 0.41365087032318115,
+      "grad_norm": 0.4197953939437866,
       "learning_rate": 8.462311557788946e-05,
-      "loss": 1.062,
+      "loss": 1.0629,
       "step": 579
     },
     {
       "epoch": 0.2580071174377224,
-      "grad_norm": 0.37708839774131775,
+      "grad_norm": 0.37437987327575684,
       "learning_rate": 8.442211055276383e-05,
-      "loss": 0.9283,
+      "loss": 0.9298,
       "step": 580
     },
     {
       "epoch": 0.2584519572953737,
-      "grad_norm": 0.40330421924591064,
+      "grad_norm": 0.40584203600883484,
       "learning_rate": 8.42211055276382e-05,
-      "loss": 0.9482,
+      "loss": 0.9493,
       "step": 581
     },
     {
       "epoch": 0.25889679715302494,
-      "grad_norm": 0.2943516969680786,
+      "grad_norm": 0.2971116900444031,
       "learning_rate": 8.402010050251256e-05,
-      "loss": 0.9211,
+      "loss": 0.9228,
       "step": 582
     },
     {
       "epoch": 0.25934163701067614,
-      "grad_norm": 0.3328346610069275,
+      "grad_norm": 0.33302900195121765,
       "learning_rate": 8.381909547738695e-05,
-      "loss": 0.9264,
+      "loss": 0.9277,
       "step": 583
     },
     {
       "epoch": 0.2597864768683274,
-      "grad_norm": 0.4220837354660034,
+      "grad_norm": 0.4172103703022003,
       "learning_rate": 8.36180904522613e-05,
-      "loss": 0.9598,
+      "loss": 0.9604,
       "step": 584
     },
     {
       "epoch": 0.26023131672597866,
-      "grad_norm": 0.36552560329437256,
+      "grad_norm": 0.36361220479011536,
       "learning_rate": 8.341708542713568e-05,
-      "loss": 0.9275,
+      "loss": 0.927,
       "step": 585
     },
     {
       "epoch": 0.2606761565836299,
-      "grad_norm": 0.36409202218055725,
+      "grad_norm": 0.365182489156723,
       "learning_rate": 8.321608040201005e-05,
       "loss": 0.9246,
       "step": 586
     },
     {
       "epoch": 0.2611209964412811,
-      "grad_norm": 0.3256109952926636,
+      "grad_norm": 0.32461798191070557,
       "learning_rate": 8.301507537688443e-05,
-      "loss": 0.9725,
+      "loss": 0.9727,
       "step": 587
     },
     {
       "epoch": 0.2615658362989324,
-      "grad_norm": 0.31421205401420593,
+      "grad_norm": 0.3159072995185852,
       "learning_rate": 8.28140703517588e-05,
-      "loss": 0.9276,
+      "loss": 0.9286,
       "step": 588
     },
     {
       "epoch": 0.26201067615658363,
-      "grad_norm": 0.33592212200164795,
+      "grad_norm": 0.32971808314323425,
       "learning_rate": 8.261306532663317e-05,
-      "loss": 0.964,
+      "loss": 0.9639,
       "step": 589
     },
     {
       "epoch": 0.2624555160142349,
-      "grad_norm": 0.49559640884399414,
+      "grad_norm": 0.5026357173919678,
       "learning_rate": 8.241206030150754e-05,
-      "loss": 0.9688,
+      "loss": 0.9706,
       "step": 590
     },
     {
       "epoch": 0.2629003558718861,
-      "grad_norm": 0.37904122471809387,
+      "grad_norm": 0.37912416458129883,
       "learning_rate": 8.22110552763819e-05,
-      "loss": 0.8688,
+      "loss": 0.8684,
       "step": 591
     },
     {
       "epoch": 0.26334519572953735,
-      "grad_norm": 0.38418570160865784,
+      "grad_norm": 0.3844146430492401,
       "learning_rate": 8.201005025125629e-05,
-      "loss": 0.966,
+      "loss": 0.9659,
       "step": 592
     },
     {
       "epoch": 0.2637900355871886,
-      "grad_norm": 0.3689751923084259,
+      "grad_norm": 0.3652094006538391,
       "learning_rate": 8.180904522613065e-05,
-      "loss": 0.8973,
+      "loss": 0.8978,
       "step": 593
     },
     {
       "epoch": 0.26423487544483987,
-      "grad_norm": 0.3859994113445282,
+      "grad_norm": 0.3851691484451294,
       "learning_rate": 8.160804020100504e-05,
-      "loss": 1.034,
+      "loss": 1.0336,
       "step": 594
     },
     {
       "epoch": 0.2646797153024911,
-      "grad_norm": 0.4068204164505005,
+      "grad_norm": 0.3992913067340851,
       "learning_rate": 8.14070351758794e-05,
-      "loss": 0.931,
+      "loss": 0.93,
       "step": 595
     },
     {
       "epoch": 0.26512455516014233,
-      "grad_norm": 0.3974169194698334,
+      "grad_norm": 0.39918407797813416,
       "learning_rate": 8.120603015075378e-05,
-      "loss": 1.055,
+      "loss": 1.0553,
       "step": 596
     },
     {
       "epoch": 0.2655693950177936,
-      "grad_norm": 0.3818078637123108,
+      "grad_norm": 0.369907021522522,
       "learning_rate": 8.100502512562814e-05,
-      "loss": 1.0652,
+      "loss": 1.0672,
       "step": 597
     },
     {
       "epoch": 0.26601423487544484,
-      "grad_norm": 0.3733369708061218,
+      "grad_norm": 0.3794615566730499,
       "learning_rate": 8.080402010050251e-05,
-      "loss": 0.9981,
+      "loss": 0.9986,
       "step": 598
     },
     {
       "epoch": 0.2664590747330961,
-      "grad_norm": 0.3980778455734253,
+      "grad_norm": 0.4018343389034271,
       "learning_rate": 8.060301507537689e-05,
-      "loss": 0.982,
+      "loss": 0.9832,
       "step": 599
     },
     {
       "epoch": 0.2669039145907473,
-      "grad_norm": 0.3596290051937103,
+      "grad_norm": 0.3550480902194977,
       "learning_rate": 8.040201005025126e-05,
-      "loss": 0.896,
+      "loss": 0.8965,
       "step": 600
     },
     {
       "epoch": 0.26734875444839856,
-      "grad_norm": 0.330400675535202,
+      "grad_norm": 0.3323322832584381,
       "learning_rate": 8.020100502512563e-05,
-      "loss": 0.9776,
+      "loss": 0.9779,
       "step": 601
     },
     {
       "epoch": 0.2677935943060498,
-      "grad_norm": 0.3300771415233612,
+      "grad_norm": 0.3290577828884125,
       "learning_rate": 8e-05,
-      "loss": 1.0576,
+      "loss": 1.0571,
       "step": 602
     },
     {
       "epoch": 0.2682384341637011,
-      "grad_norm": 0.36290502548217773,
+      "grad_norm": 0.36351272463798523,
       "learning_rate": 7.979899497487438e-05,
-      "loss": 0.9378,
+      "loss": 0.9393,
       "step": 603
     },
     {
       "epoch": 0.26868327402135234,
-      "grad_norm": 0.3487003743648529,
+      "grad_norm": 0.3494581878185272,
       "learning_rate": 7.959798994974875e-05,
-      "loss": 0.9146,
+      "loss": 0.9157,
       "step": 604
     },
     {
       "epoch": 0.26912811387900354,
-      "grad_norm": 0.3635631501674652,
+      "grad_norm": 0.36106282472610474,
       "learning_rate": 7.939698492462313e-05,
-      "loss": 0.9758,
+      "loss": 0.9772,
       "step": 605
     },
     {
       "epoch": 0.2695729537366548,
-      "grad_norm": 0.35547807812690735,
+      "grad_norm": 0.35360780358314514,
       "learning_rate": 7.91959798994975e-05,
-      "loss": 1.0026,
+      "loss": 1.0008,
       "step": 606
     },
     {
       "epoch": 0.27001779359430605,
-      "grad_norm": 0.33966726064682007,
+      "grad_norm": 0.3501545786857605,
       "learning_rate": 7.899497487437186e-05,
-      "loss": 0.9059,
+      "loss": 0.9061,
       "step": 607
     },
     {
       "epoch": 0.2704626334519573,
-      "grad_norm": 0.3759270906448364,
+      "grad_norm": 0.37978798151016235,
       "learning_rate": 7.879396984924623e-05,
-      "loss": 1.0088,
+      "loss": 1.0096,
       "step": 608
     },
     {
       "epoch": 0.2709074733096085,
-      "grad_norm": 0.36231812834739685,
+      "grad_norm": 0.3628638684749603,
       "learning_rate": 7.85929648241206e-05,
-      "loss": 0.9903,
+      "loss": 0.9912,
       "step": 609
     },
     {
       "epoch": 0.2713523131672598,
-      "grad_norm": 0.389737069606781,
+      "grad_norm": 0.3882206976413727,
       "learning_rate": 7.839195979899498e-05,
-      "loss": 0.8706,
+      "loss": 0.8698,
       "step": 610
     },
     {
       "epoch": 0.27179715302491103,
-      "grad_norm": 0.43408992886543274,
+      "grad_norm": 0.44139206409454346,
       "learning_rate": 7.819095477386935e-05,
-      "loss": 0.9437,
+      "loss": 0.9444,
       "step": 611
     },
     {
       "epoch": 0.2722419928825623,
-      "grad_norm": 0.29707393050193787,
+      "grad_norm": 0.2937663793563843,
       "learning_rate": 7.798994974874372e-05,
-      "loss": 0.8361,
+      "loss": 0.8367,
       "step": 612
     },
     {
       "epoch": 0.27268683274021355,
-      "grad_norm": 0.3469286561012268,
+      "grad_norm": 0.34553638100624084,
       "learning_rate": 7.77889447236181e-05,
-      "loss": 1.0085,
+      "loss": 1.0092,
       "step": 613
     },
     {
       "epoch": 0.27313167259786475,
-      "grad_norm": 0.4895261824131012,
+      "grad_norm": 0.4855239689350128,
       "learning_rate": 7.758793969849247e-05,
-      "loss": 0.8068,
+      "loss": 0.8105,
       "step": 614
     },
     {
       "epoch": 0.273576512455516,
-      "grad_norm": 0.3655518591403961,
+      "grad_norm": 0.359298974275589,
       "learning_rate": 7.738693467336684e-05,
-      "loss": 0.9248,
+      "loss": 0.9225,
       "step": 615
     },
     {
       "epoch": 0.27402135231316727,
-      "grad_norm": 0.36705151200294495,
+      "grad_norm": 0.3697519600391388,
       "learning_rate": 7.71859296482412e-05,
-      "loss": 1.0244,
+      "loss": 1.0232,
       "step": 616
     },
     {
       "epoch": 0.2744661921708185,
-      "grad_norm": 0.3587312400341034,
+      "grad_norm": 0.3558047115802765,
       "learning_rate": 7.698492462311559e-05,
-      "loss": 1.0522,
+      "loss": 1.0508,
       "step": 617
     },
     {
       "epoch": 0.2749110320284697,
-      "grad_norm": 0.3758445680141449,
+      "grad_norm": 0.375429630279541,
       "learning_rate": 7.678391959798995e-05,
-      "loss": 0.9904,
+      "loss": 0.9915,
       "step": 618
     },
     {
       "epoch": 0.275355871886121,
-      "grad_norm": 0.31862393021583557,
+      "grad_norm": 0.32131966948509216,
       "learning_rate": 7.658291457286433e-05,
-      "loss": 0.9216,
+      "loss": 0.9213,
       "step": 619
     },
     {
       "epoch": 0.27580071174377224,
-      "grad_norm": 0.4419485926628113,
+      "grad_norm": 0.4405272305011749,
       "learning_rate": 7.638190954773869e-05,
-      "loss": 0.9049,
+      "loss": 0.9044,
       "step": 620
     },
     {
       "epoch": 0.2762455516014235,
-      "grad_norm": 0.38519737124443054,
+      "grad_norm": 0.38022780418395996,
       "learning_rate": 7.618090452261307e-05,
-      "loss": 0.9116,
+      "loss": 0.9114,
       "step": 621
     },
     {
       "epoch": 0.27669039145907476,
-      "grad_norm": 0.37228310108184814,
+      "grad_norm": 0.375396728515625,
       "learning_rate": 7.597989949748744e-05,
-      "loss": 0.9491,
+      "loss": 0.9496,
       "step": 622
     },
     {
       "epoch": 0.27713523131672596,
-      "grad_norm": 0.3866081237792969,
+      "grad_norm": 0.38898783922195435,
       "learning_rate": 7.577889447236181e-05,
-      "loss": 0.9477,
+      "loss": 0.9488,
       "step": 623
     },
     {
       "epoch": 0.2775800711743772,
-      "grad_norm": 0.32213568687438965,
+      "grad_norm": 0.32414594292640686,
       "learning_rate": 7.557788944723618e-05,
-      "loss": 0.9049,
+      "loss": 0.9054,
       "step": 624
     },
     {
       "epoch": 0.2780249110320285,
-      "grad_norm": 0.34658104181289673,
+      "grad_norm": 0.34590399265289307,
       "learning_rate": 7.537688442211056e-05,
-      "loss": 0.9486,
+      "loss": 0.9491,
       "step": 625
     },
     {
       "epoch": 0.27846975088967973,
-      "grad_norm": 0.42399680614471436,
+      "grad_norm": 0.4231082499027252,
       "learning_rate": 7.517587939698493e-05,
-      "loss": 1.0776,
+      "loss": 1.0757,
       "step": 626
     },
     {
       "epoch": 0.27891459074733094,
-      "grad_norm": 0.3882311284542084,
+      "grad_norm": 0.3876168131828308,
       "learning_rate": 7.49748743718593e-05,
-      "loss": 0.9711,
+      "loss": 0.9705,
       "step": 627
     },
     {
       "epoch": 0.2793594306049822,
-      "grad_norm": 0.34879353642463684,
+      "grad_norm": 0.3473493158817291,
       "learning_rate": 7.477386934673368e-05,
-      "loss": 0.9253,
+      "loss": 0.9262,
       "step": 628
     },
     {
       "epoch": 0.27980427046263345,
-      "grad_norm": 0.3509921431541443,
+      "grad_norm": 0.34793248772621155,
       "learning_rate": 7.457286432160805e-05,
-      "loss": 0.9711,
+      "loss": 0.9728,
       "step": 629
     },
     {
       "epoch": 0.2802491103202847,
-      "grad_norm": 0.32417482137680054,
+      "grad_norm": 0.3288039267063141,
       "learning_rate": 7.437185929648241e-05,
-      "loss": 0.9612,
+      "loss": 0.9609,
       "step": 630
     },
     {
       "epoch": 0.28069395017793597,
-      "grad_norm": 0.3800636827945709,
+      "grad_norm": 0.37984830141067505,
       "learning_rate": 7.417085427135678e-05,
-      "loss": 1.0243,
+      "loss": 1.0235,
       "step": 631
     },
     {
       "epoch": 0.28113879003558717,
-      "grad_norm": 0.38342657685279846,
+      "grad_norm": 0.382462739944458,
       "learning_rate": 7.396984924623115e-05,
-      "loss": 0.9895,
+      "loss": 0.9898,
       "step": 632
     },
     {
       "epoch": 0.28158362989323843,
-      "grad_norm": 0.40875983238220215,
+      "grad_norm": 0.406568318605423,
       "learning_rate": 7.376884422110553e-05,
-      "loss": 0.9428,
+      "loss": 0.9442,
       "step": 633
     },
     {
       "epoch": 0.2820284697508897,
-      "grad_norm": 0.34101223945617676,
+      "grad_norm": 0.33671334385871887,
       "learning_rate": 7.35678391959799e-05,
-      "loss": 0.977,
+      "loss": 0.9771,
       "step": 634
     },
     {
       "epoch": 0.28247330960854095,
-      "grad_norm": 0.3537449836730957,
+      "grad_norm": 0.34497055411338806,
       "learning_rate": 7.336683417085427e-05,
-      "loss": 0.9458,
+      "loss": 0.9452,
       "step": 635
     },
     {
       "epoch": 0.28291814946619215,
-      "grad_norm": 0.3816027343273163,
+      "grad_norm": 0.37973660230636597,
       "learning_rate": 7.316582914572865e-05,
-      "loss": 1.0429,
+      "loss": 1.0431,
       "step": 636
     },
     {
       "epoch": 0.2833629893238434,
-      "grad_norm": 0.39149439334869385,
+      "grad_norm": 0.391178160905838,
       "learning_rate": 7.296482412060302e-05,
-      "loss": 0.974,
+      "loss": 0.9732,
       "step": 637
     },
     {
       "epoch": 0.28380782918149466,
-      "grad_norm": 0.4129406809806824,
+      "grad_norm": 0.41452303528785706,
       "learning_rate": 7.276381909547739e-05,
-      "loss": 0.988,
+      "loss": 0.986,
       "step": 638
     },
     {
       "epoch": 0.2842526690391459,
-      "grad_norm": 0.36804699897766113,
+      "grad_norm": 0.388171911239624,
       "learning_rate": 7.256281407035177e-05,
-      "loss": 0.9375,
+      "loss": 0.9368,
       "step": 639
     },
     {
       "epoch": 0.2846975088967972,
-      "grad_norm": 0.40363621711730957,
+      "grad_norm": 0.40021994709968567,
       "learning_rate": 7.236180904522614e-05,
-      "loss": 0.9214,
+      "loss": 0.9213,
       "step": 640
     },
     {
       "epoch": 0.2851423487544484,
-      "grad_norm": 0.3833264708518982,
+      "grad_norm": 0.380087286233902,
       "learning_rate": 7.21608040201005e-05,
-      "loss": 0.9454,
+      "loss": 0.9452,
       "step": 641
     },
     {
       "epoch": 0.28558718861209964,
-      "grad_norm": 0.38440853357315063,
+      "grad_norm": 0.3855552673339844,
       "learning_rate": 7.195979899497488e-05,
       "loss": 0.8841,
       "step": 642
     },
     {
       "epoch": 0.2860320284697509,
-      "grad_norm": 0.42536038160324097,
+      "grad_norm": 0.4310576319694519,
       "learning_rate": 7.175879396984924e-05,
-      "loss": 0.9665,
+      "loss": 0.9666,
       "step": 643
     },
     {
       "epoch": 0.28647686832740216,
-      "grad_norm": 0.33145928382873535,
+      "grad_norm": 0.3338180482387543,
       "learning_rate": 7.155778894472363e-05,
-      "loss": 0.9118,
+      "loss": 0.9127,
       "step": 644
     },
     {
       "epoch": 0.28692170818505336,
-      "grad_norm": 0.37018072605133057,
+      "grad_norm": 0.373388409614563,
       "learning_rate": 7.135678391959799e-05,
-      "loss": 0.9424,
+      "loss": 0.9409,
       "step": 645
     },
     {
       "epoch": 0.2873665480427046,
-      "grad_norm": 0.37953630089759827,
+      "grad_norm": 0.37845322489738464,
       "learning_rate": 7.115577889447236e-05,
-      "loss": 0.9889,
+      "loss": 0.9885,
       "step": 646
     },
     {
       "epoch": 0.2878113879003559,
-      "grad_norm": 0.3908143937587738,
+      "grad_norm": 0.39277783036231995,
       "learning_rate": 7.095477386934674e-05,
-      "loss": 0.9815,
+      "loss": 0.9819,
       "step": 647
     },
     {
       "epoch": 0.28825622775800713,
-      "grad_norm": 0.38611525297164917,
+      "grad_norm": 0.3995840549468994,
       "learning_rate": 7.075376884422111e-05,
-      "loss": 1.0179,
+      "loss": 1.018,
       "step": 648
     },
     {
       "epoch": 0.28870106761565834,
-      "grad_norm": 0.4868585169315338,
+      "grad_norm": 0.48846814036369324,
       "learning_rate": 7.055276381909548e-05,
-      "loss": 1.0551,
+      "loss": 1.0557,
       "step": 649
     },
     {
       "epoch": 0.2891459074733096,
-      "grad_norm": 0.4477649927139282,
+      "grad_norm": 0.44867080450057983,
       "learning_rate": 7.035175879396985e-05,
-      "loss": 0.8795,
+      "loss": 0.8785,
       "step": 650
     },
     {
       "epoch": 0.28959074733096085,
-      "grad_norm": 0.3561191260814667,
+      "grad_norm": 0.3561117649078369,
       "learning_rate": 7.015075376884423e-05,
-      "loss": 0.8927,
+      "loss": 0.8914,
       "step": 651
     },
     {
       "epoch": 0.2900355871886121,
-      "grad_norm": 0.4050782024860382,
+      "grad_norm": 0.4034915268421173,
       "learning_rate": 6.99497487437186e-05,
-      "loss": 0.9032,
+      "loss": 0.9038,
       "step": 652
     },
     {
       "epoch": 0.29048042704626337,
-      "grad_norm": 0.3366767466068268,
+      "grad_norm": 0.33660322427749634,
       "learning_rate": 6.974874371859297e-05,
-      "loss": 0.9763,
+      "loss": 0.9757,
       "step": 653
     },
     {
       "epoch": 0.29092526690391457,
-      "grad_norm": 0.33202803134918213,
+      "grad_norm": 0.3358772099018097,
       "learning_rate": 6.954773869346733e-05,
-      "loss": 0.8999,
+      "loss": 0.9003,
       "step": 654
     },
     {
       "epoch": 0.29137010676156583,
-      "grad_norm": 0.38702261447906494,
+      "grad_norm": 0.40661585330963135,
       "learning_rate": 6.93467336683417e-05,
-      "loss": 0.8882,
+      "loss": 0.8901,
       "step": 655
     },
     {
       "epoch": 0.2918149466192171,
-      "grad_norm": 0.39489811658859253,
+      "grad_norm": 0.3901599943637848,
       "learning_rate": 6.914572864321608e-05,
-      "loss": 0.8615,
+      "loss": 0.8622,
       "step": 656
     },
     {
       "epoch": 0.29225978647686834,
-      "grad_norm": 0.407174289226532,
+      "grad_norm": 0.4064997732639313,
       "learning_rate": 6.894472361809045e-05,
-      "loss": 0.8805,
+      "loss": 0.8799,
       "step": 657
     },
     {
       "epoch": 0.29270462633451955,
-      "grad_norm": 0.3787648379802704,
+      "grad_norm": 0.3740297555923462,
       "learning_rate": 6.874371859296482e-05,
-      "loss": 0.9016,
+      "loss": 0.9026,
       "step": 658
     },
     {
       "epoch": 0.2931494661921708,
-      "grad_norm": 0.3571087121963501,
+      "grad_norm": 0.3562834858894348,
       "learning_rate": 6.85427135678392e-05,
-      "loss": 0.9484,
+      "loss": 0.949,
       "step": 659
     },
     {
       "epoch": 0.29359430604982206,
-      "grad_norm": 0.41471996903419495,
+      "grad_norm": 0.41133421659469604,
       "learning_rate": 6.834170854271357e-05,
       "loss": 1.0558,
       "step": 660
     },
     {
       "epoch": 0.2940391459074733,
-      "grad_norm": 0.4297381639480591,
+      "grad_norm": 0.4356454014778137,
       "learning_rate": 6.814070351758794e-05,
-      "loss": 0.9687,
+      "loss": 0.9691,
       "step": 661
     },
     {
       "epoch": 0.2944839857651246,
-      "grad_norm": 0.33687537908554077,
+      "grad_norm": 0.3359556496143341,
       "learning_rate": 6.793969849246232e-05,
       "loss": 0.9266,
       "step": 662
     },
     {
       "epoch": 0.2949288256227758,
-      "grad_norm": 0.34857362508773804,
+      "grad_norm": 0.34760648012161255,
       "learning_rate": 6.773869346733669e-05,
-      "loss": 0.8977,
+      "loss": 0.8974,
       "step": 663
     },
     {
       "epoch": 0.29537366548042704,
-      "grad_norm": 0.4323669672012329,
+      "grad_norm": 0.42510053515434265,
       "learning_rate": 6.753768844221105e-05,
-      "loss": 0.9535,
+      "loss": 0.9537,
       "step": 664
     },
     {
       "epoch": 0.2958185053380783,
-      "grad_norm": 0.36021220684051514,
+      "grad_norm": 0.35934481024742126,
       "learning_rate": 6.733668341708544e-05,
-      "loss": 0.9457,
+      "loss": 0.9455,
       "step": 665
     },
     {
       "epoch": 0.29626334519572955,
-      "grad_norm": 0.3574046492576599,
+      "grad_norm": 0.3559943735599518,
       "learning_rate": 6.71356783919598e-05,
-      "loss": 0.9377,
+      "loss": 0.9363,
       "step": 666
     },
     {
       "epoch": 0.29670818505338076,
-      "grad_norm": 0.3637276291847229,
+      "grad_norm": 0.36470353603363037,
       "learning_rate": 6.693467336683418e-05,
-      "loss": 0.9769,
+      "loss": 0.978,
       "step": 667
     },
     {
       "epoch": 0.297153024911032,
-      "grad_norm": 0.4049037992954254,
+      "grad_norm": 0.4014419913291931,
       "learning_rate": 6.673366834170854e-05,
-      "loss": 1.0499,
+      "loss": 1.0503,
       "step": 668
     },
     {
       "epoch": 0.2975978647686833,
-      "grad_norm": 0.4278649687767029,
+      "grad_norm": 0.4242320656776428,
       "learning_rate": 6.653266331658293e-05,
-      "loss": 0.8835,
+      "loss": 0.8848,
       "step": 669
     },
     {
       "epoch": 0.29804270462633453,
-      "grad_norm": 0.37466055154800415,
+      "grad_norm": 0.3722448945045471,
       "learning_rate": 6.633165829145729e-05,
-      "loss": 0.9493,
+      "loss": 0.9487,
       "step": 670
     },
     {
       "epoch": 0.2984875444839858,
-      "grad_norm": 0.3321222960948944,
+      "grad_norm": 0.33282628655433655,
       "learning_rate": 6.613065326633166e-05,
-      "loss": 0.9484,
+      "loss": 0.9483,
       "step": 671
     },
     {
       "epoch": 0.298932384341637,
-      "grad_norm": 0.4597262740135193,
+      "grad_norm": 0.45830950140953064,
       "learning_rate": 6.592964824120603e-05,
-      "loss": 1.0418,
+      "loss": 1.0413,
       "step": 672
     },
     {
       "epoch": 0.29937722419928825,
-      "grad_norm": 0.3227452337741852,
+      "grad_norm": 0.32155314087867737,
       "learning_rate": 6.57286432160804e-05,
-      "loss": 0.9323,
+      "loss": 0.9299,
       "step": 673
     },
     {
       "epoch": 0.2998220640569395,
-      "grad_norm": 0.3192083239555359,
+      "grad_norm": 0.319416344165802,
       "learning_rate": 6.552763819095478e-05,
-      "loss": 0.9021,
+      "loss": 0.9,
       "step": 674
     },
     {
       "epoch": 0.30026690391459077,
-      "grad_norm": 0.4145206809043884,
+      "grad_norm": 0.4341578781604767,
       "learning_rate": 6.532663316582915e-05,
-      "loss": 0.9431,
+      "loss": 0.9455,
       "step": 675
     },
     {
       "epoch": 0.30071174377224197,
-      "grad_norm": 0.4133388102054596,
+      "grad_norm": 0.4117160737514496,
       "learning_rate": 6.512562814070352e-05,
-      "loss": 0.9673,
+      "loss": 0.9682,
       "step": 676
     },
     {
       "epoch": 0.3011565836298932,
-      "grad_norm": 0.4261665940284729,
+      "grad_norm": 0.42781034111976624,
       "learning_rate": 6.492462311557788e-05,
       "loss": 1.0128,
       "step": 677
     },
     {
       "epoch": 0.3016014234875445,
-      "grad_norm": 0.33553653955459595,
+      "grad_norm": 0.3355860710144043,
       "learning_rate": 6.472361809045227e-05,
-      "loss": 0.9103,
+      "loss": 0.911,
       "step": 678
     },
     {
       "epoch": 0.30204626334519574,
-      "grad_norm": 0.378233939409256,
+      "grad_norm": 0.3754599690437317,
       "learning_rate": 6.452261306532663e-05,
-      "loss": 1.1017,
+      "loss": 1.0985,
       "step": 679
     },
     {
       "epoch": 0.302491103202847,
-      "grad_norm": 0.39459675550460815,
+      "grad_norm": 0.3943292200565338,
       "learning_rate": 6.4321608040201e-05,
-      "loss": 0.9056,
+      "loss": 0.9051,
       "step": 680
     },
     {
       "epoch": 0.3029359430604982,
-      "grad_norm": 0.31103044748306274,
+      "grad_norm": 0.3109300434589386,
       "learning_rate": 6.412060301507538e-05,
-      "loss": 0.896,
+      "loss": 0.8964,
       "step": 681
     },
     {
       "epoch": 0.30338078291814946,
-      "grad_norm": 0.32393592596054077,
+      "grad_norm": 0.32145023345947266,
       "learning_rate": 6.391959798994975e-05,
-      "loss": 0.9233,
+      "loss": 0.9225,
       "step": 682
     },
     {
       "epoch": 0.3038256227758007,
-      "grad_norm": 0.37299537658691406,
+      "grad_norm": 0.37335747480392456,
       "learning_rate": 6.371859296482412e-05,
-      "loss": 0.9587,
+      "loss": 0.9586,
       "step": 683
     },
     {
       "epoch": 0.304270462633452,
-      "grad_norm": 0.3141055703163147,
+      "grad_norm": 0.3109886646270752,
       "learning_rate": 6.35175879396985e-05,
-      "loss": 0.9314,
+      "loss": 0.9308,
       "step": 684
     },
     {
       "epoch": 0.3047153024911032,
-      "grad_norm": 0.32686251401901245,
+      "grad_norm": 0.3259667456150055,
       "learning_rate": 6.331658291457287e-05,
-      "loss": 0.9325,
+      "loss": 0.9322,
       "step": 685
     },
     {
       "epoch": 0.30516014234875444,
-      "grad_norm": 0.35383760929107666,
+      "grad_norm": 0.35218545794487,
       "learning_rate": 6.311557788944724e-05,
-      "loss": 0.9834,
+      "loss": 0.9853,
       "step": 686
     },
     {
       "epoch": 0.3056049822064057,
-      "grad_norm": 0.30796653032302856,
+      "grad_norm": 0.3094191551208496,
       "learning_rate": 6.291457286432161e-05,
-      "loss": 0.8842,
+      "loss": 0.8857,
       "step": 687
     },
     {
       "epoch": 0.30604982206405695,
-      "grad_norm": 0.31963440775871277,
+      "grad_norm": 0.3218885064125061,
       "learning_rate": 6.271356783919599e-05,
-      "loss": 0.9372,
+      "loss": 0.9363,
       "step": 688
     },
     {
       "epoch": 0.3064946619217082,
-      "grad_norm": 0.43244582414627075,
+      "grad_norm": 0.4279980957508087,
       "learning_rate": 6.251256281407035e-05,
-      "loss": 0.939,
+      "loss": 0.9375,
       "step": 689
     },
     {
       "epoch": 0.3069395017793594,
-      "grad_norm": 0.32725387811660767,
+      "grad_norm": 0.3290557861328125,
       "learning_rate": 6.231155778894473e-05,
-      "loss": 0.8819,
+      "loss": 0.8822,
       "step": 690
     },
     {
       "epoch": 0.30738434163701067,
-      "grad_norm": 0.3326573371887207,
+      "grad_norm": 0.3323034942150116,
       "learning_rate": 6.211055276381909e-05,
-      "loss": 1.0157,
+      "loss": 1.016,
       "step": 691
     },
     {
       "epoch": 0.30782918149466193,
-      "grad_norm": 0.3267054855823517,
+      "grad_norm": 0.31911730766296387,
       "learning_rate": 6.190954773869348e-05,
-      "loss": 0.8806,
+      "loss": 0.8814,
       "step": 692
     },
     {
       "epoch": 0.3082740213523132,
-      "grad_norm": 0.3328150808811188,
+      "grad_norm": 0.3330342471599579,
       "learning_rate": 6.170854271356784e-05,
-      "loss": 0.9752,
+      "loss": 0.9745,
       "step": 693
     },
     {
       "epoch": 0.3087188612099644,
-      "grad_norm": 0.42537713050842285,
+      "grad_norm": 0.42443540692329407,
       "learning_rate": 6.150753768844222e-05,
-      "loss": 0.8548,
+      "loss": 0.8555,
       "step": 694
     },
     {
       "epoch": 0.30916370106761565,
-      "grad_norm": 0.339568167924881,
+      "grad_norm": 0.3406936526298523,
       "learning_rate": 6.130653266331658e-05,
-      "loss": 0.9964,
+      "loss": 0.9958,
       "step": 695
     },
     {
       "epoch": 0.3096085409252669,
-      "grad_norm": 0.40449681878089905,
+      "grad_norm": 0.4078681766986847,
       "learning_rate": 6.110552763819096e-05,
-      "loss": 0.8698,
+      "loss": 0.8692,
       "step": 696
     },
     {
       "epoch": 0.31005338078291816,
-      "grad_norm": 0.42647475004196167,
+      "grad_norm": 0.4206138849258423,
       "learning_rate": 6.090452261306533e-05,
-      "loss": 0.9248,
+      "loss": 0.9255,
       "step": 697
     },
     {
       "epoch": 0.3104982206405694,
-      "grad_norm": 0.3729760944843292,
+      "grad_norm": 0.37396302819252014,
       "learning_rate": 6.070351758793971e-05,
-      "loss": 0.9945,
+      "loss": 0.9942,
       "step": 698
     },
     {
       "epoch": 0.3109430604982206,
-      "grad_norm": 0.34256407618522644,
+      "grad_norm": 0.3433186411857605,
       "learning_rate": 6.0502512562814076e-05,
-      "loss": 0.8589,
+      "loss": 0.8593,
       "step": 699
     },
     {
       "epoch": 0.3113879003558719,
-      "grad_norm": 0.37434864044189453,
+      "grad_norm": 0.37683218717575073,
       "learning_rate": 6.030150753768844e-05,
       "loss": 0.8921,
       "step": 700
     },
     {
       "epoch": 0.31183274021352314,
-      "grad_norm": 0.3740074336528778,
+      "grad_norm": 0.3779331147670746,
       "learning_rate": 6.0100502512562815e-05,
-      "loss": 0.9538,
+      "loss": 0.9528,
       "step": 701
     },
     {
       "epoch": 0.3122775800711744,
-      "grad_norm": 0.2955409288406372,
+      "grad_norm": 0.296854168176651,
       "learning_rate": 5.989949748743718e-05,
-      "loss": 0.8796,
+      "loss": 0.8811,
       "step": 702
     },
     {
       "epoch": 0.3127224199288256,
-      "grad_norm": 0.41662129759788513,
+      "grad_norm": 0.4115276336669922,
       "learning_rate": 5.969849246231156e-05,
-      "loss": 0.976,
+      "loss": 0.9749,
       "step": 703
     },
     {
       "epoch": 0.31316725978647686,
-      "grad_norm": 0.3664419651031494,
+      "grad_norm": 0.36785462498664856,
       "learning_rate": 5.949748743718593e-05,
       "loss": 0.9813,
       "step": 704
     },
     {
       "epoch": 0.3136120996441281,
-      "grad_norm": 0.4082026779651642,
+      "grad_norm": 0.4077514410018921,
       "learning_rate": 5.929648241206031e-05,
-      "loss": 0.9604,
+      "loss": 0.9611,
       "step": 705
     },
     {
       "epoch": 0.3140569395017794,
-      "grad_norm": 0.34381023049354553,
+      "grad_norm": 0.34254419803619385,
       "learning_rate": 5.909547738693467e-05,
-      "loss": 0.9438,
+      "loss": 0.9411,
       "step": 706
     },
     {
       "epoch": 0.3145017793594306,
-      "grad_norm": 0.32259446382522583,
+      "grad_norm": 0.3231055736541748,
       "learning_rate": 5.889447236180905e-05,
-      "loss": 0.8939,
+      "loss": 0.8937,
       "step": 707
     },
     {
       "epoch": 0.31494661921708184,
-      "grad_norm": 0.4534759819507599,
+      "grad_norm": 0.46129921078681946,
       "learning_rate": 5.869346733668342e-05,
-      "loss": 1.0139,
+      "loss": 1.0143,
       "step": 708
     },
     {
       "epoch": 0.3153914590747331,
-      "grad_norm": 0.3195643723011017,
+      "grad_norm": 0.31858929991722107,
       "learning_rate": 5.849246231155779e-05,
-      "loss": 0.8886,
+      "loss": 0.8875,
       "step": 709
     },
     {
       "epoch": 0.31583629893238435,
-      "grad_norm": 0.34111788868904114,
+      "grad_norm": 0.34170815348625183,
       "learning_rate": 5.829145728643216e-05,
-      "loss": 0.9006,
+      "loss": 0.9015,
       "step": 710
     },
     {
       "epoch": 0.3162811387900356,
-      "grad_norm": 0.32145076990127563,
+      "grad_norm": 0.3193584084510803,
       "learning_rate": 5.809045226130654e-05,
-      "loss": 0.9042,
+      "loss": 0.9041,
       "step": 711
     },
     {
       "epoch": 0.3167259786476868,
-      "grad_norm": 0.38871321082115173,
+      "grad_norm": 0.3801109790802002,
       "learning_rate": 5.7889447236180904e-05,
-      "loss": 0.8285,
+      "loss": 0.8286,
       "step": 712
     },
     {
       "epoch": 0.31717081850533807,
-      "grad_norm": 0.34648001194000244,
+      "grad_norm": 0.3580700159072876,
       "learning_rate": 5.7688442211055284e-05,
-      "loss": 0.8948,
+      "loss": 0.8942,
       "step": 713
     },
     {
       "epoch": 0.31761565836298933,
-      "grad_norm": 0.42825043201446533,
+      "grad_norm": 0.4236672818660736,
       "learning_rate": 5.748743718592965e-05,
-      "loss": 0.9665,
+      "loss": 0.9645,
       "step": 714
     },
     {
       "epoch": 0.3180604982206406,
-      "grad_norm": 0.4382406771183014,
+      "grad_norm": 0.43990832567214966,
       "learning_rate": 5.728643216080403e-05,
-      "loss": 1.0006,
+      "loss": 1.0032,
       "step": 715
     },
     {
       "epoch": 0.3185053380782918,
-      "grad_norm": 0.3384600579738617,
+      "grad_norm": 0.34689345955848694,
       "learning_rate": 5.7085427135678396e-05,
-      "loss": 0.893,
+      "loss": 0.8941,
       "step": 716
     },
     {
       "epoch": 0.31895017793594305,
-      "grad_norm": 0.329862505197525,
+      "grad_norm": 0.31815850734710693,
       "learning_rate": 5.688442211055277e-05,
-      "loss": 0.8064,
+      "loss": 0.8067,
       "step": 717
     },
     {
       "epoch": 0.3193950177935943,
-      "grad_norm": 0.36745166778564453,
+      "grad_norm": 0.36961331963539124,
       "learning_rate": 5.6683417085427135e-05,
-      "loss": 0.9621,
+      "loss": 0.9644,
       "step": 718
     },
     {
       "epoch": 0.31983985765124556,
-      "grad_norm": 0.36240148544311523,
+      "grad_norm": 0.35958558320999146,
       "learning_rate": 5.6482412060301515e-05,
-      "loss": 0.8902,
+      "loss": 0.89,
       "step": 719
     },
     {
       "epoch": 0.3202846975088968,
-      "grad_norm": 0.3071519136428833,
+      "grad_norm": 0.3142717480659485,
       "learning_rate": 5.628140703517588e-05,
       "loss": 0.9572,
       "step": 720
     },
     {
       "epoch": 0.320729537366548,
-      "grad_norm": 0.3393983840942383,
+      "grad_norm": 0.3404165506362915,
       "learning_rate": 5.608040201005026e-05,
-      "loss": 0.9021,
+      "loss": 0.9028,
       "step": 721
     },
     {
       "epoch": 0.3211743772241993,
-      "grad_norm": 0.3375062942504883,
+      "grad_norm": 0.33231121301651,
       "learning_rate": 5.587939698492463e-05,
-      "loss": 0.8505,
+      "loss": 0.8523,
       "step": 722
     },
     {
       "epoch": 0.32161921708185054,
-      "grad_norm": 0.36166590452194214,
+      "grad_norm": 0.3641085624694824,
       "learning_rate": 5.567839195979899e-05,
-      "loss": 0.9217,
+      "loss": 0.9232,
       "step": 723
     },
     {
       "epoch": 0.3220640569395018,
-      "grad_norm": 0.423556923866272,
+      "grad_norm": 0.4227713644504547,
       "learning_rate": 5.547738693467337e-05,
-      "loss": 0.8722,
+      "loss": 0.872,
       "step": 724
     },
     {
       "epoch": 0.322508896797153,
-      "grad_norm": 0.3881194293498993,
+      "grad_norm": 0.3884482979774475,
       "learning_rate": 5.527638190954774e-05,
-      "loss": 0.9246,
+      "loss": 0.9237,
       "step": 725
     },
     {
       "epoch": 0.32295373665480426,
-      "grad_norm": 0.41174158453941345,
+      "grad_norm": 0.4094702899456024,
       "learning_rate": 5.507537688442211e-05,
-      "loss": 0.9962,
+      "loss": 0.9966,
       "step": 726
     },
     {
       "epoch": 0.3233985765124555,
-      "grad_norm": 0.3954409062862396,
+      "grad_norm": 0.39778733253479004,
       "learning_rate": 5.487437185929648e-05,
-      "loss": 0.8989,
+      "loss": 0.8983,
       "step": 727
     },
     {
       "epoch": 0.3238434163701068,
-      "grad_norm": 0.371358722448349,
+      "grad_norm": 0.3704332709312439,
       "learning_rate": 5.467336683417086e-05,
-      "loss": 0.9655,
+      "loss": 0.9645,
       "step": 728
     },
     {
       "epoch": 0.32428825622775803,
-      "grad_norm": 0.4016554355621338,
+      "grad_norm": 0.39981192350387573,
       "learning_rate": 5.4472361809045224e-05,
-      "loss": 0.9298,
+      "loss": 0.9289,
       "step": 729
     },
     {
       "epoch": 0.32473309608540923,
-      "grad_norm": 0.45096760988235474,
+      "grad_norm": 0.44905000925064087,
       "learning_rate": 5.4271356783919604e-05,
-      "loss": 1.0633,
+      "loss": 1.0632,
       "step": 730
     },
     {
       "epoch": 0.3251779359430605,
-      "grad_norm": 0.42669543623924255,
+      "grad_norm": 0.4241604208946228,
       "learning_rate": 5.407035175879397e-05,
-      "loss": 0.9897,
+      "loss": 0.9888,
       "step": 731
     },
     {
       "epoch": 0.32562277580071175,
-      "grad_norm": 0.3072949945926666,
+      "grad_norm": 0.30892229080200195,
       "learning_rate": 5.386934673366835e-05,
-      "loss": 0.8538,
+      "loss": 0.8545,
       "step": 732
     },
     {
       "epoch": 0.326067615658363,
-      "grad_norm": 0.3681629002094269,
+      "grad_norm": 0.36549416184425354,
       "learning_rate": 5.3668341708542716e-05,
       "loss": 0.9532,
       "step": 733
     },
     {
       "epoch": 0.3265124555160142,
-      "grad_norm": 0.3667491674423218,
+      "grad_norm": 0.36795175075531006,
       "learning_rate": 5.346733668341709e-05,
-      "loss": 0.9971,
+      "loss": 0.9992,
       "step": 734
     },
     {
       "epoch": 0.32695729537366547,
-      "grad_norm": 0.3561696410179138,
+      "grad_norm": 0.3399661183357239,
       "learning_rate": 5.3266331658291455e-05,
-      "loss": 0.9332,
+      "loss": 0.9349,
       "step": 735
     },
     {
       "epoch": 0.3274021352313167,
-      "grad_norm": 0.51035076379776,
+      "grad_norm": 0.5193089842796326,
       "learning_rate": 5.3065326633165835e-05,
-      "loss": 0.9938,
+      "loss": 0.9922,
       "step": 736
     },
     {
       "epoch": 0.327846975088968,
-      "grad_norm": 0.30791398882865906,
+      "grad_norm": 0.3065936863422394,
       "learning_rate": 5.28643216080402e-05,
-      "loss": 0.9195,
+      "loss": 0.9197,
       "step": 737
     },
     {
       "epoch": 0.32829181494661924,
-      "grad_norm": 0.37262293696403503,
+      "grad_norm": 0.37638553977012634,
       "learning_rate": 5.266331658291458e-05,
-      "loss": 0.8857,
+      "loss": 0.8854,
       "step": 738
     },
     {
       "epoch": 0.32873665480427045,
-      "grad_norm": 0.3665037751197815,
+      "grad_norm": 0.36696720123291016,
       "learning_rate": 5.246231155778895e-05,
-      "loss": 0.8634,
+      "loss": 0.8642,
       "step": 739
     },
     {
       "epoch": 0.3291814946619217,
-      "grad_norm": 0.3333563804626465,
+      "grad_norm": 0.33234134316444397,
       "learning_rate": 5.226130653266332e-05,
       "loss": 0.9507,
       "step": 740
     },
     {
       "epoch": 0.32962633451957296,
-      "grad_norm": 0.3755170404911041,
+      "grad_norm": 0.3777725398540497,
       "learning_rate": 5.206030150753769e-05,
-      "loss": 0.9742,
+      "loss": 0.9751,
       "step": 741
     },
     {
       "epoch": 0.3300711743772242,
-      "grad_norm": 0.3400593101978302,
+      "grad_norm": 0.3410782516002655,
       "learning_rate": 5.1859296482412066e-05,
-      "loss": 0.8169,
+      "loss": 0.8185,
       "step": 742
     },
     {
       "epoch": 0.3305160142348754,
-      "grad_norm": 0.5976528525352478,
+      "grad_norm": 0.5856947302818298,
       "learning_rate": 5.165829145728643e-05,
-      "loss": 1.0596,
+      "loss": 1.0592,
       "step": 743
     },
     {
       "epoch": 0.3309608540925267,
-      "grad_norm": 0.3495856523513794,
+      "grad_norm": 0.35289299488067627,
       "learning_rate": 5.145728643216081e-05,
-      "loss": 1.1273,
+      "loss": 1.1304,
       "step": 744
     },
     {
       "epoch": 0.33140569395017794,
-      "grad_norm": 0.3898337483406067,
+      "grad_norm": 0.4018556475639343,
       "learning_rate": 5.125628140703518e-05,
-      "loss": 1.0512,
+      "loss": 1.0516,
       "step": 745
     },
     {
       "epoch": 0.3318505338078292,
-      "grad_norm": 0.4057525396347046,
+      "grad_norm": 0.402352899312973,
       "learning_rate": 5.1055276381909544e-05,
-      "loss": 1.0474,
+      "loss": 1.0456,
       "step": 746
     },
     {
       "epoch": 0.33229537366548045,
-      "grad_norm": 0.365829199552536,
+      "grad_norm": 0.36698055267333984,
       "learning_rate": 5.0854271356783924e-05,
-      "loss": 1.0095,
+      "loss": 1.0091,
       "step": 747
     },
     {
       "epoch": 0.33274021352313166,
-      "grad_norm": 0.3495554029941559,
+      "grad_norm": 0.34098100662231445,
       "learning_rate": 5.065326633165829e-05,
-      "loss": 0.8723,
+      "loss": 0.8725,
       "step": 748
     },
     {
       "epoch": 0.3331850533807829,
-      "grad_norm": 0.3964870572090149,
+      "grad_norm": 0.3870490789413452,
       "learning_rate": 5.045226130653266e-05,
       "loss": 0.9775,
       "step": 749
     },
     {
       "epoch": 0.33362989323843417,
-      "grad_norm": 0.3591833710670471,
+      "grad_norm": 0.3566199839115143,
       "learning_rate": 5.0251256281407036e-05,
-      "loss": 0.8438,
+      "loss": 0.842,
       "step": 750
     },
     {
       "epoch": 0.33407473309608543,
-      "grad_norm": 0.4429285526275635,
+      "grad_norm": 0.4538438618183136,
       "learning_rate": 5.005025125628141e-05,
-      "loss": 0.9144,
+      "loss": 0.9158,
       "step": 751
     },
     {
       "epoch": 0.33451957295373663,
-      "grad_norm": 0.3427349328994751,
+      "grad_norm": 0.3462452292442322,
       "learning_rate": 4.984924623115578e-05,
-      "loss": 0.8728,
+      "loss": 0.8715,
       "step": 752
     },
     {
       "epoch": 0.3349644128113879,
-      "grad_norm": 0.4042606055736542,
+      "grad_norm": 0.40105798840522766,
       "learning_rate": 4.9648241206030155e-05,
-      "loss": 1.0932,
+      "loss": 1.0943,
       "step": 753
     },
     {
       "epoch": 0.33540925266903915,
-      "grad_norm": 0.38601040840148926,
+      "grad_norm": 0.38309648633003235,
       "learning_rate": 4.944723618090453e-05,
-      "loss": 0.8947,
+      "loss": 0.8961,
       "step": 754
     },
     {
       "epoch": 0.3358540925266904,
-      "grad_norm": 0.31526607275009155,
+      "grad_norm": 0.31808745861053467,
       "learning_rate": 4.92462311557789e-05,
-      "loss": 0.8783,
+      "loss": 0.8785,
       "step": 755
     },
     {
       "epoch": 0.33629893238434166,
-      "grad_norm": 0.3799164593219757,
+      "grad_norm": 0.3787521421909332,
       "learning_rate": 4.9045226130653274e-05,
-      "loss": 0.894,
+      "loss": 0.8932,
       "step": 756
     },
     {
       "epoch": 0.33674377224199287,
-      "grad_norm": 0.36352670192718506,
+      "grad_norm": 0.36229708790779114,
       "learning_rate": 4.884422110552764e-05,
-      "loss": 1.0077,
+      "loss": 1.01,
       "step": 757
     },
     {
       "epoch": 0.3371886120996441,
-      "grad_norm": 0.37971585988998413,
+      "grad_norm": 0.3797127604484558,
       "learning_rate": 4.864321608040201e-05,
-      "loss": 0.8504,
+      "loss": 0.8492,
       "step": 758
     },
     {
       "epoch": 0.3376334519572954,
-      "grad_norm": 0.3846040666103363,
+      "grad_norm": 0.3839523196220398,
       "learning_rate": 4.844221105527638e-05,
-      "loss": 0.9923,
+      "loss": 0.9927,
       "step": 759
     },
     {
       "epoch": 0.33807829181494664,
-      "grad_norm": 0.3028171956539154,
+      "grad_norm": 0.3037039637565613,
       "learning_rate": 4.824120603015075e-05,
-      "loss": 0.9328,
+      "loss": 0.9326,
       "step": 760
     },
     {
       "epoch": 0.33852313167259784,
-      "grad_norm": 0.35201942920684814,
+      "grad_norm": 0.35209789872169495,
       "learning_rate": 4.8040201005025125e-05,
-      "loss": 0.8622,
+      "loss": 0.8634,
       "step": 761
     },
     {
       "epoch": 0.3389679715302491,
-      "grad_norm": 0.3402314782142639,
+      "grad_norm": 0.338007390499115,
       "learning_rate": 4.78391959798995e-05,
-      "loss": 1.0578,
+      "loss": 1.0579,
       "step": 762
     },
     {
       "epoch": 0.33941281138790036,
-      "grad_norm": 0.43474265933036804,
+      "grad_norm": 0.43554559350013733,
       "learning_rate": 4.763819095477387e-05,
-      "loss": 1.0589,
+      "loss": 1.0572,
       "step": 763
     },
     {
       "epoch": 0.3398576512455516,
-      "grad_norm": 0.4756290912628174,
+      "grad_norm": 0.47670868039131165,
       "learning_rate": 4.7437185929648244e-05,
-      "loss": 1.1027,
+      "loss": 1.1046,
       "step": 764
     },
     {
       "epoch": 0.3403024911032028,
-      "grad_norm": 0.3077346682548523,
+      "grad_norm": 0.3101906180381775,
       "learning_rate": 4.723618090452262e-05,
-      "loss": 0.9635,
+      "loss": 0.9641,
       "step": 765
     },
     {
       "epoch": 0.3407473309608541,
-      "grad_norm": 0.36826518177986145,
+      "grad_norm": 0.3647597134113312,
       "learning_rate": 4.703517587939698e-05,
-      "loss": 0.9687,
+      "loss": 0.9678,
       "step": 766
     },
     {
       "epoch": 0.34119217081850534,
-      "grad_norm": 0.4034232199192047,
+      "grad_norm": 0.40179431438446045,
       "learning_rate": 4.6834170854271356e-05,
-      "loss": 0.9209,
+      "loss": 0.9187,
       "step": 767
     },
     {
       "epoch": 0.3416370106761566,
-      "grad_norm": 0.41342729330062866,
+      "grad_norm": 0.41022738814353943,
       "learning_rate": 4.663316582914573e-05,
-      "loss": 0.9213,
+      "loss": 0.9218,
       "step": 768
     },
     {
       "epoch": 0.34208185053380785,
-      "grad_norm": 0.3749755620956421,
+      "grad_norm": 0.3737107515335083,
       "learning_rate": 4.64321608040201e-05,
-      "loss": 0.9386,
+      "loss": 0.9396,
       "step": 769
     },
     {
       "epoch": 0.34252669039145905,
-      "grad_norm": 0.3411119878292084,
+      "grad_norm": 0.34264299273490906,
       "learning_rate": 4.6231155778894475e-05,
-      "loss": 0.8805,
+      "loss": 0.8803,
       "step": 770
     },
     {
       "epoch": 0.3429715302491103,
-      "grad_norm": 0.3668997883796692,
+      "grad_norm": 0.363738089799881,
       "learning_rate": 4.603015075376885e-05,
-      "loss": 0.9887,
+      "loss": 0.9905,
       "step": 771
     },
     {
       "epoch": 0.34341637010676157,
-      "grad_norm": 0.45816823840141296,
+      "grad_norm": 0.4605758488178253,
       "learning_rate": 4.582914572864322e-05,
-      "loss": 0.9031,
+      "loss": 0.904,
       "step": 772
     },
     {
       "epoch": 0.34386120996441283,
-      "grad_norm": 0.3586982786655426,
+      "grad_norm": 0.3742416799068451,
       "learning_rate": 4.5628140703517594e-05,
-      "loss": 0.9774,
+      "loss": 0.9772,
       "step": 773
     },
     {
       "epoch": 0.34430604982206403,
-      "grad_norm": 0.42974361777305603,
+      "grad_norm": 0.4309648871421814,
       "learning_rate": 4.542713567839196e-05,
-      "loss": 1.0421,
+      "loss": 1.0439,
       "step": 774
     },
     {
       "epoch": 0.3447508896797153,
-      "grad_norm": 0.36137211322784424,
+      "grad_norm": 0.3657279908657074,
       "learning_rate": 4.522613065326633e-05,
-      "loss": 1.0171,
+      "loss": 1.0162,
       "step": 775
     },
     {
       "epoch": 0.34519572953736655,
-      "grad_norm": 0.413507342338562,
+      "grad_norm": 0.4098432958126068,
       "learning_rate": 4.5025125628140706e-05,
-      "loss": 0.9786,
+      "loss": 0.9788,
       "step": 776
     },
     {
       "epoch": 0.3456405693950178,
-      "grad_norm": 0.42441946268081665,
+      "grad_norm": 0.4177950918674469,
       "learning_rate": 4.482412060301508e-05,
-      "loss": 1.01,
+      "loss": 1.0105,
       "step": 777
     },
     {
       "epoch": 0.34608540925266906,
-      "grad_norm": 0.35072627663612366,
+      "grad_norm": 0.35204648971557617,
       "learning_rate": 4.462311557788945e-05,
-      "loss": 0.9709,
+      "loss": 0.9713,
       "step": 778
     },
     {
       "epoch": 0.34653024911032027,
-      "grad_norm": 0.37332162261009216,
+      "grad_norm": 0.3725673258304596,
       "learning_rate": 4.4422110552763825e-05,
-      "loss": 0.9844,
+      "loss": 0.9843,
       "step": 779
     },
     {
       "epoch": 0.3469750889679715,
-      "grad_norm": 0.388919860124588,
+      "grad_norm": 0.38592076301574707,
       "learning_rate": 4.42211055276382e-05,
-      "loss": 0.8525,
+      "loss": 0.8528,
       "step": 780
     },
     {
       "epoch": 0.3474199288256228,
-      "grad_norm": 0.3552108407020569,
+      "grad_norm": 0.35379818081855774,
       "learning_rate": 4.4020100502512564e-05,
-      "loss": 0.9713,
+      "loss": 0.9712,
       "step": 781
     },
     {
       "epoch": 0.34786476868327404,
-      "grad_norm": 0.3655258119106293,
+      "grad_norm": 0.3713836371898651,
       "learning_rate": 4.381909547738694e-05,
-      "loss": 0.9532,
+      "loss": 0.9531,
       "step": 782
     },
     {
       "epoch": 0.34830960854092524,
-      "grad_norm": 0.38755300641059875,
+      "grad_norm": 0.3799002766609192,
       "learning_rate": 4.3618090452261303e-05,
-      "loss": 0.9346,
+      "loss": 0.9348,
       "step": 783
     },
     {
       "epoch": 0.3487544483985765,
-      "grad_norm": 0.4547751843929291,
+      "grad_norm": 0.34504106640815735,
       "learning_rate": 4.3417085427135676e-05,
-      "loss": 0.8872,
+      "loss": 0.8863,
       "step": 784
     },
     {
       "epoch": 0.34919928825622776,
-      "grad_norm": 0.2990303039550781,
+      "grad_norm": 0.2971758544445038,
       "learning_rate": 4.321608040201005e-05,
-      "loss": 0.8979,
+      "loss": 0.8983,
       "step": 785
     },
     {
       "epoch": 0.349644128113879,
-      "grad_norm": 0.3581852614879608,
+      "grad_norm": 0.34722504019737244,
       "learning_rate": 4.301507537688442e-05,
-      "loss": 0.9466,
+      "loss": 0.9456,
       "step": 786
     },
     {
       "epoch": 0.3500889679715303,
-      "grad_norm": 0.33662378787994385,
+      "grad_norm": 0.3370276987552643,
       "learning_rate": 4.2814070351758795e-05,
-      "loss": 0.9274,
+      "loss": 0.9288,
       "step": 787
     },
     {
       "epoch": 0.3505338078291815,
-      "grad_norm": 0.34652361273765564,
+      "grad_norm": 0.3471807837486267,
       "learning_rate": 4.261306532663317e-05,
-      "loss": 0.9791,
+      "loss": 0.98,
       "step": 788
     },
     {
       "epoch": 0.35097864768683273,
-      "grad_norm": 0.39401039481163025,
+      "grad_norm": 0.3932972252368927,
       "learning_rate": 4.241206030150754e-05,
-      "loss": 0.9485,
+      "loss": 0.9479,
       "step": 789
     },
     {
       "epoch": 0.351423487544484,
-      "grad_norm": 0.36272600293159485,
+      "grad_norm": 0.36573097109794617,
       "learning_rate": 4.2211055276381914e-05,
-      "loss": 0.9341,
+      "loss": 0.9339,
       "step": 790
     },
     {
       "epoch": 0.35186832740213525,
-      "grad_norm": 0.3438786566257477,
+      "grad_norm": 0.34445399045944214,
       "learning_rate": 4.201005025125628e-05,
-      "loss": 0.8061,
+      "loss": 0.807,
       "step": 791
     },
     {
       "epoch": 0.35231316725978645,
-      "grad_norm": 0.396484911441803,
+      "grad_norm": 0.39604124426841736,
       "learning_rate": 4.180904522613065e-05,
-      "loss": 0.9506,
+      "loss": 0.9503,
       "step": 792
     },
     {
       "epoch": 0.3527580071174377,
-      "grad_norm": 0.31982484459877014,
+      "grad_norm": 0.3157517910003662,
       "learning_rate": 4.1608040201005026e-05,
-      "loss": 0.9372,
+      "loss": 0.9381,
       "step": 793
     },
     {
       "epoch": 0.35320284697508897,
-      "grad_norm": 0.323798805475235,
+      "grad_norm": 0.3242207467556,
       "learning_rate": 4.14070351758794e-05,
-      "loss": 0.9342,
+      "loss": 0.934,
       "step": 794
     },
     {
       "epoch": 0.3536476868327402,
-      "grad_norm": 0.3632016181945801,
+      "grad_norm": 0.3632892966270447,
       "learning_rate": 4.120603015075377e-05,
-      "loss": 0.9682,
+      "loss": 0.9653,
       "step": 795
     },
     {
       "epoch": 0.3540925266903915,
-      "grad_norm": 0.3542345464229584,
+      "grad_norm": 0.35393667221069336,
       "learning_rate": 4.1005025125628145e-05,
-      "loss": 0.9129,
+      "loss": 0.9132,
       "step": 796
     },
     {
       "epoch": 0.3545373665480427,
-      "grad_norm": 0.48540955781936646,
+      "grad_norm": 0.36036303639411926,
       "learning_rate": 4.080402010050252e-05,
-      "loss": 1.0319,
+      "loss": 1.0329,
       "step": 797
     },
     {
       "epoch": 0.35498220640569395,
-      "grad_norm": 0.37020549178123474,
+      "grad_norm": 0.3710475265979767,
       "learning_rate": 4.060301507537689e-05,
-      "loss": 0.8719,
+      "loss": 0.8732,
       "step": 798
     },
     {
       "epoch": 0.3554270462633452,
-      "grad_norm": 0.3740653097629547,
+      "grad_norm": 0.37702277302742004,
       "learning_rate": 4.040201005025126e-05,
-      "loss": 0.9464,
+      "loss": 0.9467,
       "step": 799
     },
     {
       "epoch": 0.35587188612099646,
-      "grad_norm": 0.3594771921634674,
+      "grad_norm": 0.3579627275466919,
       "learning_rate": 4.020100502512563e-05,
-      "loss": 0.9592,
+      "loss": 0.9603,
       "step": 800
     },
     {
       "epoch": 0.35631672597864766,
-      "grad_norm": 0.4490291476249695,
+      "grad_norm": 0.44332355260849,
       "learning_rate": 4e-05,
-      "loss": 0.9624,
+      "loss": 0.9611,
       "step": 801
     },
     {
       "epoch": 0.3567615658362989,
-      "grad_norm": 0.3087805509567261,
+      "grad_norm": 0.30799567699432373,
       "learning_rate": 3.9798994974874376e-05,
-      "loss": 0.9905,
+      "loss": 0.9908,
       "step": 802
     },
     {
       "epoch": 0.3572064056939502,
-      "grad_norm": 0.39388784766197205,
+      "grad_norm": 0.38729211688041687,
       "learning_rate": 3.959798994974875e-05,
-      "loss": 0.9644,
+      "loss": 0.9629,
       "step": 803
     },
     {
       "epoch": 0.35765124555160144,
-      "grad_norm": 0.411502867937088,
+      "grad_norm": 0.4034986197948456,
       "learning_rate": 3.9396984924623115e-05,
-      "loss": 0.9547,
+      "loss": 0.9545,
       "step": 804
     },
     {
       "epoch": 0.3580960854092527,
-      "grad_norm": 0.3525221049785614,
+      "grad_norm": 0.35203176736831665,
       "learning_rate": 3.919597989949749e-05,
-      "loss": 0.8875,
+      "loss": 0.8863,
       "step": 805
     },
     {
       "epoch": 0.3585409252669039,
-      "grad_norm": 0.3365210294723511,
+      "grad_norm": 0.3381657004356384,
       "learning_rate": 3.899497487437186e-05,
-      "loss": 0.913,
+      "loss": 0.9144,
       "step": 806
     },
     {
       "epoch": 0.35898576512455516,
-      "grad_norm": 0.3405802249908447,
+      "grad_norm": 0.34250691533088684,
       "learning_rate": 3.8793969849246234e-05,
-      "loss": 0.8679,
+      "loss": 0.8687,
       "step": 807
     },
     {
       "epoch": 0.3594306049822064,
-      "grad_norm": 0.33175504207611084,
+      "grad_norm": 0.3308602273464203,
       "learning_rate": 3.85929648241206e-05,
-      "loss": 0.929,
+      "loss": 0.9292,
       "step": 808
     },
     {
       "epoch": 0.35987544483985767,
-      "grad_norm": 0.3878396451473236,
+      "grad_norm": 0.3888717591762543,
       "learning_rate": 3.8391959798994973e-05,
-      "loss": 0.9765,
+      "loss": 0.9774,
       "step": 809
     },
     {
       "epoch": 0.3603202846975089,
-      "grad_norm": 0.3880787491798401,
+      "grad_norm": 0.37949973344802856,
       "learning_rate": 3.8190954773869346e-05,
-      "loss": 0.9282,
+      "loss": 0.9278,
       "step": 810
     },
     {
       "epoch": 0.36076512455516013,
-      "grad_norm": 0.3783409297466278,
+      "grad_norm": 0.3511112332344055,
       "learning_rate": 3.798994974874372e-05,
-      "loss": 0.9704,
+      "loss": 0.9714,
       "step": 811
     },
     {
       "epoch": 0.3612099644128114,
-      "grad_norm": 0.37962606549263,
+      "grad_norm": 0.3813224732875824,
       "learning_rate": 3.778894472361809e-05,
-      "loss": 1.0464,
+      "loss": 1.0487,
       "step": 812
     },
     {
       "epoch": 0.36165480427046265,
-      "grad_norm": 0.4055241644382477,
+      "grad_norm": 0.4005330502986908,
       "learning_rate": 3.7587939698492465e-05,
-      "loss": 0.9668,
+      "loss": 0.9673,
       "step": 813
     },
     {
       "epoch": 0.3620996441281139,
-      "grad_norm": 0.3240121304988861,
+      "grad_norm": 0.3238542675971985,
       "learning_rate": 3.738693467336684e-05,
-      "loss": 0.9415,
+      "loss": 0.9409,
       "step": 814
     },
     {
       "epoch": 0.3625444839857651,
-      "grad_norm": 0.29463499784469604,
+      "grad_norm": 0.29547253251075745,
       "learning_rate": 3.7185929648241204e-05,
-      "loss": 0.9489,
+      "loss": 0.9494,
       "step": 815
     },
     {
       "epoch": 0.36298932384341637,
-      "grad_norm": 0.4620193839073181,
+      "grad_norm": 0.45984339714050293,
       "learning_rate": 3.698492462311558e-05,
-      "loss": 0.8513,
+      "loss": 0.8507,
       "step": 816
     },
     {
       "epoch": 0.3634341637010676,
-      "grad_norm": 0.3686508536338806,
+      "grad_norm": 0.3637581169605255,
       "learning_rate": 3.678391959798995e-05,
-      "loss": 1.016,
+      "loss": 1.0145,
       "step": 817
     },
     {
       "epoch": 0.3638790035587189,
-      "grad_norm": 0.366397500038147,
+      "grad_norm": 0.3606102764606476,
       "learning_rate": 3.658291457286432e-05,
-      "loss": 0.8772,
+      "loss": 0.877,
       "step": 818
     },
     {
       "epoch": 0.3643238434163701,
-      "grad_norm": 0.3382261097431183,
+      "grad_norm": 0.33879634737968445,
       "learning_rate": 3.6381909547738696e-05,
-      "loss": 0.9158,
+      "loss": 0.9165,
       "step": 819
     },
     {
       "epoch": 0.36476868327402134,
-      "grad_norm": 0.3870478570461273,
+      "grad_norm": 0.3897345960140228,
       "learning_rate": 3.618090452261307e-05,
-      "loss": 0.9776,
+      "loss": 0.9762,
       "step": 820
     },
     {
       "epoch": 0.3652135231316726,
-      "grad_norm": 0.40006503462791443,
+      "grad_norm": 0.39814460277557373,
       "learning_rate": 3.597989949748744e-05,
-      "loss": 0.9127,
+      "loss": 0.9111,
       "step": 821
     },
     {
       "epoch": 0.36565836298932386,
-      "grad_norm": 0.41942328214645386,
+      "grad_norm": 0.40529152750968933,
       "learning_rate": 3.5778894472361815e-05,
-      "loss": 0.8786,
+      "loss": 0.8778,
       "step": 822
     },
     {
       "epoch": 0.36610320284697506,
-      "grad_norm": 0.34577417373657227,
+      "grad_norm": 0.34325361251831055,
       "learning_rate": 3.557788944723618e-05,
-      "loss": 0.9029,
+      "loss": 0.903,
       "step": 823
     },
     {
       "epoch": 0.3665480427046263,
-      "grad_norm": 0.3666742146015167,
+      "grad_norm": 0.36211341619491577,
       "learning_rate": 3.5376884422110554e-05,
-      "loss": 0.9616,
+      "loss": 0.963,
       "step": 824
     },
     {
       "epoch": 0.3669928825622776,
-      "grad_norm": 0.41306397318840027,
+      "grad_norm": 0.4107413589954376,
       "learning_rate": 3.517587939698493e-05,
-      "loss": 0.8994,
+      "loss": 0.9007,
       "step": 825
     },
     {
       "epoch": 0.36743772241992884,
-      "grad_norm": 0.354464054107666,
+      "grad_norm": 0.35432523488998413,
       "learning_rate": 3.49748743718593e-05,
-      "loss": 0.8398,
+      "loss": 0.8406,
       "step": 826
     },
     {
       "epoch": 0.3678825622775801,
-      "grad_norm": 0.40960076451301575,
+      "grad_norm": 0.4027344286441803,
       "learning_rate": 3.4773869346733667e-05,
       "loss": 1.0062,
       "step": 827
     },
     {
       "epoch": 0.3683274021352313,
-      "grad_norm": 0.35255295038223267,
+      "grad_norm": 0.34778210520744324,
       "learning_rate": 3.457286432160804e-05,
-      "loss": 1.013,
+      "loss": 1.0124,
       "step": 828
     },
     {
       "epoch": 0.36877224199288255,
-      "grad_norm": 0.320544570684433,
+      "grad_norm": 0.3145458996295929,
       "learning_rate": 3.437185929648241e-05,
-      "loss": 0.8456,
+      "loss": 0.8455,
       "step": 829
     },
     {
       "epoch": 0.3692170818505338,
-      "grad_norm": 0.3942032754421234,
+      "grad_norm": 0.39053332805633545,
       "learning_rate": 3.4170854271356785e-05,
-      "loss": 0.9627,
+      "loss": 0.9635,
       "step": 830
     },
     {
       "epoch": 0.36966192170818507,
-      "grad_norm": 0.34646615386009216,
+      "grad_norm": 0.3456934988498688,
       "learning_rate": 3.396984924623116e-05,
-      "loss": 0.9397,
+      "loss": 0.9399,
       "step": 831
     },
     {
       "epoch": 0.3701067615658363,
-      "grad_norm": 0.3462013006210327,
+      "grad_norm": 0.34614813327789307,
       "learning_rate": 3.3768844221105525e-05,
-      "loss": 0.9722,
+      "loss": 0.9724,
       "step": 832
     },
     {
       "epoch": 0.37055160142348753,
-      "grad_norm": 0.44638511538505554,
+      "grad_norm": 0.45311570167541504,
       "learning_rate": 3.35678391959799e-05,
-      "loss": 1.0433,
+      "loss": 1.0435,
       "step": 833
     },
     {
       "epoch": 0.3709964412811388,
-      "grad_norm": 0.3865186870098114,
+      "grad_norm": 0.3952670395374298,
       "learning_rate": 3.336683417085427e-05,
-      "loss": 0.9243,
+      "loss": 0.9251,
       "step": 834
     },
     {
       "epoch": 0.37144128113879005,
-      "grad_norm": 0.3224859833717346,
+      "grad_norm": 0.3246530592441559,
       "learning_rate": 3.3165829145728643e-05,
-      "loss": 0.9197,
+      "loss": 0.9194,
       "step": 835
     },
     {
       "epoch": 0.3718861209964413,
-      "grad_norm": 0.3511156439781189,
+      "grad_norm": 0.3489208221435547,
       "learning_rate": 3.2964824120603016e-05,
-      "loss": 0.951,
+      "loss": 0.9522,
       "step": 836
     },
     {
       "epoch": 0.3723309608540925,
-      "grad_norm": 0.3146522343158722,
+      "grad_norm": 0.3140431046485901,
       "learning_rate": 3.276381909547739e-05,
-      "loss": 0.9514,
+      "loss": 0.9511,
       "step": 837
     },
     {
       "epoch": 0.37277580071174377,
-      "grad_norm": 0.3094431757926941,
+      "grad_norm": 0.3103282153606415,
       "learning_rate": 3.256281407035176e-05,
-      "loss": 0.8729,
+      "loss": 0.8739,
       "step": 838
     },
     {
       "epoch": 0.373220640569395,
-      "grad_norm": 0.3724329471588135,
+      "grad_norm": 0.3995440602302551,
       "learning_rate": 3.2361809045226135e-05,
-      "loss": 0.9897,
+      "loss": 0.9918,
       "step": 839
     },
     {
       "epoch": 0.3736654804270463,
-      "grad_norm": 0.3881984353065491,
+      "grad_norm": 0.39125266671180725,
       "learning_rate": 3.21608040201005e-05,
-      "loss": 0.9755,
+      "loss": 0.9751,
       "step": 840
     },
     {
       "epoch": 0.3741103202846975,
-      "grad_norm": 0.30279484391212463,
+      "grad_norm": 0.3043217658996582,
       "learning_rate": 3.1959798994974875e-05,
-      "loss": 0.8903,
+      "loss": 0.8899,
       "step": 841
     },
     {
       "epoch": 0.37455516014234874,
-      "grad_norm": 0.42050573229789734,
+      "grad_norm": 0.4185977280139923,
       "learning_rate": 3.175879396984925e-05,
-      "loss": 0.9163,
+      "loss": 0.9169,
       "step": 842
     },
     {
       "epoch": 0.375,
-      "grad_norm": 0.3830379545688629,
+      "grad_norm": 0.3753814697265625,
       "learning_rate": 3.155778894472362e-05,
-      "loss": 0.9519,
+      "loss": 0.951,
       "step": 843
     },
     {
       "epoch": 0.37544483985765126,
-      "grad_norm": 0.32881635427474976,
+      "grad_norm": 0.33144545555114746,
       "learning_rate": 3.1356783919597993e-05,
-      "loss": 0.9225,
+      "loss": 0.9228,
       "step": 844
     },
     {
       "epoch": 0.3758896797153025,
-      "grad_norm": 0.30189692974090576,
+      "grad_norm": 0.3024055063724518,
       "learning_rate": 3.1155778894472366e-05,
-      "loss": 0.916,
+      "loss": 0.9157,
       "step": 845
     },
     {
       "epoch": 0.3763345195729537,
-      "grad_norm": 0.37163209915161133,
+      "grad_norm": 0.3643604815006256,
       "learning_rate": 3.095477386934674e-05,
-      "loss": 0.9399,
+      "loss": 0.9387,
       "step": 846
     },
     {
       "epoch": 0.376779359430605,
-      "grad_norm": 0.43418970704078674,
+      "grad_norm": 0.426024854183197,
       "learning_rate": 3.075376884422111e-05,
-      "loss": 1.0015,
+      "loss": 1.0013,
       "step": 847
     },
     {
       "epoch": 0.37722419928825623,
-      "grad_norm": 0.3178066313266754,
+      "grad_norm": 0.3174848258495331,
       "learning_rate": 3.055276381909548e-05,
-      "loss": 0.9131,
+      "loss": 0.914,
       "step": 848
     },
     {
       "epoch": 0.3776690391459075,
-      "grad_norm": 0.3209547996520996,
+      "grad_norm": 0.3318672478199005,
       "learning_rate": 3.0351758793969855e-05,
-      "loss": 0.9498,
+      "loss": 0.9502,
       "step": 849
     },
     {
       "epoch": 0.3781138790035587,
-      "grad_norm": 0.4426248371601105,
+      "grad_norm": 0.4423961937427521,
       "learning_rate": 3.015075376884422e-05,
-      "loss": 0.9335,
+      "loss": 0.9341,
       "step": 850
     },
     {
       "epoch": 0.37855871886120995,
-      "grad_norm": 0.41413775086402893,
+      "grad_norm": 0.40427151322364807,
       "learning_rate": 2.994974874371859e-05,
-      "loss": 0.8906,
+      "loss": 0.8887,
       "step": 851
     },
     {
       "epoch": 0.3790035587188612,
-      "grad_norm": 0.3665112555027008,
+      "grad_norm": 0.3664209246635437,
       "learning_rate": 2.9748743718592964e-05,
-      "loss": 0.8849,
+      "loss": 0.8858,
       "step": 852
     },
     {
       "epoch": 0.37944839857651247,
-      "grad_norm": 0.36841294169425964,
+      "grad_norm": 0.3712497353553772,
       "learning_rate": 2.9547738693467337e-05,
       "loss": 0.8004,
       "step": 853
     },
     {
       "epoch": 0.3798932384341637,
-      "grad_norm": 0.4228864014148712,
+      "grad_norm": 0.4214048981666565,
       "learning_rate": 2.934673366834171e-05,
-      "loss": 1.0018,
+      "loss": 1.0013,
       "step": 854
     },
     {
       "epoch": 0.38033807829181493,
-      "grad_norm": 0.34962666034698486,
+      "grad_norm": 0.3496900498867035,
       "learning_rate": 2.914572864321608e-05,
-      "loss": 0.9105,
+      "loss": 0.908,
       "step": 855
     },
     {
       "epoch": 0.3807829181494662,
-      "grad_norm": 0.4448448121547699,
+      "grad_norm": 0.4434766471385956,
       "learning_rate": 2.8944723618090452e-05,
-      "loss": 0.9391,
+      "loss": 0.9394,
       "step": 856
     },
     {
       "epoch": 0.38122775800711745,
-      "grad_norm": 0.36843836307525635,
+      "grad_norm": 0.37649184465408325,
       "learning_rate": 2.8743718592964825e-05,
-      "loss": 0.9911,
+      "loss": 0.9892,
       "step": 857
     },
     {
       "epoch": 0.3816725978647687,
-      "grad_norm": 0.35289087891578674,
+      "grad_norm": 0.35058286786079407,
       "learning_rate": 2.8542713567839198e-05,
-      "loss": 0.9575,
+      "loss": 0.9576,
       "step": 858
     },
     {
       "epoch": 0.3821174377224199,
-      "grad_norm": 0.39545342326164246,
+      "grad_norm": 0.3940383493900299,
       "learning_rate": 2.8341708542713568e-05,
-      "loss": 0.9074,
+      "loss": 0.9078,
       "step": 859
     },
     {
       "epoch": 0.38256227758007116,
-      "grad_norm": 0.3101719319820404,
+      "grad_norm": 0.31289801001548767,
       "learning_rate": 2.814070351758794e-05,
-      "loss": 0.8479,
+      "loss": 0.8471,
       "step": 860
     },
     {
       "epoch": 0.3830071174377224,
-      "grad_norm": 0.38031789660453796,
+      "grad_norm": 0.3773019313812256,
       "learning_rate": 2.7939698492462314e-05,
-      "loss": 1.0281,
+      "loss": 1.0291,
       "step": 861
     },
     {
       "epoch": 0.3834519572953737,
-      "grad_norm": 0.38144952058792114,
+      "grad_norm": 0.38689449429512024,
       "learning_rate": 2.7738693467336686e-05,
-      "loss": 0.9668,
+      "loss": 0.965,
       "step": 862
     },
     {
       "epoch": 0.38389679715302494,
-      "grad_norm": 0.41572901606559753,
+      "grad_norm": 0.4001306891441345,
       "learning_rate": 2.7537688442211056e-05,
-      "loss": 0.9258,
+      "loss": 0.9251,
       "step": 863
     },
     {
       "epoch": 0.38434163701067614,
-      "grad_norm": 0.2982839345932007,
+      "grad_norm": 0.2970896363258362,
       "learning_rate": 2.733668341708543e-05,
-      "loss": 0.8807,
+      "loss": 0.8795,
       "step": 864
     },
     {
       "epoch": 0.3847864768683274,
-      "grad_norm": 0.3221079111099243,
+      "grad_norm": 0.3230406939983368,
       "learning_rate": 2.7135678391959802e-05,
-      "loss": 0.9178,
+      "loss": 0.9162,
       "step": 865
     },
     {
       "epoch": 0.38523131672597866,
-      "grad_norm": 0.38755419850349426,
+      "grad_norm": 0.38683274388313293,
       "learning_rate": 2.6934673366834175e-05,
-      "loss": 0.9183,
+      "loss": 0.921,
       "step": 866
     },
     {
       "epoch": 0.3856761565836299,
-      "grad_norm": 0.36158278584480286,
+      "grad_norm": 0.36258265376091003,
       "learning_rate": 2.6733668341708545e-05,
-      "loss": 0.9274,
+      "loss": 0.9282,
       "step": 867
     },
     {
       "epoch": 0.3861209964412811,
-      "grad_norm": 0.308855801820755,
+      "grad_norm": 0.30888402462005615,
       "learning_rate": 2.6532663316582917e-05,
-      "loss": 0.9158,
+      "loss": 0.9154,
       "step": 868
     },
     {
       "epoch": 0.3865658362989324,
-      "grad_norm": 0.42090025544166565,
+      "grad_norm": 0.4258480370044708,
       "learning_rate": 2.633165829145729e-05,
-      "loss": 0.9392,
+      "loss": 0.9405,
       "step": 869
     },
     {
       "epoch": 0.38701067615658363,
-      "grad_norm": 0.3310891389846802,
+      "grad_norm": 0.33099818229675293,
       "learning_rate": 2.613065326633166e-05,
-      "loss": 0.9232,
+      "loss": 0.923,
       "step": 870
     },
     {
       "epoch": 0.3874555160142349,
-      "grad_norm": 0.3915763199329376,
+      "grad_norm": 0.3900362253189087,
       "learning_rate": 2.5929648241206033e-05,
-      "loss": 0.8819,
+      "loss": 0.8813,
       "step": 871
     },
     {
       "epoch": 0.3879003558718861,
-      "grad_norm": 0.3351011872291565,
+      "grad_norm": 0.3388621509075165,
       "learning_rate": 2.5728643216080406e-05,
-      "loss": 0.9757,
+      "loss": 0.9765,
       "step": 872
     },
     {
       "epoch": 0.38834519572953735,
-      "grad_norm": 0.32709652185440063,
+      "grad_norm": 0.32633256912231445,
       "learning_rate": 2.5527638190954772e-05,
-      "loss": 0.8823,
+      "loss": 0.8824,
       "step": 873
     },
     {
       "epoch": 0.3887900355871886,
-      "grad_norm": 0.34310227632522583,
+      "grad_norm": 0.34141066670417786,
       "learning_rate": 2.5326633165829145e-05,
-      "loss": 0.8623,
+      "loss": 0.862,
       "step": 874
     },
     {
       "epoch": 0.38923487544483987,
-      "grad_norm": 0.3940928876399994,
+      "grad_norm": 0.39966803789138794,
       "learning_rate": 2.5125628140703518e-05,
-      "loss": 0.9409,
+      "loss": 0.9406,
       "step": 875
     },
     {
       "epoch": 0.3896797153024911,
-      "grad_norm": 0.34565237164497375,
+      "grad_norm": 0.34703922271728516,
       "learning_rate": 2.492462311557789e-05,
-      "loss": 0.9369,
+      "loss": 0.9392,
       "step": 876
     },
     {
       "epoch": 0.39012455516014233,
-      "grad_norm": 0.37526604533195496,
+      "grad_norm": 0.3692745268344879,
       "learning_rate": 2.4723618090452264e-05,
-      "loss": 0.8391,
+      "loss": 0.838,
       "step": 877
     },
     {
       "epoch": 0.3905693950177936,
-      "grad_norm": 0.3760671019554138,
+      "grad_norm": 0.3813033401966095,
       "learning_rate": 2.4522613065326637e-05,
-      "loss": 0.9549,
+      "loss": 0.9557,
       "step": 878
     },
     {
       "epoch": 0.39101423487544484,
-      "grad_norm": 0.35706883668899536,
+      "grad_norm": 0.361794650554657,
       "learning_rate": 2.4321608040201007e-05,
-      "loss": 0.7821,
+      "loss": 0.7851,
       "step": 879
     },
     {
       "epoch": 0.3914590747330961,
-      "grad_norm": 0.37825968861579895,
+      "grad_norm": 0.3740581274032593,
       "learning_rate": 2.4120603015075376e-05,
-      "loss": 1.0344,
+      "loss": 1.0353,
       "step": 880
     },
     {
       "epoch": 0.3919039145907473,
-      "grad_norm": 0.3388988673686981,
+      "grad_norm": 0.3482462465763092,
       "learning_rate": 2.391959798994975e-05,
-      "loss": 0.9347,
+      "loss": 0.9339,
       "step": 881
     },
     {
       "epoch": 0.39234875444839856,
-      "grad_norm": 0.35239583253860474,
+      "grad_norm": 0.35234349966049194,
       "learning_rate": 2.3718592964824122e-05,
-      "loss": 0.8896,
+      "loss": 0.8892,
       "step": 882
     },
     {
       "epoch": 0.3927935943060498,
-      "grad_norm": 0.34962180256843567,
+      "grad_norm": 0.349231481552124,
       "learning_rate": 2.351758793969849e-05,
-      "loss": 0.9186,
+      "loss": 0.9183,
       "step": 883
     },
     {
       "epoch": 0.3932384341637011,
-      "grad_norm": 0.3848302960395813,
+      "grad_norm": 0.3892892599105835,
       "learning_rate": 2.3316582914572865e-05,
-      "loss": 1.0796,
+      "loss": 1.0786,
       "step": 884
     },
     {
       "epoch": 0.39368327402135234,
-      "grad_norm": 0.4643072783946991,
+      "grad_norm": 0.4677392840385437,
       "learning_rate": 2.3115577889447238e-05,
-      "loss": 0.9133,
+      "loss": 0.9138,
       "step": 885
     },
     {
       "epoch": 0.39412811387900354,
-      "grad_norm": 0.3217661678791046,
+      "grad_norm": 0.32070422172546387,
       "learning_rate": 2.291457286432161e-05,
-      "loss": 0.9077,
+      "loss": 0.9076,
       "step": 886
     },
     {
       "epoch": 0.3945729537366548,
-      "grad_norm": 0.307915061712265,
+      "grad_norm": 0.30801644921302795,
       "learning_rate": 2.271356783919598e-05,
-      "loss": 0.9465,
+      "loss": 0.9457,
       "step": 887
     },
     {
       "epoch": 0.39501779359430605,
-      "grad_norm": 0.41546520590782166,
+      "grad_norm": 0.4170681834220886,
       "learning_rate": 2.2512562814070353e-05,
-      "loss": 0.9453,
+      "loss": 0.9452,
       "step": 888
     },
     {
       "epoch": 0.3954626334519573,
-      "grad_norm": 0.44572389125823975,
+      "grad_norm": 0.44136953353881836,
       "learning_rate": 2.2311557788944726e-05,
-      "loss": 1.1163,
+      "loss": 1.1165,
       "step": 889
     },
     {
       "epoch": 0.3959074733096085,
-      "grad_norm": 0.3532145619392395,
+      "grad_norm": 0.35362014174461365,
       "learning_rate": 2.21105527638191e-05,
-      "loss": 0.9142,
+      "loss": 0.9155,
       "step": 890
     },
     {
       "epoch": 0.3963523131672598,
-      "grad_norm": 0.3812106251716614,
+      "grad_norm": 0.3813976049423218,
       "learning_rate": 2.190954773869347e-05,
-      "loss": 0.8557,
+      "loss": 0.8541,
       "step": 891
     },
     {
       "epoch": 0.39679715302491103,
-      "grad_norm": 0.40992099046707153,
+      "grad_norm": 0.41842445731163025,
       "learning_rate": 2.1708542713567838e-05,
-      "loss": 0.9685,
+      "loss": 0.9675,
       "step": 892
     },
     {
       "epoch": 0.3972419928825623,
-      "grad_norm": 0.35445836186408997,
+      "grad_norm": 0.3513477146625519,
       "learning_rate": 2.150753768844221e-05,
-      "loss": 0.9416,
+      "loss": 0.9387,
       "step": 893
     },
     {
       "epoch": 0.39768683274021355,
-      "grad_norm": 0.34018823504447937,
+      "grad_norm": 0.3313136398792267,
       "learning_rate": 2.1306532663316584e-05,
-      "loss": 0.9203,
+      "loss": 0.9186,
       "step": 894
     },
     {
       "epoch": 0.39813167259786475,
-      "grad_norm": 0.37122175097465515,
+      "grad_norm": 0.370280921459198,
       "learning_rate": 2.1105527638190957e-05,
-      "loss": 1.1101,
+      "loss": 1.1071,
       "step": 895
     },
     {
       "epoch": 0.398576512455516,
-      "grad_norm": 0.3946356177330017,
+      "grad_norm": 0.3955540657043457,
       "learning_rate": 2.0904522613065327e-05,
-      "loss": 0.932,
+      "loss": 0.9314,
       "step": 896
     },
     {
       "epoch": 0.39902135231316727,
-      "grad_norm": 0.3920304775238037,
+      "grad_norm": 0.394826203584671,
       "learning_rate": 2.07035175879397e-05,
-      "loss": 0.9307,
+      "loss": 0.9315,
       "step": 897
     },
     {
       "epoch": 0.3994661921708185,
-      "grad_norm": 0.35796669125556946,
+      "grad_norm": 0.35463854670524597,
       "learning_rate": 2.0502512562814073e-05,
-      "loss": 0.9004,
+      "loss": 0.9011,
       "step": 898
     },
     {
       "epoch": 0.3999110320284697,
-      "grad_norm": 0.37431228160858154,
+      "grad_norm": 0.3725610673427582,
       "learning_rate": 2.0301507537688446e-05,
-      "loss": 0.9462,
+      "loss": 0.9455,
       "step": 899
     },
     {
       "epoch": 0.400355871886121,
-      "grad_norm": 0.43452969193458557,
+      "grad_norm": 0.4204149842262268,
       "learning_rate": 2.0100502512562815e-05,
-      "loss": 0.9755,
+      "loss": 0.973,
       "step": 900
     },
     {
       "epoch": 0.40080071174377224,
-      "grad_norm": 0.308403342962265,
+      "grad_norm": 0.312836617231369,
       "learning_rate": 1.9899497487437188e-05,
-      "loss": 0.8736,
+      "loss": 0.875,
       "step": 901
     },
     {
       "epoch": 0.4012455516014235,
-      "grad_norm": 0.425761342048645,
+      "grad_norm": 0.4284355044364929,
       "learning_rate": 1.9698492462311558e-05,
-      "loss": 1.0335,
+      "loss": 1.0345,
       "step": 902
     },
     {
       "epoch": 0.40169039145907476,
-      "grad_norm": 0.4178042709827423,
+      "grad_norm": 0.41245731711387634,
       "learning_rate": 1.949748743718593e-05,
-      "loss": 0.8853,
+      "loss": 0.884,
       "step": 903
     },
     {
       "epoch": 0.40213523131672596,
-      "grad_norm": 0.31192678213119507,
+      "grad_norm": 0.31140172481536865,
       "learning_rate": 1.92964824120603e-05,
-      "loss": 0.8916,
+      "loss": 0.8928,
       "step": 904
     },
     {
       "epoch": 0.4025800711743772,
-      "grad_norm": 0.4430267810821533,
+      "grad_norm": 0.4398500323295593,
       "learning_rate": 1.9095477386934673e-05,
-      "loss": 0.962,
+      "loss": 0.9624,
       "step": 905
     },
     {
       "epoch": 0.4030249110320285,
-      "grad_norm": 0.34475943446159363,
+      "grad_norm": 0.3438228964805603,
       "learning_rate": 1.8894472361809046e-05,
-      "loss": 0.8855,
+      "loss": 0.8849,
       "step": 906
     },
     {
       "epoch": 0.40346975088967973,
-      "grad_norm": 0.36281871795654297,
+      "grad_norm": 0.3663855195045471,
       "learning_rate": 1.869346733668342e-05,
-      "loss": 0.8925,
+      "loss": 0.8945,
       "step": 907
     },
     {
       "epoch": 0.40391459074733094,
-      "grad_norm": 0.39103028178215027,
+      "grad_norm": 0.3942212164402008,
       "learning_rate": 1.849246231155779e-05,
-      "loss": 0.86,
+      "loss": 0.8601,
       "step": 908
     },
     {
       "epoch": 0.4043594306049822,
-      "grad_norm": 0.3842105567455292,
+      "grad_norm": 0.3836296498775482,
       "learning_rate": 1.829145728643216e-05,
-      "loss": 1.0245,
+      "loss": 1.0224,
       "step": 909
     },
     {
       "epoch": 0.40480427046263345,
-      "grad_norm": 0.4096086025238037,
+      "grad_norm": 0.41282856464385986,
       "learning_rate": 1.8090452261306535e-05,
-      "loss": 0.8985,
+      "loss": 0.9005,
       "step": 910
     },
     {
       "epoch": 0.4052491103202847,
-      "grad_norm": 0.3068806827068329,
+      "grad_norm": 0.3127693235874176,
       "learning_rate": 1.7889447236180908e-05,
-      "loss": 0.804,
+      "loss": 0.8051,
       "step": 911
     },
     {
       "epoch": 0.40569395017793597,
-      "grad_norm": 0.28686726093292236,
+      "grad_norm": 0.2867036759853363,
       "learning_rate": 1.7688442211055277e-05,
-      "loss": 1.0117,
+      "loss": 1.0131,
       "step": 912
     },
     {
       "epoch": 0.40613879003558717,
-      "grad_norm": 0.2880958616733551,
+      "grad_norm": 0.28900963068008423,
       "learning_rate": 1.748743718592965e-05,
-      "loss": 0.8703,
+      "loss": 0.8713,
       "step": 913
     },
     {
       "epoch": 0.40658362989323843,
-      "grad_norm": 0.3917515277862549,
+      "grad_norm": 0.39211320877075195,
       "learning_rate": 1.728643216080402e-05,
-      "loss": 0.968,
+      "loss": 0.9688,
       "step": 914
     },
     {
       "epoch": 0.4070284697508897,
-      "grad_norm": 0.3814576268196106,
+      "grad_norm": 0.39720427989959717,
       "learning_rate": 1.7085427135678393e-05,
-      "loss": 0.9305,
+      "loss": 0.9303,
       "step": 915
     },
     {
       "epoch": 0.40747330960854095,
-      "grad_norm": 0.3277512192726135,
+      "grad_norm": 0.3245285749435425,
       "learning_rate": 1.6884422110552762e-05,
-      "loss": 0.8769,
+      "loss": 0.8772,
       "step": 916
     },
     {
       "epoch": 0.40791814946619215,
-      "grad_norm": 0.3101690709590912,
+      "grad_norm": 0.3102715015411377,
       "learning_rate": 1.6683417085427135e-05,
       "loss": 0.8974,
       "step": 917
     },
     {
       "epoch": 0.4083629893238434,
-      "grad_norm": 0.34977224469184875,
+      "grad_norm": 0.34904980659484863,
       "learning_rate": 1.6482412060301508e-05,
-      "loss": 0.9463,
+      "loss": 0.9458,
       "step": 918
     },
     {
       "epoch": 0.40880782918149466,
-      "grad_norm": 0.35917431116104126,
+      "grad_norm": 0.3596024215221405,
       "learning_rate": 1.628140703517588e-05,
-      "loss": 0.9637,
+      "loss": 0.9627,
       "step": 919
     },
     {
       "epoch": 0.4092526690391459,
-      "grad_norm": 0.36873766779899597,
+      "grad_norm": 0.37325575947761536,
       "learning_rate": 1.608040201005025e-05,
-      "loss": 0.7935,
+      "loss": 0.7928,
       "step": 920
     },
     {
       "epoch": 0.4096975088967972,
-      "grad_norm": 0.348622590303421,
+      "grad_norm": 0.3483446538448334,
       "learning_rate": 1.5879396984924624e-05,
-      "loss": 0.9165,
+      "loss": 0.9163,
       "step": 921
     },
     {
       "epoch": 0.4101423487544484,
-      "grad_norm": 0.3866550028324127,
+      "grad_norm": 0.3493581712245941,
       "learning_rate": 1.5678391959798997e-05,
-      "loss": 0.9679,
+      "loss": 0.9661,
       "step": 922
     },
     {
       "epoch": 0.41058718861209964,
-      "grad_norm": 0.3473438620567322,
+      "grad_norm": 0.3486252725124359,
       "learning_rate": 1.547738693467337e-05,
-      "loss": 0.9228,
+      "loss": 0.9236,
       "step": 923
     },
     {
       "epoch": 0.4110320284697509,
-      "grad_norm": 0.43389132618904114,
+      "grad_norm": 0.40376630425453186,
       "learning_rate": 1.527638190954774e-05,
-      "loss": 0.9541,
+      "loss": 0.9523,
       "step": 924
     },
     {
       "epoch": 0.41147686832740216,
-      "grad_norm": 0.40540170669555664,
+      "grad_norm": 0.39947953820228577,
       "learning_rate": 1.507537688442211e-05,
       "loss": 0.9814,
       "step": 925
     },
     {
       "epoch": 0.41192170818505336,
-      "grad_norm": 0.3193083107471466,
+      "grad_norm": 0.3233617842197418,
       "learning_rate": 1.4874371859296482e-05,
-      "loss": 0.9914,
+      "loss": 0.9911,
       "step": 926
     },
     {
       "epoch": 0.4123665480427046,
-      "grad_norm": 0.3414134383201599,
+      "grad_norm": 0.3426443040370941,
       "learning_rate": 1.4673366834170855e-05,
-      "loss": 1.0247,
+      "loss": 1.0236,
       "step": 927
     },
     {
       "epoch": 0.4128113879003559,
-      "grad_norm": 0.3683622181415558,
+      "grad_norm": 0.3716176450252533,
       "learning_rate": 1.4472361809045226e-05,
-      "loss": 0.997,
+      "loss": 0.9969,
       "step": 928
     },
     {
       "epoch": 0.41325622775800713,
-      "grad_norm": 0.37435421347618103,
+      "grad_norm": 0.37307268381118774,
       "learning_rate": 1.4271356783919599e-05,
-      "loss": 0.902,
+      "loss": 0.9022,
       "step": 929
     },
     {
       "epoch": 0.41370106761565834,
-      "grad_norm": 0.31382328271865845,
+      "grad_norm": 0.31498992443084717,
       "learning_rate": 1.407035175879397e-05,
-      "loss": 0.9035,
+      "loss": 0.9017,
       "step": 930
     },
     {
       "epoch": 0.4141459074733096,
-      "grad_norm": 0.35533666610717773,
+      "grad_norm": 0.3574257493019104,
       "learning_rate": 1.3869346733668343e-05,
-      "loss": 1.0198,
+      "loss": 1.0195,
       "step": 931
     },
     {
       "epoch": 0.41459074733096085,
-      "grad_norm": 0.3287508189678192,
+      "grad_norm": 0.3232157230377197,
       "learning_rate": 1.3668341708542715e-05,
-      "loss": 0.8756,
+      "loss": 0.8746,
       "step": 932
     },
     {
       "epoch": 0.4150355871886121,
-      "grad_norm": 0.3902789056301117,
+      "grad_norm": 0.3905941843986511,
       "learning_rate": 1.3467336683417087e-05,
-      "loss": 0.8743,
+      "loss": 0.8746,
       "step": 933
     },
     {
       "epoch": 0.41548042704626337,
-      "grad_norm": 0.3326834738254547,
+      "grad_norm": 0.3263537883758545,
       "learning_rate": 1.3266331658291459e-05,
-      "loss": 0.8924,
+      "loss": 0.8922,
       "step": 934
     },
     {
       "epoch": 0.41592526690391457,
-      "grad_norm": 0.3254244923591614,
+      "grad_norm": 0.3259488642215729,
       "learning_rate": 1.306532663316583e-05,
-      "loss": 0.9863,
+      "loss": 0.9866,
       "step": 935
     },
     {
       "epoch": 0.41637010676156583,
-      "grad_norm": 0.41529661417007446,
+      "grad_norm": 0.4093850255012512,
       "learning_rate": 1.2864321608040203e-05,
-      "loss": 0.8844,
+      "loss": 0.8843,
       "step": 936
     },
     {
       "epoch": 0.4168149466192171,
-      "grad_norm": 0.31171804666519165,
+      "grad_norm": 0.3108372390270233,
       "learning_rate": 1.2663316582914573e-05,
-      "loss": 0.9245,
+      "loss": 0.9248,
       "step": 937
     },
     {
       "epoch": 0.41725978647686834,
-      "grad_norm": 0.33847880363464355,
+      "grad_norm": 0.3368922770023346,
       "learning_rate": 1.2462311557788946e-05,
-      "loss": 0.9414,
+      "loss": 0.9424,
       "step": 938
     },
     {
       "epoch": 0.41770462633451955,
-      "grad_norm": 0.35839855670928955,
+      "grad_norm": 0.35581493377685547,
       "learning_rate": 1.2261306532663318e-05,
-      "loss": 0.924,
+      "loss": 0.9238,
       "step": 939
     },
     {
       "epoch": 0.4181494661921708,
-      "grad_norm": 0.3934418857097626,
+      "grad_norm": 0.392605185508728,
       "learning_rate": 1.2060301507537688e-05,
       "loss": 0.9718,
       "step": 940
     },
     {
       "epoch": 0.41859430604982206,
-      "grad_norm": 0.35172393918037415,
+      "grad_norm": 0.33466875553131104,
       "learning_rate": 1.1859296482412061e-05,
-      "loss": 0.8878,
+      "loss": 0.8881,
       "step": 941
     },
     {
       "epoch": 0.4190391459074733,
-      "grad_norm": 0.4257389307022095,
+      "grad_norm": 0.4258723258972168,
       "learning_rate": 1.1658291457286432e-05,
-      "loss": 0.9914,
+      "loss": 0.9908,
       "step": 942
     },
     {
       "epoch": 0.4194839857651246,
-      "grad_norm": 0.3889011740684509,
+      "grad_norm": 0.3950963318347931,
       "learning_rate": 1.1457286432160805e-05,
-      "loss": 0.8339,
+      "loss": 0.8359,
       "step": 943
     },
     {
       "epoch": 0.4199288256227758,
-      "grad_norm": 0.35240477323532104,
+      "grad_norm": 0.35699567198753357,
       "learning_rate": 1.1256281407035177e-05,
-      "loss": 0.9792,
+      "loss": 0.9785,
       "step": 944
     },
     {
       "epoch": 0.42037366548042704,
-      "grad_norm": 0.3817644715309143,
+      "grad_norm": 0.3818075954914093,
       "learning_rate": 1.105527638190955e-05,
-      "loss": 0.992,
+      "loss": 0.9927,
       "step": 945
     },
     {
       "epoch": 0.4208185053380783,
-      "grad_norm": 0.36161094903945923,
+      "grad_norm": 0.3606509566307068,
       "learning_rate": 1.0854271356783919e-05,
-      "loss": 0.8468,
+      "loss": 0.8477,
       "step": 946
     },
     {
       "epoch": 0.42126334519572955,
-      "grad_norm": 0.3254898488521576,
+      "grad_norm": 0.3209396004676819,
       "learning_rate": 1.0653266331658292e-05,
-      "loss": 0.9998,
+      "loss": 0.9992,
       "step": 947
     },
     {
       "epoch": 0.42170818505338076,
-      "grad_norm": 0.3513511121273041,
+      "grad_norm": 0.35111239552497864,
       "learning_rate": 1.0452261306532663e-05,
-      "loss": 0.8928,
+      "loss": 0.8933,
       "step": 948
     },
     {
       "epoch": 0.422153024911032,
-      "grad_norm": 0.3499341309070587,
+      "grad_norm": 0.3492172658443451,
       "learning_rate": 1.0251256281407036e-05,
-      "loss": 0.9137,
+      "loss": 0.9133,
       "step": 949
     },
     {
       "epoch": 0.4225978647686833,
-      "grad_norm": 0.3495025932788849,
+      "grad_norm": 0.34597399830818176,
       "learning_rate": 1.0050251256281408e-05,
-      "loss": 0.8818,
+      "loss": 0.8814,
       "step": 950
     },
     {
       "epoch": 0.42304270462633453,
-      "grad_norm": 0.3569670021533966,
+      "grad_norm": 0.35809046030044556,
       "learning_rate": 9.849246231155779e-06,
-      "loss": 0.9912,
+      "loss": 0.9919,
       "step": 951
     },
     {
       "epoch": 0.4234875444839858,
-      "grad_norm": 0.394522100687027,
+      "grad_norm": 0.3955031931400299,
       "learning_rate": 9.64824120603015e-06,
-      "loss": 0.863,
+      "loss": 0.8611,
       "step": 952
     },
     {
       "epoch": 0.423932384341637,
-      "grad_norm": 0.35868117213249207,
+      "grad_norm": 0.35886260867118835,
       "learning_rate": 9.447236180904523e-06,
-      "loss": 0.9022,
+      "loss": 0.9025,
       "step": 953
     },
     {
       "epoch": 0.42437722419928825,
-      "grad_norm": 0.33881279826164246,
+      "grad_norm": 0.340167760848999,
       "learning_rate": 9.246231155778894e-06,
-      "loss": 0.9634,
+      "loss": 0.9633,
       "step": 954
     },
     {
       "epoch": 0.4248220640569395,
-      "grad_norm": 0.36202844977378845,
+      "grad_norm": 0.3560841977596283,
       "learning_rate": 9.045226130653267e-06,
-      "loss": 0.9441,
+      "loss": 0.9444,
       "step": 955
     },
     {
       "epoch": 0.42526690391459077,
-      "grad_norm": 0.39875316619873047,
+      "grad_norm": 0.39469799399375916,
       "learning_rate": 8.844221105527639e-06,
-      "loss": 0.9702,
+      "loss": 0.9697,
       "step": 956
     },
     {
       "epoch": 0.42571174377224197,
-      "grad_norm": 0.3653241693973541,
+      "grad_norm": 0.35771381855010986,
       "learning_rate": 8.64321608040201e-06,
-      "loss": 0.9436,
+      "loss": 0.944,
       "step": 957
     },
     {
       "epoch": 0.4261565836298932,
-      "grad_norm": 0.31130754947662354,
+      "grad_norm": 0.3121163845062256,
       "learning_rate": 8.442211055276381e-06,
-      "loss": 0.8579,
+      "loss": 0.8582,
       "step": 958
     },
     {
       "epoch": 0.4266014234875445,
-      "grad_norm": 0.3955715000629425,
+      "grad_norm": 0.39604029059410095,
       "learning_rate": 8.241206030150754e-06,
-      "loss": 0.9889,
+      "loss": 0.9885,
       "step": 959
     },
     {
       "epoch": 0.42704626334519574,
-      "grad_norm": 0.35064584016799927,
+      "grad_norm": 0.35038042068481445,
       "learning_rate": 8.040201005025125e-06,
-      "loss": 0.8877,
+      "loss": 0.8873,
       "step": 960
     },
     {
       "epoch": 0.427491103202847,
-      "grad_norm": 0.38837116956710815,
+      "grad_norm": 0.39221829175949097,
       "learning_rate": 7.839195979899498e-06,
-      "loss": 0.9255,
+      "loss": 0.9252,
       "step": 961
     },
     {
       "epoch": 0.4279359430604982,
-      "grad_norm": 0.3994256556034088,
+      "grad_norm": 0.39741194248199463,
       "learning_rate": 7.63819095477387e-06,
-      "loss": 0.9613,
+      "loss": 0.9643,
       "step": 962
     },
     {
       "epoch": 0.42838078291814946,
-      "grad_norm": 0.5106455087661743,
+      "grad_norm": 0.511210024356842,
       "learning_rate": 7.437185929648241e-06,
-      "loss": 0.9656,
+      "loss": 0.9651,
       "step": 963
     },
     {
       "epoch": 0.4288256227758007,
-      "grad_norm": 0.46163278818130493,
+      "grad_norm": 0.4568634331226349,
       "learning_rate": 7.236180904522613e-06,
-      "loss": 1.0239,
+      "loss": 1.0238,
       "step": 964
     },
     {
       "epoch": 0.429270462633452,
-      "grad_norm": 0.37351593375205994,
+      "grad_norm": 0.36889776587486267,
       "learning_rate": 7.035175879396985e-06,
-      "loss": 0.9637,
+      "loss": 0.964,
       "step": 965
     },
     {
       "epoch": 0.4297153024911032,
-      "grad_norm": 0.34434255957603455,
+      "grad_norm": 0.34706658124923706,
       "learning_rate": 6.834170854271357e-06,
       "loss": 0.9929,
       "step": 966
     },
     {
       "epoch": 0.43016014234875444,
-      "grad_norm": 0.3309576213359833,
+      "grad_norm": 0.33485299348831177,
       "learning_rate": 6.633165829145729e-06,
-      "loss": 0.9139,
+      "loss": 0.9141,
       "step": 967
     },
     {
       "epoch": 0.4306049822064057,
-      "grad_norm": 0.5205715298652649,
+      "grad_norm": 0.5187619924545288,
       "learning_rate": 6.4321608040201015e-06,
-      "loss": 1.0866,
+      "loss": 1.084,
       "step": 968
     },
     {
       "epoch": 0.43104982206405695,
-      "grad_norm": 0.417214035987854,
+      "grad_norm": 0.41246598958969116,
       "learning_rate": 6.231155778894473e-06,
-      "loss": 0.9765,
+      "loss": 0.9763,
       "step": 969
     },
     {
       "epoch": 0.4314946619217082,
-      "grad_norm": 0.3578605651855469,
+      "grad_norm": 0.35957226157188416,
       "learning_rate": 6.030150753768844e-06,
-      "loss": 0.9261,
+      "loss": 0.9251,
       "step": 970
     },
     {
       "epoch": 0.4319395017793594,
-      "grad_norm": 0.35047677159309387,
+      "grad_norm": 0.3485277593135834,
       "learning_rate": 5.829145728643216e-06,
-      "loss": 0.8318,
+      "loss": 0.832,
       "step": 971
     },
     {
       "epoch": 0.43238434163701067,
-      "grad_norm": 0.30359870195388794,
+      "grad_norm": 0.2965943217277527,
       "learning_rate": 5.628140703517588e-06,
-      "loss": 0.9497,
+      "loss": 0.9496,
       "step": 972
     },
     {
       "epoch": 0.43282918149466193,
-      "grad_norm": 0.32898378372192383,
+      "grad_norm": 0.3268524706363678,
       "learning_rate": 5.4271356783919595e-06,
-      "loss": 0.9873,
+      "loss": 0.9876,
       "step": 973
     },
     {
       "epoch": 0.4332740213523132,
-      "grad_norm": 0.33104610443115234,
+      "grad_norm": 0.33093732595443726,
       "learning_rate": 5.226130653266332e-06,
-      "loss": 0.9082,
+      "loss": 0.9083,
       "step": 974
     },
     {
       "epoch": 0.4337188612099644,
-      "grad_norm": 0.3245652914047241,
+      "grad_norm": 0.3334566652774811,
       "learning_rate": 5.025125628140704e-06,
-      "loss": 0.8933,
+      "loss": 0.8941,
       "step": 975
     },
     {
       "epoch": 0.43416370106761565,
-      "grad_norm": 0.3264075517654419,
+      "grad_norm": 0.327716201543808,
       "learning_rate": 4.824120603015075e-06,
-      "loss": 0.9778,
+      "loss": 0.9783,
       "step": 976
     },
     {
       "epoch": 0.4346085409252669,
-      "grad_norm": 0.44732666015625,
+      "grad_norm": 0.4024101793766022,
       "learning_rate": 4.623115577889447e-06,
-      "loss": 0.9509,
+      "loss": 0.9505,
       "step": 977
     },
     {
       "epoch": 0.43505338078291816,
-      "grad_norm": 0.35666903853416443,
+      "grad_norm": 0.3500884175300598,
       "learning_rate": 4.422110552763819e-06,
-      "loss": 0.9392,
+      "loss": 0.9397,
       "step": 978
     },
     {
       "epoch": 0.4354982206405694,
-      "grad_norm": 0.3573732078075409,
+      "grad_norm": 0.3603179156780243,
       "learning_rate": 4.2211055276381906e-06,
-      "loss": 0.9323,
+      "loss": 0.9335,
       "step": 979
     },
     {
       "epoch": 0.4359430604982206,
-      "grad_norm": 0.3136303126811981,
+      "grad_norm": 0.3110332190990448,
       "learning_rate": 4.020100502512563e-06,
-      "loss": 0.959,
+      "loss": 0.9589,
       "step": 980
     },
     {
       "epoch": 0.4363879003558719,
-      "grad_norm": 0.36791837215423584,
+      "grad_norm": 0.3665446937084198,
       "learning_rate": 3.819095477386935e-06,
-      "loss": 0.9151,
+      "loss": 0.9139,
       "step": 981
     },
     {
       "epoch": 0.43683274021352314,
-      "grad_norm": 0.3033190965652466,
+      "grad_norm": 0.30451545119285583,
       "learning_rate": 3.6180904522613065e-06,
-      "loss": 0.9223,
+      "loss": 0.9229,
       "step": 982
     },
     {
       "epoch": 0.4372775800711744,
-      "grad_norm": 0.35878488421440125,
+      "grad_norm": 0.359427809715271,
       "learning_rate": 3.4170854271356786e-06,
-      "loss": 0.9418,
+      "loss": 0.9409,
       "step": 983
     },
     {
       "epoch": 0.4377224199288256,
-      "grad_norm": 0.3010809123516083,
+      "grad_norm": 0.29950541257858276,
       "learning_rate": 3.2160804020100507e-06,
-      "loss": 0.8418,
+      "loss": 0.8438,
       "step": 984
     },
     {
       "epoch": 0.43816725978647686,
-      "grad_norm": 0.4400476813316345,
+      "grad_norm": 0.44397425651550293,
       "learning_rate": 3.015075376884422e-06,
-      "loss": 0.9641,
+      "loss": 0.9639,
       "step": 985
     },
     {
       "epoch": 0.4386120996441281,
-      "grad_norm": 0.34438541531562805,
+      "grad_norm": 0.3420177400112152,
       "learning_rate": 2.814070351758794e-06,
-      "loss": 0.8904,
+      "loss": 0.8911,
       "step": 986
     },
     {
       "epoch": 0.4390569395017794,
-      "grad_norm": 0.39025887846946716,
+      "grad_norm": 0.3904622793197632,
       "learning_rate": 2.613065326633166e-06,
-      "loss": 0.9406,
+      "loss": 0.9407,
       "step": 987
     },
     {
       "epoch": 0.4395017793594306,
-      "grad_norm": 0.29612797498703003,
+      "grad_norm": 0.2967797517776489,
       "learning_rate": 2.4120603015075375e-06,
-      "loss": 0.9153,
+      "loss": 0.9157,
       "step": 988
     },
     {
       "epoch": 0.43994661921708184,
-      "grad_norm": 0.3811575174331665,
+      "grad_norm": 0.3801586925983429,
       "learning_rate": 2.2110552763819096e-06,
       "loss": 0.8314,
       "step": 989
     },
     {
       "epoch": 0.4403914590747331,
-      "grad_norm": 0.3829675614833832,
+      "grad_norm": 0.3838663101196289,
       "learning_rate": 2.0100502512562813e-06,
-      "loss": 0.8798,
+      "loss": 0.8797,
       "step": 990
     },
     {
       "epoch": 0.44083629893238435,
-      "grad_norm": 0.33976423740386963,
+      "grad_norm": 0.3395996689796448,
       "learning_rate": 1.8090452261306533e-06,
-      "loss": 0.8793,
+      "loss": 0.8795,
       "step": 991
     },
     {
       "epoch": 0.4412811387900356,
-      "grad_norm": 0.3907322883605957,
+      "grad_norm": 0.38413113355636597,
       "learning_rate": 1.6080402010050254e-06,
-      "loss": 0.9893,
+      "loss": 0.9914,
       "step": 992
     },
     {
       "epoch": 0.4417259786476868,
-      "grad_norm": 0.33631375432014465,
+      "grad_norm": 0.32779523730278015,
       "learning_rate": 1.407035175879397e-06,
-      "loss": 0.8798,
+      "loss": 0.8799,
       "step": 993
     },
     {
       "epoch": 0.44217081850533807,
-      "grad_norm": 0.41036009788513184,
+      "grad_norm": 0.4115554094314575,
       "learning_rate": 1.2060301507537688e-06,
-      "loss": 0.843,
+      "loss": 0.8444,
       "step": 994
     },
     {
       "epoch": 0.44261565836298933,
-      "grad_norm": 0.3805226683616638,
+      "grad_norm": 0.3810219168663025,
       "learning_rate": 1.0050251256281407e-06,
-      "loss": 1.0452,
+      "loss": 1.0453,
       "step": 995
     },
     {
       "epoch": 0.4430604982206406,
-      "grad_norm": 0.40261736512184143,
+      "grad_norm": 0.40686285495758057,
       "learning_rate": 8.040201005025127e-07,
-      "loss": 0.8972,
+      "loss": 0.8967,
       "step": 996
     },
     {
       "epoch": 0.4435053380782918,
-      "grad_norm": 0.3762563467025757,
+      "grad_norm": 0.3694639801979065,
       "learning_rate": 6.030150753768844e-07,
-      "loss": 0.8555,
+      "loss": 0.8554,
       "step": 997
     },
     {
       "epoch": 0.44395017793594305,
-      "grad_norm": 0.3261839747428894,
+      "grad_norm": 0.32546064257621765,
       "learning_rate": 4.0201005025125634e-07,
       "loss": 0.8795,
       "step": 998
     },
     {
       "epoch": 0.4443950177935943,
-      "grad_norm": 0.309962660074234,
+      "grad_norm": 0.3117218613624573,
       "learning_rate": 2.0100502512562817e-07,
-      "loss": 0.9488,
+      "loss": 0.9495,
       "step": 999
     },
     {
       "epoch": 0.44483985765124556,
-      "grad_norm": 0.3620010018348694,
+      "grad_norm": 0.36267247796058655,
       "learning_rate": 0.0,
-      "loss": 0.953,
+      "loss": 0.9526,
       "step": 1000
     }
   ],