diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6191 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9994856848962796,
+  "eval_steps": 500,
+  "global_step": 4374,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006857534716269502,
+      "grad_norm": 5.500612367122994e+18,
+      "learning_rate": 2.2831050228310502e-08,
+      "loss": 1.1976,
+      "step": 1
+    },
+    {
+      "epoch": 0.0034287673581347507,
+      "grad_norm": 324.38059468358585,
+      "learning_rate": 1.1415525114155251e-07,
+      "loss": 1.175,
+      "step": 5
+    },
+    {
+      "epoch": 0.006857534716269501,
+      "grad_norm": 265.1750103761022,
+      "learning_rate": 2.2831050228310502e-07,
+      "loss": 1.1601,
+      "step": 10
+    },
+    {
+      "epoch": 0.010286302074404252,
+      "grad_norm": 88.79558048967972,
+      "learning_rate": 3.4246575342465755e-07,
+      "loss": 1.1645,
+      "step": 15
+    },
+    {
+      "epoch": 0.013715069432539003,
+      "grad_norm": 997.1728503325202,
+      "learning_rate": 4.5662100456621004e-07,
+      "loss": 1.1547,
+      "step": 20
+    },
+    {
+      "epoch": 0.01714383679067375,
+      "grad_norm": 138.0901381156419,
+      "learning_rate": 5.707762557077626e-07,
+      "loss": 1.1146,
+      "step": 25
+    },
+    {
+      "epoch": 0.020572604148808505,
+      "grad_norm": 3.6970703101396913,
+      "learning_rate": 6.849315068493151e-07,
+      "loss": 1.0849,
+      "step": 30
+    },
+    {
+      "epoch": 0.024001371506943255,
+      "grad_norm": 2.837595372313584,
+      "learning_rate": 7.990867579908676e-07,
+      "loss": 1.0569,
+      "step": 35
+    },
+    {
+      "epoch": 0.027430138865078006,
+      "grad_norm": 6.412430536312489,
+      "learning_rate": 9.132420091324201e-07,
+      "loss": 1.0229,
+      "step": 40
+    },
+    {
+      "epoch": 0.030858906223212756,
+      "grad_norm": 1.429707929152891,
+      "learning_rate": 1.0273972602739727e-06,
+      "loss": 0.9816,
+      "step": 45
+    },
+    {
+      "epoch": 0.0342876735813475,
+      "grad_norm": 1.1893517204804187,
+      "learning_rate": 1.1415525114155251e-06,
+      "loss": 0.9501,
+      "step": 50
+    },
+    {
+      "epoch": 0.03771644093948225,
+      "grad_norm": 1.316763932645065,
+      "learning_rate": 1.2557077625570776e-06,
+      "loss": 0.9212,
+      "step": 55
+    },
+    {
+      "epoch": 0.04114520829761701,
+      "grad_norm": 1.2662857030430075,
+      "learning_rate": 1.3698630136986302e-06,
+      "loss": 0.9013,
+      "step": 60
+    },
+    {
+      "epoch": 0.04457397565575176,
+      "grad_norm": 1.1097798063025857,
+      "learning_rate": 1.4840182648401826e-06,
+      "loss": 0.8951,
+      "step": 65
+    },
+    {
+      "epoch": 0.04800274301388651,
+      "grad_norm": 1.1990919640253435,
+      "learning_rate": 1.5981735159817353e-06,
+      "loss": 0.8941,
+      "step": 70
+    },
+    {
+      "epoch": 0.05143151037202126,
+      "grad_norm": 1.154454828366115,
+      "learning_rate": 1.7123287671232877e-06,
+      "loss": 0.8742,
+      "step": 75
+    },
+    {
+      "epoch": 0.05486027773015601,
+      "grad_norm": 1.1710287337749217,
+      "learning_rate": 1.8264840182648401e-06,
+      "loss": 0.8719,
+      "step": 80
+    },
+    {
+      "epoch": 0.05828904508829076,
+      "grad_norm": 1.1235385190661975,
+      "learning_rate": 1.9406392694063926e-06,
+      "loss": 0.8612,
+      "step": 85
+    },
+    {
+      "epoch": 0.06171781244642551,
+      "grad_norm": 1.3849808686622438,
+      "learning_rate": 2.0547945205479454e-06,
+      "loss": 0.8373,
+      "step": 90
+    },
+    {
+      "epoch": 0.06514657980456026,
+      "grad_norm": 1.2705169052299314,
+      "learning_rate": 2.168949771689498e-06,
+      "loss": 0.846,
+      "step": 95
+    },
+    {
+      "epoch": 0.068575347162695,
+      "grad_norm": 1.2257177989089036,
+      "learning_rate": 2.2831050228310503e-06,
+      "loss": 0.8327,
+      "step": 100
+    },
+    {
+      "epoch": 0.07200411452082976,
+      "grad_norm": 1.3499311190223524,
+      "learning_rate": 2.3972602739726027e-06,
+      "loss": 0.8223,
+      "step": 105
+    },
+    {
+      "epoch": 0.0754328818789645,
+      "grad_norm": 1.1551811574767,
+      "learning_rate": 2.511415525114155e-06,
+      "loss": 0.8075,
+      "step": 110
+    },
+    {
+      "epoch": 0.07886164923709926,
+      "grad_norm": 1.1152359835046965,
+      "learning_rate": 2.625570776255708e-06,
+      "loss": 0.8072,
+      "step": 115
+    },
+    {
+      "epoch": 0.08229041659523402,
+      "grad_norm": 1.128721063171254,
+      "learning_rate": 2.7397260273972604e-06,
+      "loss": 0.799,
+      "step": 120
+    },
+    {
+      "epoch": 0.08571918395336876,
+      "grad_norm": 1.0606734214287294,
+      "learning_rate": 2.8538812785388133e-06,
+      "loss": 0.8028,
+      "step": 125
+    },
+    {
+      "epoch": 0.08914795131150352,
+      "grad_norm": 1.0207053836629467,
+      "learning_rate": 2.9680365296803653e-06,
+      "loss": 0.7793,
+      "step": 130
+    },
+    {
+      "epoch": 0.09257671866963826,
+      "grad_norm": 0.9678836166894507,
+      "learning_rate": 3.082191780821918e-06,
+      "loss": 0.7818,
+      "step": 135
+    },
+    {
+      "epoch": 0.09600548602777302,
+      "grad_norm": 0.958558687009426,
+      "learning_rate": 3.1963470319634706e-06,
+      "loss": 0.78,
+      "step": 140
+    },
+    {
+      "epoch": 0.09943425338590776,
+      "grad_norm": 0.9103829103840835,
+      "learning_rate": 3.310502283105023e-06,
+      "loss": 0.7745,
+      "step": 145
+    },
+    {
+      "epoch": 0.10286302074404252,
+      "grad_norm": 1.072709628680254,
+      "learning_rate": 3.4246575342465754e-06,
+      "loss": 0.76,
+      "step": 150
+    },
+    {
+      "epoch": 0.10629178810217726,
+      "grad_norm": 0.87005113272475,
+      "learning_rate": 3.5388127853881283e-06,
+      "loss": 0.7583,
+      "step": 155
+    },
+    {
+      "epoch": 0.10972055546031202,
+      "grad_norm": 0.7285413995681538,
+      "learning_rate": 3.6529680365296803e-06,
+      "loss": 0.7624,
+      "step": 160
+    },
+    {
+      "epoch": 0.11314932281844677,
+      "grad_norm": 0.8949714023624847,
+      "learning_rate": 3.767123287671233e-06,
+      "loss": 0.7475,
+      "step": 165
+    },
+    {
+      "epoch": 0.11657809017658152,
+      "grad_norm": 0.721003961805146,
+      "learning_rate": 3.881278538812785e-06,
+      "loss": 0.7564,
+      "step": 170
+    },
+    {
+      "epoch": 0.12000685753471627,
+      "grad_norm": 0.7605946035325595,
+      "learning_rate": 3.995433789954338e-06,
+      "loss": 0.7383,
+      "step": 175
+    },
+    {
+      "epoch": 0.12343562489285102,
+      "grad_norm": 1.0401397098538707,
+      "learning_rate": 4.109589041095891e-06,
+      "loss": 0.7488,
+      "step": 180
+    },
+    {
+      "epoch": 0.12686439225098578,
+      "grad_norm": 0.9619516801856884,
+      "learning_rate": 4.223744292237444e-06,
+      "loss": 0.7361,
+      "step": 185
+    },
+    {
+      "epoch": 0.13029315960912052,
+      "grad_norm": 0.7114806408995153,
+      "learning_rate": 4.337899543378996e-06,
+      "loss": 0.7413,
+      "step": 190
+    },
+    {
+      "epoch": 0.13372192696725527,
+      "grad_norm": 0.7788966170765217,
+      "learning_rate": 4.4520547945205486e-06,
+      "loss": 0.7288,
+      "step": 195
+    },
+    {
+      "epoch": 0.13715069432539,
+      "grad_norm": 0.7788077365287874,
+      "learning_rate": 4.566210045662101e-06,
+      "loss": 0.7241,
+      "step": 200
+    },
+    {
+      "epoch": 0.14057946168352478,
+      "grad_norm": 0.713495426998104,
+      "learning_rate": 4.6803652968036534e-06,
+      "loss": 0.7335,
+      "step": 205
+    },
+    {
+      "epoch": 0.14400822904165952,
+      "grad_norm": 0.7672558645856348,
+      "learning_rate": 4.7945205479452054e-06,
+      "loss": 0.7262,
+      "step": 210
+    },
+    {
+      "epoch": 0.14743699639979427,
+      "grad_norm": 0.7374473797133105,
+      "learning_rate": 4.908675799086758e-06,
+      "loss": 0.721,
+      "step": 215
+    },
+    {
+      "epoch": 0.150865763757929,
+      "grad_norm": 0.7501307226424111,
+      "learning_rate": 5.02283105022831e-06,
+      "loss": 0.7186,
+      "step": 220
+    },
+    {
+      "epoch": 0.15429453111606378,
+      "grad_norm": 0.6971360520708836,
+      "learning_rate": 5.136986301369864e-06,
+      "loss": 0.7194,
+      "step": 225
+    },
+    {
+      "epoch": 0.15772329847419853,
+      "grad_norm": 0.6367123001071819,
+      "learning_rate": 5.251141552511416e-06,
+      "loss": 0.719,
+      "step": 230
+    },
+    {
+      "epoch": 0.16115206583233327,
+      "grad_norm": 1.6312076754734648,
+      "learning_rate": 5.365296803652969e-06,
+      "loss": 0.7139,
+      "step": 235
+    },
+    {
+      "epoch": 0.16458083319046804,
+      "grad_norm": 0.8262096198704073,
+      "learning_rate": 5.479452054794521e-06,
+      "loss": 0.7089,
+      "step": 240
+    },
+    {
+      "epoch": 0.16800960054860278,
+      "grad_norm": 0.9448491152723366,
+      "learning_rate": 5.593607305936074e-06,
+      "loss": 0.7103,
+      "step": 245
+    },
+    {
+      "epoch": 0.17143836790673753,
+      "grad_norm": 0.8493857772402849,
+      "learning_rate": 5.7077625570776266e-06,
+      "loss": 0.7011,
+      "step": 250
+    },
+    {
+      "epoch": 0.17486713526487227,
+      "grad_norm": 0.7340850975823853,
+      "learning_rate": 5.821917808219179e-06,
+      "loss": 0.704,
+      "step": 255
+    },
+    {
+      "epoch": 0.17829590262300704,
+      "grad_norm": 0.713122055338853,
+      "learning_rate": 5.936073059360731e-06,
+      "loss": 0.7025,
+      "step": 260
+    },
+    {
+      "epoch": 0.18172466998114178,
+      "grad_norm": 0.7862099070131026,
+      "learning_rate": 6.050228310502284e-06,
+      "loss": 0.7023,
+      "step": 265
+    },
+    {
+      "epoch": 0.18515343733927653,
+      "grad_norm": 0.8003249221177992,
+      "learning_rate": 6.164383561643836e-06,
+      "loss": 0.6985,
+      "step": 270
+    },
+    {
+      "epoch": 0.18858220469741127,
+      "grad_norm": 0.72870713699412,
+      "learning_rate": 6.278538812785388e-06,
+      "loss": 0.6896,
+      "step": 275
+    },
+    {
+      "epoch": 0.19201097205554604,
+      "grad_norm": 0.68892271391433,
+      "learning_rate": 6.392694063926941e-06,
+      "loss": 0.6937,
+      "step": 280
+    },
+    {
+      "epoch": 0.19543973941368079,
+      "grad_norm": 0.8188477395535939,
+      "learning_rate": 6.506849315068494e-06,
+      "loss": 0.6927,
+      "step": 285
+    },
+    {
+      "epoch": 0.19886850677181553,
+      "grad_norm": 0.7692820637205647,
+      "learning_rate": 6.621004566210046e-06,
+      "loss": 0.7009,
+      "step": 290
+    },
+    {
+      "epoch": 0.20229727412995027,
+      "grad_norm": 0.6627944117356517,
+      "learning_rate": 6.735159817351599e-06,
+      "loss": 0.6876,
+      "step": 295
+    },
+    {
+      "epoch": 0.20572604148808504,
+      "grad_norm": 0.7266007369078403,
+      "learning_rate": 6.849315068493151e-06,
+      "loss": 0.6853,
+      "step": 300
+    },
+    {
+      "epoch": 0.2091548088462198,
+      "grad_norm": 0.6424680301484417,
+      "learning_rate": 6.9634703196347046e-06,
+      "loss": 0.6938,
+      "step": 305
+    },
+    {
+      "epoch": 0.21258357620435453,
+      "grad_norm": 0.7129680875859563,
+      "learning_rate": 7.077625570776257e-06,
+      "loss": 0.6909,
+      "step": 310
+    },
+    {
+      "epoch": 0.21601234356248927,
+      "grad_norm": 0.9106015540335944,
+      "learning_rate": 7.191780821917809e-06,
+      "loss": 0.6916,
+      "step": 315
+    },
+    {
+      "epoch": 0.21944111092062404,
+      "grad_norm": 0.6760706144438374,
+      "learning_rate": 7.305936073059361e-06,
+      "loss": 0.6851,
+      "step": 320
+    },
+    {
+      "epoch": 0.2228698782787588,
+      "grad_norm": 0.6512553623989397,
+      "learning_rate": 7.420091324200914e-06,
+      "loss": 0.6897,
+      "step": 325
+    },
+    {
+      "epoch": 0.22629864563689353,
+      "grad_norm": 0.7850669023798967,
+      "learning_rate": 7.534246575342466e-06,
+      "loss": 0.6835,
+      "step": 330
+    },
+    {
+      "epoch": 0.22972741299502827,
+      "grad_norm": 1.3429850514334625,
+      "learning_rate": 7.648401826484018e-06,
+      "loss": 0.6862,
+      "step": 335
+    },
+    {
+      "epoch": 0.23315618035316305,
+      "grad_norm": 1.0601962354732333,
+      "learning_rate": 7.76255707762557e-06,
+      "loss": 0.6865,
+      "step": 340
+    },
+    {
+      "epoch": 0.2365849477112978,
+      "grad_norm": 0.7310633594389766,
+      "learning_rate": 7.876712328767124e-06,
+      "loss": 0.6796,
+      "step": 345
+    },
+    {
+      "epoch": 0.24001371506943253,
+      "grad_norm": 1.2791540029610797,
+      "learning_rate": 7.990867579908676e-06,
+      "loss": 0.678,
+      "step": 350
+    },
+    {
+      "epoch": 0.2434424824275673,
+      "grad_norm": 1.0514021597488898,
+      "learning_rate": 8.105022831050228e-06,
+      "loss": 0.6773,
+      "step": 355
+    },
+    {
+      "epoch": 0.24687124978570205,
+      "grad_norm": 0.6590297903145794,
+      "learning_rate": 8.219178082191782e-06,
+      "loss": 0.6787,
+      "step": 360
+    },
+    {
+      "epoch": 0.2503000171438368,
+      "grad_norm": 1.4407668581292676,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.6747,
+      "step": 365
+    },
+    {
+      "epoch": 0.25372878450197156,
+      "grad_norm": 0.5783708488581324,
+      "learning_rate": 8.447488584474887e-06,
+      "loss": 0.6761,
+      "step": 370
+    },
+    {
+      "epoch": 0.2571575518601063,
+      "grad_norm": 0.7196205508845748,
+      "learning_rate": 8.56164383561644e-06,
+      "loss": 0.6741,
+      "step": 375
+    },
+    {
+      "epoch": 0.26058631921824105,
+      "grad_norm": 0.8426560938281862,
+      "learning_rate": 8.675799086757991e-06,
+      "loss": 0.6783,
+      "step": 380
+    },
+    {
+      "epoch": 0.2640150865763758,
+      "grad_norm": 0.6248407108995583,
+      "learning_rate": 8.789954337899545e-06,
+      "loss": 0.6812,
+      "step": 385
+    },
+    {
+      "epoch": 0.26744385393451053,
+      "grad_norm": 0.625929290973614,
+      "learning_rate": 8.904109589041097e-06,
+      "loss": 0.6722,
+      "step": 390
+    },
+    {
+      "epoch": 0.2708726212926453,
+      "grad_norm": 0.8071769890986592,
+      "learning_rate": 9.01826484018265e-06,
+      "loss": 0.6777,
+      "step": 395
+    },
+    {
+      "epoch": 0.27430138865078,
+      "grad_norm": 0.9295641422813844,
+      "learning_rate": 9.132420091324201e-06,
+      "loss": 0.6744,
+      "step": 400
+    },
+    {
+      "epoch": 0.2777301560089148,
+      "grad_norm": 0.7634843175176039,
+      "learning_rate": 9.246575342465755e-06,
+      "loss": 0.6718,
+      "step": 405
+    },
+    {
+      "epoch": 0.28115892336704956,
+      "grad_norm": 0.8056800028036205,
+      "learning_rate": 9.360730593607307e-06,
+      "loss": 0.6685,
+      "step": 410
+    },
+    {
+      "epoch": 0.2845876907251843,
+      "grad_norm": 0.7359796110052708,
+      "learning_rate": 9.474885844748859e-06,
+      "loss": 0.6686,
+      "step": 415
+    },
+    {
+      "epoch": 0.28801645808331905,
+      "grad_norm": 0.8275083411284923,
+      "learning_rate": 9.589041095890411e-06,
+      "loss": 0.6713,
+      "step": 420
+    },
+    {
+      "epoch": 0.2914452254414538,
+      "grad_norm": 0.9767666036041072,
+      "learning_rate": 9.703196347031965e-06,
+      "loss": 0.6644,
+      "step": 425
+    },
+    {
+      "epoch": 0.29487399279958854,
+      "grad_norm": 0.6970783265751548,
+      "learning_rate": 9.817351598173517e-06,
+      "loss": 0.6692,
+      "step": 430
+    },
+    {
+      "epoch": 0.2983027601577233,
+      "grad_norm": 0.584165558382207,
+      "learning_rate": 9.931506849315069e-06,
+      "loss": 0.6659,
+      "step": 435
+    },
+    {
+      "epoch": 0.301731527515858,
+      "grad_norm": 0.6245810626396433,
+      "learning_rate": 9.999993629265979e-06,
+      "loss": 0.6647,
+      "step": 440
+    },
+    {
+      "epoch": 0.3051602948739928,
+      "grad_norm": 0.7335612100903357,
+      "learning_rate": 9.999921958694681e-06,
+      "loss": 0.667,
+      "step": 445
+    },
+    {
+      "epoch": 0.30858906223212756,
+      "grad_norm": 0.6843530498214124,
+      "learning_rate": 9.999770655279843e-06,
+      "loss": 0.6683,
+      "step": 450
+    },
+    {
+      "epoch": 0.3120178295902623,
+      "grad_norm": 0.6124751321352965,
+      "learning_rate": 9.999539721431253e-06,
+      "loss": 0.6619,
+      "step": 455
+    },
+    {
+      "epoch": 0.31544659694839705,
+      "grad_norm": 0.8614237317784202,
+      "learning_rate": 9.999229160826947e-06,
+      "loss": 0.6637,
+      "step": 460
+    },
+    {
+      "epoch": 0.3188753643065318,
+      "grad_norm": 0.6836210220640591,
+      "learning_rate": 9.998838978413167e-06,
+      "loss": 0.664,
+      "step": 465
+    },
+    {
+      "epoch": 0.32230413166466654,
+      "grad_norm": 0.7142979223973136,
+      "learning_rate": 9.998369180404283e-06,
+      "loss": 0.6682,
+      "step": 470
+    },
+    {
+      "epoch": 0.3257328990228013,
+      "grad_norm": 0.6970601000506839,
+      "learning_rate": 9.997819774282676e-06,
+      "loss": 0.6654,
+      "step": 475
+    },
+    {
+      "epoch": 0.3291616663809361,
+      "grad_norm": 0.7855321640617767,
+      "learning_rate": 9.997190768798639e-06,
+      "loss": 0.6628,
+      "step": 480
+    },
+    {
+      "epoch": 0.3325904337390708,
+      "grad_norm": 0.8068114924796952,
+      "learning_rate": 9.996482173970227e-06,
+      "loss": 0.6532,
+      "step": 485
+    },
+    {
+      "epoch": 0.33601920109720557,
+      "grad_norm": 0.6190829521241801,
+      "learning_rate": 9.995694001083103e-06,
+      "loss": 0.6625,
+      "step": 490
+    },
+    {
+      "epoch": 0.3394479684553403,
+      "grad_norm": 0.6559745100197821,
+      "learning_rate": 9.994826262690347e-06,
+      "loss": 0.6601,
+      "step": 495
+    },
+    {
+      "epoch": 0.34287673581347505,
+      "grad_norm": 0.6960576238421567,
+      "learning_rate": 9.993878972612276e-06,
+      "loss": 0.6575,
+      "step": 500
+    },
+    {
+      "epoch": 0.3463055031716098,
+      "grad_norm": 0.8092838621716005,
+      "learning_rate": 9.992852145936202e-06,
+      "loss": 0.6562,
+      "step": 505
+    },
+    {
+      "epoch": 0.34973427052974454,
+      "grad_norm": 0.7436128976237549,
+      "learning_rate": 9.991745799016206e-06,
+      "loss": 0.6514,
+      "step": 510
+    },
+    {
+      "epoch": 0.3531630378878793,
+      "grad_norm": 0.754160742003698,
+      "learning_rate": 9.990559949472875e-06,
+      "loss": 0.6579,
+      "step": 515
+    },
+    {
+      "epoch": 0.3565918052460141,
+      "grad_norm": 0.6596249438947815,
+      "learning_rate": 9.989294616193018e-06,
+      "loss": 0.6619,
+      "step": 520
+    },
+    {
+      "epoch": 0.3600205726041488,
+      "grad_norm": 0.7268392727602604,
+      "learning_rate": 9.987949819329366e-06,
+      "loss": 0.6569,
+      "step": 525
+    },
+    {
+      "epoch": 0.36344933996228357,
+      "grad_norm": 0.7908918422796353,
+      "learning_rate": 9.986525580300253e-06,
+      "loss": 0.6602,
+      "step": 530
+    },
+    {
+      "epoch": 0.3668781073204183,
+      "grad_norm": 0.731825676141184,
+      "learning_rate": 9.985021921789274e-06,
+      "loss": 0.659,
+      "step": 535
+    },
+    {
+      "epoch": 0.37030687467855306,
+      "grad_norm": 0.6826536905758885,
+      "learning_rate": 9.983438867744923e-06,
+      "loss": 0.6501,
+      "step": 540
+    },
+    {
+      "epoch": 0.3737356420366878,
+      "grad_norm": 0.6873425512137389,
+      "learning_rate": 9.981776443380214e-06,
+      "loss": 0.653,
+      "step": 545
+    },
+    {
+      "epoch": 0.37716440939482254,
+      "grad_norm": 0.6087207671115116,
+      "learning_rate": 9.980034675172274e-06,
+      "loss": 0.6501,
+      "step": 550
+    },
+    {
+      "epoch": 0.3805931767529573,
+      "grad_norm": 0.6758600678083865,
+      "learning_rate": 9.97821359086193e-06,
+      "loss": 0.6457,
+      "step": 555
+    },
+    {
+      "epoch": 0.3840219441110921,
+      "grad_norm": 0.6996994398315339,
+      "learning_rate": 9.976313219453255e-06,
+      "loss": 0.6516,
+      "step": 560
+    },
+    {
+      "epoch": 0.3874507114692268,
+      "grad_norm": 0.5956097888192097,
+      "learning_rate": 9.97433359121312e-06,
+      "loss": 0.6561,
+      "step": 565
+    },
+    {
+      "epoch": 0.39087947882736157,
+      "grad_norm": 0.5809586858657639,
+      "learning_rate": 9.972274737670702e-06,
+      "loss": 0.6515,
+      "step": 570
+    },
+    {
+      "epoch": 0.3943082461854963,
+      "grad_norm": 0.6234744336805654,
+      "learning_rate": 9.970136691616985e-06,
+      "loss": 0.6567,
+      "step": 575
+    },
+    {
+      "epoch": 0.39773701354363106,
+      "grad_norm": 0.7019194825681323,
+      "learning_rate": 9.967919487104237e-06,
+      "loss": 0.6407,
+      "step": 580
+    },
+    {
+      "epoch": 0.4011657809017658,
+      "grad_norm": 0.6464500473501631,
+      "learning_rate": 9.965623159445471e-06,
+      "loss": 0.6478,
+      "step": 585
+    },
+    {
+      "epoch": 0.40459454825990054,
+      "grad_norm": 0.6397619177468029,
+      "learning_rate": 9.963247745213876e-06,
+      "loss": 0.6438,
+      "step": 590
+    },
+    {
+      "epoch": 0.40802331561803534,
+      "grad_norm": 0.6910890699251881,
+      "learning_rate": 9.960793282242243e-06,
+      "loss": 0.649,
+      "step": 595
+    },
+    {
+      "epoch": 0.4114520829761701,
+      "grad_norm": 0.7163950332830918,
+      "learning_rate": 9.958259809622353e-06,
+      "loss": 0.6455,
+      "step": 600
+    },
+    {
+      "epoch": 0.41488085033430483,
+      "grad_norm": 0.8206374824286882,
+      "learning_rate": 9.955647367704362e-06,
+      "loss": 0.6569,
+      "step": 605
+    },
+    {
+      "epoch": 0.4183096176924396,
+      "grad_norm": 0.7207309045108575,
+      "learning_rate": 9.952955998096155e-06,
+      "loss": 0.6446,
+      "step": 610
+    },
+    {
+      "epoch": 0.4217383850505743,
+      "grad_norm": 0.5651258904128029,
+      "learning_rate": 9.950185743662685e-06,
+      "loss": 0.6444,
+      "step": 615
+    },
+    {
+      "epoch": 0.42516715240870906,
+      "grad_norm": 0.5443692863606336,
+      "learning_rate": 9.94733664852529e-06,
+      "loss": 0.6418,
+      "step": 620
+    },
+    {
+      "epoch": 0.4285959197668438,
+      "grad_norm": 0.6042279487491544,
+      "learning_rate": 9.944408758060982e-06,
+      "loss": 0.6463,
+      "step": 625
+    },
+    {
+      "epoch": 0.43202468712497855,
+      "grad_norm": 0.5656088824328167,
+      "learning_rate": 9.941402118901743e-06,
+      "loss": 0.6407,
+      "step": 630
+    },
+    {
+      "epoch": 0.43545345448311334,
+      "grad_norm": 0.658014001778976,
+      "learning_rate": 9.938316778933763e-06,
+      "loss": 0.6468,
+      "step": 635
+    },
+    {
+      "epoch": 0.4388822218412481,
+      "grad_norm": 0.6639637664489146,
+      "learning_rate": 9.935152787296689e-06,
+      "loss": 0.6494,
+      "step": 640
+    },
+    {
+      "epoch": 0.44231098919938283,
+      "grad_norm": 0.5783566760560125,
+      "learning_rate": 9.931910194382836e-06,
+      "loss": 0.6463,
+      "step": 645
+    },
+    {
+      "epoch": 0.4457397565575176,
+      "grad_norm": 0.6490847428513987,
+      "learning_rate": 9.928589051836392e-06,
+      "loss": 0.6438,
+      "step": 650
+    },
+    {
+      "epoch": 0.4491685239156523,
+      "grad_norm": 0.5672865850934122,
+      "learning_rate": 9.925189412552585e-06,
+      "loss": 0.6333,
+      "step": 655
+    },
+    {
+      "epoch": 0.45259729127378706,
+      "grad_norm": 0.569087352718797,
+      "learning_rate": 9.921711330676848e-06,
+      "loss": 0.6478,
+      "step": 660
+    },
+    {
+      "epoch": 0.4560260586319218,
+      "grad_norm": 0.6207854393319554,
+      "learning_rate": 9.918154861603959e-06,
+      "loss": 0.6431,
+      "step": 665
+    },
+    {
+      "epoch": 0.45945482599005655,
+      "grad_norm": 0.586346598273814,
+      "learning_rate": 9.91452006197715e-06,
+      "loss": 0.6383,
+      "step": 670
+    },
+    {
+      "epoch": 0.46288359334819135,
+      "grad_norm": 0.5045583675197564,
+      "learning_rate": 9.910806989687207e-06,
+      "loss": 0.6365,
+      "step": 675
+    },
+    {
+      "epoch": 0.4663123607063261,
+      "grad_norm": 0.6674046560237406,
+      "learning_rate": 9.907015703871558e-06,
+      "loss": 0.6446,
+      "step": 680
+    },
+    {
+      "epoch": 0.46974112806446083,
+      "grad_norm": 0.6966429050410609,
+      "learning_rate": 9.903146264913318e-06,
+      "loss": 0.6416,
+      "step": 685
+    },
+    {
+      "epoch": 0.4731698954225956,
+      "grad_norm": 0.6302591461200223,
+      "learning_rate": 9.899198734440335e-06,
+      "loss": 0.6371,
+      "step": 690
+    },
+    {
+      "epoch": 0.4765986627807303,
+      "grad_norm": 0.8144573920281895,
+      "learning_rate": 9.895173175324205e-06,
+      "loss": 0.6342,
+      "step": 695
+    },
+    {
+      "epoch": 0.48002743013886506,
+      "grad_norm": 0.7886436043083611,
+      "learning_rate": 9.891069651679273e-06,
+      "loss": 0.6355,
+      "step": 700
+    },
+    {
+      "epoch": 0.4834561974969998,
+      "grad_norm": 0.8615238941842659,
+      "learning_rate": 9.886888228861608e-06,
+      "loss": 0.6364,
+      "step": 705
+    },
+    {
+      "epoch": 0.4868849648551346,
+      "grad_norm": 0.6441681216184715,
+      "learning_rate": 9.882628973467972e-06,
+      "loss": 0.6365,
+      "step": 710
+    },
+    {
+      "epoch": 0.49031373221326935,
+      "grad_norm": 0.539258830386699,
+      "learning_rate": 9.878291953334744e-06,
+      "loss": 0.6409,
+      "step": 715
+    },
+    {
+      "epoch": 0.4937424995714041,
+      "grad_norm": 0.5498193436814545,
+      "learning_rate": 9.873877237536854e-06,
+      "loss": 0.6394,
+      "step": 720
+    },
+    {
+      "epoch": 0.49717126692953884,
+      "grad_norm": 0.574880600021989,
+      "learning_rate": 9.869384896386669e-06,
+      "loss": 0.6413,
+      "step": 725
+    },
+    {
+      "epoch": 0.5006000342876736,
+      "grad_norm": 0.618195186077868,
+      "learning_rate": 9.86481500143289e-06,
+      "loss": 0.6416,
+      "step": 730
+    },
+    {
+      "epoch": 0.5040288016458083,
+      "grad_norm": 0.5418237252054477,
+      "learning_rate": 9.860167625459398e-06,
+      "loss": 0.6372,
+      "step": 735
+    },
+    {
+      "epoch": 0.5074575690039431,
+      "grad_norm": 0.5863545275273223,
+      "learning_rate": 9.855442842484101e-06,
+      "loss": 0.6345,
+      "step": 740
+    },
+    {
+      "epoch": 0.5108863363620778,
+      "grad_norm": 0.553229939288754,
+      "learning_rate": 9.850640727757753e-06,
+      "loss": 0.6347,
+      "step": 745
+    },
+    {
+      "epoch": 0.5143151037202126,
+      "grad_norm": 0.6152186401867638,
+      "learning_rate": 9.84576135776276e-06,
+      "loss": 0.6355,
+      "step": 750
+    },
+    {
+      "epoch": 0.5177438710783473,
+      "grad_norm": 0.6357690608918045,
+      "learning_rate": 9.840804810211954e-06,
+      "loss": 0.6347,
+      "step": 755
+    },
+    {
+      "epoch": 0.5211726384364821,
+      "grad_norm": 0.5531025379600504,
+      "learning_rate": 9.835771164047365e-06,
+      "loss": 0.6367,
+      "step": 760
+    },
+    {
+      "epoch": 0.5246014057946168,
+      "grad_norm": 0.5598603097452307,
+      "learning_rate": 9.830660499438955e-06,
+      "loss": 0.6322,
+      "step": 765
+    },
+    {
+      "epoch": 0.5280301731527516,
+      "grad_norm": 0.5973164400797644,
+      "learning_rate": 9.825472897783344e-06,
+      "loss": 0.6363,
+      "step": 770
+    },
+    {
+      "epoch": 0.5314589405108864,
+      "grad_norm": 0.6319771585799853,
+      "learning_rate": 9.820208441702516e-06,
+      "loss": 0.632,
+      "step": 775
+    },
+    {
+      "epoch": 0.5348877078690211,
+      "grad_norm": 0.6648742107102181,
+      "learning_rate": 9.814867215042503e-06,
+      "loss": 0.6379,
+      "step": 780
+    },
+    {
+      "epoch": 0.5383164752271559,
+      "grad_norm": 0.7342225799559353,
+      "learning_rate": 9.809449302872045e-06,
+      "loss": 0.6405,
+      "step": 785
+    },
+    {
+      "epoch": 0.5417452425852906,
+      "grad_norm": 0.7688881238342142,
+      "learning_rate": 9.803954791481239e-06,
+      "loss": 0.634,
+      "step": 790
+    },
+    {
+      "epoch": 0.5451740099434254,
+      "grad_norm": 0.639917210365233,
+      "learning_rate": 9.798383768380164e-06,
+      "loss": 0.6348,
+      "step": 795
+    },
+    {
+      "epoch": 0.54860277730156,
+      "grad_norm": 0.5045187940762229,
+      "learning_rate": 9.792736322297489e-06,
+      "loss": 0.6362,
+      "step": 800
+    },
+    {
+      "epoch": 0.5520315446596948,
+      "grad_norm": 0.5381061183966045,
+      "learning_rate": 9.787012543179053e-06,
+      "loss": 0.6391,
+      "step": 805
+    },
+    {
+      "epoch": 0.5554603120178296,
+      "grad_norm": 0.5429280857380482,
+      "learning_rate": 9.781212522186442e-06,
+      "loss": 0.6307,
+      "step": 810
+    },
+    {
+      "epoch": 0.5588890793759643,
+      "grad_norm": 0.594664535184356,
+      "learning_rate": 9.77533635169553e-06,
+      "loss": 0.6298,
+      "step": 815
+    },
+    {
+      "epoch": 0.5623178467340991,
+      "grad_norm": 0.6702565938406244,
+      "learning_rate": 9.769384125295012e-06,
+      "loss": 0.6342,
+      "step": 820
+    },
+    {
+      "epoch": 0.5657466140922338,
+      "grad_norm": 0.7024332099849335,
+      "learning_rate": 9.763355937784908e-06,
+      "loss": 0.6261,
+      "step": 825
+    },
+    {
+      "epoch": 0.5691753814503686,
+      "grad_norm": 0.6596631748942144,
+      "learning_rate": 9.757251885175063e-06,
+      "loss": 0.6377,
+      "step": 830
+    },
+    {
+      "epoch": 0.5726041488085033,
+      "grad_norm": 0.580750320713107,
+      "learning_rate": 9.751072064683604e-06,
+      "loss": 0.638,
+      "step": 835
+    },
+    {
+      "epoch": 0.5760329161666381,
+      "grad_norm": 0.5865851818669229,
+      "learning_rate": 9.744816574735405e-06,
+      "loss": 0.6257,
+      "step": 840
+    },
+    {
+      "epoch": 0.5794616835247729,
+      "grad_norm": 0.6490256085654721,
+      "learning_rate": 9.73848551496051e-06,
+      "loss": 0.6268,
+      "step": 845
+    },
+    {
+      "epoch": 0.5828904508829076,
+      "grad_norm": 0.5699242035532047,
+      "learning_rate": 9.732078986192552e-06,
+      "loss": 0.6343,
+      "step": 850
+    },
+    {
+      "epoch": 0.5863192182410424,
+      "grad_norm": 0.5773888152890198,
+      "learning_rate": 9.725597090467145e-06,
+      "loss": 0.6335,
+      "step": 855
+    },
+    {
+      "epoch": 0.5897479855991771,
+      "grad_norm": 0.5030181039807039,
+      "learning_rate": 9.719039931020258e-06,
+      "loss": 0.6297,
+      "step": 860
+    },
+    {
+      "epoch": 0.5931767529573119,
+      "grad_norm": 0.6927945011578892,
+      "learning_rate": 9.712407612286568e-06,
+      "loss": 0.6317,
+      "step": 865
+    },
+    {
+      "epoch": 0.5966055203154466,
+      "grad_norm": 0.6863261538586617,
+      "learning_rate": 9.705700239897809e-06,
+      "loss": 0.6349,
+      "step": 870
+    },
+    {
+      "epoch": 0.6000342876735814,
+      "grad_norm": 0.5863644745576692,
+      "learning_rate": 9.698917920681072e-06,
+      "loss": 0.6336,
+      "step": 875
+    },
+    {
+      "epoch": 0.603463055031716,
+      "grad_norm": 0.569507321040072,
+      "learning_rate": 9.692060762657118e-06,
+      "loss": 0.6311,
+      "step": 880
+    },
+    {
+      "epoch": 0.6068918223898508,
+      "grad_norm": 0.5320598410866568,
+      "learning_rate": 9.685128875038648e-06,
+      "loss": 0.628,
+      "step": 885
+    },
+    {
+      "epoch": 0.6103205897479856,
+      "grad_norm": 0.6614795760343849,
+      "learning_rate": 9.678122368228571e-06,
+      "loss": 0.6273,
+      "step": 890
+    },
+    {
+      "epoch": 0.6137493571061203,
+      "grad_norm": 0.5899705788099368,
+      "learning_rate": 9.67104135381824e-06,
+      "loss": 0.626,
+      "step": 895
+    },
+    {
+      "epoch": 0.6171781244642551,
+      "grad_norm": 0.6244814955846902,
+      "learning_rate": 9.66388594458568e-06,
+      "loss": 0.6202,
+      "step": 900
+    },
+    {
+      "epoch": 0.6206068918223898,
+      "grad_norm": 0.5468113327023256,
+      "learning_rate": 9.656656254493783e-06,
+      "loss": 0.6247,
+      "step": 905
+    },
+    {
+      "epoch": 0.6240356591805246,
+      "grad_norm": 0.5197583062967639,
+      "learning_rate": 9.649352398688506e-06,
+      "loss": 0.6341,
+      "step": 910
+    },
+    {
+      "epoch": 0.6274644265386593,
+      "grad_norm": 0.578711261845795,
+      "learning_rate": 9.641974493497024e-06,
+      "loss": 0.6274,
+      "step": 915
+    },
+    {
+      "epoch": 0.6308931938967941,
+      "grad_norm": 0.536259271020302,
+      "learning_rate": 9.634522656425885e-06,
+      "loss": 0.638,
+      "step": 920
+    },
+    {
+      "epoch": 0.6343219612549289,
+      "grad_norm": 0.5108361690063496,
+      "learning_rate": 9.626997006159135e-06,
+      "loss": 0.6268,
+      "step": 925
+    },
+    {
+      "epoch": 0.6377507286130636,
+      "grad_norm": 0.5258629436603048,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.6306,
+      "step": 930
+    },
+    {
+      "epoch": 0.6411794959711984,
+      "grad_norm": 0.5457258768955178,
+      "learning_rate": 9.61172474665114e-06,
+      "loss": 0.6286,
+      "step": 935
+    },
+    {
+      "epoch": 0.6446082633293331,
+      "grad_norm": 0.6608031766605356,
+      "learning_rate": 9.603978380648375e-06,
+      "loss": 0.6295,
+      "step": 940
+    },
+    {
+      "epoch": 0.6480370306874679,
+      "grad_norm": 0.5276661545510922,
+      "learning_rate": 9.596158687923105e-06,
+      "loss": 0.6263,
+      "step": 945
+    },
+    {
+      "epoch": 0.6514657980456026,
+      "grad_norm": 0.530790337719698,
+      "learning_rate": 9.588265793018141e-06,
+      "loss": 0.625,
+      "step": 950
+    },
+    {
+      "epoch": 0.6548945654037374,
+      "grad_norm": 0.562840339494226,
+      "learning_rate": 9.58029982164218e-06,
+      "loss": 0.6286,
+      "step": 955
+    },
+    {
+      "epoch": 0.6583233327618722,
+      "grad_norm": 0.4947380099055408,
+      "learning_rate": 9.572260900667794e-06,
+      "loss": 0.6236,
+      "step": 960
+    },
+    {
+      "epoch": 0.6617521001200068,
+      "grad_norm": 0.47265335918997425,
+      "learning_rate": 9.564149158129406e-06,
+      "loss": 0.6296,
+      "step": 965
+    },
+    {
+      "epoch": 0.6651808674781416,
+      "grad_norm": 0.5775895832290389,
+      "learning_rate": 9.555964723221258e-06,
+      "loss": 0.6235,
+      "step": 970
+    },
+    {
+      "epoch": 0.6686096348362763,
+      "grad_norm": 0.5561593222147068,
+      "learning_rate": 9.54770772629535e-06,
+      "loss": 0.6203,
+      "step": 975
+    },
+    {
+      "epoch": 0.6720384021944111,
+      "grad_norm": 0.5564531018664398,
+      "learning_rate": 9.539378298859365e-06,
+      "loss": 0.6131,
+      "step": 980
+    },
+    {
+      "epoch": 0.6754671695525458,
+      "grad_norm": 0.6036545672788025,
+      "learning_rate": 9.53097657357457e-06,
+      "loss": 0.6166,
+      "step": 985
+    },
+    {
+      "epoch": 0.6788959369106806,
+      "grad_norm": 0.5409723710972858,
+      "learning_rate": 9.522502684253709e-06,
+      "loss": 0.6229,
+      "step": 990
+    },
+    {
+      "epoch": 0.6823247042688153,
+      "grad_norm": 0.509669382435153,
+      "learning_rate": 9.51395676585887e-06,
+      "loss": 0.6222,
+      "step": 995
+    },
+    {
+      "epoch": 0.6857534716269501,
+      "grad_norm": 0.6082473746180043,
+      "learning_rate": 9.505338954499332e-06,
+      "loss": 0.6148,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6891822389850849,
+      "grad_norm": 0.5381865337988982,
+      "learning_rate": 9.496649387429405e-06,
+      "loss": 0.6253,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6926110063432196,
+      "grad_norm": 0.5519342026599645,
+      "learning_rate": 9.487888203046232e-06,
+      "loss": 0.6272,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6960397737013544,
+      "grad_norm": 0.5016332652349211,
+      "learning_rate": 9.479055540887599e-06,
+      "loss": 0.6228,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6994685410594891,
+      "grad_norm": 0.6079273353999184,
+      "learning_rate": 9.4701515416297e-06,
+      "loss": 0.6208,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7028973084176239,
+      "grad_norm": 0.700304976409857,
+      "learning_rate": 9.461176347084909e-06,
+      "loss": 0.6214,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7063260757757586,
+      "grad_norm": 0.5534413781804028,
+      "learning_rate": 9.452130100199504e-06,
+      "loss": 0.6191,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7097548431338934,
+      "grad_norm": 0.5985964197962274,
+      "learning_rate": 9.44301294505141e-06,
+      "loss": 0.6232,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7131836104920282,
+      "grad_norm": 0.5866063492223905,
+      "learning_rate": 9.433825026847891e-06,
+      "loss": 0.6233,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7166123778501629,
+      "grad_norm": 0.6647059350986494,
+      "learning_rate": 9.42456649192324e-06,
+      "loss": 0.6197,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7200411452082977,
+      "grad_norm": 0.6902179520870335,
+      "learning_rate": 9.415237487736452e-06,
+      "loss": 0.6162,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7234699125664323,
+      "grad_norm": 0.5946749998378368,
+      "learning_rate": 9.405838162868874e-06,
+      "loss": 0.6169,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7268986799245671,
+      "grad_norm": 0.5785836386179494,
+      "learning_rate": 9.396368667021835e-06,
+      "loss": 0.6204,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7303274472827018,
+      "grad_norm": 0.5715009908016999,
+      "learning_rate": 9.386829151014262e-06,
+      "loss": 0.6161,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7337562146408366,
+      "grad_norm": 0.5493040557418012,
+      "learning_rate": 9.377219766780288e-06,
+      "loss": 0.6177,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7371849819989714,
+      "grad_norm": 0.5614424899020751,
+      "learning_rate": 9.367540667366816e-06,
+      "loss": 0.6213,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7406137493571061,
+      "grad_norm": 0.5955493656424166,
+      "learning_rate": 9.3577920069311e-06,
+      "loss": 0.6115,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7440425167152409,
+      "grad_norm": 0.6005939054835772,
+      "learning_rate": 9.347973940738266e-06,
+      "loss": 0.6146,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7474712840733756,
+      "grad_norm": 0.5179120177804368,
+      "learning_rate": 9.338086625158867e-06,
+      "loss": 0.6095,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7509000514315104,
+      "grad_norm": 0.5736625803401397,
+      "learning_rate": 9.328130217666366e-06,
+      "loss": 0.6184,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7543288187896451,
+      "grad_norm": 0.48238347864003633,
+      "learning_rate": 9.318104876834652e-06,
+      "loss": 0.6147,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7577575861477799,
+      "grad_norm": 0.5972927862243746,
+      "learning_rate": 9.308010762335492e-06,
+      "loss": 0.6239,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7611863535059146,
+      "grad_norm": 0.7711586637339868,
+      "learning_rate": 9.297848034936007e-06,
+      "loss": 0.622,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7646151208640494,
+      "grad_norm": 0.617930863519868,
+      "learning_rate": 9.287616856496097e-06,
+      "loss": 0.6049,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7680438882221842,
+      "grad_norm": 0.562401219116577,
+      "learning_rate": 9.277317389965871e-06,
+      "loss": 0.6174,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7714726555803189,
+      "grad_norm": 0.6288444408114011,
+      "learning_rate": 9.266949799383053e-06,
+      "loss": 0.6132,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7749014229384537,
+      "grad_norm": 0.5138761921522274,
+      "learning_rate": 9.256514249870366e-06,
+      "loss": 0.6144,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7783301902965883,
+      "grad_norm": 0.5557394107732245,
+      "learning_rate": 9.246010907632894e-06,
+      "loss": 0.625,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7817589576547231,
+      "grad_norm": 0.4785123138439094,
+      "learning_rate": 9.235439939955458e-06,
+      "loss": 0.6129,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7851877250128578,
+      "grad_norm": 0.4845330276832426,
+      "learning_rate": 9.22480151519992e-06,
+      "loss": 0.6249,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7886164923709926,
+      "grad_norm": 0.5198630830752909,
+      "learning_rate": 9.214095802802533e-06,
+      "loss": 0.6181,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7920452597291274,
+      "grad_norm": 0.5032353282124736,
+      "learning_rate": 9.203322973271223e-06,
+      "loss": 0.6208,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7954740270872621,
+      "grad_norm": 0.710468651962232,
+      "learning_rate": 9.192483198182876e-06,
+      "loss": 0.6148,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7989027944453969,
+      "grad_norm": 0.5960153692293567,
+      "learning_rate": 9.181576650180606e-06,
+      "loss": 0.6216,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8023315618035316,
+      "grad_norm": 0.5541889269998591,
+      "learning_rate": 9.170603502971017e-06,
+      "loss": 0.6181,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8057603291616664,
+      "grad_norm": 0.4842737593004163,
+      "learning_rate": 9.159563931321416e-06,
+      "loss": 0.6118,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8091890965198011,
+      "grad_norm": 0.5527421600120501,
+      "learning_rate": 9.148458111057043e-06,
+      "loss": 0.617,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8126178638779359,
+      "grad_norm": 0.517599668993277,
+      "learning_rate": 9.13728621905827e-06,
+      "loss": 0.6081,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8160466312360707,
+      "grad_norm": 0.5442628844464218,
+      "learning_rate": 9.12604843325778e-06,
+      "loss": 0.6211,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8194753985942054,
+      "grad_norm": 0.5027283276915621,
+      "learning_rate": 9.114744932637734e-06,
+      "loss": 0.6127,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8229041659523402,
+      "grad_norm": 0.5296307123252137,
+      "learning_rate": 9.103375897226919e-06,
+      "loss": 0.613,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8263329333104749,
+      "grad_norm": 0.5490731527916699,
+      "learning_rate": 9.091941508097886e-06,
+      "loss": 0.6141,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8297617006686097,
+      "grad_norm": 0.48494095011709015,
+      "learning_rate": 9.080441947364065e-06,
+      "loss": 0.6137,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8331904680267443,
+      "grad_norm": 0.5200558540203623,
+      "learning_rate": 9.068877398176854e-06,
+      "loss": 0.6086,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8366192353848791,
+      "grad_norm": 0.5307240302552572,
+      "learning_rate": 9.057248044722718e-06,
+      "loss": 0.6186,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8400480027430138,
+      "grad_norm": 0.6928206617626188,
+      "learning_rate": 9.045554072220241e-06,
+      "loss": 0.616,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8434767701011486,
+      "grad_norm": 0.6694845253440784,
+      "learning_rate": 9.033795666917191e-06,
+      "loss": 0.6105,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8469055374592834,
+      "grad_norm": 0.48405743005066365,
+      "learning_rate": 9.02197301608754e-06,
+      "loss": 0.606,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8503343048174181,
+      "grad_norm": 0.5821371319328923,
+      "learning_rate": 9.010086308028487e-06,
+      "loss": 0.6129,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8537630721755529,
+      "grad_norm": 0.6308021249120006,
+      "learning_rate": 8.998135732057458e-06,
+      "loss": 0.6101,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8571918395336876,
+      "grad_norm": 0.6393999209710391,
+      "learning_rate": 8.986121478509096e-06,
+      "loss": 0.6099,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8606206068918224,
+      "grad_norm": 0.5210274848669988,
+      "learning_rate": 8.97404373873222e-06,
+      "loss": 0.6136,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8640493742499571,
+      "grad_norm": 0.5493385621402502,
+      "learning_rate": 8.961902705086785e-06,
+      "loss": 0.6114,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8674781416080919,
+      "grad_norm": 0.565383947144019,
+      "learning_rate": 8.949698570940816e-06,
+      "loss": 0.6117,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8709069089662267,
+      "grad_norm": 0.4809292251179887,
+      "learning_rate": 8.937431530667329e-06,
+      "loss": 0.6163,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8743356763243614,
+      "grad_norm": 0.6569955954243558,
+      "learning_rate": 8.925101779641232e-06,
+      "loss": 0.6136,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8777644436824962,
+      "grad_norm": 0.7586122995884393,
+      "learning_rate": 8.912709514236218e-06,
+      "loss": 0.6152,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8811932110406309,
+      "grad_norm": 0.5863591455341749,
+      "learning_rate": 8.900254931821632e-06,
+      "loss": 0.6105,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8846219783987657,
+      "grad_norm": 0.5808595100846887,
+      "learning_rate": 8.887738230759334e-06,
+      "loss": 0.6103,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8880507457569004,
+      "grad_norm": 0.6637416525469315,
+      "learning_rate": 8.875159610400535e-06,
+      "loss": 0.6151,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8914795131150351,
+      "grad_norm": 0.635843590965775,
+      "learning_rate": 8.862519271082624e-06,
+      "loss": 0.6122,
+      "step": 1300
+    },
+    {
+      "epoch": 0.89490828047317,
+      "grad_norm": 0.48644669408598906,
+      "learning_rate": 8.849817414125973e-06,
+      "loss": 0.6107,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8983370478313046,
+      "grad_norm": 0.7387834738899194,
+      "learning_rate": 8.83705424183074e-06,
+      "loss": 0.6144,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9017658151894394,
+      "grad_norm": 0.47318607768496285,
+      "learning_rate": 8.824229957473638e-06,
+      "loss": 0.615,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9051945825475741,
+      "grad_norm": 0.6099273713462994,
+      "learning_rate": 8.811344765304698e-06,
+      "loss": 0.6126,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9086233499057089,
+      "grad_norm": 0.5213938292230176,
+      "learning_rate": 8.798398870544027e-06,
+      "loss": 0.6034,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9120521172638436,
+      "grad_norm": 0.5394625940524758,
+      "learning_rate": 8.785392479378522e-06,
+      "loss": 0.6039,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9154808846219784,
+      "grad_norm": 0.48958440285150634,
+      "learning_rate": 8.772325798958597e-06,
+      "loss": 0.6029,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9189096519801131,
+      "grad_norm": 0.49051315574848703,
+      "learning_rate": 8.759199037394888e-06,
+      "loss": 0.6007,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9223384193382479,
+      "grad_norm": 0.5004329030521917,
+      "learning_rate": 8.746012403754924e-06,
+      "loss": 0.6079,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9257671866963827,
+      "grad_norm": 0.5018875978088632,
+      "learning_rate": 8.732766108059814e-06,
+      "loss": 0.614,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9291959540545174,
+      "grad_norm": 0.5162939197438497,
+      "learning_rate": 8.719460361280888e-06,
+      "loss": 0.6161,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9326247214126522,
+      "grad_norm": 0.4971369395606836,
+      "learning_rate": 8.70609537533634e-06,
+      "loss": 0.605,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9360534887707869,
+      "grad_norm": 0.5016579726115805,
+      "learning_rate": 8.692671363087864e-06,
+      "loss": 0.6173,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9394822561289217,
+      "grad_norm": 0.5580053234884178,
+      "learning_rate": 8.679188538337248e-06,
+      "loss": 0.6087,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9429110234870564,
+      "grad_norm": 0.4577425988677661,
+      "learning_rate": 8.665647115822981e-06,
+      "loss": 0.5988,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9463397908451912,
+      "grad_norm": 0.45683981074908053,
+      "learning_rate": 8.652047311216823e-06,
+      "loss": 0.6064,
+      "step": 1380
+    },
+    {
+      "epoch": 0.949768558203326,
+      "grad_norm": 0.5288094567173458,
+      "learning_rate": 8.638389341120378e-06,
+      "loss": 0.6089,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9531973255614606,
+      "grad_norm": 0.5037863596617661,
+      "learning_rate": 8.62467342306164e-06,
+      "loss": 0.6085,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9566260929195954,
+      "grad_norm": 0.4706120011800052,
+      "learning_rate": 8.610899775491531e-06,
+      "loss": 0.6023,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9600548602777301,
+      "grad_norm": 0.5245905876570266,
+      "learning_rate": 8.597068617780419e-06,
+      "loss": 0.6152,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9634836276358649,
+      "grad_norm": 0.5294292049294829,
+      "learning_rate": 8.583180170214625e-06,
+      "loss": 0.6087,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9669123949939996,
+      "grad_norm": 0.5012747145599434,
+      "learning_rate": 8.569234653992916e-06,
+      "loss": 0.6074,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9703411623521344,
+      "grad_norm": 0.5519787441985884,
+      "learning_rate": 8.555232291222982e-06,
+      "loss": 0.608,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9737699297102692,
+      "grad_norm": 0.566179243123642,
+      "learning_rate": 8.541173304917895e-06,
+      "loss": 0.6057,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9771986970684039,
+      "grad_norm": 0.4833724946956298,
+      "learning_rate": 8.527057918992564e-06,
+      "loss": 0.6048,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9806274644265387,
+      "grad_norm": 0.5444527229286149,
+      "learning_rate": 8.512886358260162e-06,
+      "loss": 0.6074,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9840562317846734,
+      "grad_norm": 0.6265580889574098,
+      "learning_rate": 8.498658848428541e-06,
+      "loss": 0.6104,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9874849991428082,
+      "grad_norm": 0.511037692990781,
+      "learning_rate": 8.484375616096658e-06,
+      "loss": 0.6111,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9909137665009429,
+      "grad_norm": 0.618617261146988,
+      "learning_rate": 8.470036888750941e-06,
+      "loss": 0.6059,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9943425338590777,
+      "grad_norm": 0.6912314169042373,
+      "learning_rate": 8.455642894761684e-06,
+      "loss": 0.6109,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9977713012172124,
+      "grad_norm": 0.6536131654185897,
+      "learning_rate": 8.441193863379396e-06,
+      "loss": 0.6117,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9998285616320932,
+      "eval_loss": 0.6076797842979431,
+      "eval_runtime": 204.4178,
+      "eval_samples_per_second": 48.073,
+      "eval_steps_per_second": 1.507,
+      "step": 1458
+    },
+    {
+      "epoch": 1.0012000685753473,
+      "grad_norm": 0.5744253109468758,
+      "learning_rate": 8.426690024731161e-06,
+      "loss": 0.5965,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0046288359334818,
+      "grad_norm": 0.6126234725401744,
+      "learning_rate": 8.412131609816968e-06,
+      "loss": 0.5426,
+      "step": 1465
+    },
+    {
+      "epoch": 1.0080576032916166,
+      "grad_norm": 0.5502946067152649,
+      "learning_rate": 8.39751885050603e-06,
+      "loss": 0.5483,
+      "step": 1470
+    },
+    {
+      "epoch": 1.0114863706497514,
+      "grad_norm": 0.5203632756557842,
+      "learning_rate": 8.382851979533095e-06,
+      "loss": 0.5429,
+      "step": 1475
+    },
+    {
+      "epoch": 1.0149151380078862,
+      "grad_norm": 0.5256969786092444,
+      "learning_rate": 8.36813123049474e-06,
+      "loss": 0.5416,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0183439053660208,
+      "grad_norm": 0.5199622468399526,
+      "learning_rate": 8.353356837845642e-06,
+      "loss": 0.5497,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0217726727241556,
+      "grad_norm": 0.5409282245995303,
+      "learning_rate": 8.338529036894855e-06,
+      "loss": 0.549,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0252014400822904,
+      "grad_norm": 0.4884152712675349,
+      "learning_rate": 8.323648063802056e-06,
+      "loss": 0.5444,
+      "step": 1495
+    },
+    {
+      "epoch": 1.0286302074404252,
+      "grad_norm": 0.5014268314366117,
+      "learning_rate": 8.308714155573785e-06,
+      "loss": 0.5495,
+      "step": 1500
+    },
+    {
+      "epoch": 1.03205897479856,
+      "grad_norm": 0.4846536766697211,
+      "learning_rate": 8.293727550059668e-06,
+      "loss": 0.5457,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0354877421566946,
+      "grad_norm": 0.5390111306520632,
+      "learning_rate": 8.278688485948634e-06,
+      "loss": 0.5479,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0389165095148294,
+      "grad_norm": 0.5668492664534543,
+      "learning_rate": 8.263597202765109e-06,
+      "loss": 0.5412,
+      "step": 1515
+    },
+    {
+      "epoch": 1.0423452768729642,
+      "grad_norm": 0.49509927353990013,
+      "learning_rate": 8.248453940865204e-06,
+      "loss": 0.5509,
+      "step": 1520
+    },
+    {
+      "epoch": 1.045774044231099,
+      "grad_norm": 0.5658638682509576,
+      "learning_rate": 8.233258941432883e-06,
+      "loss": 0.5466,
+      "step": 1525
+    },
+    {
+      "epoch": 1.0492028115892336,
+      "grad_norm": 0.599624817317566,
+      "learning_rate": 8.218012446476128e-06,
+      "loss": 0.5456,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.5811724248156606,
+      "learning_rate": 8.202714698823078e-06,
+      "loss": 0.5478,
+      "step": 1535
+    },
+    {
+      "epoch": 1.0560603463055032,
+      "grad_norm": 0.5829817578564964,
+      "learning_rate": 8.187365942118162e-06,
+      "loss": 0.5447,
+      "step": 1540
+    },
+    {
+      "epoch": 1.059489113663638,
+      "grad_norm": 0.6195689312853232,
+      "learning_rate": 8.171966420818227e-06,
+      "loss": 0.5462,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0629178810217728,
+      "grad_norm": 0.5753478735487544,
+      "learning_rate": 8.156516380188635e-06,
+      "loss": 0.5461,
+      "step": 1550
+    },
+    {
+      "epoch": 1.0663466483799073,
+      "grad_norm": 0.7022154988882346,
+      "learning_rate": 8.14101606629936e-06,
+      "loss": 0.5499,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0697754157380421,
+      "grad_norm": 0.6777327419801819,
+      "learning_rate": 8.125465726021068e-06,
+      "loss": 0.5499,
+      "step": 1560
+    },
+    {
+      "epoch": 1.073204183096177,
+      "grad_norm": 0.5453482327699332,
+      "learning_rate": 8.10986560702119e-06,
+      "loss": 0.5503,
+      "step": 1565
+    },
+    {
+      "epoch": 1.0766329504543117,
+      "grad_norm": 0.5825073699218344,
+      "learning_rate": 8.09421595775997e-06,
+      "loss": 0.5451,
+      "step": 1570
+    },
+    {
+      "epoch": 1.0800617178124465,
+      "grad_norm": 0.5964267696290744,
+      "learning_rate": 8.07851702748651e-06,
+      "loss": 0.5497,
+      "step": 1575
+    },
+    {
+      "epoch": 1.083490485170581,
+      "grad_norm": 0.5945145884647497,
+      "learning_rate": 8.062769066234807e-06,
+      "loss": 0.5467,
+      "step": 1580
+    },
+    {
+      "epoch": 1.086919252528716,
+      "grad_norm": 0.5596939286945352,
+      "learning_rate": 8.046972324819762e-06,
+      "loss": 0.5469,
+      "step": 1585
+    },
+    {
+      "epoch": 1.0903480198868507,
+      "grad_norm": 0.5511374508562613,
+      "learning_rate": 8.031127054833192e-06,
+      "loss": 0.5473,
+      "step": 1590
+    },
+    {
+      "epoch": 1.0937767872449855,
+      "grad_norm": 0.5908893744280362,
+      "learning_rate": 8.015233508639811e-06,
+      "loss": 0.5504,
+      "step": 1595
+    },
+    {
+      "epoch": 1.09720555460312,
+      "grad_norm": 0.6257657254754231,
+      "learning_rate": 7.999291939373232e-06,
+      "loss": 0.5473,
+      "step": 1600
+    },
+    {
+      "epoch": 1.1006343219612549,
+      "grad_norm": 0.6187428942884239,
+      "learning_rate": 7.983302600931911e-06,
+      "loss": 0.5509,
+      "step": 1605
+    },
+    {
+      "epoch": 1.1040630893193897,
+      "grad_norm": 0.5783570762769669,
+      "learning_rate": 7.967265747975124e-06,
+      "loss": 0.5507,
+      "step": 1610
+    },
+    {
+      "epoch": 1.1074918566775245,
+      "grad_norm": 0.6002185456345444,
+      "learning_rate": 7.951181635918895e-06,
+      "loss": 0.5462,
+      "step": 1615
+    },
+    {
+      "epoch": 1.1109206240356593,
+      "grad_norm": 0.5515256307910629,
+      "learning_rate": 7.93505052093194e-06,
+      "loss": 0.5533,
+      "step": 1620
+    },
+    {
+      "epoch": 1.1143493913937939,
+      "grad_norm": 0.6078936624704993,
+      "learning_rate": 7.91887265993158e-06,
+      "loss": 0.5472,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1177781587519287,
+      "grad_norm": 0.5442571913199759,
+      "learning_rate": 7.90264831057965e-06,
+      "loss": 0.5486,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1212069261100635,
+      "grad_norm": 0.5524504533522675,
+      "learning_rate": 7.8863777312784e-06,
+      "loss": 0.5456,
+      "step": 1635
+    },
+    {
+      "epoch": 1.1246356934681982,
+      "grad_norm": 0.543132690088679,
+      "learning_rate": 7.870061181166372e-06,
+      "loss": 0.5541,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1280644608263328,
+      "grad_norm": 0.5329328262846712,
+      "learning_rate": 7.853698920114281e-06,
+      "loss": 0.5528,
+      "step": 1645
+    },
+    {
+      "epoch": 1.1314932281844676,
+      "grad_norm": 0.5139681859606589,
+      "learning_rate": 7.837291208720867e-06,
+      "loss": 0.5539,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1349219955426024,
+      "grad_norm": 0.5328884447403797,
+      "learning_rate": 7.820838308308751e-06,
+      "loss": 0.5489,
+      "step": 1655
+    },
+    {
+      "epoch": 1.1383507629007372,
+      "grad_norm": 0.6114939943677603,
+      "learning_rate": 7.804340480920274e-06,
+      "loss": 0.5487,
+      "step": 1660
+    },
+    {
+      "epoch": 1.141779530258872,
+      "grad_norm": 0.6051189586236244,
+      "learning_rate": 7.787797989313317e-06,
+      "loss": 0.5488,
+      "step": 1665
+    },
+    {
+      "epoch": 1.1452082976170066,
+      "grad_norm": 0.5408630245450696,
+      "learning_rate": 7.771211096957125e-06,
+      "loss": 0.552,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1486370649751414,
+      "grad_norm": 0.5456867423021694,
+      "learning_rate": 7.754580068028096e-06,
+      "loss": 0.5469,
+      "step": 1675
+    },
+    {
+      "epoch": 1.1520658323332762,
+      "grad_norm": 0.5474469138093225,
+      "learning_rate": 7.737905167405596e-06,
+      "loss": 0.5527,
+      "step": 1680
+    },
+    {
+      "epoch": 1.155494599691411,
+      "grad_norm": 0.521490633503822,
+      "learning_rate": 7.721186660667715e-06,
+      "loss": 0.5476,
+      "step": 1685
+    },
+    {
+      "epoch": 1.1589233670495456,
+      "grad_norm": 0.6191340910966903,
+      "learning_rate": 7.704424814087056e-06,
+      "loss": 0.5466,
+      "step": 1690
+    },
+    {
+      "epoch": 1.1623521344076804,
+      "grad_norm": 0.5052183730102225,
+      "learning_rate": 7.687619894626493e-06,
+      "loss": 0.555,
+      "step": 1695
+    },
+    {
+      "epoch": 1.1657809017658152,
+      "grad_norm": 0.5722854768006315,
+      "learning_rate": 7.670772169934902e-06,
+      "loss": 0.5513,
+      "step": 1700
+    },
+    {
+      "epoch": 1.16920966912395,
+      "grad_norm": 0.5067922499590597,
+      "learning_rate": 7.653881908342916e-06,
+      "loss": 0.5489,
+      "step": 1705
+    },
+    {
+      "epoch": 1.1726384364820848,
+      "grad_norm": 0.5604900851332252,
+      "learning_rate": 7.636949378858647e-06,
+      "loss": 0.5389,
+      "step": 1710
+    },
+    {
+      "epoch": 1.1760672038402193,
+      "grad_norm": 0.5018446215200031,
+      "learning_rate": 7.6199748511634005e-06,
+      "loss": 0.5463,
+      "step": 1715
+    },
+    {
+      "epoch": 1.1794959711983541,
+      "grad_norm": 0.48196349070673195,
+      "learning_rate": 7.602958595607375e-06,
+      "loss": 0.549,
+      "step": 1720
+    },
+    {
+      "epoch": 1.182924738556489,
+      "grad_norm": 0.5951894113068615,
+      "learning_rate": 7.5859008832053636e-06,
+      "loss": 0.5484,
+      "step": 1725
+    },
+    {
+      "epoch": 1.1863535059146237,
+      "grad_norm": 0.6230046151022545,
+      "learning_rate": 7.568801985632439e-06,
+      "loss": 0.5495,
+      "step": 1730
+    },
+    {
+      "epoch": 1.1897822732727585,
+      "grad_norm": 0.5119551401124481,
+      "learning_rate": 7.551662175219614e-06,
+      "loss": 0.5466,
+      "step": 1735
+    },
+    {
+      "epoch": 1.1932110406308931,
+      "grad_norm": 0.5302998077477141,
+      "learning_rate": 7.5344817249495195e-06,
+      "loss": 0.539,
+      "step": 1740
+    },
+    {
+      "epoch": 1.196639807989028,
+      "grad_norm": 0.5804990070353486,
+      "learning_rate": 7.51726090845205e-06,
+      "loss": 0.5503,
+      "step": 1745
+    },
+    {
+      "epoch": 1.2000685753471627,
+      "grad_norm": 0.5861147333730155,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5481,
+      "step": 1750
+    },
+    {
+      "epoch": 1.2034973427052975,
+      "grad_norm": 0.5966836470331061,
+      "learning_rate": 7.482699274504708e-06,
+      "loss": 0.5438,
+      "step": 1755
+    },
+    {
+      "epoch": 1.2069261100634323,
+      "grad_norm": 0.5203444560983739,
+      "learning_rate": 7.465359007511667e-06,
+      "loss": 0.5476,
+      "step": 1760
+    },
+    {
+      "epoch": 1.2103548774215669,
+      "grad_norm": 0.4994889019945669,
+      "learning_rate": 7.44797947519614e-06,
+      "loss": 0.5526,
+      "step": 1765
+    },
+    {
+      "epoch": 1.2137836447797017,
+      "grad_norm": 0.5743096754445195,
+      "learning_rate": 7.430560954358764e-06,
+      "loss": 0.5484,
+      "step": 1770
+    },
+    {
+      "epoch": 1.2172124121378365,
+      "grad_norm": 0.5032391615298711,
+      "learning_rate": 7.413103722421139e-06,
+      "loss": 0.5431,
+      "step": 1775
+    },
+    {
+      "epoch": 1.2206411794959713,
+      "grad_norm": 0.5072263050309479,
+      "learning_rate": 7.395608057421406e-06,
+      "loss": 0.5415,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2240699468541059,
+      "grad_norm": 0.5302212227201574,
+      "learning_rate": 7.378074238009827e-06,
+      "loss": 0.5375,
+      "step": 1785
+    },
+    {
+      "epoch": 1.2274987142122407,
+      "grad_norm": 0.5042827018544307,
+      "learning_rate": 7.360502543444339e-06,
+      "loss": 0.5493,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2309274815703755,
+      "grad_norm": 0.5067327603110284,
+      "learning_rate": 7.342893253586112e-06,
+      "loss": 0.5547,
+      "step": 1795
+    },
+    {
+      "epoch": 1.2343562489285103,
+      "grad_norm": 0.5008933072194933,
+      "learning_rate": 7.325246648895089e-06,
+      "loss": 0.541,
+      "step": 1800
+    },
+    {
+      "epoch": 1.237785016286645,
+      "grad_norm": 0.5065899468061212,
+      "learning_rate": 7.307563010425517e-06,
+      "loss": 0.5497,
+      "step": 1805
+    },
+    {
+      "epoch": 1.2412137836447796,
+      "grad_norm": 0.5456582166819376,
+      "learning_rate": 7.289842619821475e-06,
+      "loss": 0.5501,
+      "step": 1810
+    },
+    {
+      "epoch": 1.2446425510029144,
+      "grad_norm": 0.625178112160367,
+      "learning_rate": 7.272085759312389e-06,
+      "loss": 0.5478,
+      "step": 1815
+    },
+    {
+      "epoch": 1.2480713183610492,
+      "grad_norm": 0.5527992246013412,
+      "learning_rate": 7.254292711708529e-06,
+      "loss": 0.5527,
+      "step": 1820
+    },
+    {
+      "epoch": 1.251500085719184,
+      "grad_norm": 0.6681908362750082,
+      "learning_rate": 7.236463760396516e-06,
+      "loss": 0.5627,
+      "step": 1825
+    },
+    {
+      "epoch": 1.2549288530773186,
+      "grad_norm": 0.5414761906585449,
+      "learning_rate": 7.218599189334799e-06,
+      "loss": 0.5437,
+      "step": 1830
+    },
+    {
+      "epoch": 1.2583576204354534,
+      "grad_norm": 0.545519378308768,
+      "learning_rate": 7.200699283049138e-06,
+      "loss": 0.5458,
+      "step": 1835
+    },
+    {
+      "epoch": 1.2617863877935882,
+      "grad_norm": 0.5175128490097378,
+      "learning_rate": 7.182764326628068e-06,
+      "loss": 0.5471,
+      "step": 1840
+    },
+    {
+      "epoch": 1.265215155151723,
+      "grad_norm": 0.5436775659977866,
+      "learning_rate": 7.164794605718366e-06,
+      "loss": 0.5515,
+      "step": 1845
+    },
+    {
+      "epoch": 1.2686439225098578,
+      "grad_norm": 0.6110723983754518,
+      "learning_rate": 7.146790406520491e-06,
+      "loss": 0.5456,
+      "step": 1850
+    },
+    {
+      "epoch": 1.2720726898679924,
+      "grad_norm": 0.5770751655663271,
+      "learning_rate": 7.128752015784036e-06,
+      "loss": 0.5443,
+      "step": 1855
+    },
+    {
+      "epoch": 1.2755014572261272,
+      "grad_norm": 0.4962264740154398,
+      "learning_rate": 7.1106797208031554e-06,
+      "loss": 0.5455,
+      "step": 1860
+    },
+    {
+      "epoch": 1.278930224584262,
+      "grad_norm": 0.5486471659395239,
+      "learning_rate": 7.0925738094119865e-06,
+      "loss": 0.5519,
+      "step": 1865
+    },
+    {
+      "epoch": 1.2823589919423968,
+      "grad_norm": 0.5503873404673431,
+      "learning_rate": 7.0744345699800755e-06,
+      "loss": 0.5478,
+      "step": 1870
+    },
+    {
+      "epoch": 1.2857877593005314,
+      "grad_norm": 0.5042441453375152,
+      "learning_rate": 7.056262291407773e-06,
+      "loss": 0.5495,
+      "step": 1875
+    },
+    {
+      "epoch": 1.2892165266586662,
+      "grad_norm": 0.5245616326038465,
+      "learning_rate": 7.038057263121639e-06,
+      "loss": 0.5397,
+      "step": 1880
+    },
+    {
+      "epoch": 1.292645294016801,
+      "grad_norm": 0.5294940877618698,
+      "learning_rate": 7.019819775069834e-06,
+      "loss": 0.5403,
+      "step": 1885
+    },
+    {
+      "epoch": 1.2960740613749357,
+      "grad_norm": 0.6147918213943395,
+      "learning_rate": 7.001550117717499e-06,
+      "loss": 0.5405,
+      "step": 1890
+    },
+    {
+      "epoch": 1.2995028287330705,
+      "grad_norm": 0.5184070224956804,
+      "learning_rate": 6.983248582042126e-06,
+      "loss": 0.5481,
+      "step": 1895
+    },
+    {
+      "epoch": 1.3029315960912053,
+      "grad_norm": 0.5295411569445538,
+      "learning_rate": 6.9649154595289326e-06,
+      "loss": 0.5466,
+      "step": 1900
+    },
+    {
+      "epoch": 1.30636036344934,
+      "grad_norm": 0.49956769450406724,
+      "learning_rate": 6.94655104216621e-06,
+      "loss": 0.5461,
+      "step": 1905
+    },
+    {
+      "epoch": 1.3097891308074747,
+      "grad_norm": 0.5529813856201232,
+      "learning_rate": 6.92815562244068e-06,
+      "loss": 0.5456,
+      "step": 1910
+    },
+    {
+      "epoch": 1.3132178981656095,
+      "grad_norm": 0.48747897180543187,
+      "learning_rate": 6.90972949333283e-06,
+      "loss": 0.5506,
+      "step": 1915
+    },
+    {
+      "epoch": 1.316646665523744,
+      "grad_norm": 0.5423084499887999,
+      "learning_rate": 6.891272948312251e-06,
+      "loss": 0.5497,
+      "step": 1920
+    },
+    {
+      "epoch": 1.320075432881879,
+      "grad_norm": 0.4982712510800406,
+      "learning_rate": 6.872786281332965e-06,
+      "loss": 0.5493,
+      "step": 1925
+    },
+    {
+      "epoch": 1.3235042002400137,
+      "grad_norm": 0.47720857931691807,
+      "learning_rate": 6.854269786828741e-06,
+      "loss": 0.5471,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3269329675981485,
+      "grad_norm": 0.5568806463136045,
+      "learning_rate": 6.8357237597084015e-06,
+      "loss": 0.5527,
+      "step": 1935
+    },
+    {
+      "epoch": 1.3303617349562833,
+      "grad_norm": 0.510634543124876,
+      "learning_rate": 6.817148495351131e-06,
+      "loss": 0.5495,
+      "step": 1940
+    },
+    {
+      "epoch": 1.333790502314418,
+      "grad_norm": 0.5164556362855032,
+      "learning_rate": 6.7985442896017765e-06,
+      "loss": 0.5463,
+      "step": 1945
+    },
+    {
+      "epoch": 1.3372192696725527,
+      "grad_norm": 0.5254656454551451,
+      "learning_rate": 6.779911438766117e-06,
+      "loss": 0.5426,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3406480370306875,
+      "grad_norm": 0.49029587603019725,
+      "learning_rate": 6.7612502396061685e-06,
+      "loss": 0.5429,
+      "step": 1955
+    },
+    {
+      "epoch": 1.3440768043888223,
+      "grad_norm": 0.4789585618635431,
+      "learning_rate": 6.742560989335438e-06,
+      "loss": 0.543,
+      "step": 1960
+    },
+    {
+      "epoch": 1.3475055717469568,
+      "grad_norm": 0.5351724574021284,
+      "learning_rate": 6.723843985614201e-06,
+      "loss": 0.549,
+      "step": 1965
+    },
+    {
+      "epoch": 1.3509343391050916,
+      "grad_norm": 0.5285309854529648,
+      "learning_rate": 6.705099526544757e-06,
+      "loss": 0.5491,
+      "step": 1970
+    },
+    {
+      "epoch": 1.3543631064632264,
+      "grad_norm": 0.5744541149885962,
+      "learning_rate": 6.686327910666679e-06,
+      "loss": 0.5469,
+      "step": 1975
+    },
+    {
+      "epoch": 1.3577918738213612,
+      "grad_norm": 0.5276476840483824,
+      "learning_rate": 6.667529436952064e-06,
+      "loss": 0.5501,
+      "step": 1980
+    },
+    {
+      "epoch": 1.361220641179496,
+      "grad_norm": 0.5136271937577194,
+      "learning_rate": 6.6487044048007696e-06,
+      "loss": 0.5457,
+      "step": 1985
+    },
+    {
+      "epoch": 1.3646494085376308,
+      "grad_norm": 0.46138577500457834,
+      "learning_rate": 6.629853114035643e-06,
+      "loss": 0.5461,
+      "step": 1990
+    },
+    {
+      "epoch": 1.3680781758957654,
+      "grad_norm": 0.4895982866432622,
+      "learning_rate": 6.610975864897746e-06,
+      "loss": 0.5392,
+      "step": 1995
+    },
+    {
+      "epoch": 1.3715069432539002,
+      "grad_norm": 0.5042414486514584,
+      "learning_rate": 6.5920729580415795e-06,
+      "loss": 0.5463,
+      "step": 2000
+    },
+    {
+      "epoch": 1.374935710612035,
+      "grad_norm": 0.507137387593949,
+      "learning_rate": 6.573144694530287e-06,
+      "loss": 0.5486,
+      "step": 2005
+    },
+    {
+      "epoch": 1.3783644779701696,
+      "grad_norm": 0.5035112896994087,
+      "learning_rate": 6.554191375830861e-06,
+      "loss": 0.5399,
+      "step": 2010
+    },
+    {
+      "epoch": 1.3817932453283044,
+      "grad_norm": 0.5191434631626274,
+      "learning_rate": 6.535213303809349e-06,
+      "loss": 0.5462,
+      "step": 2015
+    },
+    {
+      "epoch": 1.3852220126864392,
+      "grad_norm": 0.5215645909213554,
+      "learning_rate": 6.516210780726032e-06,
+      "loss": 0.5492,
+      "step": 2020
+    },
+    {
+      "epoch": 1.388650780044574,
+      "grad_norm": 0.4955063426711122,
+      "learning_rate": 6.497184109230628e-06,
+      "loss": 0.547,
+      "step": 2025
+    },
+    {
+      "epoch": 1.3920795474027088,
+      "grad_norm": 0.5083904799973671,
+      "learning_rate": 6.478133592357455e-06,
+      "loss": 0.5495,
+      "step": 2030
+    },
+    {
+      "epoch": 1.3955083147608436,
+      "grad_norm": 0.5413459772025052,
+      "learning_rate": 6.4590595335206154e-06,
+      "loss": 0.5471,
+      "step": 2035
+    },
+    {
+      "epoch": 1.3989370821189782,
+      "grad_norm": 0.5159096563706932,
+      "learning_rate": 6.43996223650916e-06,
+      "loss": 0.5475,
+      "step": 2040
+    },
+    {
+      "epoch": 1.402365849477113,
+      "grad_norm": 0.5232964962054851,
+      "learning_rate": 6.420842005482248e-06,
+      "loss": 0.5526,
+      "step": 2045
+    },
+    {
+      "epoch": 1.4057946168352478,
+      "grad_norm": 0.4879763438255916,
+      "learning_rate": 6.401699144964306e-06,
+      "loss": 0.5563,
+      "step": 2050
+    },
+    {
+      "epoch": 1.4092233841933826,
+      "grad_norm": 0.49746751110206966,
+      "learning_rate": 6.382533959840178e-06,
+      "loss": 0.5401,
+      "step": 2055
+    },
+    {
+      "epoch": 1.4126521515515171,
+      "grad_norm": 0.551660200659416,
+      "learning_rate": 6.3633467553502625e-06,
+      "loss": 0.5464,
+      "step": 2060
+    },
+    {
+      "epoch": 1.416080918909652,
+      "grad_norm": 0.5127978902578759,
+      "learning_rate": 6.344137837085662e-06,
+      "loss": 0.5436,
+      "step": 2065
+    },
+    {
+      "epoch": 1.4195096862677867,
+      "grad_norm": 0.528021714081405,
+      "learning_rate": 6.32490751098331e-06,
+      "loss": 0.5475,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4229384536259215,
+      "grad_norm": 0.5235426406662949,
+      "learning_rate": 6.305656083321092e-06,
+      "loss": 0.5495,
+      "step": 2075
+    },
+    {
+      "epoch": 1.4263672209840563,
+      "grad_norm": 0.5932716644747056,
+      "learning_rate": 6.286383860712982e-06,
+      "loss": 0.5437,
+      "step": 2080
+    },
+    {
+      "epoch": 1.429795988342191,
+      "grad_norm": 0.5329514416163247,
+      "learning_rate": 6.2670911501041495e-06,
+      "loss": 0.5537,
+      "step": 2085
+    },
+    {
+      "epoch": 1.4332247557003257,
+      "grad_norm": 0.5203312437593606,
+      "learning_rate": 6.247778258766069e-06,
+      "loss": 0.5453,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4366535230584605,
+      "grad_norm": 0.5530831064402718,
+      "learning_rate": 6.228445494291635e-06,
+      "loss": 0.5441,
+      "step": 2095
+    },
+    {
+      "epoch": 1.4400822904165953,
+      "grad_norm": 0.5356248838373525,
+      "learning_rate": 6.209093164590253e-06,
+      "loss": 0.5467,
+      "step": 2100
+    },
+    {
+      "epoch": 1.4435110577747299,
+      "grad_norm": 0.507854588281778,
+      "learning_rate": 6.189721577882942e-06,
+      "loss": 0.5473,
+      "step": 2105
+    },
+    {
+      "epoch": 1.4469398251328647,
+      "grad_norm": 0.5020349063977403,
+      "learning_rate": 6.170331042697425e-06,
+      "loss": 0.5439,
+      "step": 2110
+    },
+    {
+      "epoch": 1.4503685924909995,
+      "grad_norm": 0.5299904655120313,
+      "learning_rate": 6.150921867863215e-06,
+      "loss": 0.5429,
+      "step": 2115
+    },
+    {
+      "epoch": 1.4537973598491343,
+      "grad_norm": 0.5599366493504166,
+      "learning_rate": 6.131494362506693e-06,
+      "loss": 0.5479,
+      "step": 2120
+    },
+    {
+      "epoch": 1.457226127207269,
+      "grad_norm": 0.5419693437182285,
+      "learning_rate": 6.112048836046185e-06,
+      "loss": 0.5515,
+      "step": 2125
+    },
+    {
+      "epoch": 1.4606548945654039,
+      "grad_norm": 0.5619194156719064,
+      "learning_rate": 6.09258559818704e-06,
+      "loss": 0.5493,
+      "step": 2130
+    },
+    {
+      "epoch": 1.4640836619235384,
+      "grad_norm": 0.5180975989810822,
+      "learning_rate": 6.0731049589166895e-06,
+      "loss": 0.5447,
+      "step": 2135
+    },
+    {
+      "epoch": 1.4675124292816732,
+      "grad_norm": 0.47862105476236594,
+      "learning_rate": 6.053607228499719e-06,
+      "loss": 0.5411,
+      "step": 2140
+    },
+    {
+      "epoch": 1.470941196639808,
+      "grad_norm": 0.5019570636347335,
+      "learning_rate": 6.034092717472917e-06,
+      "loss": 0.5479,
+      "step": 2145
+    },
+    {
+      "epoch": 1.4743699639979426,
+      "grad_norm": 0.5005596379353178,
+      "learning_rate": 6.014561736640334e-06,
+      "loss": 0.5435,
+      "step": 2150
+    },
+    {
+      "epoch": 1.4777987313560774,
+      "grad_norm": 0.4759889724446824,
+      "learning_rate": 5.9950145970683375e-06,
+      "loss": 0.5434,
+      "step": 2155
+    },
+    {
+      "epoch": 1.4812274987142122,
+      "grad_norm": 0.47708542568255435,
+      "learning_rate": 5.975451610080643e-06,
+      "loss": 0.5382,
+      "step": 2160
+    },
+    {
+      "epoch": 1.484656266072347,
+      "grad_norm": 0.507797233049438,
+      "learning_rate": 5.955873087253371e-06,
+      "loss": 0.5488,
+      "step": 2165
+    },
+    {
+      "epoch": 1.4880850334304818,
+      "grad_norm": 0.48901155035523547,
+      "learning_rate": 5.936279340410082e-06,
+      "loss": 0.5379,
+      "step": 2170
+    },
+    {
+      "epoch": 1.4915138007886166,
+      "grad_norm": 0.5482350970529439,
+      "learning_rate": 5.916670681616797e-06,
+      "loss": 0.5444,
+      "step": 2175
+    },
+    {
+      "epoch": 1.4949425681467512,
+      "grad_norm": 0.48909311036019854,
+      "learning_rate": 5.8970474231770445e-06,
+      "loss": 0.5438,
+      "step": 2180
+    },
+    {
+      "epoch": 1.498371335504886,
+      "grad_norm": 0.5029129982744719,
+      "learning_rate": 5.877409877626876e-06,
+      "loss": 0.543,
+      "step": 2185
+    },
+    {
+      "epoch": 1.5018001028630208,
+      "grad_norm": 0.49508908810255237,
+      "learning_rate": 5.857758357729892e-06,
+      "loss": 0.5401,
+      "step": 2190
+    },
+    {
+      "epoch": 1.5052288702211554,
+      "grad_norm": 0.5186990415007219,
+      "learning_rate": 5.838093176472259e-06,
+      "loss": 0.544,
+      "step": 2195
+    },
+    {
+      "epoch": 1.5086576375792902,
+      "grad_norm": 0.5236949094459556,
+      "learning_rate": 5.8184146470577265e-06,
+      "loss": 0.5334,
+      "step": 2200
+    },
+    {
+      "epoch": 1.512086404937425,
+      "grad_norm": 0.4893285394654287,
+      "learning_rate": 5.798723082902636e-06,
+      "loss": 0.5433,
+      "step": 2205
+    },
+    {
+      "epoch": 1.5155151722955598,
+      "grad_norm": 0.4907763180175273,
+      "learning_rate": 5.779018797630934e-06,
+      "loss": 0.543,
+      "step": 2210
+    },
+    {
+      "epoch": 1.5189439396536946,
+      "grad_norm": 0.5637821736714953,
+      "learning_rate": 5.75930210506917e-06,
+      "loss": 0.5406,
+      "step": 2215
+    },
+    {
+      "epoch": 1.5223727070118294,
+      "grad_norm": 0.5100910825032349,
+      "learning_rate": 5.739573319241505e-06,
+      "loss": 0.5397,
+      "step": 2220
+    },
+    {
+      "epoch": 1.5258014743699642,
+      "grad_norm": 0.45580474474329813,
+      "learning_rate": 5.719832754364707e-06,
+      "loss": 0.5468,
+      "step": 2225
+    },
+    {
+      "epoch": 1.5292302417280987,
+      "grad_norm": 0.5045515176992598,
+      "learning_rate": 5.7000807248431466e-06,
+      "loss": 0.5367,
+      "step": 2230
+    },
+    {
+      "epoch": 1.5326590090862335,
+      "grad_norm": 0.4933966502093522,
+      "learning_rate": 5.680317545263786e-06,
+      "loss": 0.5461,
+      "step": 2235
+    },
+    {
+      "epoch": 1.5360877764443681,
+      "grad_norm": 0.5029727396999837,
+      "learning_rate": 5.66054353039118e-06,
+      "loss": 0.5419,
+      "step": 2240
+    },
+    {
+      "epoch": 1.539516543802503,
+      "grad_norm": 0.5034621640442088,
+      "learning_rate": 5.640758995162446e-06,
+      "loss": 0.5425,
+      "step": 2245
+    },
+    {
+      "epoch": 1.5429453111606377,
+      "grad_norm": 0.5084425747704971,
+      "learning_rate": 5.620964254682267e-06,
+      "loss": 0.5414,
+      "step": 2250
+    },
+    {
+      "epoch": 1.5463740785187725,
+      "grad_norm": 0.505196202702042,
+      "learning_rate": 5.601159624217854e-06,
+      "loss": 0.5474,
+      "step": 2255
+    },
+    {
+      "epoch": 1.5498028458769073,
+      "grad_norm": 0.5308371720246889,
+      "learning_rate": 5.58134541919394e-06,
+      "loss": 0.5433,
+      "step": 2260
+    },
+    {
+      "epoch": 1.553231613235042,
+      "grad_norm": 0.4934713800815739,
+      "learning_rate": 5.5615219551877476e-06,
+      "loss": 0.5504,
+      "step": 2265
+    },
+    {
+      "epoch": 1.556660380593177,
+      "grad_norm": 0.46471141316877956,
+      "learning_rate": 5.5416895479239665e-06,
+      "loss": 0.5413,
+      "step": 2270
+    },
+    {
+      "epoch": 1.5600891479513115,
+      "grad_norm": 0.47931849486997197,
+      "learning_rate": 5.521848513269723e-06,
+      "loss": 0.5373,
+      "step": 2275
+    },
+    {
+      "epoch": 1.5635179153094463,
+      "grad_norm": 0.4684090998446449,
+      "learning_rate": 5.501999167229554e-06,
+      "loss": 0.5419,
+      "step": 2280
+    },
+    {
+      "epoch": 1.5669466826675809,
+      "grad_norm": 0.5051008793716584,
+      "learning_rate": 5.482141825940363e-06,
+      "loss": 0.539,
+      "step": 2285
+    },
+    {
+      "epoch": 1.5703754500257157,
+      "grad_norm": 0.5270612668963798,
+      "learning_rate": 5.4622768056664e-06,
+      "loss": 0.552,
+      "step": 2290
+    },
+    {
+      "epoch": 1.5738042173838505,
+      "grad_norm": 0.5102700734031906,
+      "learning_rate": 5.442404422794212e-06,
+      "loss": 0.5413,
+      "step": 2295
+    },
+    {
+      "epoch": 1.5772329847419853,
+      "grad_norm": 0.4651045699311542,
+      "learning_rate": 5.42252499382761e-06,
+      "loss": 0.5457,
+      "step": 2300
+    },
+    {
+      "epoch": 1.58066175210012,
+      "grad_norm": 0.48408561746393897,
+      "learning_rate": 5.402638835382628e-06,
+      "loss": 0.5454,
+      "step": 2305
+    },
+    {
+      "epoch": 1.5840905194582549,
+      "grad_norm": 0.4886707605239818,
+      "learning_rate": 5.38274626418248e-06,
+      "loss": 0.5404,
+      "step": 2310
+    },
+    {
+      "epoch": 1.5875192868163897,
+      "grad_norm": 0.5358058957587754,
+      "learning_rate": 5.362847597052509e-06,
+      "loss": 0.5407,
+      "step": 2315
+    },
+    {
+      "epoch": 1.5909480541745242,
+      "grad_norm": 0.4969840985996426,
+      "learning_rate": 5.3429431509151515e-06,
+      "loss": 0.5408,
+      "step": 2320
+    },
+    {
+      "epoch": 1.594376821532659,
+      "grad_norm": 0.4903079059609584,
+      "learning_rate": 5.323033242784889e-06,
+      "loss": 0.5368,
+      "step": 2325
+    },
+    {
+      "epoch": 1.5978055888907936,
+      "grad_norm": 0.4738390208380356,
+      "learning_rate": 5.303118189763187e-06,
+      "loss": 0.5482,
+      "step": 2330
+    },
+    {
+      "epoch": 1.6012343562489284,
+      "grad_norm": 0.5096387167156792,
+      "learning_rate": 5.283198309033457e-06,
+      "loss": 0.5476,
+      "step": 2335
+    },
+    {
+      "epoch": 1.6046631236070632,
+      "grad_norm": 0.5265182211890799,
+      "learning_rate": 5.263273917856e-06,
+      "loss": 0.5363,
+      "step": 2340
+    },
+    {
+      "epoch": 1.608091890965198,
+      "grad_norm": 0.5159782309776882,
+      "learning_rate": 5.243345333562954e-06,
+      "loss": 0.5453,
+      "step": 2345
+    },
+    {
+      "epoch": 1.6115206583233328,
+      "grad_norm": 0.4866586786251063,
+      "learning_rate": 5.22341287355324e-06,
+      "loss": 0.5366,
+      "step": 2350
+    },
+    {
+      "epoch": 1.6149494256814676,
+      "grad_norm": 0.5069139373526034,
+      "learning_rate": 5.203476855287507e-06,
+      "loss": 0.5424,
+      "step": 2355
+    },
+    {
+      "epoch": 1.6183781930396024,
+      "grad_norm": 0.47559808993135405,
+      "learning_rate": 5.183537596283075e-06,
+      "loss": 0.5437,
+      "step": 2360
+    },
+    {
+      "epoch": 1.621806960397737,
+      "grad_norm": 0.4777303119256066,
+      "learning_rate": 5.1635954141088815e-06,
+      "loss": 0.5432,
+      "step": 2365
+    },
+    {
+      "epoch": 1.6252357277558718,
+      "grad_norm": 0.46998224709647424,
+      "learning_rate": 5.143650626380417e-06,
+      "loss": 0.5487,
+      "step": 2370
+    },
+    {
+      "epoch": 1.6286644951140063,
+      "grad_norm": 0.5020515855916554,
+      "learning_rate": 5.123703550754669e-06,
+      "loss": 0.5437,
+      "step": 2375
+    },
+    {
+      "epoch": 1.6320932624721411,
+      "grad_norm": 0.4986048548791825,
+      "learning_rate": 5.103754504925071e-06,
+      "loss": 0.548,
+      "step": 2380
+    },
+    {
+      "epoch": 1.635522029830276,
+      "grad_norm": 0.49818017928688363,
+      "learning_rate": 5.083803806616428e-06,
+      "loss": 0.5407,
+      "step": 2385
+    },
+    {
+      "epoch": 1.6389507971884107,
+      "grad_norm": 0.4687036183917943,
+      "learning_rate": 5.06385177357987e-06,
+      "loss": 0.5418,
+      "step": 2390
+    },
+    {
+      "epoch": 1.6423795645465455,
+      "grad_norm": 0.5165271708659772,
+      "learning_rate": 5.043898723587779e-06,
+      "loss": 0.5473,
+      "step": 2395
+    },
+    {
+      "epoch": 1.6458083319046803,
+      "grad_norm": 0.583421869027607,
+      "learning_rate": 5.023944974428739e-06,
+      "loss": 0.5383,
+      "step": 2400
+    },
+    {
+      "epoch": 1.6492370992628151,
+      "grad_norm": 0.508317228928292,
+      "learning_rate": 5.003990843902463e-06,
+      "loss": 0.5386,
+      "step": 2405
+    },
+    {
+      "epoch": 1.6526658666209497,
+      "grad_norm": 0.5359492047497594,
+      "learning_rate": 4.9840366498147495e-06,
+      "loss": 0.5425,
+      "step": 2410
+    },
+    {
+      "epoch": 1.6560946339790845,
+      "grad_norm": 0.5183122191069008,
+      "learning_rate": 4.964082709972394e-06,
+      "loss": 0.5502,
+      "step": 2415
+    },
+    {
+      "epoch": 1.6595234013372193,
+      "grad_norm": 0.5124930538710774,
+      "learning_rate": 4.944129342178156e-06,
+      "loss": 0.54,
+      "step": 2420
+    },
+    {
+      "epoch": 1.662952168695354,
+      "grad_norm": 0.5037854005505844,
+      "learning_rate": 4.924176864225678e-06,
+      "loss": 0.538,
+      "step": 2425
+    },
+    {
+      "epoch": 1.6663809360534887,
+      "grad_norm": 0.5034524367425028,
+      "learning_rate": 4.90422559389443e-06,
+      "loss": 0.5342,
+      "step": 2430
+    },
+    {
+      "epoch": 1.6698097034116235,
+      "grad_norm": 0.48847928157114906,
+      "learning_rate": 4.88427584894465e-06,
+      "loss": 0.5414,
+      "step": 2435
+    },
+    {
+      "epoch": 1.6732384707697583,
+      "grad_norm": 0.5197993207547689,
+      "learning_rate": 4.864327947112281e-06,
+      "loss": 0.5437,
+      "step": 2440
+    },
+    {
+      "epoch": 1.676667238127893,
+      "grad_norm": 0.4979107929661232,
+      "learning_rate": 4.84438220610391e-06,
+      "loss": 0.541,
+      "step": 2445
+    },
+    {
+      "epoch": 1.6800960054860279,
+      "grad_norm": 0.47357100832452387,
+      "learning_rate": 4.82443894359171e-06,
+      "loss": 0.5375,
+      "step": 2450
+    },
+    {
+      "epoch": 1.6835247728441627,
+      "grad_norm": 0.47499915823273264,
+      "learning_rate": 4.804498477208382e-06,
+      "loss": 0.5435,
+      "step": 2455
+    },
+    {
+      "epoch": 1.6869535402022973,
+      "grad_norm": 0.5246936291111205,
+      "learning_rate": 4.784561124542088e-06,
+      "loss": 0.5399,
+      "step": 2460
+    },
+    {
+      "epoch": 1.690382307560432,
+      "grad_norm": 0.5396195065092629,
+      "learning_rate": 4.764627203131401e-06,
+      "loss": 0.5323,
+      "step": 2465
+    },
+    {
+      "epoch": 1.6938110749185666,
+      "grad_norm": 0.49550562745484766,
+      "learning_rate": 4.744697030460248e-06,
+      "loss": 0.54,
+      "step": 2470
+    },
+    {
+      "epoch": 1.6972398422767014,
+      "grad_norm": 0.4834798755139004,
+      "learning_rate": 4.724770923952844e-06,
+      "loss": 0.5441,
+      "step": 2475
+    },
+    {
+      "epoch": 1.7006686096348362,
+      "grad_norm": 0.5285955350607044,
+      "learning_rate": 4.7048492009686525e-06,
+      "loss": 0.5391,
+      "step": 2480
+    },
+    {
+      "epoch": 1.704097376992971,
+      "grad_norm": 0.5121886893362284,
+      "learning_rate": 4.6849321787973116e-06,
+      "loss": 0.5406,
+      "step": 2485
+    },
+    {
+      "epoch": 1.7075261443511058,
+      "grad_norm": 0.5563183347246416,
+      "learning_rate": 4.6650201746535926e-06,
+      "loss": 0.5381,
+      "step": 2490
+    },
+    {
+      "epoch": 1.7109549117092406,
+      "grad_norm": 0.49534162399833953,
+      "learning_rate": 4.645113505672353e-06,
+      "loss": 0.5359,
+      "step": 2495
+    },
+    {
+      "epoch": 1.7143836790673754,
+      "grad_norm": 0.4754638729290019,
+      "learning_rate": 4.625212488903467e-06,
+      "loss": 0.5411,
+      "step": 2500
+    },
+    {
+      "epoch": 1.71781244642551,
+      "grad_norm": 0.4701836831162324,
+      "learning_rate": 4.605317441306798e-06,
+      "loss": 0.5362,
+      "step": 2505
+    },
+    {
+      "epoch": 1.7212412137836448,
+      "grad_norm": 0.48575768441353867,
+      "learning_rate": 4.585428679747133e-06,
+      "loss": 0.5419,
+      "step": 2510
+    },
+    {
+      "epoch": 1.7246699811417794,
+      "grad_norm": 0.561176281859862,
+      "learning_rate": 4.565546520989139e-06,
+      "loss": 0.5361,
+      "step": 2515
+    },
+    {
+      "epoch": 1.7280987484999142,
+      "grad_norm": 0.4991619958671278,
+      "learning_rate": 4.545671281692331e-06,
+      "loss": 0.5379,
+      "step": 2520
+    },
+    {
+      "epoch": 1.731527515858049,
+      "grad_norm": 0.4894062203682473,
+      "learning_rate": 4.525803278406011e-06,
+      "loss": 0.5359,
+      "step": 2525
+    },
+    {
+      "epoch": 1.7349562832161838,
+      "grad_norm": 0.4811044764793809,
+      "learning_rate": 4.505942827564242e-06,
+      "loss": 0.5426,
+      "step": 2530
+    },
+    {
+      "epoch": 1.7383850505743186,
+      "grad_norm": 0.48626341909810544,
+      "learning_rate": 4.4860902454807905e-06,
+      "loss": 0.5336,
+      "step": 2535
+    },
+    {
+      "epoch": 1.7418138179324534,
+      "grad_norm": 0.4677680916526279,
+      "learning_rate": 4.466245848344106e-06,
+      "loss": 0.5408,
+      "step": 2540
+    },
+    {
+      "epoch": 1.7452425852905882,
+      "grad_norm": 0.49984747393620377,
+      "learning_rate": 4.446409952212278e-06,
+      "loss": 0.5347,
+      "step": 2545
+    },
+    {
+      "epoch": 1.7486713526487228,
+      "grad_norm": 0.4932054681232252,
+      "learning_rate": 4.426582873007999e-06,
+      "loss": 0.5451,
+      "step": 2550
+    },
+    {
+      "epoch": 1.7521001200068576,
+      "grad_norm": 0.4794644104456506,
+      "learning_rate": 4.406764926513536e-06,
+      "loss": 0.5403,
+      "step": 2555
+    },
+    {
+      "epoch": 1.7555288873649921,
+      "grad_norm": 0.45717917978063133,
+      "learning_rate": 4.386956428365701e-06,
+      "loss": 0.5379,
+      "step": 2560
+    },
+    {
+      "epoch": 1.758957654723127,
+      "grad_norm": 0.46935105917323766,
+      "learning_rate": 4.36715769405083e-06,
+      "loss": 0.5366,
+      "step": 2565
+    },
+    {
+      "epoch": 1.7623864220812617,
+      "grad_norm": 0.4839602278611378,
+      "learning_rate": 4.347369038899744e-06,
+      "loss": 0.5403,
+      "step": 2570
+    },
+    {
+      "epoch": 1.7658151894393965,
+      "grad_norm": 0.4996673443974444,
+      "learning_rate": 4.327590778082734e-06,
+      "loss": 0.5413,
+      "step": 2575
+    },
+    {
+      "epoch": 1.7692439567975313,
+      "grad_norm": 0.47519860849274637,
+      "learning_rate": 4.307823226604555e-06,
+      "loss": 0.5402,
+      "step": 2580
+    },
+    {
+      "epoch": 1.7726727241556661,
+      "grad_norm": 0.47545446063997776,
+      "learning_rate": 4.28806669929938e-06,
+      "loss": 0.5374,
+      "step": 2585
+    },
+    {
+      "epoch": 1.776101491513801,
+      "grad_norm": 0.5148035850870979,
+      "learning_rate": 4.2683215108258145e-06,
+      "loss": 0.5392,
+      "step": 2590
+    },
+    {
+      "epoch": 1.7795302588719355,
+      "grad_norm": 0.49138867529600283,
+      "learning_rate": 4.2485879756618685e-06,
+      "loss": 0.5377,
+      "step": 2595
+    },
+    {
+      "epoch": 1.7829590262300703,
+      "grad_norm": 0.4888120553433302,
+      "learning_rate": 4.228866408099945e-06,
+      "loss": 0.5415,
+      "step": 2600
+    },
+    {
+      "epoch": 1.7863877935882049,
+      "grad_norm": 0.46500050166415424,
+      "learning_rate": 4.2091571222418546e-06,
+      "loss": 0.536,
+      "step": 2605
+    },
+    {
+      "epoch": 1.7898165609463397,
+      "grad_norm": 0.4829554221776432,
+      "learning_rate": 4.189460431993788e-06,
+      "loss": 0.5451,
+      "step": 2610
+    },
+    {
+      "epoch": 1.7932453283044745,
+      "grad_norm": 0.4917384629382616,
+      "learning_rate": 4.169776651061334e-06,
+      "loss": 0.5409,
+      "step": 2615
+    },
+    {
+      "epoch": 1.7966740956626093,
+      "grad_norm": 0.49620621093693373,
+      "learning_rate": 4.150106092944475e-06,
+      "loss": 0.541,
+      "step": 2620
+    },
+    {
+      "epoch": 1.800102863020744,
+      "grad_norm": 0.467340353977474,
+      "learning_rate": 4.130449070932594e-06,
+      "loss": 0.5407,
+      "step": 2625
+    },
+    {
+      "epoch": 1.8035316303788789,
+      "grad_norm": 0.49470765077176454,
+      "learning_rate": 4.110805898099492e-06,
+      "loss": 0.5416,
+      "step": 2630
+    },
+    {
+      "epoch": 1.8069603977370137,
+      "grad_norm": 0.4666535437410526,
+      "learning_rate": 4.091176887298392e-06,
+      "loss": 0.5406,
+      "step": 2635
+    },
+    {
+      "epoch": 1.8103891650951482,
+      "grad_norm": 0.46143547716397726,
+      "learning_rate": 4.071562351156966e-06,
+      "loss": 0.5344,
+      "step": 2640
+    },
+    {
+      "epoch": 1.813817932453283,
+      "grad_norm": 0.46910481572596163,
+      "learning_rate": 4.051962602072343e-06,
+      "loss": 0.5364,
+      "step": 2645
+    },
+    {
+      "epoch": 1.8172466998114178,
+      "grad_norm": 0.4694928382743332,
+      "learning_rate": 4.032377952206148e-06,
+      "loss": 0.5413,
+      "step": 2650
+    },
+    {
+      "epoch": 1.8206754671695524,
+      "grad_norm": 0.47422238985591536,
+      "learning_rate": 4.012808713479523e-06,
+      "loss": 0.5435,
+      "step": 2655
+    },
+    {
+      "epoch": 1.8241042345276872,
+      "grad_norm": 0.48396135634665843,
+      "learning_rate": 3.993255197568154e-06,
+      "loss": 0.5336,
+      "step": 2660
+    },
+    {
+      "epoch": 1.827533001885822,
+      "grad_norm": 0.4873226150393351,
+      "learning_rate": 3.9737177158973185e-06,
+      "loss": 0.5309,
+      "step": 2665
+    },
+    {
+      "epoch": 1.8309617692439568,
+      "grad_norm": 0.4576384143762062,
+      "learning_rate": 3.954196579636918e-06,
+      "loss": 0.5444,
+      "step": 2670
+    },
+    {
+      "epoch": 1.8343905366020916,
+      "grad_norm": 0.49454868144127,
+      "learning_rate": 3.93469209969652e-06,
+      "loss": 0.5353,
+      "step": 2675
+    },
+    {
+      "epoch": 1.8378193039602264,
+      "grad_norm": 0.4490405182223933,
+      "learning_rate": 3.91520458672042e-06,
+      "loss": 0.5352,
+      "step": 2680
+    },
+    {
+      "epoch": 1.8412480713183612,
+      "grad_norm": 0.4892412353541936,
+      "learning_rate": 3.895734351082668e-06,
+      "loss": 0.5364,
+      "step": 2685
+    },
+    {
+      "epoch": 1.8446768386764958,
+      "grad_norm": 0.44948826305297584,
+      "learning_rate": 3.876281702882156e-06,
+      "loss": 0.5382,
+      "step": 2690
+    },
+    {
+      "epoch": 1.8481056060346306,
+      "grad_norm": 0.46974192317523716,
+      "learning_rate": 3.8568469519376585e-06,
+      "loss": 0.534,
+      "step": 2695
+    },
+    {
+      "epoch": 1.8515343733927652,
+      "grad_norm": 0.4541143574340376,
+      "learning_rate": 3.837430407782896e-06,
+      "loss": 0.5426,
+      "step": 2700
+    },
+    {
+      "epoch": 1.8549631407509,
+      "grad_norm": 0.4749792696125447,
+      "learning_rate": 3.818032379661626e-06,
+      "loss": 0.5396,
+      "step": 2705
+    },
+    {
+      "epoch": 1.8583919081090348,
+      "grad_norm": 0.46058829711339094,
+      "learning_rate": 3.7986531765226965e-06,
+      "loss": 0.5385,
+      "step": 2710
+    },
+    {
+      "epoch": 1.8618206754671696,
+      "grad_norm": 0.45724757617193085,
+      "learning_rate": 3.779293107015136e-06,
+      "loss": 0.5325,
+      "step": 2715
+    },
+    {
+      "epoch": 1.8652494428253044,
+      "grad_norm": 0.4613637136202585,
+      "learning_rate": 3.759952479483232e-06,
+      "loss": 0.5332,
+      "step": 2720
+    },
+    {
+      "epoch": 1.8686782101834392,
+      "grad_norm": 0.45902944748623037,
+      "learning_rate": 3.7406316019616297e-06,
+      "loss": 0.5402,
+      "step": 2725
+    },
+    {
+      "epoch": 1.872106977541574,
+      "grad_norm": 0.4810427911391462,
+      "learning_rate": 3.7213307821704115e-06,
+      "loss": 0.5397,
+      "step": 2730
+    },
+    {
+      "epoch": 1.8755357448997085,
+      "grad_norm": 0.5021212230417156,
+      "learning_rate": 3.7020503275102095e-06,
+      "loss": 0.5478,
+      "step": 2735
+    },
+    {
+      "epoch": 1.8789645122578433,
+      "grad_norm": 0.4836776357761823,
+      "learning_rate": 3.6827905450573022e-06,
+      "loss": 0.5358,
+      "step": 2740
+    },
+    {
+      "epoch": 1.882393279615978,
+      "grad_norm": 0.5101566373577191,
+      "learning_rate": 3.6635517415587264e-06,
+      "loss": 0.5299,
+      "step": 2745
+    },
+    {
+      "epoch": 1.8858220469741127,
+      "grad_norm": 0.4734869935919425,
+      "learning_rate": 3.6443342234273905e-06,
+      "loss": 0.5364,
+      "step": 2750
+    },
+    {
+      "epoch": 1.8892508143322475,
+      "grad_norm": 0.4977350264147008,
+      "learning_rate": 3.6251382967371938e-06,
+      "loss": 0.5344,
+      "step": 2755
+    },
+    {
+      "epoch": 1.8926795816903823,
+      "grad_norm": 0.45745712112103115,
+      "learning_rate": 3.6059642672181537e-06,
+      "loss": 0.5338,
+      "step": 2760
+    },
+    {
+      "epoch": 1.896108349048517,
+      "grad_norm": 0.5104324084550051,
+      "learning_rate": 3.586812440251537e-06,
+      "loss": 0.5347,
+      "step": 2765
+    },
+    {
+      "epoch": 1.899537116406652,
+      "grad_norm": 0.5335392069435976,
+      "learning_rate": 3.5676831208649887e-06,
+      "loss": 0.5346,
+      "step": 2770
+    },
+    {
+      "epoch": 1.9029658837647867,
+      "grad_norm": 0.4895513127909529,
+      "learning_rate": 3.5485766137276894e-06,
+      "loss": 0.5418,
+      "step": 2775
+    },
+    {
+      "epoch": 1.9063946511229213,
+      "grad_norm": 0.4542167715320498,
+      "learning_rate": 3.5294932231454838e-06,
+      "loss": 0.5385,
+      "step": 2780
+    },
+    {
+      "epoch": 1.909823418481056,
+      "grad_norm": 0.4744313937102546,
+      "learning_rate": 3.510433253056045e-06,
+      "loss": 0.5391,
+      "step": 2785
+    },
+    {
+      "epoch": 1.9132521858391907,
+      "grad_norm": 0.4427663775130247,
+      "learning_rate": 3.4913970070240388e-06,
+      "loss": 0.5444,
+      "step": 2790
+    },
+    {
+      "epoch": 1.9166809531973255,
+      "grad_norm": 0.47529978421135,
+      "learning_rate": 3.4723847882362755e-06,
+      "loss": 0.5373,
+      "step": 2795
+    },
+    {
+      "epoch": 1.9201097205554603,
+      "grad_norm": 0.4561099449022098,
+      "learning_rate": 3.4533968994968913e-06,
+      "loss": 0.5332,
+      "step": 2800
+    },
+    {
+      "epoch": 1.923538487913595,
+      "grad_norm": 0.47185420315317955,
+      "learning_rate": 3.4344336432225205e-06,
+      "loss": 0.5277,
+      "step": 2805
+    },
+    {
+      "epoch": 1.9269672552717299,
+      "grad_norm": 0.4730035039099479,
+      "learning_rate": 3.41549532143748e-06,
+      "loss": 0.5377,
+      "step": 2810
+    },
+    {
+      "epoch": 1.9303960226298646,
+      "grad_norm": 0.48079447901439726,
+      "learning_rate": 3.3965822357689626e-06,
+      "loss": 0.5349,
+      "step": 2815
+    },
+    {
+      "epoch": 1.9338247899879994,
+      "grad_norm": 0.458616349152276,
+      "learning_rate": 3.3776946874422268e-06,
+      "loss": 0.5312,
+      "step": 2820
+    },
+    {
+      "epoch": 1.937253557346134,
+      "grad_norm": 0.4555844676968641,
+      "learning_rate": 3.3588329772758054e-06,
+      "loss": 0.5359,
+      "step": 2825
+    },
+    {
+      "epoch": 1.9406823247042688,
+      "grad_norm": 0.4916927479459658,
+      "learning_rate": 3.3399974056767095e-06,
+      "loss": 0.5343,
+      "step": 2830
+    },
+    {
+      "epoch": 1.9441110920624034,
+      "grad_norm": 0.47636277234450725,
+      "learning_rate": 3.3211882726356447e-06,
+      "loss": 0.5363,
+      "step": 2835
+    },
+    {
+      "epoch": 1.9475398594205382,
+      "grad_norm": 0.46891572696488587,
+      "learning_rate": 3.30240587772224e-06,
+      "loss": 0.537,
+      "step": 2840
+    },
+    {
+      "epoch": 1.950968626778673,
+      "grad_norm": 0.4845143780415882,
+      "learning_rate": 3.283650520080265e-06,
+      "loss": 0.5367,
+      "step": 2845
+    },
+    {
+      "epoch": 1.9543973941368078,
+      "grad_norm": 0.4714569513171776,
+      "learning_rate": 3.2649224984228756e-06,
+      "loss": 0.5394,
+      "step": 2850
+    },
+    {
+      "epoch": 1.9578261614949426,
+      "grad_norm": 0.46470157896584324,
+      "learning_rate": 3.246222111027849e-06,
+      "loss": 0.5404,
+      "step": 2855
+    },
+    {
+      "epoch": 1.9612549288530774,
+      "grad_norm": 0.46189223769008775,
+      "learning_rate": 3.227549655732843e-06,
+      "loss": 0.5394,
+      "step": 2860
+    },
+    {
+      "epoch": 1.9646836962112122,
+      "grad_norm": 0.459933595776418,
+      "learning_rate": 3.2089054299306376e-06,
+      "loss": 0.5351,
+      "step": 2865
+    },
+    {
+      "epoch": 1.9681124635693468,
+      "grad_norm": 0.4881070886064201,
+      "learning_rate": 3.19028973056441e-06,
+      "loss": 0.5326,
+      "step": 2870
+    },
+    {
+      "epoch": 1.9715412309274816,
+      "grad_norm": 0.4518011695478976,
+      "learning_rate": 3.171702854123008e-06,
+      "loss": 0.5423,
+      "step": 2875
+    },
+    {
+      "epoch": 1.9749699982856164,
+      "grad_norm": 0.45756039660819475,
+      "learning_rate": 3.153145096636211e-06,
+      "loss": 0.5339,
+      "step": 2880
+    },
+    {
+      "epoch": 1.978398765643751,
+      "grad_norm": 0.4544722411338075,
+      "learning_rate": 3.134616753670036e-06,
+      "loss": 0.5308,
+      "step": 2885
+    },
+    {
+      "epoch": 1.9818275330018857,
+      "grad_norm": 0.4592115025179121,
+      "learning_rate": 3.1161181203220146e-06,
+      "loss": 0.5353,
+      "step": 2890
+    },
+    {
+      "epoch": 1.9852563003600205,
+      "grad_norm": 0.46992937248193806,
+      "learning_rate": 3.0976494912165e-06,
+      "loss": 0.5369,
+      "step": 2895
+    },
+    {
+      "epoch": 1.9886850677181553,
+      "grad_norm": 0.4675123294101481,
+      "learning_rate": 3.079211160499975e-06,
+      "loss": 0.5326,
+      "step": 2900
+    },
+    {
+      "epoch": 1.9921138350762901,
+      "grad_norm": 0.480213421331211,
+      "learning_rate": 3.060803421836363e-06,
+      "loss": 0.5392,
+      "step": 2905
+    },
+    {
+      "epoch": 1.995542602434425,
+      "grad_norm": 0.4628568848487098,
+      "learning_rate": 3.0424265684023556e-06,
+      "loss": 0.5321,
+      "step": 2910
+    },
+    {
+      "epoch": 1.9989713697925597,
+      "grad_norm": 0.49388511981349187,
+      "learning_rate": 3.0240808928827397e-06,
+      "loss": 0.5287,
+      "step": 2915
+    },
+    {
+      "epoch": 1.9996571232641864,
+      "eval_loss": 0.5881877541542053,
+      "eval_runtime": 204.4729,
+      "eval_samples_per_second": 48.06,
+      "eval_steps_per_second": 1.506,
+      "step": 2916
+    },
+    {
+      "epoch": 2.0024001371506945,
+      "grad_norm": 0.874753190390515,
+      "learning_rate": 3.0057666874657365e-06,
+      "loss": 0.4981,
+      "step": 2920
+    },
+    {
+      "epoch": 2.005828904508829,
+      "grad_norm": 0.5805650180046337,
+      "learning_rate": 2.9874842438383508e-06,
+      "loss": 0.4781,
+      "step": 2925
+    },
+    {
+      "epoch": 2.0092576718669637,
+      "grad_norm": 0.5290132742038788,
+      "learning_rate": 2.9692338531817205e-06,
+      "loss": 0.473,
+      "step": 2930
+    },
+    {
+      "epoch": 2.0126864392250985,
+      "grad_norm": 0.5751295049236377,
+      "learning_rate": 2.951015806166485e-06,
+      "loss": 0.4788,
+      "step": 2935
+    },
+    {
+      "epoch": 2.0161152065832333,
+      "grad_norm": 0.5442073882893506,
+      "learning_rate": 2.9328303929481507e-06,
+      "loss": 0.4756,
+      "step": 2940
+    },
+    {
+      "epoch": 2.019543973941368,
+      "grad_norm": 0.5046978835481467,
+      "learning_rate": 2.9146779031624683e-06,
+      "loss": 0.4747,
+      "step": 2945
+    },
+    {
+      "epoch": 2.022972741299503,
+      "grad_norm": 0.4694017759591277,
+      "learning_rate": 2.8965586259208295e-06,
+      "loss": 0.4787,
+      "step": 2950
+    },
+    {
+      "epoch": 2.0264015086576377,
+      "grad_norm": 0.49182764229033304,
+      "learning_rate": 2.878472849805651e-06,
+      "loss": 0.4817,
+      "step": 2955
+    },
+    {
+      "epoch": 2.0298302760157725,
+      "grad_norm": 0.5353082540824908,
+      "learning_rate": 2.860420862865787e-06,
+      "loss": 0.4814,
+      "step": 2960
+    },
+    {
+      "epoch": 2.0332590433739073,
+      "grad_norm": 0.4849568396076438,
+      "learning_rate": 2.842402952611931e-06,
+      "loss": 0.4774,
+      "step": 2965
+    },
+    {
+      "epoch": 2.0366878107320416,
+      "grad_norm": 0.455032309231727,
+      "learning_rate": 2.82441940601205e-06,
+      "loss": 0.488,
+      "step": 2970
+    },
+    {
+      "epoch": 2.0401165780901764,
+      "grad_norm": 0.47238583732747885,
+      "learning_rate": 2.8064705094868074e-06,
+      "loss": 0.4847,
+      "step": 2975
+    },
+    {
+      "epoch": 2.0435453454483112,
+      "grad_norm": 0.46589829010871375,
+      "learning_rate": 2.7885565489049948e-06,
+      "loss": 0.4745,
+      "step": 2980
+    },
+    {
+      "epoch": 2.046974112806446,
+      "grad_norm": 0.5097253834071658,
+      "learning_rate": 2.7706778095789905e-06,
+      "loss": 0.4822,
+      "step": 2985
+    },
+    {
+      "epoch": 2.050402880164581,
+      "grad_norm": 0.4744363699843961,
+      "learning_rate": 2.7528345762602125e-06,
+      "loss": 0.4837,
+      "step": 2990
+    },
+    {
+      "epoch": 2.0538316475227156,
+      "grad_norm": 0.48096853101811016,
+      "learning_rate": 2.7350271331345733e-06,
+      "loss": 0.4689,
+      "step": 2995
+    },
+    {
+      "epoch": 2.0572604148808504,
+      "grad_norm": 0.45888264058390604,
+      "learning_rate": 2.7172557638179674e-06,
+      "loss": 0.4814,
+      "step": 3000
+    },
+    {
+      "epoch": 2.0606891822389852,
+      "grad_norm": 0.5074519719478089,
+      "learning_rate": 2.6995207513517484e-06,
+      "loss": 0.4835,
+      "step": 3005
+    },
+    {
+      "epoch": 2.06411794959712,
+      "grad_norm": 0.46736698036035224,
+      "learning_rate": 2.681822378198221e-06,
+      "loss": 0.4821,
+      "step": 3010
+    },
+    {
+      "epoch": 2.0675467169552544,
+      "grad_norm": 0.4677318714005963,
+      "learning_rate": 2.6641609262361343e-06,
+      "loss": 0.4825,
+      "step": 3015
+    },
+    {
+      "epoch": 2.070975484313389,
+      "grad_norm": 0.47774880777961237,
+      "learning_rate": 2.6465366767562162e-06,
+      "loss": 0.4748,
+      "step": 3020
+    },
+    {
+      "epoch": 2.074404251671524,
+      "grad_norm": 0.45924654536691306,
+      "learning_rate": 2.628949910456663e-06,
+      "loss": 0.4767,
+      "step": 3025
+    },
+    {
+      "epoch": 2.077833019029659,
+      "grad_norm": 0.47391360563192103,
+      "learning_rate": 2.611400907438685e-06,
+      "loss": 0.4775,
+      "step": 3030
+    },
+    {
+      "epoch": 2.0812617863877936,
+      "grad_norm": 0.49156177694024655,
+      "learning_rate": 2.5938899472020545e-06,
+      "loss": 0.4789,
+      "step": 3035
+    },
+    {
+      "epoch": 2.0846905537459284,
+      "grad_norm": 0.4723037751330721,
+      "learning_rate": 2.5764173086406306e-06,
+      "loss": 0.4829,
+      "step": 3040
+    },
+    {
+      "epoch": 2.088119321104063,
+      "grad_norm": 0.46114598626056696,
+      "learning_rate": 2.558983270037937e-06,
+      "loss": 0.4743,
+      "step": 3045
+    },
+    {
+      "epoch": 2.091548088462198,
+      "grad_norm": 0.4718425543262081,
+      "learning_rate": 2.5415881090627227e-06,
+      "loss": 0.4785,
+      "step": 3050
+    },
+    {
+      "epoch": 2.0949768558203328,
+      "grad_norm": 0.48700567548085505,
+      "learning_rate": 2.5242321027645354e-06,
+      "loss": 0.4731,
+      "step": 3055
+    },
+    {
+      "epoch": 2.098405623178467,
+      "grad_norm": 0.47428262784263286,
+      "learning_rate": 2.506915527569318e-06,
+      "loss": 0.4805,
+      "step": 3060
+    },
+    {
+      "epoch": 2.101834390536602,
+      "grad_norm": 0.482699000253055,
+      "learning_rate": 2.4896386592749994e-06,
+      "loss": 0.4809,
+      "step": 3065
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.48854325554987943,
+      "learning_rate": 2.472401773047107e-06,
+      "loss": 0.4753,
+      "step": 3070
+    },
+    {
+      "epoch": 2.1086919252528715,
+      "grad_norm": 0.5091234833700028,
+      "learning_rate": 2.455205143414375e-06,
+      "loss": 0.477,
+      "step": 3075
+    },
+    {
+      "epoch": 2.1121206926110063,
+      "grad_norm": 0.474218431216323,
+      "learning_rate": 2.438049044264382e-06,
+      "loss": 0.4779,
+      "step": 3080
+    },
+    {
+      "epoch": 2.115549459969141,
+      "grad_norm": 0.4762105524360361,
+      "learning_rate": 2.4209337488391886e-06,
+      "loss": 0.4763,
+      "step": 3085
+    },
+    {
+      "epoch": 2.118978227327276,
+      "grad_norm": 0.4708922326438511,
+      "learning_rate": 2.4038595297309712e-06,
+      "loss": 0.4803,
+      "step": 3090
+    },
+    {
+      "epoch": 2.1224069946854107,
+      "grad_norm": 0.48521692596512245,
+      "learning_rate": 2.386826658877706e-06,
+      "loss": 0.482,
+      "step": 3095
+    },
+    {
+      "epoch": 2.1258357620435455,
+      "grad_norm": 0.494787221830382,
+      "learning_rate": 2.3698354075588105e-06,
+      "loss": 0.4803,
+      "step": 3100
+    },
+    {
+      "epoch": 2.12926452940168,
+      "grad_norm": 0.48222263673783083,
+      "learning_rate": 2.3528860463908375e-06,
+      "loss": 0.472,
+      "step": 3105
+    },
+    {
+      "epoch": 2.1326932967598147,
+      "grad_norm": 0.4824015117741457,
+      "learning_rate": 2.3359788453231723e-06,
+      "loss": 0.4797,
+      "step": 3110
+    },
+    {
+      "epoch": 2.1361220641179495,
+      "grad_norm": 0.49605929290504214,
+      "learning_rate": 2.319114073633712e-06,
+      "loss": 0.4788,
+      "step": 3115
+    },
+    {
+      "epoch": 2.1395508314760843,
+      "grad_norm": 0.474543321413953,
+      "learning_rate": 2.3022919999245964e-06,
+      "loss": 0.4773,
+      "step": 3120
+    },
+    {
+      "epoch": 2.142979598834219,
+      "grad_norm": 0.501540938539211,
+      "learning_rate": 2.285512892117921e-06,
+      "loss": 0.4822,
+      "step": 3125
+    },
+    {
+      "epoch": 2.146408366192354,
+      "grad_norm": 0.5038191337466582,
+      "learning_rate": 2.2687770174514674e-06,
+      "loss": 0.4791,
+      "step": 3130
+    },
+    {
+      "epoch": 2.1498371335504887,
+      "grad_norm": 0.4648773455909595,
+      "learning_rate": 2.2520846424744545e-06,
+      "loss": 0.4828,
+      "step": 3135
+    },
+    {
+      "epoch": 2.1532659009086235,
+      "grad_norm": 0.4620270257517909,
+      "learning_rate": 2.23543603304329e-06,
+      "loss": 0.479,
+      "step": 3140
+    },
+    {
+      "epoch": 2.1566946682667583,
+      "grad_norm": 0.4950092604387871,
+      "learning_rate": 2.218831454317336e-06,
+      "loss": 0.4746,
+      "step": 3145
+    },
+    {
+      "epoch": 2.160123435624893,
+      "grad_norm": 0.46350733624535256,
+      "learning_rate": 2.20227117075468e-06,
+      "loss": 0.4817,
+      "step": 3150
+    },
+    {
+      "epoch": 2.1635522029830274,
+      "grad_norm": 0.4842429161440888,
+      "learning_rate": 2.1857554461079356e-06,
+      "loss": 0.4829,
+      "step": 3155
+    },
+    {
+      "epoch": 2.166980970341162,
+      "grad_norm": 0.48989158037943475,
+      "learning_rate": 2.1692845434200323e-06,
+      "loss": 0.477,
+      "step": 3160
+    },
+    {
+      "epoch": 2.170409737699297,
+      "grad_norm": 0.444636329908828,
+      "learning_rate": 2.152858725020025e-06,
+      "loss": 0.4762,
+      "step": 3165
+    },
+    {
+      "epoch": 2.173838505057432,
+      "grad_norm": 0.46845184246079136,
+      "learning_rate": 2.136478252518924e-06,
+      "loss": 0.4867,
+      "step": 3170
+    },
+    {
+      "epoch": 2.1772672724155666,
+      "grad_norm": 0.5158451108463605,
+      "learning_rate": 2.1201433868055223e-06,
+      "loss": 0.4773,
+      "step": 3175
+    },
+    {
+      "epoch": 2.1806960397737014,
+      "grad_norm": 0.484433856048475,
+      "learning_rate": 2.103854388042243e-06,
+      "loss": 0.4808,
+      "step": 3180
+    },
+    {
+      "epoch": 2.184124807131836,
+      "grad_norm": 0.4689424096794033,
+      "learning_rate": 2.08761151566099e-06,
+      "loss": 0.4851,
+      "step": 3185
+    },
+    {
+      "epoch": 2.187553574489971,
+      "grad_norm": 0.46853301342354137,
+      "learning_rate": 2.071415028359026e-06,
+      "loss": 0.4854,
+      "step": 3190
+    },
+    {
+      "epoch": 2.190982341848106,
+      "grad_norm": 0.46534481693626095,
+      "learning_rate": 2.0552651840948473e-06,
+      "loss": 0.4789,
+      "step": 3195
+    },
+    {
+      "epoch": 2.19441110920624,
+      "grad_norm": 0.47458944353516924,
+      "learning_rate": 2.0391622400840665e-06,
+      "loss": 0.4801,
+      "step": 3200
+    },
+    {
+      "epoch": 2.197839876564375,
+      "grad_norm": 0.46593527956244135,
+      "learning_rate": 2.023106452795339e-06,
+      "loss": 0.4726,
+      "step": 3205
+    },
+    {
+      "epoch": 2.2012686439225098,
+      "grad_norm": 0.4773652150154532,
+      "learning_rate": 2.0070980779462513e-06,
+      "loss": 0.4806,
+      "step": 3210
+    },
+    {
+      "epoch": 2.2046974112806446,
+      "grad_norm": 0.4647727192391881,
+      "learning_rate": 1.9911373704992617e-06,
+      "loss": 0.4806,
+      "step": 3215
+    },
+    {
+      "epoch": 2.2081261786387794,
+      "grad_norm": 0.4794479506202543,
+      "learning_rate": 1.975224584657648e-06,
+      "loss": 0.4791,
+      "step": 3220
+    },
+    {
+      "epoch": 2.211554945996914,
+      "grad_norm": 0.49928694289640674,
+      "learning_rate": 1.9593599738614382e-06,
+      "loss": 0.4812,
+      "step": 3225
+    },
+    {
+      "epoch": 2.214983713355049,
+      "grad_norm": 0.489572736256811,
+      "learning_rate": 1.943543790783392e-06,
+      "loss": 0.4822,
+      "step": 3230
+    },
+    {
+      "epoch": 2.2184124807131838,
+      "grad_norm": 0.460165152565946,
+      "learning_rate": 1.927776287324969e-06,
+      "loss": 0.4845,
+      "step": 3235
+    },
+    {
+      "epoch": 2.2218412480713186,
+      "grad_norm": 0.4741945403355791,
+      "learning_rate": 1.9120577146123125e-06,
+      "loss": 0.4753,
+      "step": 3240
+    },
+    {
+      "epoch": 2.225270015429453,
+      "grad_norm": 0.4816051218900053,
+      "learning_rate": 1.8963883229922603e-06,
+      "loss": 0.4759,
+      "step": 3245
+    },
+    {
+      "epoch": 2.2286987827875877,
+      "grad_norm": 0.4865842140542505,
+      "learning_rate": 1.8807683620283496e-06,
+      "loss": 0.4887,
+      "step": 3250
+    },
+    {
+      "epoch": 2.2321275501457225,
+      "grad_norm": 0.49327525598298544,
+      "learning_rate": 1.8651980804968466e-06,
+      "loss": 0.4828,
+      "step": 3255
+    },
+    {
+      "epoch": 2.2355563175038573,
+      "grad_norm": 0.4710718086720219,
+      "learning_rate": 1.8496777263827775e-06,
+      "loss": 0.4805,
+      "step": 3260
+    },
+    {
+      "epoch": 2.238985084861992,
+      "grad_norm": 0.4790195567473799,
+      "learning_rate": 1.834207546875988e-06,
+      "loss": 0.4761,
+      "step": 3265
+    },
+    {
+      "epoch": 2.242413852220127,
+      "grad_norm": 0.47138983012088903,
+      "learning_rate": 1.8187877883672024e-06,
+      "loss": 0.4809,
+      "step": 3270
+    },
+    {
+      "epoch": 2.2458426195782617,
+      "grad_norm": 0.48900189445082154,
+      "learning_rate": 1.8034186964440937e-06,
+      "loss": 0.482,
+      "step": 3275
+    },
+    {
+      "epoch": 2.2492713869363965,
+      "grad_norm": 0.4723460008561715,
+      "learning_rate": 1.7881005158873826e-06,
+      "loss": 0.478,
+      "step": 3280
+    },
+    {
+      "epoch": 2.2527001542945313,
+      "grad_norm": 0.45431679994556473,
+      "learning_rate": 1.7728334906669342e-06,
+      "loss": 0.4828,
+      "step": 3285
+    },
+    {
+      "epoch": 2.2561289216526657,
+      "grad_norm": 0.45761440455845676,
+      "learning_rate": 1.757617863937865e-06,
+      "loss": 0.4707,
+      "step": 3290
+    },
+    {
+      "epoch": 2.2595576890108005,
+      "grad_norm": 0.48498068790945004,
+      "learning_rate": 1.7424538780366884e-06,
+      "loss": 0.4724,
+      "step": 3295
+    },
+    {
+      "epoch": 2.2629864563689353,
+      "grad_norm": 0.45783948313254996,
+      "learning_rate": 1.7273417744774323e-06,
+      "loss": 0.4774,
+      "step": 3300
+    },
+    {
+      "epoch": 2.26641522372707,
+      "grad_norm": 0.46625983570633794,
+      "learning_rate": 1.7122817939478104e-06,
+      "loss": 0.4781,
+      "step": 3305
+    },
+    {
+      "epoch": 2.269843991085205,
+      "grad_norm": 0.46981244219286544,
+      "learning_rate": 1.6972741763053835e-06,
+      "loss": 0.4771,
+      "step": 3310
+    },
+    {
+      "epoch": 2.2732727584433396,
+      "grad_norm": 0.47749785750136037,
+      "learning_rate": 1.6823191605737316e-06,
+      "loss": 0.4812,
+      "step": 3315
+    },
+    {
+      "epoch": 2.2767015258014744,
+      "grad_norm": 0.4781030507544295,
+      "learning_rate": 1.6674169849386606e-06,
+      "loss": 0.4745,
+      "step": 3320
+    },
+    {
+      "epoch": 2.2801302931596092,
+      "grad_norm": 0.47642004107759217,
+      "learning_rate": 1.6525678867443989e-06,
+      "loss": 0.4782,
+      "step": 3325
+    },
+    {
+      "epoch": 2.283559060517744,
+      "grad_norm": 0.4595236798861582,
+      "learning_rate": 1.6377721024898214e-06,
+      "loss": 0.4846,
+      "step": 3330
+    },
+    {
+      "epoch": 2.286987827875879,
+      "grad_norm": 0.4607279352731881,
+      "learning_rate": 1.623029867824678e-06,
+      "loss": 0.4783,
+      "step": 3335
+    },
+    {
+      "epoch": 2.290416595234013,
+      "grad_norm": 0.4857822760894794,
+      "learning_rate": 1.608341417545849e-06,
+      "loss": 0.4774,
+      "step": 3340
+    },
+    {
+      "epoch": 2.293845362592148,
+      "grad_norm": 0.4577566161740047,
+      "learning_rate": 1.593706985593599e-06,
+      "loss": 0.4727,
+      "step": 3345
+    },
+    {
+      "epoch": 2.297274129950283,
+      "grad_norm": 0.4573512037078888,
+      "learning_rate": 1.5791268050478487e-06,
+      "loss": 0.4746,
+      "step": 3350
+    },
+    {
+      "epoch": 2.3007028973084176,
+      "grad_norm": 0.47439501092754294,
+      "learning_rate": 1.5646011081244717e-06,
+      "loss": 0.4865,
+      "step": 3355
+    },
+    {
+      "epoch": 2.3041316646665524,
+      "grad_norm": 0.49446637460132087,
+      "learning_rate": 1.5501301261715896e-06,
+      "loss": 0.4751,
+      "step": 3360
+    },
+    {
+      "epoch": 2.307560432024687,
+      "grad_norm": 0.4952360092448791,
+      "learning_rate": 1.535714089665889e-06,
+      "loss": 0.4815,
+      "step": 3365
+    },
+    {
+      "epoch": 2.310989199382822,
+      "grad_norm": 0.4731157255344,
+      "learning_rate": 1.5213532282089466e-06,
+      "loss": 0.4792,
+      "step": 3370
+    },
+    {
+      "epoch": 2.314417966740957,
+      "grad_norm": 0.4538368226200657,
+      "learning_rate": 1.5070477705235803e-06,
+      "loss": 0.4818,
+      "step": 3375
+    },
+    {
+      "epoch": 2.317846734099091,
+      "grad_norm": 0.45299648856103863,
+      "learning_rate": 1.4927979444502028e-06,
+      "loss": 0.4761,
+      "step": 3380
+    },
+    {
+      "epoch": 2.321275501457226,
+      "grad_norm": 0.48133972008005105,
+      "learning_rate": 1.478603976943186e-06,
+      "loss": 0.4796,
+      "step": 3385
+    },
+    {
+      "epoch": 2.3247042688153607,
+      "grad_norm": 0.46559397796529,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.4833,
+      "step": 3390
+    },
+    {
+      "epoch": 2.3281330361734955,
+      "grad_norm": 0.46261472772428197,
+      "learning_rate": 1.4503845209939083e-06,
+      "loss": 0.4866,
+      "step": 3395
+    },
+    {
+      "epoch": 2.3315618035316303,
+      "grad_norm": 0.48901577194657203,
+      "learning_rate": 1.4363594819977606e-06,
+      "loss": 0.4802,
+      "step": 3400
+    },
+    {
+      "epoch": 2.334990570889765,
+      "grad_norm": 0.47026562258797505,
+      "learning_rate": 1.4223912004530593e-06,
+      "loss": 0.4842,
+      "step": 3405
+    },
+    {
+      "epoch": 2.3384193382479,
+      "grad_norm": 0.46476370601090344,
+      "learning_rate": 1.4084798988300684e-06,
+      "loss": 0.4746,
+      "step": 3410
+    },
+    {
+      "epoch": 2.3418481056060347,
+      "grad_norm": 0.4736014629302977,
+      "learning_rate": 1.3946257986915486e-06,
+      "loss": 0.4811,
+      "step": 3415
+    },
+    {
+      "epoch": 2.3452768729641695,
+      "grad_norm": 0.4734489622060502,
+      "learning_rate": 1.3808291206892232e-06,
+      "loss": 0.4773,
+      "step": 3420
+    },
+    {
+      "epoch": 2.3487056403223043,
+      "grad_norm": 0.4629066750832987,
+      "learning_rate": 1.3670900845602585e-06,
+      "loss": 0.4829,
+      "step": 3425
+    },
+    {
+      "epoch": 2.3521344076804387,
+      "grad_norm": 0.4681227467302218,
+      "learning_rate": 1.3534089091237757e-06,
+      "loss": 0.476,
+      "step": 3430
+    },
+    {
+      "epoch": 2.3555631750385735,
+      "grad_norm": 0.454191418129244,
+      "learning_rate": 1.3397858122773544e-06,
+      "loss": 0.4747,
+      "step": 3435
+    },
+    {
+      "epoch": 2.3589919423967083,
+      "grad_norm": 0.4579452618693956,
+      "learning_rate": 1.3262210109935719e-06,
+      "loss": 0.4832,
+      "step": 3440
+    },
+    {
+      "epoch": 2.362420709754843,
+      "grad_norm": 0.45074466888367914,
+      "learning_rate": 1.312714721316537e-06,
+      "loss": 0.4755,
+      "step": 3445
+    },
+    {
+      "epoch": 2.365849477112978,
+      "grad_norm": 0.4723349908020595,
+      "learning_rate": 1.2992671583584587e-06,
+      "loss": 0.4862,
+      "step": 3450
+    },
+    {
+      "epoch": 2.3692782444711127,
+      "grad_norm": 0.4708208262941508,
+      "learning_rate": 1.2858785362962166e-06,
+      "loss": 0.4777,
+      "step": 3455
+    },
+    {
+      "epoch": 2.3727070118292475,
+      "grad_norm": 0.47375162486732203,
+      "learning_rate": 1.2725490683679458e-06,
+      "loss": 0.4729,
+      "step": 3460
+    },
+    {
+      "epoch": 2.3761357791873823,
+      "grad_norm": 0.4678384719988464,
+      "learning_rate": 1.2592789668696482e-06,
+      "loss": 0.4787,
+      "step": 3465
+    },
+    {
+      "epoch": 2.379564546545517,
+      "grad_norm": 0.4702212675672681,
+      "learning_rate": 1.2460684431518055e-06,
+      "loss": 0.4813,
+      "step": 3470
+    },
+    {
+      "epoch": 2.3829933139036514,
+      "grad_norm": 0.4532398056344406,
+      "learning_rate": 1.232917707616017e-06,
+      "loss": 0.4735,
+      "step": 3475
+    },
+    {
+      "epoch": 2.3864220812617862,
+      "grad_norm": 0.45797439884914765,
+      "learning_rate": 1.2198269697116416e-06,
+      "loss": 0.4778,
+      "step": 3480
+    },
+    {
+      "epoch": 2.389850848619921,
+      "grad_norm": 0.4723176003130105,
+      "learning_rate": 1.206796437932472e-06,
+      "loss": 0.4774,
+      "step": 3485
+    },
+    {
+      "epoch": 2.393279615978056,
+      "grad_norm": 0.46906434827216664,
+      "learning_rate": 1.1938263198134087e-06,
+      "loss": 0.4808,
+      "step": 3490
+    },
+    {
+      "epoch": 2.3967083833361906,
+      "grad_norm": 0.4678613569139536,
+      "learning_rate": 1.1809168219271488e-06,
+      "loss": 0.4795,
+      "step": 3495
+    },
+    {
+      "epoch": 2.4001371506943254,
+      "grad_norm": 0.47248188702171634,
+      "learning_rate": 1.168068149880912e-06,
+      "loss": 0.4781,
+      "step": 3500
+    },
+    {
+      "epoch": 2.4035659180524602,
+      "grad_norm": 0.46791595503949973,
+      "learning_rate": 1.1552805083131468e-06,
+      "loss": 0.4786,
+      "step": 3505
+    },
+    {
+      "epoch": 2.406994685410595,
+      "grad_norm": 0.45897414399547337,
+      "learning_rate": 1.1425541008902852e-06,
+      "loss": 0.4775,
+      "step": 3510
+    },
+    {
+      "epoch": 2.41042345276873,
+      "grad_norm": 0.4610154781111884,
+      "learning_rate": 1.1298891303034942e-06,
+      "loss": 0.4823,
+      "step": 3515
+    },
+    {
+      "epoch": 2.4138522201268646,
+      "grad_norm": 0.46406483424598494,
+      "learning_rate": 1.1172857982654445e-06,
+      "loss": 0.4761,
+      "step": 3520
+    },
+    {
+      "epoch": 2.417280987484999,
+      "grad_norm": 0.4475457001447567,
+      "learning_rate": 1.1047443055071033e-06,
+      "loss": 0.4784,
+      "step": 3525
+    },
+    {
+      "epoch": 2.4207097548431338,
+      "grad_norm": 0.46031864575357934,
+      "learning_rate": 1.092264851774536e-06,
+      "loss": 0.4741,
+      "step": 3530
+    },
+    {
+      "epoch": 2.4241385222012686,
+      "grad_norm": 0.4642032024999473,
+      "learning_rate": 1.0798476358257198e-06,
+      "loss": 0.4763,
+      "step": 3535
+    },
+    {
+      "epoch": 2.4275672895594034,
+      "grad_norm": 0.46462241022456946,
+      "learning_rate": 1.067492855427385e-06,
+      "loss": 0.4787,
+      "step": 3540
+    },
+    {
+      "epoch": 2.430996056917538,
+      "grad_norm": 0.4769725916465071,
+      "learning_rate": 1.0552007073518628e-06,
+      "loss": 0.4826,
+      "step": 3545
+    },
+    {
+      "epoch": 2.434424824275673,
+      "grad_norm": 0.46174302133938927,
+      "learning_rate": 1.0429713873739505e-06,
+      "loss": 0.4847,
+      "step": 3550
+    },
+    {
+      "epoch": 2.4378535916338078,
+      "grad_norm": 0.45270102445912286,
+      "learning_rate": 1.0308050902677907e-06,
+      "loss": 0.4793,
+      "step": 3555
+    },
+    {
+      "epoch": 2.4412823589919426,
+      "grad_norm": 0.46466350347691254,
+      "learning_rate": 1.0187020098037759e-06,
+      "loss": 0.4807,
+      "step": 3560
+    },
+    {
+      "epoch": 2.444711126350077,
+      "grad_norm": 0.4615768268062226,
+      "learning_rate": 1.0066623387454582e-06,
+      "loss": 0.4811,
+      "step": 3565
+    },
+    {
+      "epoch": 2.4481398937082117,
+      "grad_norm": 0.4544451248675266,
+      "learning_rate": 9.946862688464753e-07,
+      "loss": 0.4767,
+      "step": 3570
+    },
+    {
+      "epoch": 2.4515686610663465,
+      "grad_norm": 0.4742062204892906,
+      "learning_rate": 9.827739908475099e-07,
+      "loss": 0.4776,
+      "step": 3575
+    },
+    {
+      "epoch": 2.4549974284244813,
+      "grad_norm": 0.4672145082984867,
+      "learning_rate": 9.709256944732343e-07,
+      "loss": 0.4804,
+      "step": 3580
+    },
+    {
+      "epoch": 2.458426195782616,
+      "grad_norm": 0.44899579410081497,
+      "learning_rate": 9.591415684292975e-07,
+      "loss": 0.4778,
+      "step": 3585
+    },
+    {
+      "epoch": 2.461854963140751,
+      "grad_norm": 0.46844243005424735,
+      "learning_rate": 9.474218003993275e-07,
+      "loss": 0.4753,
+      "step": 3590
+    },
+    {
+      "epoch": 2.4652837304988857,
+      "grad_norm": 0.4654191148385016,
+      "learning_rate": 9.357665770419244e-07,
+      "loss": 0.482,
+      "step": 3595
+    },
+    {
+      "epoch": 2.4687124978570205,
+      "grad_norm": 0.45942731935750447,
+      "learning_rate": 9.241760839877023e-07,
+      "loss": 0.478,
+      "step": 3600
+    },
+    {
+      "epoch": 2.4721412652151553,
+      "grad_norm": 0.4666227072724768,
+      "learning_rate": 9.126505058363283e-07,
+      "loss": 0.4753,
+      "step": 3605
+    },
+    {
+      "epoch": 2.47557003257329,
+      "grad_norm": 0.4519254146818986,
+      "learning_rate": 9.011900261535767e-07,
+      "loss": 0.4783,
+      "step": 3610
+    },
+    {
+      "epoch": 2.4789987999314245,
+      "grad_norm": 0.463358478501908,
+      "learning_rate": 8.897948274684154e-07,
+      "loss": 0.4739,
+      "step": 3615
+    },
+    {
+      "epoch": 2.4824275672895593,
+      "grad_norm": 0.45511211330917706,
+      "learning_rate": 8.784650912700909e-07,
+      "loss": 0.4729,
+      "step": 3620
+    },
+    {
+      "epoch": 2.485856334647694,
+      "grad_norm": 0.47432321264965754,
+      "learning_rate": 8.672009980052442e-07,
+      "loss": 0.4811,
+      "step": 3625
+    },
+    {
+      "epoch": 2.489285102005829,
+      "grad_norm": 0.4673877749252423,
+      "learning_rate": 8.560027270750276e-07,
+      "loss": 0.4814,
+      "step": 3630
+    },
+    {
+      "epoch": 2.4927138693639637,
+      "grad_norm": 0.4579464435065071,
+      "learning_rate": 8.44870456832258e-07,
+      "loss": 0.4806,
+      "step": 3635
+    },
+    {
+      "epoch": 2.4961426367220985,
+      "grad_norm": 0.476376159360163,
+      "learning_rate": 8.338043645785698e-07,
+      "loss": 0.4823,
+      "step": 3640
+    },
+    {
+      "epoch": 2.4995714040802333,
+      "grad_norm": 0.46673661354408297,
+      "learning_rate": 8.228046265615908e-07,
+      "loss": 0.4759,
+      "step": 3645
+    },
+    {
+      "epoch": 2.503000171438368,
+      "grad_norm": 0.4689234252839364,
+      "learning_rate": 8.118714179721404e-07,
+      "loss": 0.4781,
+      "step": 3650
+    },
+    {
+      "epoch": 2.5064289387965024,
+      "grad_norm": 0.457577939504501,
+      "learning_rate": 8.010049129414333e-07,
+      "loss": 0.4754,
+      "step": 3655
+    },
+    {
+      "epoch": 2.509857706154637,
+      "grad_norm": 0.46767489107398075,
+      "learning_rate": 7.902052845383112e-07,
+      "loss": 0.4832,
+      "step": 3660
+    },
+    {
+      "epoch": 2.513286473512772,
+      "grad_norm": 0.47806526391479376,
+      "learning_rate": 7.794727047664807e-07,
+      "loss": 0.4777,
+      "step": 3665
+    },
+    {
+      "epoch": 2.516715240870907,
+      "grad_norm": 0.44998017534315693,
+      "learning_rate": 7.6880734456178e-07,
+      "loss": 0.4728,
+      "step": 3670
+    },
+    {
+      "epoch": 2.5201440082290416,
+      "grad_norm": 0.4477060965035502,
+      "learning_rate": 7.58209373789453e-07,
+      "loss": 0.4734,
+      "step": 3675
+    },
+    {
+      "epoch": 2.5235727755871764,
+      "grad_norm": 0.46217386088270807,
+      "learning_rate": 7.476789612414414e-07,
+      "loss": 0.4846,
+      "step": 3680
+    },
+    {
+      "epoch": 2.527001542945311,
+      "grad_norm": 0.4563904666440247,
+      "learning_rate": 7.372162746337059e-07,
+      "loss": 0.4816,
+      "step": 3685
+    },
+    {
+      "epoch": 2.530430310303446,
+      "grad_norm": 0.44520759689704725,
+      "learning_rate": 7.268214806035423e-07,
+      "loss": 0.4768,
+      "step": 3690
+    },
+    {
+      "epoch": 2.533859077661581,
+      "grad_norm": 0.4542091051330333,
+      "learning_rate": 7.164947447069343e-07,
+      "loss": 0.4798,
+      "step": 3695
+    },
+    {
+      "epoch": 2.5372878450197156,
+      "grad_norm": 0.45146271333603777,
+      "learning_rate": 7.062362314159211e-07,
+      "loss": 0.474,
+      "step": 3700
+    },
+    {
+      "epoch": 2.5407166123778504,
+      "grad_norm": 0.4843676817399376,
+      "learning_rate": 6.960461041159677e-07,
+      "loss": 0.4826,
+      "step": 3705
+    },
+    {
+      "epoch": 2.5441453797359848,
+      "grad_norm": 0.450384975906084,
+      "learning_rate": 6.859245251033697e-07,
+      "loss": 0.4768,
+      "step": 3710
+    },
+    {
+      "epoch": 2.5475741470941196,
+      "grad_norm": 0.4460709873735881,
+      "learning_rate": 6.758716555826683e-07,
+      "loss": 0.4733,
+      "step": 3715
+    },
+    {
+      "epoch": 2.5510029144522544,
+      "grad_norm": 0.4521937918515753,
+      "learning_rate": 6.658876556640781e-07,
+      "loss": 0.4769,
+      "step": 3720
+    },
+    {
+      "epoch": 2.554431681810389,
+      "grad_norm": 0.4490013681026186,
+      "learning_rate": 6.559726843609421e-07,
+      "loss": 0.4805,
+      "step": 3725
+    },
+    {
+      "epoch": 2.557860449168524,
+      "grad_norm": 0.45604478361558604,
+      "learning_rate": 6.461268995871967e-07,
+      "loss": 0.472,
+      "step": 3730
+    },
+    {
+      "epoch": 2.5612892165266588,
+      "grad_norm": 0.4684730083561354,
+      "learning_rate": 6.36350458154859e-07,
+      "loss": 0.4823,
+      "step": 3735
+    },
+    {
+      "epoch": 2.5647179838847936,
+      "grad_norm": 0.44901484120067714,
+      "learning_rate": 6.266435157715222e-07,
+      "loss": 0.4799,
+      "step": 3740
+    },
+    {
+      "epoch": 2.568146751242928,
+      "grad_norm": 0.45631182390582853,
+      "learning_rate": 6.170062270378862e-07,
+      "loss": 0.4762,
+      "step": 3745
+    },
+    {
+      "epoch": 2.5715755186010627,
+      "grad_norm": 0.4602387391697606,
+      "learning_rate": 6.074387454452891e-07,
+      "loss": 0.4713,
+      "step": 3750
+    },
+    {
+      "epoch": 2.5750042859591975,
+      "grad_norm": 0.4491465188344547,
+      "learning_rate": 5.979412233732578e-07,
+      "loss": 0.4751,
+      "step": 3755
+    },
+    {
+      "epoch": 2.5784330533173323,
+      "grad_norm": 0.46805364957632967,
+      "learning_rate": 5.885138120870965e-07,
+      "loss": 0.4789,
+      "step": 3760
+    },
+    {
+      "epoch": 2.581861820675467,
+      "grad_norm": 0.4531331451215189,
+      "learning_rate": 5.791566617354599e-07,
+      "loss": 0.4754,
+      "step": 3765
+    },
+    {
+      "epoch": 2.585290588033602,
+      "grad_norm": 0.4652091913696696,
+      "learning_rate": 5.698699213479697e-07,
+      "loss": 0.4763,
+      "step": 3770
+    },
+    {
+      "epoch": 2.5887193553917367,
+      "grad_norm": 0.4542601341667562,
+      "learning_rate": 5.606537388328481e-07,
+      "loss": 0.4789,
+      "step": 3775
+    },
+    {
+      "epoch": 2.5921481227498715,
+      "grad_norm": 0.4447156024624712,
+      "learning_rate": 5.515082609745465e-07,
+      "loss": 0.476,
+      "step": 3780
+    },
+    {
+      "epoch": 2.5955768901080063,
+      "grad_norm": 0.45794390952267283,
+      "learning_rate": 5.424336334314212e-07,
+      "loss": 0.4781,
+      "step": 3785
+    },
+    {
+      "epoch": 2.599005657466141,
+      "grad_norm": 0.45160699771932494,
+      "learning_rate": 5.334300007334065e-07,
+      "loss": 0.4771,
+      "step": 3790
+    },
+    {
+      "epoch": 2.602434424824276,
+      "grad_norm": 0.4618426839968327,
+      "learning_rate": 5.244975062797176e-07,
+      "loss": 0.4728,
+      "step": 3795
+    },
+    {
+      "epoch": 2.6058631921824107,
+      "grad_norm": 0.4486150363583273,
+      "learning_rate": 5.156362923365587e-07,
+      "loss": 0.4748,
+      "step": 3800
+    },
+    {
+      "epoch": 2.609291959540545,
+      "grad_norm": 0.4545115387932617,
+      "learning_rate": 5.068465000348666e-07,
+      "loss": 0.4739,
+      "step": 3805
+    },
+    {
+      "epoch": 2.61272072689868,
+      "grad_norm": 0.4623008940487107,
+      "learning_rate": 4.981282693680584e-07,
+      "loss": 0.4783,
+      "step": 3810
+    },
+    {
+      "epoch": 2.6161494942568146,
+      "grad_norm": 0.4653422446290839,
+      "learning_rate": 4.894817391897988e-07,
+      "loss": 0.4854,
+      "step": 3815
+    },
+    {
+      "epoch": 2.6195782616149494,
+      "grad_norm": 0.4515048388429379,
+      "learning_rate": 4.80907047211796e-07,
+      "loss": 0.4831,
+      "step": 3820
+    },
+    {
+      "epoch": 2.6230070289730842,
+      "grad_norm": 0.4639284889413238,
+      "learning_rate": 4.7240433000160233e-07,
+      "loss": 0.4836,
+      "step": 3825
+    },
+    {
+      "epoch": 2.626435796331219,
+      "grad_norm": 0.4465998513923373,
+      "learning_rate": 4.639737229804403e-07,
+      "loss": 0.4789,
+      "step": 3830
+    },
+    {
+      "epoch": 2.629864563689354,
+      "grad_norm": 0.46955783511456173,
+      "learning_rate": 4.556153604210484e-07,
+      "loss": 0.4823,
+      "step": 3835
+    },
+    {
+      "epoch": 2.633293331047488,
+      "grad_norm": 0.4509791198423544,
+      "learning_rate": 4.473293754455399e-07,
+      "loss": 0.477,
+      "step": 3840
+    },
+    {
+      "epoch": 2.636722098405623,
+      "grad_norm": 0.46642581073567424,
+      "learning_rate": 4.3911590002328476e-07,
+      "loss": 0.4787,
+      "step": 3845
+    },
+    {
+      "epoch": 2.640150865763758,
+      "grad_norm": 0.45567923757743767,
+      "learning_rate": 4.3097506496880325e-07,
+      "loss": 0.478,
+      "step": 3850
+    },
+    {
+      "epoch": 2.6435796331218926,
+      "grad_norm": 0.46511398506896673,
+      "learning_rate": 4.2290699993968897e-07,
+      "loss": 0.4796,
+      "step": 3855
+    },
+    {
+      "epoch": 2.6470084004800274,
+      "grad_norm": 0.4739348227964508,
+      "learning_rate": 4.149118334345403e-07,
+      "loss": 0.474,
+      "step": 3860
+    },
+    {
+      "epoch": 2.650437167838162,
+      "grad_norm": 0.46126850845642486,
+      "learning_rate": 4.0698969279091115e-07,
+      "loss": 0.4786,
+      "step": 3865
+    },
+    {
+      "epoch": 2.653865935196297,
+      "grad_norm": 0.4536599491788961,
+      "learning_rate": 3.9914070418329123e-07,
+      "loss": 0.477,
+      "step": 3870
+    },
+    {
+      "epoch": 2.657294702554432,
+      "grad_norm": 0.4620488468718019,
+      "learning_rate": 3.9136499262108716e-07,
+      "loss": 0.4808,
+      "step": 3875
+    },
+    {
+      "epoch": 2.6607234699125666,
+      "grad_norm": 0.4655535577302228,
+      "learning_rate": 3.836626819466338e-07,
+      "loss": 0.4809,
+      "step": 3880
+    },
+    {
+      "epoch": 2.6641522372707014,
+      "grad_norm": 0.5048993299270467,
+      "learning_rate": 3.760338948332293e-07,
+      "loss": 0.4731,
+      "step": 3885
+    },
+    {
+      "epoch": 2.667581004628836,
+      "grad_norm": 0.47154869975353264,
+      "learning_rate": 3.684787527831707e-07,
+      "loss": 0.477,
+      "step": 3890
+    },
+    {
+      "epoch": 2.6710097719869705,
+      "grad_norm": 0.4642713924482236,
+      "learning_rate": 3.6099737612582455e-07,
+      "loss": 0.475,
+      "step": 3895
+    },
+    {
+      "epoch": 2.6744385393451053,
+      "grad_norm": 0.4483908444332171,
+      "learning_rate": 3.53589884015712e-07,
+      "loss": 0.4753,
+      "step": 3900
+    },
+    {
+      "epoch": 2.67786730670324,
+      "grad_norm": 0.4652782765678345,
+      "learning_rate": 3.462563944306047e-07,
+      "loss": 0.4816,
+      "step": 3905
+    },
+    {
+      "epoch": 2.681296074061375,
+      "grad_norm": 0.45331932775420813,
+      "learning_rate": 3.3899702416965166e-07,
+      "loss": 0.4777,
+      "step": 3910
+    },
+    {
+      "epoch": 2.6847248414195097,
+      "grad_norm": 0.4412624964963365,
+      "learning_rate": 3.3181188885151706e-07,
+      "loss": 0.4763,
+      "step": 3915
+    },
+    {
+      "epoch": 2.6881536087776445,
+      "grad_norm": 0.45014302824757474,
+      "learning_rate": 3.247011029125391e-07,
+      "loss": 0.4786,
+      "step": 3920
+    },
+    {
+      "epoch": 2.6915823761357793,
+      "grad_norm": 0.4631051719702464,
+      "learning_rate": 3.1766477960490485e-07,
+      "loss": 0.4846,
+      "step": 3925
+    },
+    {
+      "epoch": 2.6950111434939137,
+      "grad_norm": 0.4581555189827555,
+      "learning_rate": 3.1070303099485055e-07,
+      "loss": 0.484,
+      "step": 3930
+    },
+    {
+      "epoch": 2.6984399108520485,
+      "grad_norm": 0.45350853465967644,
+      "learning_rate": 3.038159679608754e-07,
+      "loss": 0.4789,
+      "step": 3935
+    },
+    {
+      "epoch": 2.7018686782101833,
+      "grad_norm": 0.4546945454352565,
+      "learning_rate": 2.9700370019197287e-07,
+      "loss": 0.4763,
+      "step": 3940
+    },
+    {
+      "epoch": 2.705297445568318,
+      "grad_norm": 0.449234094604194,
+      "learning_rate": 2.9026633618588706e-07,
+      "loss": 0.4759,
+      "step": 3945
+    },
+    {
+      "epoch": 2.708726212926453,
+      "grad_norm": 0.45160809662086393,
+      "learning_rate": 2.8360398324738415e-07,
+      "loss": 0.4716,
+      "step": 3950
+    },
+    {
+      "epoch": 2.7121549802845877,
+      "grad_norm": 0.4482757694655283,
+      "learning_rate": 2.7701674748654206e-07,
+      "loss": 0.4774,
+      "step": 3955
+    },
+    {
+      "epoch": 2.7155837476427225,
+      "grad_norm": 0.4605988211239266,
+      "learning_rate": 2.7050473381706186e-07,
+      "loss": 0.4796,
+      "step": 3960
+    },
+    {
+      "epoch": 2.7190125150008573,
+      "grad_norm": 0.4542079439617735,
+      "learning_rate": 2.640680459545941e-07,
+      "loss": 0.4814,
+      "step": 3965
+    },
+    {
+      "epoch": 2.722441282358992,
+      "grad_norm": 0.47170501988492425,
+      "learning_rate": 2.577067864150906e-07,
+      "loss": 0.477,
+      "step": 3970
+    },
+    {
+      "epoch": 2.725870049717127,
+      "grad_norm": 0.45766219103133904,
+      "learning_rate": 2.514210565131708e-07,
+      "loss": 0.479,
+      "step": 3975
+    },
+    {
+      "epoch": 2.7292988170752617,
+      "grad_norm": 0.46091999743971984,
+      "learning_rate": 2.452109563605065e-07,
+      "loss": 0.4795,
+      "step": 3980
+    },
+    {
+      "epoch": 2.732727584433396,
+      "grad_norm": 0.46053964874575465,
+      "learning_rate": 2.3907658486422734e-07,
+      "loss": 0.4816,
+      "step": 3985
+    },
+    {
+      "epoch": 2.736156351791531,
+      "grad_norm": 0.44585333829135504,
+      "learning_rate": 2.330180397253473e-07,
+      "loss": 0.4712,
+      "step": 3990
+    },
+    {
+      "epoch": 2.7395851191496656,
+      "grad_norm": 0.45616604492168555,
+      "learning_rate": 2.2703541743720913e-07,
+      "loss": 0.4765,
+      "step": 3995
+    },
+    {
+      "epoch": 2.7430138865078004,
+      "grad_norm": 0.458862963479115,
+      "learning_rate": 2.2112881328394287e-07,
+      "loss": 0.4734,
+      "step": 4000
+    },
+    {
+      "epoch": 2.7464426538659352,
+      "grad_norm": 0.4564004005454542,
+      "learning_rate": 2.152983213389559e-07,
+      "loss": 0.4817,
+      "step": 4005
+    },
+    {
+      "epoch": 2.74987142122407,
+      "grad_norm": 0.4527978346632276,
+      "learning_rate": 2.0954403446342753e-07,
+      "loss": 0.4778,
+      "step": 4010
+    },
+    {
+      "epoch": 2.753300188582205,
+      "grad_norm": 0.44674470885527423,
+      "learning_rate": 2.0386604430483237e-07,
+      "loss": 0.4753,
+      "step": 4015
+    },
+    {
+      "epoch": 2.756728955940339,
+      "grad_norm": 0.4573693361590737,
+      "learning_rate": 1.9826444129548317e-07,
+      "loss": 0.4829,
+      "step": 4020
+    },
+    {
+      "epoch": 2.760157723298474,
+      "grad_norm": 0.46202280759285275,
+      "learning_rate": 1.927393146510881e-07,
+      "loss": 0.4794,
+      "step": 4025
+    },
+    {
+      "epoch": 2.7635864906566088,
+      "grad_norm": 0.44872435639531505,
+      "learning_rate": 1.8729075236932903e-07,
+      "loss": 0.4712,
+      "step": 4030
+    },
+    {
+      "epoch": 2.7670152580147436,
+      "grad_norm": 0.4562789796076479,
+      "learning_rate": 1.8191884122846226e-07,
+      "loss": 0.4775,
+      "step": 4035
+    },
+    {
+      "epoch": 2.7704440253728784,
+      "grad_norm": 0.4585752798754004,
+      "learning_rate": 1.7662366678593502e-07,
+      "loss": 0.4803,
+      "step": 4040
+    },
+    {
+      "epoch": 2.773872792731013,
+      "grad_norm": 0.44460297881292643,
+      "learning_rate": 1.7140531337702383e-07,
+      "loss": 0.4747,
+      "step": 4045
+    },
+    {
+      "epoch": 2.777301560089148,
+      "grad_norm": 0.4425263511378165,
+      "learning_rate": 1.6626386411348783e-07,
+      "loss": 0.4715,
+      "step": 4050
+    },
+    {
+      "epoch": 2.7807303274472828,
+      "grad_norm": 0.448778429615437,
+      "learning_rate": 1.6119940088225317e-07,
+      "loss": 0.4821,
+      "step": 4055
+    },
+    {
+      "epoch": 2.7841590948054176,
+      "grad_norm": 0.4454280991695363,
+      "learning_rate": 1.56212004344099e-07,
+      "loss": 0.4725,
+      "step": 4060
+    },
+    {
+      "epoch": 2.7875878621635524,
+      "grad_norm": 0.45389256280014806,
+      "learning_rate": 1.5130175393237744e-07,
+      "loss": 0.482,
+      "step": 4065
+    },
+    {
+      "epoch": 2.791016629521687,
+      "grad_norm": 0.44797422922275487,
+      "learning_rate": 1.4646872785175182e-07,
+      "loss": 0.4736,
+      "step": 4070
+    },
+    {
+      "epoch": 2.794445396879822,
+      "grad_norm": 0.4500167947023535,
+      "learning_rate": 1.417130030769448e-07,
+      "loss": 0.4748,
+      "step": 4075
+    },
+    {
+      "epoch": 2.7978741642379563,
+      "grad_norm": 0.4384121757610316,
+      "learning_rate": 1.3703465535151505e-07,
+      "loss": 0.4787,
+      "step": 4080
+    },
+    {
+      "epoch": 2.801302931596091,
+      "grad_norm": 0.4542996969309403,
+      "learning_rate": 1.3243375918665314e-07,
+      "loss": 0.4757,
+      "step": 4085
+    },
+    {
+      "epoch": 2.804731698954226,
+      "grad_norm": 0.44313787892697587,
+      "learning_rate": 1.2791038785999243e-07,
+      "loss": 0.4723,
+      "step": 4090
+    },
+    {
+      "epoch": 2.8081604663123607,
+      "grad_norm": 0.45221326925330546,
+      "learning_rate": 1.2346461341443962e-07,
+      "loss": 0.4739,
+      "step": 4095
+    },
+    {
+      "epoch": 2.8115892336704955,
+      "grad_norm": 0.4493204806606999,
+      "learning_rate": 1.1909650665703265e-07,
+      "loss": 0.476,
+      "step": 4100
+    },
+    {
+      "epoch": 2.8150180010286303,
+      "grad_norm": 0.4511251142367096,
+      "learning_rate": 1.1480613715781074e-07,
+      "loss": 0.4712,
+      "step": 4105
+    },
+    {
+      "epoch": 2.818446768386765,
+      "grad_norm": 0.44576223137927595,
+      "learning_rate": 1.1059357324870456e-07,
+      "loss": 0.4804,
+      "step": 4110
+    },
+    {
+      "epoch": 2.8218755357448995,
+      "grad_norm": 0.45162764316225673,
+      "learning_rate": 1.064588820224488e-07,
+      "loss": 0.4756,
+      "step": 4115
+    },
+    {
+      "epoch": 2.8253043031030343,
+      "grad_norm": 0.45919178354925205,
+      "learning_rate": 1.024021293315175e-07,
+      "loss": 0.4789,
+      "step": 4120
+    },
+    {
+      "epoch": 2.828733070461169,
+      "grad_norm": 0.45178753105354696,
+      "learning_rate": 9.842337978706707e-08,
+      "loss": 0.4712,
+      "step": 4125
+    },
+    {
+      "epoch": 2.832161837819304,
+      "grad_norm": 0.44369641899607265,
+      "learning_rate": 9.452269675791603e-08,
+      "loss": 0.4693,
+      "step": 4130
+    },
+    {
+      "epoch": 2.8355906051774387,
+      "grad_norm": 0.4397395162890691,
+      "learning_rate": 9.070014236952973e-08,
+      "loss": 0.4725,
+      "step": 4135
+    },
+    {
+      "epoch": 2.8390193725355735,
+      "grad_norm": 0.4515355683540246,
+      "learning_rate": 8.69557775030344e-08,
+      "loss": 0.4773,
+      "step": 4140
+    },
+    {
+      "epoch": 2.8424481398937083,
+      "grad_norm": 0.4403255731977004,
+      "learning_rate": 8.328966179424413e-08,
+      "loss": 0.4733,
+      "step": 4145
+    },
+    {
+      "epoch": 2.845876907251843,
+      "grad_norm": 0.4543200394332973,
+      "learning_rate": 7.970185363271432e-08,
+      "loss": 0.481,
+      "step": 4150
+    },
+    {
+      "epoch": 2.849305674609978,
+      "grad_norm": 0.4564283136704236,
+      "learning_rate": 7.619241016081081e-08,
+      "loss": 0.4761,
+      "step": 4155
+    },
+    {
+      "epoch": 2.8527344419681127,
+      "grad_norm": 0.44588624487284356,
+      "learning_rate": 7.276138727279669e-08,
+      "loss": 0.4797,
+      "step": 4160
+    },
+    {
+      "epoch": 2.8561632093262475,
+      "grad_norm": 0.45198814901899365,
+      "learning_rate": 6.940883961394917e-08,
+      "loss": 0.4798,
+      "step": 4165
+    },
+    {
+      "epoch": 2.859591976684382,
+      "grad_norm": 0.4501783006120383,
+      "learning_rate": 6.613482057968023e-08,
+      "loss": 0.4732,
+      "step": 4170
+    },
+    {
+      "epoch": 2.8630207440425166,
+      "grad_norm": 0.4373982769658886,
+      "learning_rate": 6.293938231469398e-08,
+      "loss": 0.4736,
+      "step": 4175
+    },
+    {
+      "epoch": 2.8664495114006514,
+      "grad_norm": 0.4468552193790084,
+      "learning_rate": 5.982257571215178e-08,
+      "loss": 0.4775,
+      "step": 4180
+    },
+    {
+      "epoch": 2.869878278758786,
+      "grad_norm": 0.44830321173439996,
+      "learning_rate": 5.6784450412862293e-08,
+      "loss": 0.4721,
+      "step": 4185
+    },
+    {
+      "epoch": 2.873307046116921,
+      "grad_norm": 0.44597562076766367,
+      "learning_rate": 5.382505480449274e-08,
+      "loss": 0.4757,
+      "step": 4190
+    },
+    {
+      "epoch": 2.876735813475056,
+      "grad_norm": 0.46198888054755305,
+      "learning_rate": 5.094443602079613e-08,
+      "loss": 0.4769,
+      "step": 4195
+    },
+    {
+      "epoch": 2.8801645808331906,
+      "grad_norm": 0.4481590814268269,
+      "learning_rate": 4.814263994086077e-08,
+      "loss": 0.4745,
+      "step": 4200
+    },
+    {
+      "epoch": 2.883593348191325,
+      "grad_norm": 0.44268948043488404,
+      "learning_rate": 4.541971118838196e-08,
+      "loss": 0.4802,
+      "step": 4205
+    },
+    {
+      "epoch": 2.8870221155494598,
+      "grad_norm": 0.4450782983770214,
+      "learning_rate": 4.2775693130948094e-08,
+      "loss": 0.4786,
+      "step": 4210
+    },
+    {
+      "epoch": 2.8904508829075946,
+      "grad_norm": 0.4605099335234408,
+      "learning_rate": 4.021062787935181e-08,
+      "loss": 0.4784,
+      "step": 4215
+    },
+    {
+      "epoch": 2.8938796502657294,
+      "grad_norm": 0.4481715420899884,
+      "learning_rate": 3.772455628691829e-08,
+      "loss": 0.4704,
+      "step": 4220
+    },
+    {
+      "epoch": 2.897308417623864,
+      "grad_norm": 0.4646866287413105,
+      "learning_rate": 3.531751794885574e-08,
+      "loss": 0.4799,
+      "step": 4225
+    },
+    {
+      "epoch": 2.900737184981999,
+      "grad_norm": 0.4565890986941479,
+      "learning_rate": 3.2989551201624836e-08,
+      "loss": 0.4825,
+      "step": 4230
+    },
+    {
+      "epoch": 2.9041659523401337,
+      "grad_norm": 0.44645402505000176,
+      "learning_rate": 3.0740693122325304e-08,
+      "loss": 0.4735,
+      "step": 4235
+    },
+    {
+      "epoch": 2.9075947196982685,
+      "grad_norm": 0.4507869739723372,
+      "learning_rate": 2.857097952810972e-08,
+      "loss": 0.4862,
+      "step": 4240
+    },
+    {
+      "epoch": 2.9110234870564033,
+      "grad_norm": 0.4501912072102517,
+      "learning_rate": 2.6480444975610086e-08,
+      "loss": 0.4789,
+      "step": 4245
+    },
+    {
+      "epoch": 2.914452254414538,
+      "grad_norm": 0.447987603121036,
+      "learning_rate": 2.4469122760388264e-08,
+      "loss": 0.4778,
+      "step": 4250
+    },
+    {
+      "epoch": 2.917881021772673,
+      "grad_norm": 0.44971834009936357,
+      "learning_rate": 2.2537044916405292e-08,
+      "loss": 0.4772,
+      "step": 4255
+    },
+    {
+      "epoch": 2.9213097891308077,
+      "grad_norm": 0.4495077806550815,
+      "learning_rate": 2.0684242215511797e-08,
+      "loss": 0.4749,
+      "step": 4260
+    },
+    {
+      "epoch": 2.924738556488942,
+      "grad_norm": 0.44694368287876113,
+      "learning_rate": 1.8910744166958926e-08,
+      "loss": 0.4801,
+      "step": 4265
+    },
+    {
+      "epoch": 2.928167323847077,
+      "grad_norm": 0.4497877428185916,
+      "learning_rate": 1.7216579016925415e-08,
+      "loss": 0.484,
+      "step": 4270
+    },
+    {
+      "epoch": 2.9315960912052117,
+      "grad_norm": 0.44920658823268794,
+      "learning_rate": 1.5601773748070147e-08,
+      "loss": 0.4814,
+      "step": 4275
+    },
+    {
+      "epoch": 2.9350248585633465,
+      "grad_norm": 0.4438689776804719,
+      "learning_rate": 1.4066354079101396e-08,
+      "loss": 0.4792,
+      "step": 4280
+    },
+    {
+      "epoch": 2.9384536259214813,
+      "grad_norm": 0.446097901399973,
+      "learning_rate": 1.2610344464367707e-08,
+      "loss": 0.478,
+      "step": 4285
+    },
+    {
+      "epoch": 2.941882393279616,
+      "grad_norm": 0.447115729483812,
+      "learning_rate": 1.1233768093468766e-08,
+      "loss": 0.4757,
+      "step": 4290
+    },
+    {
+      "epoch": 2.945311160637751,
+      "grad_norm": 0.4481899023438092,
+      "learning_rate": 9.936646890884582e-09,
+      "loss": 0.4739,
+      "step": 4295
+    },
+    {
+      "epoch": 2.9487399279958852,
+      "grad_norm": 0.4506621904879036,
+      "learning_rate": 8.719001515627434e-09,
+      "loss": 0.479,
+      "step": 4300
+    },
+    {
+      "epoch": 2.95216869535402,
+      "grad_norm": 0.4533955216869492,
+      "learning_rate": 7.58085136091269e-09,
+      "loss": 0.4817,
+      "step": 4305
+    },
+    {
+      "epoch": 2.955597462712155,
+      "grad_norm": 0.43437042453552366,
+      "learning_rate": 6.5222145538501595e-09,
+      "loss": 0.4821,
+      "step": 4310
+    },
+    {
+      "epoch": 2.9590262300702896,
+      "grad_norm": 0.45340023997429035,
+      "learning_rate": 5.543107955154892e-09,
+      "loss": 0.4722,
+      "step": 4315
+    },
+    {
+      "epoch": 2.9624549974284244,
+      "grad_norm": 0.4547732433450872,
+      "learning_rate": 4.643547158878492e-09,
+      "loss": 0.4783,
+      "step": 4320
+    },
+    {
+      "epoch": 2.9658837647865592,
+      "grad_norm": 0.4459015086328452,
+      "learning_rate": 3.8235464921621e-09,
+      "loss": 0.4759,
+      "step": 4325
+    },
+    {
+      "epoch": 2.969312532144694,
+      "grad_norm": 0.43573098371648133,
+      "learning_rate": 3.0831190150054646e-09,
+      "loss": 0.4793,
+      "step": 4330
+    },
+    {
+      "epoch": 2.972741299502829,
+      "grad_norm": 0.4558549928949452,
+      "learning_rate": 2.422276520062661e-09,
+      "loss": 0.4754,
+      "step": 4335
+    },
+    {
+      "epoch": 2.9761700668609636,
+      "grad_norm": 0.45059494788354154,
+      "learning_rate": 1.8410295324505778e-09,
+      "loss": 0.4808,
+      "step": 4340
+    },
+    {
+      "epoch": 2.9795988342190984,
+      "grad_norm": 0.4527487921892312,
+      "learning_rate": 1.3393873095846055e-09,
+      "loss": 0.4771,
+      "step": 4345
+    },
+    {
+      "epoch": 2.9830276015772332,
+      "grad_norm": 0.44727140352439004,
+      "learning_rate": 9.173578410281992e-10,
+      "loss": 0.4793,
+      "step": 4350
+    },
+    {
+      "epoch": 2.9864563689353676,
+      "grad_norm": 0.44819706968866224,
+      "learning_rate": 5.749478483679794e-10,
+      "loss": 0.4785,
+      "step": 4355
+    },
+    {
+      "epoch": 2.9898851362935024,
+      "grad_norm": 0.4540122997268082,
+      "learning_rate": 3.1216278510493027e-10,
+      "loss": 0.4768,
+      "step": 4360
+    },
+    {
+      "epoch": 2.993313903651637,
+      "grad_norm": 0.44168345357587324,
+      "learning_rate": 1.2900683656891232e-10,
+      "loss": 0.4688,
+      "step": 4365
+    },
+    {
+      "epoch": 2.996742671009772,
+      "grad_norm": 0.45853813294171786,
+      "learning_rate": 2.548291985149387e-11,
+      "loss": 0.4775,
+      "step": 4370
+    },
+    {
+      "epoch": 2.9994856848962796,
+      "eval_loss": 0.5953422784805298,
+      "eval_runtime": 204.761,
+      "eval_samples_per_second": 47.993,
+      "eval_steps_per_second": 1.504,
+      "step": 4374
+    },
+    {
+      "epoch": 2.9994856848962796,
+      "step": 4374,
+      "total_flos": 3663097076121600.0,
+      "train_loss": 0.5645236694000604,
+      "train_runtime": 46689.0597,
+      "train_samples_per_second": 11.993,
+      "train_steps_per_second": 0.094
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 4374,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3663097076121600.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}