diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12047 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 8576,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005830223880597014,
+      "grad_norm": 2.550195533598444,
+      "learning_rate": 5.827505827505828e-07,
+      "loss": 0.8852,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011660447761194029,
+      "grad_norm": 2.2188855019798273,
+      "learning_rate": 1.1655011655011657e-06,
+      "loss": 0.8311,
+      "step": 10
+    },
+    {
+      "epoch": 0.0017490671641791045,
+      "grad_norm": 1.7967346771825998,
+      "learning_rate": 1.7482517482517485e-06,
+      "loss": 0.8113,
+      "step": 15
+    },
+    {
+      "epoch": 0.0023320895522388058,
+      "grad_norm": 1.618619190869765,
+      "learning_rate": 2.3310023310023313e-06,
+      "loss": 0.8333,
+      "step": 20
+    },
+    {
+      "epoch": 0.0029151119402985076,
+      "grad_norm": 1.137673984585037,
+      "learning_rate": 2.9137529137529138e-06,
+      "loss": 0.7745,
+      "step": 25
+    },
+    {
+      "epoch": 0.003498134328358209,
+      "grad_norm": 1.1933766134733839,
+      "learning_rate": 3.496503496503497e-06,
+      "loss": 0.7904,
+      "step": 30
+    },
+    {
+      "epoch": 0.00408115671641791,
+      "grad_norm": 1.0157304336095063,
+      "learning_rate": 4.079254079254079e-06,
+      "loss": 0.7624,
+      "step": 35
+    },
+    {
+      "epoch": 0.0046641791044776115,
+      "grad_norm": 0.9429820002012593,
+      "learning_rate": 4.662004662004663e-06,
+      "loss": 0.7285,
+      "step": 40
+    },
+    {
+      "epoch": 0.005247201492537314,
+      "grad_norm": 0.915430165066572,
+      "learning_rate": 5.244755244755245e-06,
+      "loss": 0.7469,
+      "step": 45
+    },
+    {
+      "epoch": 0.005830223880597015,
+      "grad_norm": 0.8161946795665328,
+      "learning_rate": 5.8275058275058275e-06,
+      "loss": 0.707,
+      "step": 50
+    },
+    {
+      "epoch": 0.006413246268656717,
+      "grad_norm": 0.8339668281290502,
+      "learning_rate": 6.41025641025641e-06,
+      "loss": 0.6905,
+      "step": 55
+    },
+    {
+      "epoch": 0.006996268656716418,
+      "grad_norm": 0.8574171717816127,
+      "learning_rate": 6.993006993006994e-06,
+      "loss": 0.7698,
+      "step": 60
+    },
+    {
+      "epoch": 0.00757929104477612,
+      "grad_norm": 0.8619956369841163,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.7082,
+      "step": 65
+    },
+    {
+      "epoch": 0.00816231343283582,
+      "grad_norm": 0.8045557883502801,
+      "learning_rate": 8.158508158508159e-06,
+      "loss": 0.6587,
+      "step": 70
+    },
+    {
+      "epoch": 0.008745335820895522,
+      "grad_norm": 0.7946638318702278,
+      "learning_rate": 8.741258741258741e-06,
+      "loss": 0.6539,
+      "step": 75
+    },
+    {
+      "epoch": 0.009328358208955223,
+      "grad_norm": 0.8503979707980048,
+      "learning_rate": 9.324009324009325e-06,
+      "loss": 0.6879,
+      "step": 80
+    },
+    {
+      "epoch": 0.009911380597014926,
+      "grad_norm": 0.804711208532189,
+      "learning_rate": 9.906759906759908e-06,
+      "loss": 0.6743,
+      "step": 85
+    },
+    {
+      "epoch": 0.010494402985074628,
+      "grad_norm": 0.7951085316698083,
+      "learning_rate": 1.048951048951049e-05,
+      "loss": 0.6586,
+      "step": 90
+    },
+    {
+      "epoch": 0.011077425373134329,
+      "grad_norm": 0.793128900429868,
+      "learning_rate": 1.1072261072261073e-05,
+      "loss": 0.6842,
+      "step": 95
+    },
+    {
+      "epoch": 0.01166044776119403,
+      "grad_norm": 0.8075245421009535,
+      "learning_rate": 1.1655011655011655e-05,
+      "loss": 0.6684,
+      "step": 100
+    },
+    {
+      "epoch": 0.012243470149253732,
+      "grad_norm": 0.8581518728101114,
+      "learning_rate": 1.2237762237762239e-05,
+      "loss": 0.6207,
+      "step": 105
+    },
+    {
+      "epoch": 0.012826492537313433,
+      "grad_norm": 0.9082387952747121,
+      "learning_rate": 1.282051282051282e-05,
+      "loss": 0.6678,
+      "step": 110
+    },
+    {
+      "epoch": 0.013409514925373135,
+      "grad_norm": 0.9029958898223162,
+      "learning_rate": 1.3403263403263406e-05,
+      "loss": 0.6394,
+      "step": 115
+    },
+    {
+      "epoch": 0.013992537313432836,
+      "grad_norm": 0.8377367698700525,
+      "learning_rate": 1.3986013986013988e-05,
+      "loss": 0.6707,
+      "step": 120
+    },
+    {
+      "epoch": 0.014575559701492538,
+      "grad_norm": 0.7587966655835441,
+      "learning_rate": 1.456876456876457e-05,
+      "loss": 0.6258,
+      "step": 125
+    },
+    {
+      "epoch": 0.01515858208955224,
+      "grad_norm": 0.8173306054331542,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.6619,
+      "step": 130
+    },
+    {
+      "epoch": 0.01574160447761194,
+      "grad_norm": 0.9376337343046882,
+      "learning_rate": 1.5734265734265734e-05,
+      "loss": 0.6673,
+      "step": 135
+    },
+    {
+      "epoch": 0.01632462686567164,
+      "grad_norm": 0.9363652268784762,
+      "learning_rate": 1.6317016317016318e-05,
+      "loss": 0.668,
+      "step": 140
+    },
+    {
+      "epoch": 0.016907649253731342,
+      "grad_norm": 0.8794426221093276,
+      "learning_rate": 1.68997668997669e-05,
+      "loss": 0.6569,
+      "step": 145
+    },
+    {
+      "epoch": 0.017490671641791043,
+      "grad_norm": 0.9407491307884387,
+      "learning_rate": 1.7482517482517483e-05,
+      "loss": 0.6778,
+      "step": 150
+    },
+    {
+      "epoch": 0.018073694029850745,
+      "grad_norm": 0.8720667910368793,
+      "learning_rate": 1.8065268065268067e-05,
+      "loss": 0.6291,
+      "step": 155
+    },
+    {
+      "epoch": 0.018656716417910446,
+      "grad_norm": 0.8603987252532928,
+      "learning_rate": 1.864801864801865e-05,
+      "loss": 0.6372,
+      "step": 160
+    },
+    {
+      "epoch": 0.019239738805970148,
+      "grad_norm": 0.8449689793110765,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 0.6586,
+      "step": 165
+    },
+    {
+      "epoch": 0.019822761194029852,
+      "grad_norm": 1.0325515822144362,
+      "learning_rate": 1.9813519813519816e-05,
+      "loss": 0.6386,
+      "step": 170
+    },
+    {
+      "epoch": 0.020405783582089554,
+      "grad_norm": 0.8513517345308381,
+      "learning_rate": 2.0396270396270396e-05,
+      "loss": 0.6311,
+      "step": 175
+    },
+    {
+      "epoch": 0.020988805970149255,
+      "grad_norm": 0.8847582203013,
+      "learning_rate": 2.097902097902098e-05,
+      "loss": 0.6834,
+      "step": 180
+    },
+    {
+      "epoch": 0.021571828358208957,
+      "grad_norm": 0.9006033563159529,
+      "learning_rate": 2.156177156177156e-05,
+      "loss": 0.6381,
+      "step": 185
+    },
+    {
+      "epoch": 0.022154850746268658,
+      "grad_norm": 0.8147114381410464,
+      "learning_rate": 2.2144522144522145e-05,
+      "loss": 0.6432,
+      "step": 190
+    },
+    {
+      "epoch": 0.02273787313432836,
+      "grad_norm": 0.9560611394473829,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.6261,
+      "step": 195
+    },
+    {
+      "epoch": 0.02332089552238806,
+      "grad_norm": 0.7951106600912393,
+      "learning_rate": 2.331002331002331e-05,
+      "loss": 0.6307,
+      "step": 200
+    },
+    {
+      "epoch": 0.023903917910447763,
+      "grad_norm": 0.855140755847949,
+      "learning_rate": 2.3892773892773894e-05,
+      "loss": 0.6213,
+      "step": 205
+    },
+    {
+      "epoch": 0.024486940298507464,
+      "grad_norm": 1.0420866065102492,
+      "learning_rate": 2.4475524475524478e-05,
+      "loss": 0.6275,
+      "step": 210
+    },
+    {
+      "epoch": 0.025069962686567165,
+      "grad_norm": 0.89019303742937,
+      "learning_rate": 2.505827505827506e-05,
+      "loss": 0.6275,
+      "step": 215
+    },
+    {
+      "epoch": 0.025652985074626867,
+      "grad_norm": 0.9758673014251983,
+      "learning_rate": 2.564102564102564e-05,
+      "loss": 0.6631,
+      "step": 220
+    },
+    {
+      "epoch": 0.02623600746268657,
+      "grad_norm": 0.8596860100346014,
+      "learning_rate": 2.6223776223776224e-05,
+      "loss": 0.6216,
+      "step": 225
+    },
+    {
+      "epoch": 0.02681902985074627,
+      "grad_norm": 0.8908792386022747,
+      "learning_rate": 2.680652680652681e-05,
+      "loss": 0.6406,
+      "step": 230
+    },
+    {
+      "epoch": 0.02740205223880597,
+      "grad_norm": 0.8721610635945679,
+      "learning_rate": 2.738927738927739e-05,
+      "loss": 0.6216,
+      "step": 235
+    },
+    {
+      "epoch": 0.027985074626865673,
+      "grad_norm": 0.8671325117829104,
+      "learning_rate": 2.7972027972027976e-05,
+      "loss": 0.6713,
+      "step": 240
+    },
+    {
+      "epoch": 0.028568097014925374,
+      "grad_norm": 0.8803714665797042,
+      "learning_rate": 2.8554778554778557e-05,
+      "loss": 0.629,
+      "step": 245
+    },
+    {
+      "epoch": 0.029151119402985076,
+      "grad_norm": 0.8308844754653052,
+      "learning_rate": 2.913752913752914e-05,
+      "loss": 0.6373,
+      "step": 250
+    },
+    {
+      "epoch": 0.029734141791044777,
+      "grad_norm": 0.829842731040146,
+      "learning_rate": 2.972027972027972e-05,
+      "loss": 0.6163,
+      "step": 255
+    },
+    {
+      "epoch": 0.03031716417910448,
+      "grad_norm": 0.9311493282737088,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.6448,
+      "step": 260
+    },
+    {
+      "epoch": 0.03090018656716418,
+      "grad_norm": 0.8380478242539974,
+      "learning_rate": 3.088578088578088e-05,
+      "loss": 0.6123,
+      "step": 265
+    },
+    {
+      "epoch": 0.03148320895522388,
+      "grad_norm": 0.8612483599648676,
+      "learning_rate": 3.146853146853147e-05,
+      "loss": 0.6206,
+      "step": 270
+    },
+    {
+      "epoch": 0.03206623134328358,
+      "grad_norm": 0.7504191034824335,
+      "learning_rate": 3.205128205128206e-05,
+      "loss": 0.5956,
+      "step": 275
+    },
+    {
+      "epoch": 0.03264925373134328,
+      "grad_norm": 0.8435455726913331,
+      "learning_rate": 3.2634032634032635e-05,
+      "loss": 0.6147,
+      "step": 280
+    },
+    {
+      "epoch": 0.033232276119402986,
+      "grad_norm": 0.8228330420668449,
+      "learning_rate": 3.321678321678322e-05,
+      "loss": 0.6285,
+      "step": 285
+    },
+    {
+      "epoch": 0.033815298507462684,
+      "grad_norm": 0.7466739529712338,
+      "learning_rate": 3.37995337995338e-05,
+      "loss": 0.62,
+      "step": 290
+    },
+    {
+      "epoch": 0.03439832089552239,
+      "grad_norm": 0.8056760014927022,
+      "learning_rate": 3.438228438228439e-05,
+      "loss": 0.6216,
+      "step": 295
+    },
+    {
+      "epoch": 0.034981343283582086,
+      "grad_norm": 0.8113797852368729,
+      "learning_rate": 3.4965034965034965e-05,
+      "loss": 0.6105,
+      "step": 300
+    },
+    {
+      "epoch": 0.03556436567164179,
+      "grad_norm": 0.8890502750300378,
+      "learning_rate": 3.554778554778555e-05,
+      "loss": 0.6139,
+      "step": 305
+    },
+    {
+      "epoch": 0.03614738805970149,
+      "grad_norm": 0.7687545593968855,
+      "learning_rate": 3.613053613053613e-05,
+      "loss": 0.614,
+      "step": 310
+    },
+    {
+      "epoch": 0.036730410447761194,
+      "grad_norm": 0.8336903362213683,
+      "learning_rate": 3.671328671328672e-05,
+      "loss": 0.6278,
+      "step": 315
+    },
+    {
+      "epoch": 0.03731343283582089,
+      "grad_norm": 0.869875611794232,
+      "learning_rate": 3.72960372960373e-05,
+      "loss": 0.6645,
+      "step": 320
+    },
+    {
+      "epoch": 0.0378964552238806,
+      "grad_norm": 0.789926244606821,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.6027,
+      "step": 325
+    },
+    {
+      "epoch": 0.038479477611940295,
+      "grad_norm": 0.8392459694319648,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 0.6567,
+      "step": 330
+    },
+    {
+      "epoch": 0.0390625,
+      "grad_norm": 0.7918852769546142,
+      "learning_rate": 3.904428904428905e-05,
+      "loss": 0.6453,
+      "step": 335
+    },
+    {
+      "epoch": 0.039645522388059705,
+      "grad_norm": 0.7819447011686889,
+      "learning_rate": 3.962703962703963e-05,
+      "loss": 0.6,
+      "step": 340
+    },
+    {
+      "epoch": 0.0402285447761194,
+      "grad_norm": 0.7599671478703276,
+      "learning_rate": 4.020979020979021e-05,
+      "loss": 0.6097,
+      "step": 345
+    },
+    {
+      "epoch": 0.04081156716417911,
+      "grad_norm": 0.8770061789317196,
+      "learning_rate": 4.079254079254079e-05,
+      "loss": 0.6199,
+      "step": 350
+    },
+    {
+      "epoch": 0.041394589552238806,
+      "grad_norm": 0.8228590505028837,
+      "learning_rate": 4.1375291375291377e-05,
+      "loss": 0.63,
+      "step": 355
+    },
+    {
+      "epoch": 0.04197761194029851,
+      "grad_norm": 0.9266679768403295,
+      "learning_rate": 4.195804195804196e-05,
+      "loss": 0.6611,
+      "step": 360
+    },
+    {
+      "epoch": 0.04256063432835821,
+      "grad_norm": 0.776925533918814,
+      "learning_rate": 4.254079254079254e-05,
+      "loss": 0.6383,
+      "step": 365
+    },
+    {
+      "epoch": 0.043143656716417914,
+      "grad_norm": 0.7652459941000134,
+      "learning_rate": 4.312354312354312e-05,
+      "loss": 0.5967,
+      "step": 370
+    },
+    {
+      "epoch": 0.04372667910447761,
+      "grad_norm": 0.8438476797737248,
+      "learning_rate": 4.370629370629371e-05,
+      "loss": 0.6146,
+      "step": 375
+    },
+    {
+      "epoch": 0.044309701492537316,
+      "grad_norm": 0.8158888620815322,
+      "learning_rate": 4.428904428904429e-05,
+      "loss": 0.6237,
+      "step": 380
+    },
+    {
+      "epoch": 0.044892723880597014,
+      "grad_norm": 0.9144427364330033,
+      "learning_rate": 4.4871794871794874e-05,
+      "loss": 0.6098,
+      "step": 385
+    },
+    {
+      "epoch": 0.04547574626865672,
+      "grad_norm": 0.7260277122998825,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.6105,
+      "step": 390
+    },
+    {
+      "epoch": 0.04605876865671642,
+      "grad_norm": 0.774195742981333,
+      "learning_rate": 4.603729603729604e-05,
+      "loss": 0.6182,
+      "step": 395
+    },
+    {
+      "epoch": 0.04664179104477612,
+      "grad_norm": 0.8931158013280471,
+      "learning_rate": 4.662004662004662e-05,
+      "loss": 0.6269,
+      "step": 400
+    },
+    {
+      "epoch": 0.04722481343283582,
+      "grad_norm": 0.7219414153894123,
+      "learning_rate": 4.7202797202797204e-05,
+      "loss": 0.6159,
+      "step": 405
+    },
+    {
+      "epoch": 0.047807835820895525,
+      "grad_norm": 0.7392180125184434,
+      "learning_rate": 4.778554778554779e-05,
+      "loss": 0.6597,
+      "step": 410
+    },
+    {
+      "epoch": 0.04839085820895522,
+      "grad_norm": 0.7724562759848209,
+      "learning_rate": 4.836829836829837e-05,
+      "loss": 0.6588,
+      "step": 415
+    },
+    {
+      "epoch": 0.04897388059701493,
+      "grad_norm": 0.7698999004589875,
+      "learning_rate": 4.8951048951048956e-05,
+      "loss": 0.6431,
+      "step": 420
+    },
+    {
+      "epoch": 0.049556902985074626,
+      "grad_norm": 0.756825275252759,
+      "learning_rate": 4.9533799533799534e-05,
+      "loss": 0.6475,
+      "step": 425
+    },
+    {
+      "epoch": 0.05013992537313433,
+      "grad_norm": 0.7182404605677212,
+      "learning_rate": 4.9999998327150664e-05,
+      "loss": 0.6519,
+      "step": 430
+    },
+    {
+      "epoch": 0.05072294776119403,
+      "grad_norm": 0.7352734921536612,
+      "learning_rate": 4.999993977744981e-05,
+      "loss": 0.6367,
+      "step": 435
+    },
+    {
+      "epoch": 0.051305970149253734,
+      "grad_norm": 0.8663133271054537,
+      "learning_rate": 4.9999797585530614e-05,
+      "loss": 0.6367,
+      "step": 440
+    },
+    {
+      "epoch": 0.05188899253731343,
+      "grad_norm": 0.8007074508867144,
+      "learning_rate": 4.9999571751921666e-05,
+      "loss": 0.649,
+      "step": 445
+    },
+    {
+      "epoch": 0.05247201492537314,
+      "grad_norm": 0.8808725057777399,
+      "learning_rate": 4.999926227746247e-05,
+      "loss": 0.6407,
+      "step": 450
+    },
+    {
+      "epoch": 0.053055037313432835,
+      "grad_norm": 0.8207754194931053,
+      "learning_rate": 4.999886916330351e-05,
+      "loss": 0.6664,
+      "step": 455
+    },
+    {
+      "epoch": 0.05363805970149254,
+      "grad_norm": 0.7200644017362087,
+      "learning_rate": 4.9998392410906135e-05,
+      "loss": 0.6264,
+      "step": 460
+    },
+    {
+      "epoch": 0.05422108208955224,
+      "grad_norm": 0.7655394577960307,
+      "learning_rate": 4.9997832022042676e-05,
+      "loss": 0.6854,
+      "step": 465
+    },
+    {
+      "epoch": 0.05480410447761194,
+      "grad_norm": 1.5845328151194846,
+      "learning_rate": 4.9997187998796316e-05,
+      "loss": 0.6092,
+      "step": 470
+    },
+    {
+      "epoch": 0.05538712686567164,
+      "grad_norm": 0.763948541759687,
+      "learning_rate": 4.9996460343561184e-05,
+      "loss": 0.6601,
+      "step": 475
+    },
+    {
+      "epoch": 0.055970149253731345,
+      "grad_norm": 0.7370601652088519,
+      "learning_rate": 4.99956490590423e-05,
+      "loss": 0.6062,
+      "step": 480
+    },
+    {
+      "epoch": 0.05655317164179104,
+      "grad_norm": 0.6503724111090272,
+      "learning_rate": 4.9994754148255566e-05,
+      "loss": 0.597,
+      "step": 485
+    },
+    {
+      "epoch": 0.05713619402985075,
+      "grad_norm": 0.642633522142368,
+      "learning_rate": 4.999377561452776e-05,
+      "loss": 0.633,
+      "step": 490
+    },
+    {
+      "epoch": 0.057719216417910446,
+      "grad_norm": 0.6866312480811294,
+      "learning_rate": 4.999271346149652e-05,
+      "loss": 0.6421,
+      "step": 495
+    },
+    {
+      "epoch": 0.05830223880597015,
+      "grad_norm": 0.6472205796291965,
+      "learning_rate": 4.999156769311035e-05,
+      "loss": 0.615,
+      "step": 500
+    },
+    {
+      "epoch": 0.05888526119402985,
+      "grad_norm": 0.6977115669022897,
+      "learning_rate": 4.999033831362857e-05,
+      "loss": 0.6214,
+      "step": 505
+    },
+    {
+      "epoch": 0.059468283582089554,
+      "grad_norm": 0.6986043336376292,
+      "learning_rate": 4.998902532762132e-05,
+      "loss": 0.6193,
+      "step": 510
+    },
+    {
+      "epoch": 0.06005130597014925,
+      "grad_norm": 0.7475999210726344,
+      "learning_rate": 4.9987628739969554e-05,
+      "loss": 0.6224,
+      "step": 515
+    },
+    {
+      "epoch": 0.06063432835820896,
+      "grad_norm": 0.6813745517845872,
+      "learning_rate": 4.9986148555865016e-05,
+      "loss": 0.6177,
+      "step": 520
+    },
+    {
+      "epoch": 0.061217350746268655,
+      "grad_norm": 0.7015168996041015,
+      "learning_rate": 4.9984584780810196e-05,
+      "loss": 0.6768,
+      "step": 525
+    },
+    {
+      "epoch": 0.06180037313432836,
+      "grad_norm": 0.7037997780013406,
+      "learning_rate": 4.998293742061833e-05,
+      "loss": 0.6742,
+      "step": 530
+    },
+    {
+      "epoch": 0.06238339552238806,
+      "grad_norm": 0.7156012560309609,
+      "learning_rate": 4.998120648141338e-05,
+      "loss": 0.6304,
+      "step": 535
+    },
+    {
+      "epoch": 0.06296641791044776,
+      "grad_norm": 0.711127994617591,
+      "learning_rate": 4.997939196963004e-05,
+      "loss": 0.6765,
+      "step": 540
+    },
+    {
+      "epoch": 0.06354944029850747,
+      "grad_norm": 0.7367809530722819,
+      "learning_rate": 4.997749389201363e-05,
+      "loss": 0.6443,
+      "step": 545
+    },
+    {
+      "epoch": 0.06413246268656717,
+      "grad_norm": 0.7023254028972605,
+      "learning_rate": 4.997551225562014e-05,
+      "loss": 0.614,
+      "step": 550
+    },
+    {
+      "epoch": 0.06471548507462686,
+      "grad_norm": 0.7934049324407979,
+      "learning_rate": 4.99734470678162e-05,
+      "loss": 0.6804,
+      "step": 555
+    },
+    {
+      "epoch": 0.06529850746268656,
+      "grad_norm": 0.7211790595438915,
+      "learning_rate": 4.997129833627902e-05,
+      "loss": 0.6022,
+      "step": 560
+    },
+    {
+      "epoch": 0.06588152985074627,
+      "grad_norm": 0.7803510446535082,
+      "learning_rate": 4.996906606899639e-05,
+      "loss": 0.6324,
+      "step": 565
+    },
+    {
+      "epoch": 0.06646455223880597,
+      "grad_norm": 0.7268141035030203,
+      "learning_rate": 4.996675027426662e-05,
+      "loss": 0.6244,
+      "step": 570
+    },
+    {
+      "epoch": 0.06704757462686567,
+      "grad_norm": 1.937711458766374,
+      "learning_rate": 4.9964350960698564e-05,
+      "loss": 0.637,
+      "step": 575
+    },
+    {
+      "epoch": 0.06763059701492537,
+      "grad_norm": 0.6807880062940403,
+      "learning_rate": 4.996186813721152e-05,
+      "loss": 0.6009,
+      "step": 580
+    },
+    {
+      "epoch": 0.06821361940298508,
+      "grad_norm": 0.7031632269645188,
+      "learning_rate": 4.995930181303522e-05,
+      "loss": 0.6312,
+      "step": 585
+    },
+    {
+      "epoch": 0.06879664179104478,
+      "grad_norm": 0.6771751142969727,
+      "learning_rate": 4.995665199770986e-05,
+      "loss": 0.6604,
+      "step": 590
+    },
+    {
+      "epoch": 0.06937966417910447,
+      "grad_norm": 0.711794153251972,
+      "learning_rate": 4.995391870108595e-05,
+      "loss": 0.6527,
+      "step": 595
+    },
+    {
+      "epoch": 0.06996268656716417,
+      "grad_norm": 0.7336499566418994,
+      "learning_rate": 4.9951101933324374e-05,
+      "loss": 0.6056,
+      "step": 600
+    },
+    {
+      "epoch": 0.07054570895522388,
+      "grad_norm": 0.9123524114339076,
+      "learning_rate": 4.994820170489629e-05,
+      "loss": 0.6351,
+      "step": 605
+    },
+    {
+      "epoch": 0.07112873134328358,
+      "grad_norm": 0.6937351959939653,
+      "learning_rate": 4.9945218026583147e-05,
+      "loss": 0.6415,
+      "step": 610
+    },
+    {
+      "epoch": 0.07171175373134328,
+      "grad_norm": 0.6701402337273787,
+      "learning_rate": 4.9942150909476576e-05,
+      "loss": 0.616,
+      "step": 615
+    },
+    {
+      "epoch": 0.07229477611940298,
+      "grad_norm": 0.6737823946741938,
+      "learning_rate": 4.9939000364978424e-05,
+      "loss": 0.5916,
+      "step": 620
+    },
+    {
+      "epoch": 0.07287779850746269,
+      "grad_norm": 0.6443035255679863,
+      "learning_rate": 4.993576640480064e-05,
+      "loss": 0.593,
+      "step": 625
+    },
+    {
+      "epoch": 0.07346082089552239,
+      "grad_norm": 0.7194715850493604,
+      "learning_rate": 4.9932449040965296e-05,
+      "loss": 0.6537,
+      "step": 630
+    },
+    {
+      "epoch": 0.07404384328358209,
+      "grad_norm": 0.6638929998798874,
+      "learning_rate": 4.992904828580449e-05,
+      "loss": 0.651,
+      "step": 635
+    },
+    {
+      "epoch": 0.07462686567164178,
+      "grad_norm": 0.6403792391822621,
+      "learning_rate": 4.992556415196035e-05,
+      "loss": 0.5995,
+      "step": 640
+    },
+    {
+      "epoch": 0.0752098880597015,
+      "grad_norm": 0.7646937059991467,
+      "learning_rate": 4.9921996652384915e-05,
+      "loss": 0.6254,
+      "step": 645
+    },
+    {
+      "epoch": 0.0757929104477612,
+      "grad_norm": 0.708046339189841,
+      "learning_rate": 4.991834580034017e-05,
+      "loss": 0.6285,
+      "step": 650
+    },
+    {
+      "epoch": 0.07637593283582089,
+      "grad_norm": 0.6787038237176444,
+      "learning_rate": 4.991461160939795e-05,
+      "loss": 0.6089,
+      "step": 655
+    },
+    {
+      "epoch": 0.07695895522388059,
+      "grad_norm": 0.637822387437937,
+      "learning_rate": 4.991079409343989e-05,
+      "loss": 0.6339,
+      "step": 660
+    },
+    {
+      "epoch": 0.0775419776119403,
+      "grad_norm": 0.7266823799603278,
+      "learning_rate": 4.990689326665738e-05,
+      "loss": 0.601,
+      "step": 665
+    },
+    {
+      "epoch": 0.078125,
+      "grad_norm": 0.6501949889104538,
+      "learning_rate": 4.9902909143551516e-05,
+      "loss": 0.6016,
+      "step": 670
+    },
+    {
+      "epoch": 0.0787080223880597,
+      "grad_norm": 0.6820009767011616,
+      "learning_rate": 4.989884173893305e-05,
+      "loss": 0.6007,
+      "step": 675
+    },
+    {
+      "epoch": 0.07929104477611941,
+      "grad_norm": 0.6877686038902945,
+      "learning_rate": 4.989469106792231e-05,
+      "loss": 0.6136,
+      "step": 680
+    },
+    {
+      "epoch": 0.07987406716417911,
+      "grad_norm": 0.6085151630088098,
+      "learning_rate": 4.9890457145949186e-05,
+      "loss": 0.5881,
+      "step": 685
+    },
+    {
+      "epoch": 0.0804570895522388,
+      "grad_norm": 0.6214607197858061,
+      "learning_rate": 4.988613998875304e-05,
+      "loss": 0.6053,
+      "step": 690
+    },
+    {
+      "epoch": 0.0810401119402985,
+      "grad_norm": 0.6744013331308584,
+      "learning_rate": 4.988173961238264e-05,
+      "loss": 0.5802,
+      "step": 695
+    },
+    {
+      "epoch": 0.08162313432835822,
+      "grad_norm": 0.6808032691033227,
+      "learning_rate": 4.987725603319615e-05,
+      "loss": 0.6094,
+      "step": 700
+    },
+    {
+      "epoch": 0.08220615671641791,
+      "grad_norm": 0.6979985310436119,
+      "learning_rate": 4.987268926786098e-05,
+      "loss": 0.6323,
+      "step": 705
+    },
+    {
+      "epoch": 0.08278917910447761,
+      "grad_norm": 0.6226469149492844,
+      "learning_rate": 4.986803933335385e-05,
+      "loss": 0.6459,
+      "step": 710
+    },
+    {
+      "epoch": 0.08337220149253731,
+      "grad_norm": 0.676363656496179,
+      "learning_rate": 4.9863306246960605e-05,
+      "loss": 0.5761,
+      "step": 715
+    },
+    {
+      "epoch": 0.08395522388059702,
+      "grad_norm": 0.6685016854342901,
+      "learning_rate": 4.9858490026276226e-05,
+      "loss": 0.5988,
+      "step": 720
+    },
+    {
+      "epoch": 0.08453824626865672,
+      "grad_norm": 0.6396794014235626,
+      "learning_rate": 4.9853590689204715e-05,
+      "loss": 0.5854,
+      "step": 725
+    },
+    {
+      "epoch": 0.08512126865671642,
+      "grad_norm": 0.720334288248887,
+      "learning_rate": 4.9848608253959096e-05,
+      "loss": 0.6207,
+      "step": 730
+    },
+    {
+      "epoch": 0.08570429104477612,
+      "grad_norm": 0.6079861012766015,
+      "learning_rate": 4.984354273906127e-05,
+      "loss": 0.5953,
+      "step": 735
+    },
+    {
+      "epoch": 0.08628731343283583,
+      "grad_norm": 0.6782045394701461,
+      "learning_rate": 4.9838394163341993e-05,
+      "loss": 0.63,
+      "step": 740
+    },
+    {
+      "epoch": 0.08687033582089553,
+      "grad_norm": 0.6465534111571817,
+      "learning_rate": 4.983316254594081e-05,
+      "loss": 0.6006,
+      "step": 745
+    },
+    {
+      "epoch": 0.08745335820895522,
+      "grad_norm": 0.6048787945996146,
+      "learning_rate": 4.9827847906305934e-05,
+      "loss": 0.5937,
+      "step": 750
+    },
+    {
+      "epoch": 0.08803638059701492,
+      "grad_norm": 0.6704335077914664,
+      "learning_rate": 4.982245026419424e-05,
+      "loss": 0.6433,
+      "step": 755
+    },
+    {
+      "epoch": 0.08861940298507463,
+      "grad_norm": 0.6420162328599116,
+      "learning_rate": 4.981696963967116e-05,
+      "loss": 0.5806,
+      "step": 760
+    },
+    {
+      "epoch": 0.08920242537313433,
+      "grad_norm": 0.589278560052739,
+      "learning_rate": 4.981140605311057e-05,
+      "loss": 0.6049,
+      "step": 765
+    },
+    {
+      "epoch": 0.08978544776119403,
+      "grad_norm": 0.6681567808799939,
+      "learning_rate": 4.98057595251948e-05,
+      "loss": 0.6369,
+      "step": 770
+    },
+    {
+      "epoch": 0.09036847014925373,
+      "grad_norm": 0.5822554638380099,
+      "learning_rate": 4.980003007691449e-05,
+      "loss": 0.633,
+      "step": 775
+    },
+    {
+      "epoch": 0.09095149253731344,
+      "grad_norm": 0.6014893132349064,
+      "learning_rate": 4.979421772956852e-05,
+      "loss": 0.637,
+      "step": 780
+    },
+    {
+      "epoch": 0.09153451492537314,
+      "grad_norm": 0.6190152455325075,
+      "learning_rate": 4.9788322504763954e-05,
+      "loss": 0.626,
+      "step": 785
+    },
+    {
+      "epoch": 0.09211753731343283,
+      "grad_norm": 0.6043190595480243,
+      "learning_rate": 4.978234442441596e-05,
+      "loss": 0.5958,
+      "step": 790
+    },
+    {
+      "epoch": 0.09270055970149253,
+      "grad_norm": 0.7067979054927527,
+      "learning_rate": 4.977628351074769e-05,
+      "loss": 0.6391,
+      "step": 795
+    },
+    {
+      "epoch": 0.09328358208955224,
+      "grad_norm": 1.0900391421932065,
+      "learning_rate": 4.977013978629025e-05,
+      "loss": 0.5869,
+      "step": 800
+    },
+    {
+      "epoch": 0.09386660447761194,
+      "grad_norm": 0.7289575339056653,
+      "learning_rate": 4.976391327388257e-05,
+      "loss": 0.5817,
+      "step": 805
+    },
+    {
+      "epoch": 0.09444962686567164,
+      "grad_norm": 0.6772024103975534,
+      "learning_rate": 4.9757603996671354e-05,
+      "loss": 0.644,
+      "step": 810
+    },
+    {
+      "epoch": 0.09503264925373134,
+      "grad_norm": 0.6273836766329598,
+      "learning_rate": 4.975121197811096e-05,
+      "loss": 0.6343,
+      "step": 815
+    },
+    {
+      "epoch": 0.09561567164179105,
+      "grad_norm": 0.6925933519325139,
+      "learning_rate": 4.974473724196338e-05,
+      "loss": 0.6527,
+      "step": 820
+    },
+    {
+      "epoch": 0.09619869402985075,
+      "grad_norm": 0.7918262789810132,
+      "learning_rate": 4.973817981229802e-05,
+      "loss": 0.6043,
+      "step": 825
+    },
+    {
+      "epoch": 0.09678171641791045,
+      "grad_norm": 0.6915704244723462,
+      "learning_rate": 4.9731539713491776e-05,
+      "loss": 0.6101,
+      "step": 830
+    },
+    {
+      "epoch": 0.09736473880597014,
+      "grad_norm": 0.7026540645380082,
+      "learning_rate": 4.972481697022883e-05,
+      "loss": 0.6321,
+      "step": 835
+    },
+    {
+      "epoch": 0.09794776119402986,
+      "grad_norm": 0.6870682595723567,
+      "learning_rate": 4.971801160750057e-05,
+      "loss": 0.6431,
+      "step": 840
+    },
+    {
+      "epoch": 0.09853078358208955,
+      "grad_norm": 0.6762758634999102,
+      "learning_rate": 4.971112365060555e-05,
+      "loss": 0.6212,
+      "step": 845
+    },
+    {
+      "epoch": 0.09911380597014925,
+      "grad_norm": 1.0041969348534898,
+      "learning_rate": 4.970415312514936e-05,
+      "loss": 0.5934,
+      "step": 850
+    },
+    {
+      "epoch": 0.09969682835820895,
+      "grad_norm": 0.6273960276780831,
+      "learning_rate": 4.969710005704449e-05,
+      "loss": 0.6499,
+      "step": 855
+    },
+    {
+      "epoch": 0.10027985074626866,
+      "grad_norm": 0.6488940069009905,
+      "learning_rate": 4.9689964472510345e-05,
+      "loss": 0.5949,
+      "step": 860
+    },
+    {
+      "epoch": 0.10086287313432836,
+      "grad_norm": 0.763231378289182,
+      "learning_rate": 4.968274639807304e-05,
+      "loss": 0.6146,
+      "step": 865
+    },
+    {
+      "epoch": 0.10144589552238806,
+      "grad_norm": 0.6658920295533095,
+      "learning_rate": 4.967544586056532e-05,
+      "loss": 0.6049,
+      "step": 870
+    },
+    {
+      "epoch": 0.10202891791044776,
+      "grad_norm": 0.7074390644632157,
+      "learning_rate": 4.966806288712654e-05,
+      "loss": 0.6175,
+      "step": 875
+    },
+    {
+      "epoch": 0.10261194029850747,
+      "grad_norm": 0.695343426907801,
+      "learning_rate": 4.966059750520246e-05,
+      "loss": 0.591,
+      "step": 880
+    },
+    {
+      "epoch": 0.10319496268656717,
+      "grad_norm": 0.6418320757008029,
+      "learning_rate": 4.965304974254521e-05,
+      "loss": 0.6176,
+      "step": 885
+    },
+    {
+      "epoch": 0.10377798507462686,
+      "grad_norm": 0.7727923536862525,
+      "learning_rate": 4.9645419627213155e-05,
+      "loss": 0.6134,
+      "step": 890
+    },
+    {
+      "epoch": 0.10436100746268656,
+      "grad_norm": 0.6029656422873543,
+      "learning_rate": 4.96377071875708e-05,
+      "loss": 0.5654,
+      "step": 895
+    },
+    {
+      "epoch": 0.10494402985074627,
+      "grad_norm": 0.6358118150755105,
+      "learning_rate": 4.9629912452288696e-05,
+      "loss": 0.6252,
+      "step": 900
+    },
+    {
+      "epoch": 0.10552705223880597,
+      "grad_norm": 0.61344386800639,
+      "learning_rate": 4.962203545034332e-05,
+      "loss": 0.6059,
+      "step": 905
+    },
+    {
+      "epoch": 0.10611007462686567,
+      "grad_norm": 0.6135409466759565,
+      "learning_rate": 4.961407621101697e-05,
+      "loss": 0.5614,
+      "step": 910
+    },
+    {
+      "epoch": 0.10669309701492537,
+      "grad_norm": 1.0821214504665195,
+      "learning_rate": 4.960603476389765e-05,
+      "loss": 0.6162,
+      "step": 915
+    },
+    {
+      "epoch": 0.10727611940298508,
+      "grad_norm": 0.5759584165021776,
+      "learning_rate": 4.959791113887898e-05,
+      "loss": 0.6055,
+      "step": 920
+    },
+    {
+      "epoch": 0.10785914179104478,
+      "grad_norm": 0.5799139948464339,
+      "learning_rate": 4.958970536616006e-05,
+      "loss": 0.6067,
+      "step": 925
+    },
+    {
+      "epoch": 0.10844216417910447,
+      "grad_norm": 0.6049902178680088,
+      "learning_rate": 4.9581417476245365e-05,
+      "loss": 0.5808,
+      "step": 930
+    },
+    {
+      "epoch": 0.10902518656716417,
+      "grad_norm": 0.6216812883236675,
+      "learning_rate": 4.957304749994465e-05,
+      "loss": 0.5644,
+      "step": 935
+    },
+    {
+      "epoch": 0.10960820895522388,
+      "grad_norm": 0.6264774945250813,
+      "learning_rate": 4.956459546837283e-05,
+      "loss": 0.5889,
+      "step": 940
+    },
+    {
+      "epoch": 0.11019123134328358,
+      "grad_norm": 0.7584079121183279,
+      "learning_rate": 4.955606141294982e-05,
+      "loss": 0.6662,
+      "step": 945
+    },
+    {
+      "epoch": 0.11077425373134328,
+      "grad_norm": 0.6021040722392922,
+      "learning_rate": 4.954744536540048e-05,
+      "loss": 0.6075,
+      "step": 950
+    },
+    {
+      "epoch": 0.11135727611940298,
+      "grad_norm": 0.6516825177813871,
+      "learning_rate": 4.953874735775448e-05,
+      "loss": 0.6163,
+      "step": 955
+    },
+    {
+      "epoch": 0.11194029850746269,
+      "grad_norm": 0.6483316850868462,
+      "learning_rate": 4.9529967422346137e-05,
+      "loss": 0.6388,
+      "step": 960
+    },
+    {
+      "epoch": 0.11252332089552239,
+      "grad_norm": 1.0296359523919179,
+      "learning_rate": 4.952110559181437e-05,
+      "loss": 0.6187,
+      "step": 965
+    },
+    {
+      "epoch": 0.11310634328358209,
+      "grad_norm": 0.5963874485076113,
+      "learning_rate": 4.95121618991025e-05,
+      "loss": 0.653,
+      "step": 970
+    },
+    {
+      "epoch": 0.11368936567164178,
+      "grad_norm": 0.5827112927403142,
+      "learning_rate": 4.950313637745819e-05,
+      "loss": 0.6013,
+      "step": 975
+    },
+    {
+      "epoch": 0.1142723880597015,
+      "grad_norm": 0.5978368012555301,
+      "learning_rate": 4.9494029060433304e-05,
+      "loss": 0.5854,
+      "step": 980
+    },
+    {
+      "epoch": 0.1148554104477612,
+      "grad_norm": 0.6061921826097367,
+      "learning_rate": 4.9484839981883755e-05,
+      "loss": 0.6042,
+      "step": 985
+    },
+    {
+      "epoch": 0.11543843283582089,
+      "grad_norm": 0.577591628346454,
+      "learning_rate": 4.9475569175969414e-05,
+      "loss": 0.6033,
+      "step": 990
+    },
+    {
+      "epoch": 0.11602145522388059,
+      "grad_norm": 0.578876456470585,
+      "learning_rate": 4.9466216677153945e-05,
+      "loss": 0.5762,
+      "step": 995
+    },
+    {
+      "epoch": 0.1166044776119403,
+      "grad_norm": 0.6111711530967577,
+      "learning_rate": 4.945678252020475e-05,
+      "loss": 0.6078,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1171875,
+      "grad_norm": 0.5389571001747652,
+      "learning_rate": 4.9447266740192735e-05,
+      "loss": 0.586,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1177705223880597,
+      "grad_norm": 0.5665361736437953,
+      "learning_rate": 4.943766937249226e-05,
+      "loss": 0.5896,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11835354477611941,
+      "grad_norm": 0.6004663875659557,
+      "learning_rate": 4.942799045278099e-05,
+      "loss": 0.5931,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11893656716417911,
+      "grad_norm": 0.5877689595746617,
+      "learning_rate": 4.941823001703974e-05,
+      "loss": 0.5985,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1195195895522388,
+      "grad_norm": 0.5935807095087586,
+      "learning_rate": 4.940838810155237e-05,
+      "loss": 0.5991,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1201026119402985,
+      "grad_norm": 0.6656506474951656,
+      "learning_rate": 4.939846474290563e-05,
+      "loss": 0.6148,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12068563432835822,
+      "grad_norm": 0.6400226054588848,
+      "learning_rate": 4.9388459977989016e-05,
+      "loss": 0.6035,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12126865671641791,
+      "grad_norm": 0.5682054111741724,
+      "learning_rate": 4.937837384399467e-05,
+      "loss": 0.5942,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12185167910447761,
+      "grad_norm": 0.5722092103161287,
+      "learning_rate": 4.936820637841721e-05,
+      "loss": 0.6297,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12243470149253731,
+      "grad_norm": 0.5744583767360888,
+      "learning_rate": 4.935795761905359e-05,
+      "loss": 0.5799,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12301772388059702,
+      "grad_norm": 0.5955258224599197,
+      "learning_rate": 4.934762760400299e-05,
+      "loss": 0.5901,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12360074626865672,
+      "grad_norm": 0.5941450778138045,
+      "learning_rate": 4.933721637166662e-05,
+      "loss": 0.5932,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12418376865671642,
+      "grad_norm": 0.6984953623677035,
+      "learning_rate": 4.9326723960747655e-05,
+      "loss": 0.5928,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12476679104477612,
+      "grad_norm": 0.5556738969991728,
+      "learning_rate": 4.931615041025101e-05,
+      "loss": 0.5449,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12534981343283583,
+      "grad_norm": 0.6202875202064847,
+      "learning_rate": 4.9305495759483246e-05,
+      "loss": 0.6245,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1259328358208955,
+      "grad_norm": 0.6360648582792375,
+      "learning_rate": 4.929476004805241e-05,
+      "loss": 0.622,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12651585820895522,
+      "grad_norm": 0.548961800240671,
+      "learning_rate": 4.928394331586788e-05,
+      "loss": 0.5878,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12709888059701493,
+      "grad_norm": 0.6534887664001585,
+      "learning_rate": 4.927304560314023e-05,
+      "loss": 0.5985,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12768190298507462,
+      "grad_norm": 0.6051258985285789,
+      "learning_rate": 4.9262066950381074e-05,
+      "loss": 0.5815,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12826492537313433,
+      "grad_norm": 0.5940919877769313,
+      "learning_rate": 4.925100739840293e-05,
+      "loss": 0.6056,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12884794776119404,
+      "grad_norm": 0.5795416953387396,
+      "learning_rate": 4.923986698831902e-05,
+      "loss": 0.593,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12943097014925373,
+      "grad_norm": 0.6060003300391443,
+      "learning_rate": 4.922864576154318e-05,
+      "loss": 0.6293,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13001399253731344,
+      "grad_norm": 0.5926697361708111,
+      "learning_rate": 4.921734375978966e-05,
+      "loss": 0.5863,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13059701492537312,
+      "grad_norm": 0.5666399578950052,
+      "learning_rate": 4.9205961025073005e-05,
+      "loss": 0.6093,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13118003731343283,
+      "grad_norm": 0.6290080779373148,
+      "learning_rate": 4.919449759970787e-05,
+      "loss": 0.6149,
+      "step": 1125
+    },
+    {
+      "epoch": 0.13176305970149255,
+      "grad_norm": 0.6030059289618742,
+      "learning_rate": 4.9182953526308866e-05,
+      "loss": 0.5983,
+      "step": 1130
+    },
+    {
+      "epoch": 0.13234608208955223,
+      "grad_norm": 0.6967548151886129,
+      "learning_rate": 4.9171328847790416e-05,
+      "loss": 0.5979,
+      "step": 1135
+    },
+    {
+      "epoch": 0.13292910447761194,
+      "grad_norm": 0.6568770405720038,
+      "learning_rate": 4.9159623607366587e-05,
+      "loss": 0.5857,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13351212686567165,
+      "grad_norm": 0.5840444295366739,
+      "learning_rate": 4.914783784855093e-05,
+      "loss": 0.6438,
+      "step": 1145
+    },
+    {
+      "epoch": 0.13409514925373134,
+      "grad_norm": 0.7004511399791347,
+      "learning_rate": 4.913597161515633e-05,
+      "loss": 0.6184,
+      "step": 1150
+    },
+    {
+      "epoch": 0.13467817164179105,
+      "grad_norm": 0.550373143569304,
+      "learning_rate": 4.91240249512948e-05,
+      "loss": 0.567,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13526119402985073,
+      "grad_norm": 0.5313433759195957,
+      "learning_rate": 4.9111997901377373e-05,
+      "loss": 0.5855,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13584421641791045,
+      "grad_norm": 0.5655705667845206,
+      "learning_rate": 4.9099890510113924e-05,
+      "loss": 0.5661,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13642723880597016,
+      "grad_norm": 0.6195584641586247,
+      "learning_rate": 4.908770282251296e-05,
+      "loss": 0.5905,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13701026119402984,
+      "grad_norm": 0.5583698659457814,
+      "learning_rate": 4.9075434883881504e-05,
+      "loss": 0.6002,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13759328358208955,
+      "grad_norm": 0.6389070121403235,
+      "learning_rate": 4.906308673982491e-05,
+      "loss": 0.5607,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13817630597014927,
+      "grad_norm": 0.59774495236772,
+      "learning_rate": 4.905065843624668e-05,
+      "loss": 0.5678,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13875932835820895,
+      "grad_norm": 0.6173804758323792,
+      "learning_rate": 4.903815001934832e-05,
+      "loss": 0.6002,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13934235074626866,
+      "grad_norm": 0.5469869287414887,
+      "learning_rate": 4.9025561535629125e-05,
+      "loss": 0.5977,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13992537313432835,
+      "grad_norm": 0.5624438700425504,
+      "learning_rate": 4.9012893031886075e-05,
+      "loss": 0.582,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14050839552238806,
+      "grad_norm": 0.5985379982120503,
+      "learning_rate": 4.9000144555213575e-05,
+      "loss": 0.5825,
+      "step": 1205
+    },
+    {
+      "epoch": 0.14109141791044777,
+      "grad_norm": 0.5800377004502044,
+      "learning_rate": 4.898731615300336e-05,
+      "loss": 0.5706,
+      "step": 1210
+    },
+    {
+      "epoch": 0.14167444029850745,
+      "grad_norm": 0.5901666286652832,
+      "learning_rate": 4.8974407872944263e-05,
+      "loss": 0.5937,
+      "step": 1215
+    },
+    {
+      "epoch": 0.14225746268656717,
+      "grad_norm": 0.644009126842579,
+      "learning_rate": 4.8961419763022065e-05,
+      "loss": 0.5612,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14284048507462688,
+      "grad_norm": 0.5922807627602008,
+      "learning_rate": 4.894835187151931e-05,
+      "loss": 0.6067,
+      "step": 1225
+    },
+    {
+      "epoch": 0.14342350746268656,
+      "grad_norm": 0.5722199265482402,
+      "learning_rate": 4.893520424701513e-05,
+      "loss": 0.6082,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14400652985074627,
+      "grad_norm": 0.6126570623385963,
+      "learning_rate": 4.892197693838504e-05,
+      "loss": 0.5964,
+      "step": 1235
+    },
+    {
+      "epoch": 0.14458955223880596,
+      "grad_norm": 0.5973655677327272,
+      "learning_rate": 4.890866999480082e-05,
+      "loss": 0.5918,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14517257462686567,
+      "grad_norm": 0.54393899856667,
+      "learning_rate": 4.889528346573023e-05,
+      "loss": 0.5827,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14575559701492538,
+      "grad_norm": 0.5508599028743483,
+      "learning_rate": 4.888181740093693e-05,
+      "loss": 0.5966,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14633861940298507,
+      "grad_norm": 0.5748810947596694,
+      "learning_rate": 4.886827185048023e-05,
+      "loss": 0.6069,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14692164179104478,
+      "grad_norm": 0.5843720987519714,
+      "learning_rate": 4.8854646864714906e-05,
+      "loss": 0.626,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1475046641791045,
+      "grad_norm": 0.5574624014831214,
+      "learning_rate": 4.884094249429109e-05,
+      "loss": 0.5513,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14808768656716417,
+      "grad_norm": 0.6341906909347658,
+      "learning_rate": 4.882715879015396e-05,
+      "loss": 0.597,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14867070895522388,
+      "grad_norm": 0.575813430423846,
+      "learning_rate": 4.881329580354363e-05,
+      "loss": 0.6081,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 0.6445668774873804,
+      "learning_rate": 4.8799353585994954e-05,
+      "loss": 0.6087,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14983675373134328,
+      "grad_norm": 0.5859378252774211,
+      "learning_rate": 4.8785332189337325e-05,
+      "loss": 0.5602,
+      "step": 1285
+    },
+    {
+      "epoch": 0.150419776119403,
+      "grad_norm": 0.6268377383651598,
+      "learning_rate": 4.877123166569445e-05,
+      "loss": 0.621,
+      "step": 1290
+    },
+    {
+      "epoch": 0.15100279850746268,
+      "grad_norm": 0.539111443847871,
+      "learning_rate": 4.8757052067484216e-05,
+      "loss": 0.5778,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1515858208955224,
+      "grad_norm": 0.6230886477061637,
+      "learning_rate": 4.874279344741846e-05,
+      "loss": 0.6052,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1521688432835821,
+      "grad_norm": 0.5804714929817124,
+      "learning_rate": 4.872845585850277e-05,
+      "loss": 0.5704,
+      "step": 1305
+    },
+    {
+      "epoch": 0.15275186567164178,
+      "grad_norm": 0.5644557466436395,
+      "learning_rate": 4.87140393540363e-05,
+      "loss": 0.6135,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1533348880597015,
+      "grad_norm": 0.5979238448583827,
+      "learning_rate": 4.869954398761156e-05,
+      "loss": 0.5891,
+      "step": 1315
+    },
+    {
+      "epoch": 0.15391791044776118,
+      "grad_norm": 0.5684870674237956,
+      "learning_rate": 4.868496981311424e-05,
+      "loss": 0.6057,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1545009328358209,
+      "grad_norm": 0.6084433820062218,
+      "learning_rate": 4.8670316884722984e-05,
+      "loss": 0.6183,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1550839552238806,
+      "grad_norm": 0.5168917985718466,
+      "learning_rate": 4.86555852569092e-05,
+      "loss": 0.558,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1556669776119403,
+      "grad_norm": 0.6169882009143579,
+      "learning_rate": 4.864077498443687e-05,
+      "loss": 0.5894,
+      "step": 1335
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 0.5573154647203309,
+      "learning_rate": 4.8625886122362305e-05,
+      "loss": 0.6047,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1568330223880597,
+      "grad_norm": 0.5430765223123376,
+      "learning_rate": 4.861091872603399e-05,
+      "loss": 0.5936,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1574160447761194,
+      "grad_norm": 0.5952512138917507,
+      "learning_rate": 4.859587285109235e-05,
+      "loss": 0.6323,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1579990671641791,
+      "grad_norm": 0.6257733920419736,
+      "learning_rate": 4.8580748553469554e-05,
+      "loss": 0.621,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15858208955223882,
+      "grad_norm": 0.6484544777362733,
+      "learning_rate": 4.8565545889389295e-05,
+      "loss": 0.6225,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1591651119402985,
+      "grad_norm": 0.6001757890053516,
+      "learning_rate": 4.85502649153666e-05,
+      "loss": 0.7485,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15974813432835822,
+      "grad_norm": 0.661835508508043,
+      "learning_rate": 4.853490568820759e-05,
+      "loss": 0.8953,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1603311567164179,
+      "grad_norm": 0.5424000700147391,
+      "learning_rate": 4.851946826500932e-05,
+      "loss": 0.6088,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1609141791044776,
+      "grad_norm": 0.5642506132746299,
+      "learning_rate": 4.8503952703159485e-05,
+      "loss": 0.5788,
+      "step": 1380
+    },
+    {
+      "epoch": 0.16149720149253732,
+      "grad_norm": 1.5770888453223382,
+      "learning_rate": 4.8488359060336314e-05,
+      "loss": 0.6464,
+      "step": 1385
+    },
+    {
+      "epoch": 0.162080223880597,
+      "grad_norm": 0.5798567582108715,
+      "learning_rate": 4.847268739450825e-05,
+      "loss": 0.5962,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16266324626865672,
+      "grad_norm": 0.6247134338348835,
+      "learning_rate": 4.84569377639338e-05,
+      "loss": 0.5762,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16324626865671643,
+      "grad_norm": 0.6403693981471119,
+      "learning_rate": 4.84411102271613e-05,
+      "loss": 0.6033,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16382929104477612,
+      "grad_norm": 0.6328461284086258,
+      "learning_rate": 4.842520484302871e-05,
+      "loss": 0.6124,
+      "step": 1405
+    },
+    {
+      "epoch": 0.16441231343283583,
+      "grad_norm": 0.5690467644511553,
+      "learning_rate": 4.840922167066335e-05,
+      "loss": 0.6058,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1649953358208955,
+      "grad_norm": 0.5991271256192877,
+      "learning_rate": 4.8393160769481755e-05,
+      "loss": 0.5759,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16557835820895522,
+      "grad_norm": 0.5518688454925015,
+      "learning_rate": 4.8377022199189374e-05,
+      "loss": 0.5758,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16616138059701493,
+      "grad_norm": 0.8358279547212623,
+      "learning_rate": 4.836080601978043e-05,
+      "loss": 0.614,
+      "step": 1425
+    },
+    {
+      "epoch": 0.16674440298507462,
+      "grad_norm": 0.5736501078455153,
+      "learning_rate": 4.83445122915376e-05,
+      "loss": 0.6268,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16732742537313433,
+      "grad_norm": 0.501139191324822,
+      "learning_rate": 4.832814107503188e-05,
+      "loss": 0.5522,
+      "step": 1435
+    },
+    {
+      "epoch": 0.16791044776119404,
+      "grad_norm": 0.5414270989894118,
+      "learning_rate": 4.8311692431122326e-05,
+      "loss": 0.5985,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16849347014925373,
+      "grad_norm": 0.5503417116648404,
+      "learning_rate": 4.82951664209558e-05,
+      "loss": 0.5883,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16907649253731344,
+      "grad_norm": 0.5828254158920338,
+      "learning_rate": 4.82785631059668e-05,
+      "loss": 0.5973,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16965951492537312,
+      "grad_norm": 0.5750003990997431,
+      "learning_rate": 4.826188254787717e-05,
+      "loss": 0.5929,
+      "step": 1455
+    },
+    {
+      "epoch": 0.17024253731343283,
+      "grad_norm": 0.5931754407423996,
+      "learning_rate": 4.824512480869593e-05,
+      "loss": 0.584,
+      "step": 1460
+    },
+    {
+      "epoch": 0.17082555970149255,
+      "grad_norm": 0.5457054499115934,
+      "learning_rate": 4.822828995071899e-05,
+      "loss": 0.5711,
+      "step": 1465
+    },
+    {
+      "epoch": 0.17140858208955223,
+      "grad_norm": 0.5229569450197525,
+      "learning_rate": 4.821137803652896e-05,
+      "loss": 0.5459,
+      "step": 1470
+    },
+    {
+      "epoch": 0.17199160447761194,
+      "grad_norm": 0.6027439958371525,
+      "learning_rate": 4.819438912899489e-05,
+      "loss": 0.5814,
+      "step": 1475
+    },
+    {
+      "epoch": 0.17257462686567165,
+      "grad_norm": 0.5702069062273913,
+      "learning_rate": 4.8177323291272066e-05,
+      "loss": 0.6299,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17315764925373134,
+      "grad_norm": 0.5780484151492952,
+      "learning_rate": 4.8160180586801744e-05,
+      "loss": 0.5777,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17374067164179105,
+      "grad_norm": 0.575917162342149,
+      "learning_rate": 4.814296107931093e-05,
+      "loss": 0.5547,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17432369402985073,
+      "grad_norm": 0.5192602462313004,
+      "learning_rate": 4.812566483281216e-05,
+      "loss": 0.5669,
+      "step": 1495
+    },
+    {
+      "epoch": 0.17490671641791045,
+      "grad_norm": 0.5345650171129616,
+      "learning_rate": 4.81082919116032e-05,
+      "loss": 0.5917,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17548973880597016,
+      "grad_norm": 0.5475736070312467,
+      "learning_rate": 4.809084238026689e-05,
+      "loss": 0.602,
+      "step": 1505
+    },
+    {
+      "epoch": 0.17607276119402984,
+      "grad_norm": 0.5786621619994619,
+      "learning_rate": 4.8073316303670835e-05,
+      "loss": 0.5861,
+      "step": 1510
+    },
+    {
+      "epoch": 0.17665578358208955,
+      "grad_norm": 0.5448850237456587,
+      "learning_rate": 4.8055713746967216e-05,
+      "loss": 0.5857,
+      "step": 1515
+    },
+    {
+      "epoch": 0.17723880597014927,
+      "grad_norm": 0.5779873112523268,
+      "learning_rate": 4.803803477559252e-05,
+      "loss": 0.5665,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17782182835820895,
+      "grad_norm": 0.5652601574196091,
+      "learning_rate": 4.8020279455267274e-05,
+      "loss": 0.5794,
+      "step": 1525
+    },
+    {
+      "epoch": 0.17840485074626866,
+      "grad_norm": 0.5437626429311837,
+      "learning_rate": 4.800244785199588e-05,
+      "loss": 0.547,
+      "step": 1530
+    },
+    {
+      "epoch": 0.17898787313432835,
+      "grad_norm": 0.5657663832577134,
+      "learning_rate": 4.7984540032066266e-05,
+      "loss": 0.5645,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17957089552238806,
+      "grad_norm": 0.5396084565585528,
+      "learning_rate": 4.796655606204971e-05,
+      "loss": 0.5917,
+      "step": 1540
+    },
+    {
+      "epoch": 0.18015391791044777,
+      "grad_norm": 0.5659614973871961,
+      "learning_rate": 4.794849600880059e-05,
+      "loss": 0.5888,
+      "step": 1545
+    },
+    {
+      "epoch": 0.18073694029850745,
+      "grad_norm": 0.610344463972923,
+      "learning_rate": 4.793035993945609e-05,
+      "loss": 0.5973,
+      "step": 1550
+    },
+    {
+      "epoch": 0.18131996268656717,
+      "grad_norm": 0.5315364790581317,
+      "learning_rate": 4.7912147921436e-05,
+      "loss": 0.5793,
+      "step": 1555
+    },
+    {
+      "epoch": 0.18190298507462688,
+      "grad_norm": 0.5215816759572497,
+      "learning_rate": 4.789386002244244e-05,
+      "loss": 0.5773,
+      "step": 1560
+    },
+    {
+      "epoch": 0.18248600746268656,
+      "grad_norm": 0.48268753752530635,
+      "learning_rate": 4.7875496310459607e-05,
+      "loss": 0.5439,
+      "step": 1565
+    },
+    {
+      "epoch": 0.18306902985074627,
+      "grad_norm": 0.5928345158273391,
+      "learning_rate": 4.7857056853753536e-05,
+      "loss": 0.5946,
+      "step": 1570
+    },
+    {
+      "epoch": 0.18365205223880596,
+      "grad_norm": 0.5217322463974674,
+      "learning_rate": 4.783854172087183e-05,
+      "loss": 0.5633,
+      "step": 1575
+    },
+    {
+      "epoch": 0.18423507462686567,
+      "grad_norm": 0.5361438235075061,
+      "learning_rate": 4.781995098064343e-05,
+      "loss": 0.5616,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18481809701492538,
+      "grad_norm": 0.5319409918375809,
+      "learning_rate": 4.780128470217833e-05,
+      "loss": 0.5959,
+      "step": 1585
+    },
+    {
+      "epoch": 0.18540111940298507,
+      "grad_norm": 0.5656220383824416,
+      "learning_rate": 4.778254295486732e-05,
+      "loss": 0.5941,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18598414179104478,
+      "grad_norm": 0.664107972745298,
+      "learning_rate": 4.7763725808381777e-05,
+      "loss": 0.5932,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1865671641791045,
+      "grad_norm": 0.5594876020852717,
+      "learning_rate": 4.7744833332673336e-05,
+      "loss": 0.5754,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18715018656716417,
+      "grad_norm": 0.5748188546122887,
+      "learning_rate": 4.7725865597973684e-05,
+      "loss": 0.5929,
+      "step": 1605
+    },
+    {
+      "epoch": 0.18773320895522388,
+      "grad_norm": 0.5608090588962066,
+      "learning_rate": 4.770682267479427e-05,
+      "loss": 0.6,
+      "step": 1610
+    },
+    {
+      "epoch": 0.18831623134328357,
+      "grad_norm": 0.5614561936493092,
+      "learning_rate": 4.7687704633926056e-05,
+      "loss": 0.5885,
+      "step": 1615
+    },
+    {
+      "epoch": 0.18889925373134328,
+      "grad_norm": 0.5020575112535327,
+      "learning_rate": 4.766851154643924e-05,
+      "loss": 0.5766,
+      "step": 1620
+    },
+    {
+      "epoch": 0.189482276119403,
+      "grad_norm": 0.5350491170633971,
+      "learning_rate": 4.7649243483683015e-05,
+      "loss": 0.6016,
+      "step": 1625
+    },
+    {
+      "epoch": 0.19006529850746268,
+      "grad_norm": 0.5723463466212365,
+      "learning_rate": 4.762990051728529e-05,
+      "loss": 0.5938,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1906483208955224,
+      "grad_norm": 0.528038155312228,
+      "learning_rate": 4.7610482719152404e-05,
+      "loss": 0.5919,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1912313432835821,
+      "grad_norm": 0.5195292947087375,
+      "learning_rate": 4.7590990161468906e-05,
+      "loss": 0.587,
+      "step": 1640
+    },
+    {
+      "epoch": 0.19181436567164178,
+      "grad_norm": 0.5309411178378524,
+      "learning_rate": 4.757142291669724e-05,
+      "loss": 0.5584,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1923973880597015,
+      "grad_norm": 0.5235366705843203,
+      "learning_rate": 4.755178105757751e-05,
+      "loss": 0.5332,
+      "step": 1650
+    },
+    {
+      "epoch": 0.19298041044776118,
+      "grad_norm": 0.5662388478996405,
+      "learning_rate": 4.753206465712717e-05,
+      "loss": 0.6003,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1935634328358209,
+      "grad_norm": 0.541515776380167,
+      "learning_rate": 4.751227378864081e-05,
+      "loss": 0.6167,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1941464552238806,
+      "grad_norm": 0.5451263614433529,
+      "learning_rate": 4.749240852568981e-05,
+      "loss": 0.5795,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1947294776119403,
+      "grad_norm": 0.5539737195153389,
+      "learning_rate": 4.747246894212216e-05,
+      "loss": 0.6156,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1953125,
+      "grad_norm": 0.4838135257147904,
+      "learning_rate": 4.7452455112062076e-05,
+      "loss": 0.586,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1958955223880597,
+      "grad_norm": 0.6206912586835625,
+      "learning_rate": 4.743236710990982e-05,
+      "loss": 0.5835,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1964785447761194,
+      "grad_norm": 0.5575344781989854,
+      "learning_rate": 4.7412205010341385e-05,
+      "loss": 0.5615,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1970615671641791,
+      "grad_norm": 0.5465760234872273,
+      "learning_rate": 4.739196888830818e-05,
+      "loss": 0.5614,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19764458955223882,
+      "grad_norm": 0.6116107229280592,
+      "learning_rate": 4.737165881903683e-05,
+      "loss": 0.5777,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1982276119402985,
+      "grad_norm": 0.526352835776055,
+      "learning_rate": 4.735127487802882e-05,
+      "loss": 0.5499,
+      "step": 1700
+    },
+    {
+      "epoch": 0.19881063432835822,
+      "grad_norm": 0.5790768202270913,
+      "learning_rate": 4.7330817141060284e-05,
+      "loss": 0.6062,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1993936567164179,
+      "grad_norm": 0.5786125838532169,
+      "learning_rate": 4.731028568418167e-05,
+      "loss": 0.5853,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1999766791044776,
+      "grad_norm": 0.5220981929969758,
+      "learning_rate": 4.728968058371746e-05,
+      "loss": 0.5917,
+      "step": 1715
+    },
+    {
+      "epoch": 0.20055970149253732,
+      "grad_norm": 0.5962771524705726,
+      "learning_rate": 4.726900191626592e-05,
+      "loss": 0.615,
+      "step": 1720
+    },
+    {
+      "epoch": 0.201142723880597,
+      "grad_norm": 0.614179304629615,
+      "learning_rate": 4.724824975869881e-05,
+      "loss": 0.6088,
+      "step": 1725
+    },
+    {
+      "epoch": 0.20172574626865672,
+      "grad_norm": 0.5339809717905718,
+      "learning_rate": 4.722742418816106e-05,
+      "loss": 0.5856,
+      "step": 1730
+    },
+    {
+      "epoch": 0.20230876865671643,
+      "grad_norm": 0.5775559838523747,
+      "learning_rate": 4.7206525282070514e-05,
+      "loss": 0.5525,
+      "step": 1735
+    },
+    {
+      "epoch": 0.20289179104477612,
+      "grad_norm": 0.5787127160424809,
+      "learning_rate": 4.718555311811764e-05,
+      "loss": 0.5889,
+      "step": 1740
+    },
+    {
+      "epoch": 0.20347481343283583,
+      "grad_norm": 0.5507748006986873,
+      "learning_rate": 4.716450777426525e-05,
+      "loss": 0.5811,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2040578358208955,
+      "grad_norm": 0.5386322774377927,
+      "learning_rate": 4.7143389328748174e-05,
+      "loss": 0.5681,
+      "step": 1750
+    },
+    {
+      "epoch": 0.20464085820895522,
+      "grad_norm": 0.5253847435139612,
+      "learning_rate": 4.712219786007302e-05,
+      "loss": 0.5352,
+      "step": 1755
+    },
+    {
+      "epoch": 0.20522388059701493,
+      "grad_norm": 0.5229542272956548,
+      "learning_rate": 4.710093344701782e-05,
+      "loss": 0.5878,
+      "step": 1760
+    },
+    {
+      "epoch": 0.20580690298507462,
+      "grad_norm": 0.5347624824538455,
+      "learning_rate": 4.707959616863181e-05,
+      "loss": 0.5622,
+      "step": 1765
+    },
+    {
+      "epoch": 0.20638992537313433,
+      "grad_norm": 0.5561607185138324,
+      "learning_rate": 4.7058186104235086e-05,
+      "loss": 0.5797,
+      "step": 1770
+    },
+    {
+      "epoch": 0.20697294776119404,
+      "grad_norm": 0.4892165988190017,
+      "learning_rate": 4.70367033334183e-05,
+      "loss": 0.5403,
+      "step": 1775
+    },
+    {
+      "epoch": 0.20755597014925373,
+      "grad_norm": 0.5510594553982379,
+      "learning_rate": 4.701514793604242e-05,
+      "loss": 0.5559,
+      "step": 1780
+    },
+    {
+      "epoch": 0.20813899253731344,
+      "grad_norm": 0.5473696320978395,
+      "learning_rate": 4.699351999223838e-05,
+      "loss": 0.5753,
+      "step": 1785
+    },
+    {
+      "epoch": 0.20872201492537312,
+      "grad_norm": 0.4873353162223706,
+      "learning_rate": 4.697181958240679e-05,
+      "loss": 0.5492,
+      "step": 1790
+    },
+    {
+      "epoch": 0.20930503731343283,
+      "grad_norm": 0.5064007191967134,
+      "learning_rate": 4.695004678721768e-05,
+      "loss": 0.5858,
+      "step": 1795
+    },
+    {
+      "epoch": 0.20988805970149255,
+      "grad_norm": 0.5667430033926578,
+      "learning_rate": 4.692820168761014e-05,
+      "loss": 0.6062,
+      "step": 1800
+    },
+    {
+      "epoch": 0.21047108208955223,
+      "grad_norm": 0.5212510033253821,
+      "learning_rate": 4.690628436479206e-05,
+      "loss": 0.5598,
+      "step": 1805
+    },
+    {
+      "epoch": 0.21105410447761194,
+      "grad_norm": 0.5588622002334643,
+      "learning_rate": 4.688429490023982e-05,
+      "loss": 0.5763,
+      "step": 1810
+    },
+    {
+      "epoch": 0.21163712686567165,
+      "grad_norm": 0.6512372849264118,
+      "learning_rate": 4.6862233375697964e-05,
+      "loss": 0.5808,
+      "step": 1815
+    },
+    {
+      "epoch": 0.21222014925373134,
+      "grad_norm": 0.600056817158455,
+      "learning_rate": 4.684009987317894e-05,
+      "loss": 0.5929,
+      "step": 1820
+    },
+    {
+      "epoch": 0.21280317164179105,
+      "grad_norm": 0.5525047925178632,
+      "learning_rate": 4.6817894474962756e-05,
+      "loss": 0.5598,
+      "step": 1825
+    },
+    {
+      "epoch": 0.21338619402985073,
+      "grad_norm": 0.49741821795163615,
+      "learning_rate": 4.679561726359668e-05,
+      "loss": 0.5689,
+      "step": 1830
+    },
+    {
+      "epoch": 0.21396921641791045,
+      "grad_norm": 0.5369544968462594,
+      "learning_rate": 4.677326832189496e-05,
+      "loss": 0.5846,
+      "step": 1835
+    },
+    {
+      "epoch": 0.21455223880597016,
+      "grad_norm": 0.578124984252202,
+      "learning_rate": 4.675084773293848e-05,
+      "loss": 0.6068,
+      "step": 1840
+    },
+    {
+      "epoch": 0.21513526119402984,
+      "grad_norm": 0.5489389574421355,
+      "learning_rate": 4.6728355580074476e-05,
+      "loss": 0.5799,
+      "step": 1845
+    },
+    {
+      "epoch": 0.21571828358208955,
+      "grad_norm": 0.5492579167575462,
+      "learning_rate": 4.6705791946916236e-05,
+      "loss": 0.5969,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21630130597014927,
+      "grad_norm": 0.4976854096096666,
+      "learning_rate": 4.6683156917342726e-05,
+      "loss": 0.6034,
+      "step": 1855
+    },
+    {
+      "epoch": 0.21688432835820895,
+      "grad_norm": 0.600940601658767,
+      "learning_rate": 4.666045057549838e-05,
+      "loss": 0.5946,
+      "step": 1860
+    },
+    {
+      "epoch": 0.21746735074626866,
+      "grad_norm": 0.5506775932621145,
+      "learning_rate": 4.663767300579268e-05,
+      "loss": 0.5847,
+      "step": 1865
+    },
+    {
+      "epoch": 0.21805037313432835,
+      "grad_norm": 0.5271315191050433,
+      "learning_rate": 4.661482429289994e-05,
+      "loss": 0.5662,
+      "step": 1870
+    },
+    {
+      "epoch": 0.21863339552238806,
+      "grad_norm": 0.5806047125500812,
+      "learning_rate": 4.659190452175891e-05,
+      "loss": 0.5717,
+      "step": 1875
+    },
+    {
+      "epoch": 0.21921641791044777,
+      "grad_norm": 0.5548551191672546,
+      "learning_rate": 4.65689137775725e-05,
+      "loss": 0.5918,
+      "step": 1880
+    },
+    {
+      "epoch": 0.21979944029850745,
+      "grad_norm": 0.5224077396834346,
+      "learning_rate": 4.654585214580749e-05,
+      "loss": 0.5764,
+      "step": 1885
+    },
+    {
+      "epoch": 0.22038246268656717,
+      "grad_norm": 0.595212892553765,
+      "learning_rate": 4.652271971219412e-05,
+      "loss": 0.5718,
+      "step": 1890
+    },
+    {
+      "epoch": 0.22096548507462688,
+      "grad_norm": 0.5205351842902096,
+      "learning_rate": 4.6499516562725906e-05,
+      "loss": 0.5843,
+      "step": 1895
+    },
+    {
+      "epoch": 0.22154850746268656,
+      "grad_norm": 0.5193771435539847,
+      "learning_rate": 4.647624278365917e-05,
+      "loss": 0.6024,
+      "step": 1900
+    },
+    {
+      "epoch": 0.22213152985074627,
+      "grad_norm": 0.5762529929483823,
+      "learning_rate": 4.6452898461512866e-05,
+      "loss": 0.5841,
+      "step": 1905
+    },
+    {
+      "epoch": 0.22271455223880596,
+      "grad_norm": 0.5613099807801801,
+      "learning_rate": 4.642948368306814e-05,
+      "loss": 0.5909,
+      "step": 1910
+    },
+    {
+      "epoch": 0.22329757462686567,
+      "grad_norm": 0.5857098538355904,
+      "learning_rate": 4.640599853536806e-05,
+      "loss": 0.5986,
+      "step": 1915
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 0.5346551057187519,
+      "learning_rate": 4.6382443105717324e-05,
+      "loss": 0.5651,
+      "step": 1920
+    },
+    {
+      "epoch": 0.22446361940298507,
+      "grad_norm": 0.6133229780421945,
+      "learning_rate": 4.635881748168184e-05,
+      "loss": 0.6051,
+      "step": 1925
+    },
+    {
+      "epoch": 0.22504664179104478,
+      "grad_norm": 0.564856310131765,
+      "learning_rate": 4.633512175108851e-05,
+      "loss": 0.5645,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2256296641791045,
+      "grad_norm": 0.6086776045173963,
+      "learning_rate": 4.631135600202485e-05,
+      "loss": 0.611,
+      "step": 1935
+    },
+    {
+      "epoch": 0.22621268656716417,
+      "grad_norm": 0.5053374707380298,
+      "learning_rate": 4.628752032283862e-05,
+      "loss": 0.6359,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22679570895522388,
+      "grad_norm": 0.5177309796242289,
+      "learning_rate": 4.626361480213759e-05,
+      "loss": 0.5161,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22737873134328357,
+      "grad_norm": 0.5655297929571091,
+      "learning_rate": 4.623963952878914e-05,
+      "loss": 0.5215,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22796175373134328,
+      "grad_norm": 0.5007101086568317,
+      "learning_rate": 4.621559459191996e-05,
+      "loss": 0.5672,
+      "step": 1955
+    },
+    {
+      "epoch": 0.228544776119403,
+      "grad_norm": 0.5604337094818254,
+      "learning_rate": 4.619148008091569e-05,
+      "loss": 0.5495,
+      "step": 1960
+    },
+    {
+      "epoch": 0.22912779850746268,
+      "grad_norm": 0.5600231646159735,
+      "learning_rate": 4.616729608542064e-05,
+      "loss": 0.5914,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2297108208955224,
+      "grad_norm": 0.5261560439698072,
+      "learning_rate": 4.61430426953374e-05,
+      "loss": 0.5761,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2302938432835821,
+      "grad_norm": 0.5177160101117486,
+      "learning_rate": 4.611872000082654e-05,
+      "loss": 0.6105,
+      "step": 1975
+    },
+    {
+      "epoch": 0.23087686567164178,
+      "grad_norm": 0.4991176586608243,
+      "learning_rate": 4.609432809230627e-05,
+      "loss": 0.5576,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2314598880597015,
+      "grad_norm": 0.49011496350766276,
+      "learning_rate": 4.606986706045207e-05,
+      "loss": 0.5217,
+      "step": 1985
+    },
+    {
+      "epoch": 0.23204291044776118,
+      "grad_norm": 0.49863205095192953,
+      "learning_rate": 4.604533699619643e-05,
+      "loss": 0.5606,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2326259328358209,
+      "grad_norm": 0.5202522274733646,
+      "learning_rate": 4.602073799072841e-05,
+      "loss": 0.582,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2332089552238806,
+      "grad_norm": 0.5778738097193321,
+      "learning_rate": 4.5996070135493426e-05,
+      "loss": 0.5608,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2337919776119403,
+      "grad_norm": 0.5383730742466348,
+      "learning_rate": 4.597133352219275e-05,
+      "loss": 0.5833,
+      "step": 2005
+    },
+    {
+      "epoch": 0.234375,
+      "grad_norm": 0.504768235114523,
+      "learning_rate": 4.594652824278333e-05,
+      "loss": 0.5428,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2349580223880597,
+      "grad_norm": 0.5424141377382832,
+      "learning_rate": 4.592165438947734e-05,
+      "loss": 0.5234,
+      "step": 2015
+    },
+    {
+      "epoch": 0.2355410447761194,
+      "grad_norm": 0.560600079358499,
+      "learning_rate": 4.589671205474189e-05,
+      "loss": 0.5949,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2361240671641791,
+      "grad_norm": 0.5710842213852398,
+      "learning_rate": 4.587170133129867e-05,
+      "loss": 0.5675,
+      "step": 2025
+    },
+    {
+      "epoch": 0.23670708955223882,
+      "grad_norm": 0.6954740937313674,
+      "learning_rate": 4.5846622312123566e-05,
+      "loss": 0.5593,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2372901119402985,
+      "grad_norm": 0.4936861843329207,
+      "learning_rate": 4.582147509044639e-05,
+      "loss": 0.544,
+      "step": 2035
+    },
+    {
+      "epoch": 0.23787313432835822,
+      "grad_norm": 0.5990080756647657,
+      "learning_rate": 4.579625975975047e-05,
+      "loss": 0.6093,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2384561567164179,
+      "grad_norm": 0.46710498554704555,
+      "learning_rate": 4.577097641377234e-05,
+      "loss": 0.5421,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2390391791044776,
+      "grad_norm": 0.6220614730173019,
+      "learning_rate": 4.574562514650137e-05,
+      "loss": 0.5959,
+      "step": 2050
+    },
+    {
+      "epoch": 0.23962220149253732,
+      "grad_norm": 0.5736687871170686,
+      "learning_rate": 4.572020605217941e-05,
+      "loss": 0.5705,
+      "step": 2055
+    },
+    {
+      "epoch": 0.240205223880597,
+      "grad_norm": 0.5700891932767567,
+      "learning_rate": 4.569471922530048e-05,
+      "loss": 0.5812,
+      "step": 2060
+    },
+    {
+      "epoch": 0.24078824626865672,
+      "grad_norm": 0.5076933159103915,
+      "learning_rate": 4.566916476061036e-05,
+      "loss": 0.565,
+      "step": 2065
+    },
+    {
+      "epoch": 0.24137126865671643,
+      "grad_norm": 0.5485662358157128,
+      "learning_rate": 4.56435427531063e-05,
+      "loss": 0.576,
+      "step": 2070
+    },
+    {
+      "epoch": 0.24195429104477612,
+      "grad_norm": 0.5699225728909032,
+      "learning_rate": 4.5617853298036634e-05,
+      "loss": 0.5984,
+      "step": 2075
+    },
+    {
+      "epoch": 0.24253731343283583,
+      "grad_norm": 0.5970990880266768,
+      "learning_rate": 4.559209649090039e-05,
+      "loss": 0.5648,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2431203358208955,
+      "grad_norm": 0.5310811562750187,
+      "learning_rate": 4.556627242744703e-05,
+      "loss": 0.5616,
+      "step": 2085
+    },
+    {
+      "epoch": 0.24370335820895522,
+      "grad_norm": 0.4892958829299151,
+      "learning_rate": 4.5540381203675994e-05,
+      "loss": 0.5867,
+      "step": 2090
+    },
+    {
+      "epoch": 0.24428638059701493,
+      "grad_norm": 0.5227922779707119,
+      "learning_rate": 4.55144229158364e-05,
+      "loss": 0.5534,
+      "step": 2095
+    },
+    {
+      "epoch": 0.24486940298507462,
+      "grad_norm": 0.5041942400636352,
+      "learning_rate": 4.548839766042668e-05,
+      "loss": 0.5371,
+      "step": 2100
+    },
+    {
+      "epoch": 0.24545242537313433,
+      "grad_norm": 0.49660405961707166,
+      "learning_rate": 4.5462305534194204e-05,
+      "loss": 0.572,
+      "step": 2105
+    },
+    {
+      "epoch": 0.24603544776119404,
+      "grad_norm": 0.5499042134726534,
+      "learning_rate": 4.543614663413493e-05,
+      "loss": 0.5611,
+      "step": 2110
+    },
+    {
+      "epoch": 0.24661847014925373,
+      "grad_norm": 0.5496526828713953,
+      "learning_rate": 4.5409921057493064e-05,
+      "loss": 0.57,
+      "step": 2115
+    },
+    {
+      "epoch": 0.24720149253731344,
+      "grad_norm": 0.5599675685862884,
+      "learning_rate": 4.538362890176066e-05,
+      "loss": 0.5618,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24778451492537312,
+      "grad_norm": 0.5579889614380124,
+      "learning_rate": 4.535727026467727e-05,
+      "loss": 0.5682,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24836753731343283,
+      "grad_norm": 0.5300253135611827,
+      "learning_rate": 4.533084524422959e-05,
+      "loss": 0.5828,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24895055970149255,
+      "grad_norm": 0.498730689430679,
+      "learning_rate": 4.530435393865111e-05,
+      "loss": 0.5535,
+      "step": 2135
+    },
+    {
+      "epoch": 0.24953358208955223,
+      "grad_norm": 0.5177126292248756,
+      "learning_rate": 4.527779644642172e-05,
+      "loss": 0.5661,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2501166044776119,
+      "grad_norm": 0.5708740823647661,
+      "learning_rate": 4.525117286626734e-05,
+      "loss": 0.5764,
+      "step": 2145
+    },
+    {
+      "epoch": 0.25069962686567165,
+      "grad_norm": 0.5326668265886716,
+      "learning_rate": 4.522448329715959e-05,
+      "loss": 0.562,
+      "step": 2150
+    },
+    {
+      "epoch": 0.25128264925373134,
+      "grad_norm": 0.4878189090302149,
+      "learning_rate": 4.51977278383154e-05,
+      "loss": 0.5389,
+      "step": 2155
+    },
+    {
+      "epoch": 0.251865671641791,
+      "grad_norm": 0.5459043435498905,
+      "learning_rate": 4.517090658919662e-05,
+      "loss": 0.5456,
+      "step": 2160
+    },
+    {
+      "epoch": 0.25244869402985076,
+      "grad_norm": 0.5451507549234296,
+      "learning_rate": 4.5144019649509694e-05,
+      "loss": 0.5619,
+      "step": 2165
+    },
+    {
+      "epoch": 0.25303171641791045,
+      "grad_norm": 0.48817546876328066,
+      "learning_rate": 4.5117067119205256e-05,
+      "loss": 0.5328,
+      "step": 2170
+    },
+    {
+      "epoch": 0.25361473880597013,
+      "grad_norm": 0.5573481611505946,
+      "learning_rate": 4.5090049098477756e-05,
+      "loss": 0.6166,
+      "step": 2175
+    },
+    {
+      "epoch": 0.25419776119402987,
+      "grad_norm": 0.46166230865153285,
+      "learning_rate": 4.506296568776513e-05,
+      "loss": 0.5603,
+      "step": 2180
+    },
+    {
+      "epoch": 0.25478078358208955,
+      "grad_norm": 0.5255525351801149,
+      "learning_rate": 4.503581698774838e-05,
+      "loss": 0.5597,
+      "step": 2185
+    },
+    {
+      "epoch": 0.25536380597014924,
+      "grad_norm": 0.5036713071359231,
+      "learning_rate": 4.5008603099351235e-05,
+      "loss": 0.5572,
+      "step": 2190
+    },
+    {
+      "epoch": 0.255946828358209,
+      "grad_norm": 0.5629197395168302,
+      "learning_rate": 4.498132412373972e-05,
+      "loss": 0.5549,
+      "step": 2195
+    },
+    {
+      "epoch": 0.25652985074626866,
+      "grad_norm": 0.5281653518122478,
+      "learning_rate": 4.4953980162321845e-05,
+      "loss": 0.5783,
+      "step": 2200
+    },
+    {
+      "epoch": 0.25711287313432835,
+      "grad_norm": 0.52357818056316,
+      "learning_rate": 4.492657131674722e-05,
+      "loss": 0.5365,
+      "step": 2205
+    },
+    {
+      "epoch": 0.2576958955223881,
+      "grad_norm": 0.5581473844002001,
+      "learning_rate": 4.48990976889066e-05,
+      "loss": 0.5314,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25827891791044777,
+      "grad_norm": 0.5463641738254288,
+      "learning_rate": 4.487155938093163e-05,
+      "loss": 0.5678,
+      "step": 2215
+    },
+    {
+      "epoch": 0.25886194029850745,
+      "grad_norm": 0.580427123805623,
+      "learning_rate": 4.484395649519435e-05,
+      "loss": 0.5831,
+      "step": 2220
+    },
+    {
+      "epoch": 0.25944496268656714,
+      "grad_norm": 0.5515157134010146,
+      "learning_rate": 4.48162891343069e-05,
+      "loss": 0.5916,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2600279850746269,
+      "grad_norm": 0.4614774802102389,
+      "learning_rate": 4.478855740112107e-05,
+      "loss": 0.5747,
+      "step": 2230
+    },
+    {
+      "epoch": 0.26061100746268656,
+      "grad_norm": 0.5263544516058696,
+      "learning_rate": 4.476076139872797e-05,
+      "loss": 0.5873,
+      "step": 2235
+    },
+    {
+      "epoch": 0.26119402985074625,
+      "grad_norm": 0.5279147571827948,
+      "learning_rate": 4.473290123045764e-05,
+      "loss": 0.5694,
+      "step": 2240
+    },
+    {
+      "epoch": 0.261777052238806,
+      "grad_norm": 0.9742556257776407,
+      "learning_rate": 4.470497699987861e-05,
+      "loss": 0.5686,
+      "step": 2245
+    },
+    {
+      "epoch": 0.26236007462686567,
+      "grad_norm": 0.5328912557304052,
+      "learning_rate": 4.4676988810797596e-05,
+      "loss": 0.5591,
+      "step": 2250
+    },
+    {
+      "epoch": 0.26294309701492535,
+      "grad_norm": 0.5633803039126256,
+      "learning_rate": 4.464893676725906e-05,
+      "loss": 0.5728,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2635261194029851,
+      "grad_norm": 0.5134466833234131,
+      "learning_rate": 4.4620820973544866e-05,
+      "loss": 0.588,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2641091417910448,
+      "grad_norm": 0.5049188829803447,
+      "learning_rate": 4.459264153417381e-05,
+      "loss": 0.5973,
+      "step": 2265
+    },
+    {
+      "epoch": 0.26469216417910446,
+      "grad_norm": 0.5404106159275316,
+      "learning_rate": 4.4564398553901344e-05,
+      "loss": 0.5788,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2652751865671642,
+      "grad_norm": 0.6037489755180414,
+      "learning_rate": 4.4536092137719094e-05,
+      "loss": 0.5935,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2658582089552239,
+      "grad_norm": 0.5161487013459114,
+      "learning_rate": 4.450772239085452e-05,
+      "loss": 0.5371,
+      "step": 2280
+    },
+    {
+      "epoch": 0.26644123134328357,
+      "grad_norm": 0.5278496691940492,
+      "learning_rate": 4.44792894187705e-05,
+      "loss": 0.5234,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2670242537313433,
+      "grad_norm": 0.5092042449789793,
+      "learning_rate": 4.445079332716497e-05,
+      "loss": 0.5925,
+      "step": 2290
+    },
+    {
+      "epoch": 0.267607276119403,
+      "grad_norm": 0.5223498221449892,
+      "learning_rate": 4.4422234221970475e-05,
+      "loss": 0.5629,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2681902985074627,
+      "grad_norm": 0.5456470351851613,
+      "learning_rate": 4.439361220935385e-05,
+      "loss": 0.5416,
+      "step": 2300
+    },
+    {
+      "epoch": 0.26877332089552236,
+      "grad_norm": 0.5476526134142059,
+      "learning_rate": 4.436492739571575e-05,
+      "loss": 0.5499,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2693563432835821,
+      "grad_norm": 0.49505688936501147,
+      "learning_rate": 4.433617988769031e-05,
+      "loss": 0.5431,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2699393656716418,
+      "grad_norm": 0.5342065164923998,
+      "learning_rate": 4.43073697921447e-05,
+      "loss": 0.5693,
+      "step": 2315
+    },
+    {
+      "epoch": 0.27052238805970147,
+      "grad_norm": 0.5145761250584094,
+      "learning_rate": 4.4278497216178805e-05,
+      "loss": 0.5795,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2711054104477612,
+      "grad_norm": 0.5273626768859015,
+      "learning_rate": 4.4249562267124735e-05,
+      "loss": 0.5549,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2716884328358209,
+      "grad_norm": 0.5079624559473638,
+      "learning_rate": 4.422056505254648e-05,
+      "loss": 0.5746,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2722714552238806,
+      "grad_norm": 0.505058134321932,
+      "learning_rate": 4.4191505680239494e-05,
+      "loss": 0.5565,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2728544776119403,
+      "grad_norm": 0.5668879894960839,
+      "learning_rate": 4.416238425823031e-05,
+      "loss": 0.5578,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2734375,
+      "grad_norm": 0.5141499926926323,
+      "learning_rate": 4.413320089477612e-05,
+      "loss": 0.5489,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2740205223880597,
+      "grad_norm": 0.5109670063026779,
+      "learning_rate": 4.4103955698364394e-05,
+      "loss": 0.5592,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2746035447761194,
+      "grad_norm": 0.5082096864806888,
+      "learning_rate": 4.407464877771243e-05,
+      "loss": 0.5213,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2751865671641791,
+      "grad_norm": 0.5083698199510284,
+      "learning_rate": 4.4045280241767024e-05,
+      "loss": 0.5493,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2757695895522388,
+      "grad_norm": 0.5053920381989545,
+      "learning_rate": 4.401585019970397e-05,
+      "loss": 0.5686,
+      "step": 2365
+    },
+    {
+      "epoch": 0.27635261194029853,
+      "grad_norm": 0.4963501376761793,
+      "learning_rate": 4.3986358760927774e-05,
+      "loss": 0.5545,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2769356343283582,
+      "grad_norm": 0.5288831258581409,
+      "learning_rate": 4.3956806035071123e-05,
+      "loss": 0.5249,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2775186567164179,
+      "grad_norm": 0.62308836025263,
+      "learning_rate": 4.392719213199457e-05,
+      "loss": 0.6042,
+      "step": 2380
+    },
+    {
+      "epoch": 0.27810167910447764,
+      "grad_norm": 0.4832296950151894,
+      "learning_rate": 4.389751716178606e-05,
+      "loss": 0.5633,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2786847014925373,
+      "grad_norm": 0.5237890773447569,
+      "learning_rate": 4.386778123476059e-05,
+      "loss": 0.5271,
+      "step": 2390
+    },
+    {
+      "epoch": 0.279267723880597,
+      "grad_norm": 0.5595739424570599,
+      "learning_rate": 4.383798446145973e-05,
+      "loss": 0.6001,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2798507462686567,
+      "grad_norm": 0.5332931182325285,
+      "learning_rate": 4.380812695265126e-05,
+      "loss": 0.6024,
+      "step": 2400
+    },
+    {
+      "epoch": 0.28043376865671643,
+      "grad_norm": 0.4918731480131771,
+      "learning_rate": 4.3778208819328724e-05,
+      "loss": 0.56,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2810167910447761,
+      "grad_norm": 0.4625310842353903,
+      "learning_rate": 4.374823017271105e-05,
+      "loss": 0.5161,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2815998134328358,
+      "grad_norm": 0.5070557648263391,
+      "learning_rate": 4.371819112424212e-05,
+      "loss": 0.518,
+      "step": 2415
+    },
+    {
+      "epoch": 0.28218283582089554,
+      "grad_norm": 0.56213077270057,
+      "learning_rate": 4.368809178559034e-05,
+      "loss": 0.5645,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2827658582089552,
+      "grad_norm": 0.5027850751505164,
+      "learning_rate": 4.365793226864825e-05,
+      "loss": 0.5516,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2833488805970149,
+      "grad_norm": 0.496969778411097,
+      "learning_rate": 4.3627712685532104e-05,
+      "loss": 0.5661,
+      "step": 2430
+    },
+    {
+      "epoch": 0.28393190298507465,
+      "grad_norm": 0.5203901533643535,
+      "learning_rate": 4.3597433148581465e-05,
+      "loss": 0.5564,
+      "step": 2435
+    },
+    {
+      "epoch": 0.28451492537313433,
+      "grad_norm": 0.5761459191507305,
+      "learning_rate": 4.3567093770358724e-05,
+      "loss": 0.584,
+      "step": 2440
+    },
+    {
+      "epoch": 0.285097947761194,
+      "grad_norm": 0.5440010534822834,
+      "learning_rate": 4.353669466364877e-05,
+      "loss": 0.5833,
+      "step": 2445
+    },
+    {
+      "epoch": 0.28568097014925375,
+      "grad_norm": 0.5575628208489772,
+      "learning_rate": 4.3506235941458516e-05,
+      "loss": 0.5926,
+      "step": 2450
+    },
+    {
+      "epoch": 0.28626399253731344,
+      "grad_norm": 0.500211988994469,
+      "learning_rate": 4.347571771701648e-05,
+      "loss": 0.5651,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2868470149253731,
+      "grad_norm": 0.46825863619458763,
+      "learning_rate": 4.34451401037724e-05,
+      "loss": 0.5461,
+      "step": 2460
+    },
+    {
+      "epoch": 0.28743003731343286,
+      "grad_norm": 0.5521309621293441,
+      "learning_rate": 4.3414503215396776e-05,
+      "loss": 0.5659,
+      "step": 2465
+    },
+    {
+      "epoch": 0.28801305970149255,
+      "grad_norm": 0.4609786323278,
+      "learning_rate": 4.338380716578046e-05,
+      "loss": 0.5487,
+      "step": 2470
+    },
+    {
+      "epoch": 0.28859608208955223,
+      "grad_norm": 0.499738534899754,
+      "learning_rate": 4.3353052069034214e-05,
+      "loss": 0.5381,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2891791044776119,
+      "grad_norm": 0.5917418512742165,
+      "learning_rate": 4.332223803948834e-05,
+      "loss": 0.5434,
+      "step": 2480
+    },
+    {
+      "epoch": 0.28976212686567165,
+      "grad_norm": 0.48180474136483686,
+      "learning_rate": 4.3291365191692204e-05,
+      "loss": 0.5734,
+      "step": 2485
+    },
+    {
+      "epoch": 0.29034514925373134,
+      "grad_norm": 0.5007564237127459,
+      "learning_rate": 4.326043364041381e-05,
+      "loss": 0.5352,
+      "step": 2490
+    },
+    {
+      "epoch": 0.290928171641791,
+      "grad_norm": 0.47349703509695523,
+      "learning_rate": 4.3229443500639414e-05,
+      "loss": 0.5553,
+      "step": 2495
+    },
+    {
+      "epoch": 0.29151119402985076,
+      "grad_norm": 0.5192204895800268,
+      "learning_rate": 4.319839488757305e-05,
+      "loss": 0.5398,
+      "step": 2500
+    },
+    {
+      "epoch": 0.29209421641791045,
+      "grad_norm": 0.5757316112230765,
+      "learning_rate": 4.3167287916636145e-05,
+      "loss": 0.5795,
+      "step": 2505
+    },
+    {
+      "epoch": 0.29267723880597013,
+      "grad_norm": 0.5046004476804008,
+      "learning_rate": 4.3136122703467045e-05,
+      "loss": 0.5657,
+      "step": 2510
+    },
+    {
+      "epoch": 0.29326026119402987,
+      "grad_norm": 0.6116298626095651,
+      "learning_rate": 4.3104899363920616e-05,
+      "loss": 0.5892,
+      "step": 2515
+    },
+    {
+      "epoch": 0.29384328358208955,
+      "grad_norm": 0.5028293067243745,
+      "learning_rate": 4.3073618014067824e-05,
+      "loss": 0.5614,
+      "step": 2520
+    },
+    {
+      "epoch": 0.29442630597014924,
+      "grad_norm": 0.5284740774719737,
+      "learning_rate": 4.304227877019525e-05,
+      "loss": 0.5509,
+      "step": 2525
+    },
+    {
+      "epoch": 0.295009328358209,
+      "grad_norm": 0.5043452273361897,
+      "learning_rate": 4.301088174880472e-05,
+      "loss": 0.5734,
+      "step": 2530
+    },
+    {
+      "epoch": 0.29559235074626866,
+      "grad_norm": 0.5491908726146885,
+      "learning_rate": 4.297942706661283e-05,
+      "loss": 0.5994,
+      "step": 2535
+    },
+    {
+      "epoch": 0.29617537313432835,
+      "grad_norm": 0.5967635993887994,
+      "learning_rate": 4.2947914840550544e-05,
+      "loss": 0.5895,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2967583955223881,
+      "grad_norm": 0.49957976391073805,
+      "learning_rate": 4.291634518776273e-05,
+      "loss": 0.5559,
+      "step": 2545
+    },
+    {
+      "epoch": 0.29734141791044777,
+      "grad_norm": 0.586452369328343,
+      "learning_rate": 4.2884718225607736e-05,
+      "loss": 0.5987,
+      "step": 2550
+    },
+    {
+      "epoch": 0.29792444029850745,
+      "grad_norm": 0.5079777653661454,
+      "learning_rate": 4.285303407165694e-05,
+      "loss": 0.5892,
+      "step": 2555
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.45945289033221137,
+      "learning_rate": 4.282129284369436e-05,
+      "loss": 0.5353,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2990904850746269,
+      "grad_norm": 0.47293762371540526,
+      "learning_rate": 4.278949465971616e-05,
+      "loss": 0.5451,
+      "step": 2565
+    },
+    {
+      "epoch": 0.29967350746268656,
+      "grad_norm": 0.5125425687772994,
+      "learning_rate": 4.2757639637930246e-05,
+      "loss": 0.588,
+      "step": 2570
+    },
+    {
+      "epoch": 0.30025652985074625,
+      "grad_norm": 0.560469657657478,
+      "learning_rate": 4.2725727896755804e-05,
+      "loss": 0.5353,
+      "step": 2575
+    },
+    {
+      "epoch": 0.300839552238806,
+      "grad_norm": 0.5254788469907365,
+      "learning_rate": 4.269375955482287e-05,
+      "loss": 0.547,
+      "step": 2580
+    },
+    {
+      "epoch": 0.30142257462686567,
+      "grad_norm": 0.48987507837663824,
+      "learning_rate": 4.266173473097192e-05,
+      "loss": 0.5175,
+      "step": 2585
+    },
+    {
+      "epoch": 0.30200559701492535,
+      "grad_norm": 0.5484732726515896,
+      "learning_rate": 4.262965354425335e-05,
+      "loss": 0.6067,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3025886194029851,
+      "grad_norm": 0.5064490667857149,
+      "learning_rate": 4.259751611392712e-05,
+      "loss": 0.533,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3031716417910448,
+      "grad_norm": 0.5014575792637805,
+      "learning_rate": 4.256532255946226e-05,
+      "loss": 0.5913,
+      "step": 2600
+    },
+    {
+      "epoch": 0.30375466417910446,
+      "grad_norm": 0.48901348419554475,
+      "learning_rate": 4.253307300053643e-05,
+      "loss": 0.5543,
+      "step": 2605
+    },
+    {
+      "epoch": 0.3043376865671642,
+      "grad_norm": 0.5232458296317694,
+      "learning_rate": 4.25007675570355e-05,
+      "loss": 0.5759,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3049207089552239,
+      "grad_norm": 0.5070730983893377,
+      "learning_rate": 4.246840634905307e-05,
+      "loss": 0.5515,
+      "step": 2615
+    },
+    {
+      "epoch": 0.30550373134328357,
+      "grad_norm": 0.4828637362156461,
+      "learning_rate": 4.2435989496890054e-05,
+      "loss": 0.5817,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3060867537313433,
+      "grad_norm": 0.45681007096402865,
+      "learning_rate": 4.240351712105422e-05,
+      "loss": 0.5538,
+      "step": 2625
+    },
+    {
+      "epoch": 0.306669776119403,
+      "grad_norm": 0.5633919156991598,
+      "learning_rate": 4.237098934225973e-05,
+      "loss": 0.5619,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3072527985074627,
+      "grad_norm": 0.4744742163372956,
+      "learning_rate": 4.233840628142672e-05,
+      "loss": 0.5749,
+      "step": 2635
+    },
+    {
+      "epoch": 0.30783582089552236,
+      "grad_norm": 0.47221088669237776,
+      "learning_rate": 4.2305768059680806e-05,
+      "loss": 0.5689,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3084188432835821,
+      "grad_norm": 0.4762515982934417,
+      "learning_rate": 4.2273074798352706e-05,
+      "loss": 0.5507,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3090018656716418,
+      "grad_norm": 0.4718217399644616,
+      "learning_rate": 4.22403266189777e-05,
+      "loss": 0.5483,
+      "step": 2650
+    },
+    {
+      "epoch": 0.30958488805970147,
+      "grad_norm": 0.4671780211152603,
+      "learning_rate": 4.2207523643295253e-05,
+      "loss": 0.5463,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3101679104477612,
+      "grad_norm": 0.49004196187956206,
+      "learning_rate": 4.2174665993248505e-05,
+      "loss": 0.5474,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3107509328358209,
+      "grad_norm": 0.5092130863744895,
+      "learning_rate": 4.214175379098388e-05,
+      "loss": 0.5512,
+      "step": 2665
+    },
+    {
+      "epoch": 0.3113339552238806,
+      "grad_norm": 0.4925159889445034,
+      "learning_rate": 4.210878715885056e-05,
+      "loss": 0.5609,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3119169776119403,
+      "grad_norm": 0.5438536521932689,
+      "learning_rate": 4.2075766219400095e-05,
+      "loss": 0.5949,
+      "step": 2675
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.4985765329826545,
+      "learning_rate": 4.20426910953859e-05,
+      "loss": 0.5726,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3130830223880597,
+      "grad_norm": 0.5036063584187528,
+      "learning_rate": 4.200956190976284e-05,
+      "loss": 0.5255,
+      "step": 2685
+    },
+    {
+      "epoch": 0.3136660447761194,
+      "grad_norm": 0.5634159268083127,
+      "learning_rate": 4.1976378785686715e-05,
+      "loss": 0.5254,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3142490671641791,
+      "grad_norm": 0.483875275656377,
+      "learning_rate": 4.1943141846513886e-05,
+      "loss": 0.5518,
+      "step": 2695
+    },
+    {
+      "epoch": 0.3148320895522388,
+      "grad_norm": 0.5510683619678165,
+      "learning_rate": 4.190985121580071e-05,
+      "loss": 0.557,
+      "step": 2700
+    },
+    {
+      "epoch": 0.31541511194029853,
+      "grad_norm": 0.45786711381238016,
+      "learning_rate": 4.18765070173032e-05,
+      "loss": 0.521,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3159981343283582,
+      "grad_norm": 0.47954794442661985,
+      "learning_rate": 4.184310937497647e-05,
+      "loss": 0.55,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3165811567164179,
+      "grad_norm": 0.5348827941441646,
+      "learning_rate": 4.1809658412974314e-05,
+      "loss": 0.5668,
+      "step": 2715
+    },
+    {
+      "epoch": 0.31716417910447764,
+      "grad_norm": 0.5035567889348412,
+      "learning_rate": 4.177615425564872e-05,
+      "loss": 0.5584,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3177472014925373,
+      "grad_norm": 0.5014049010285576,
+      "learning_rate": 4.174259702754947e-05,
+      "loss": 0.5538,
+      "step": 2725
+    },
+    {
+      "epoch": 0.318330223880597,
+      "grad_norm": 0.7108080618041146,
+      "learning_rate": 4.17089868534236e-05,
+      "loss": 0.593,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3189132462686567,
+      "grad_norm": 0.4901344322857948,
+      "learning_rate": 4.1675323858214975e-05,
+      "loss": 0.5504,
+      "step": 2735
+    },
+    {
+      "epoch": 0.31949626865671643,
+      "grad_norm": 0.5455638212871261,
+      "learning_rate": 4.164160816706383e-05,
+      "loss": 0.5781,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3200792910447761,
+      "grad_norm": 0.4691865337288382,
+      "learning_rate": 4.160783990530629e-05,
+      "loss": 0.5349,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3206623134328358,
+      "grad_norm": 0.5397164117756076,
+      "learning_rate": 4.157401919847389e-05,
+      "loss": 0.5635,
+      "step": 2750
+    },
+    {
+      "epoch": 0.32124533582089554,
+      "grad_norm": 0.49203133885978545,
+      "learning_rate": 4.1540146172293154e-05,
+      "loss": 0.553,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3218283582089552,
+      "grad_norm": 0.6956383718262041,
+      "learning_rate": 4.150622095268508e-05,
+      "loss": 0.5454,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3224113805970149,
+      "grad_norm": 0.5236479255481173,
+      "learning_rate": 4.1472243665764715e-05,
+      "loss": 0.546,
+      "step": 2765
+    },
+    {
+      "epoch": 0.32299440298507465,
+      "grad_norm": 0.5226139094665005,
+      "learning_rate": 4.1438214437840625e-05,
+      "loss": 0.5685,
+      "step": 2770
+    },
+    {
+      "epoch": 0.32357742537313433,
+      "grad_norm": 0.519532395995944,
+      "learning_rate": 4.140413339541451e-05,
+      "loss": 0.5621,
+      "step": 2775
+    },
+    {
+      "epoch": 0.324160447761194,
+      "grad_norm": 0.5221117984856967,
+      "learning_rate": 4.137000066518065e-05,
+      "loss": 0.5945,
+      "step": 2780
+    },
+    {
+      "epoch": 0.32474347014925375,
+      "grad_norm": 0.4662764439552845,
+      "learning_rate": 4.13358163740255e-05,
+      "loss": 0.5439,
+      "step": 2785
+    },
+    {
+      "epoch": 0.32532649253731344,
+      "grad_norm": 0.4825709779338081,
+      "learning_rate": 4.1301580649027154e-05,
+      "loss": 0.5487,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3259095149253731,
+      "grad_norm": 0.49595525241615773,
+      "learning_rate": 4.126729361745495e-05,
+      "loss": 0.5456,
+      "step": 2795
+    },
+    {
+      "epoch": 0.32649253731343286,
+      "grad_norm": 0.46259611525800753,
+      "learning_rate": 4.1232955406768925e-05,
+      "loss": 0.5385,
+      "step": 2800
+    },
+    {
+      "epoch": 0.32707555970149255,
+      "grad_norm": 0.48140441921381905,
+      "learning_rate": 4.119856614461938e-05,
+      "loss": 0.5289,
+      "step": 2805
+    },
+    {
+      "epoch": 0.32765858208955223,
+      "grad_norm": 0.4878844018028054,
+      "learning_rate": 4.11641259588464e-05,
+      "loss": 0.5698,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3282416044776119,
+      "grad_norm": 0.521050501093044,
+      "learning_rate": 4.1129634977479375e-05,
+      "loss": 0.5609,
+      "step": 2815
+    },
+    {
+      "epoch": 0.32882462686567165,
+      "grad_norm": 0.49828928878222534,
+      "learning_rate": 4.109509332873653e-05,
+      "loss": 0.5421,
+      "step": 2820
+    },
+    {
+      "epoch": 0.32940764925373134,
+      "grad_norm": 0.5526407883676059,
+      "learning_rate": 4.106050114102443e-05,
+      "loss": 0.5789,
+      "step": 2825
+    },
+    {
+      "epoch": 0.329990671641791,
+      "grad_norm": 0.5516248967750637,
+      "learning_rate": 4.102585854293751e-05,
+      "loss": 0.526,
+      "step": 2830
+    },
+    {
+      "epoch": 0.33057369402985076,
+      "grad_norm": 0.5461716849773546,
+      "learning_rate": 4.0991165663257636e-05,
+      "loss": 0.542,
+      "step": 2835
+    },
+    {
+      "epoch": 0.33115671641791045,
+      "grad_norm": 0.4524230892035543,
+      "learning_rate": 4.095642263095356e-05,
+      "loss": 0.5429,
+      "step": 2840
+    },
+    {
+      "epoch": 0.33173973880597013,
+      "grad_norm": 0.4907608846635581,
+      "learning_rate": 4.0921629575180485e-05,
+      "loss": 0.5536,
+      "step": 2845
+    },
+    {
+      "epoch": 0.33232276119402987,
+      "grad_norm": 0.5284875026264199,
+      "learning_rate": 4.088678662527959e-05,
+      "loss": 0.6164,
+      "step": 2850
+    },
+    {
+      "epoch": 0.33290578358208955,
+      "grad_norm": 0.48638662847344827,
+      "learning_rate": 4.085189391077749e-05,
+      "loss": 0.576,
+      "step": 2855
+    },
+    {
+      "epoch": 0.33348880597014924,
+      "grad_norm": 0.5695349098133777,
+      "learning_rate": 4.0816951561385836e-05,
+      "loss": 0.5521,
+      "step": 2860
+    },
+    {
+      "epoch": 0.334071828358209,
+      "grad_norm": 0.44821729017419637,
+      "learning_rate": 4.078195970700079e-05,
+      "loss": 0.526,
+      "step": 2865
+    },
+    {
+      "epoch": 0.33465485074626866,
+      "grad_norm": 0.5289515772805433,
+      "learning_rate": 4.074691847770251e-05,
+      "loss": 0.5505,
+      "step": 2870
+    },
+    {
+      "epoch": 0.33523787313432835,
+      "grad_norm": 0.5491964687999824,
+      "learning_rate": 4.0711828003754764e-05,
+      "loss": 0.5795,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3358208955223881,
+      "grad_norm": 0.5184535389920294,
+      "learning_rate": 4.067668841560432e-05,
+      "loss": 0.5864,
+      "step": 2880
+    },
+    {
+      "epoch": 0.33640391791044777,
+      "grad_norm": 0.4640726695845034,
+      "learning_rate": 4.064149984388057e-05,
+      "loss": 0.5377,
+      "step": 2885
+    },
+    {
+      "epoch": 0.33698694029850745,
+      "grad_norm": 0.5223555480526277,
+      "learning_rate": 4.060626241939499e-05,
+      "loss": 0.5528,
+      "step": 2890
+    },
+    {
+      "epoch": 0.33756996268656714,
+      "grad_norm": 0.47444935336532806,
+      "learning_rate": 4.057097627314063e-05,
+      "loss": 0.5366,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3381529850746269,
+      "grad_norm": 0.49936360299242055,
+      "learning_rate": 4.0535641536291725e-05,
+      "loss": 0.5981,
+      "step": 2900
+    },
+    {
+      "epoch": 0.33873600746268656,
+      "grad_norm": 0.49567180693590535,
+      "learning_rate": 4.050025834020307e-05,
+      "loss": 0.564,
+      "step": 2905
+    },
+    {
+      "epoch": 0.33931902985074625,
+      "grad_norm": 0.48231305889253284,
+      "learning_rate": 4.046482681640967e-05,
+      "loss": 0.5559,
+      "step": 2910
+    },
+    {
+      "epoch": 0.339902052238806,
+      "grad_norm": 0.5713231807843953,
+      "learning_rate": 4.042934709662613e-05,
+      "loss": 0.6046,
+      "step": 2915
+    },
+    {
+      "epoch": 0.34048507462686567,
+      "grad_norm": 0.5599418550405636,
+      "learning_rate": 4.039381931274626e-05,
+      "loss": 0.5459,
+      "step": 2920
+    },
+    {
+      "epoch": 0.34106809701492535,
+      "grad_norm": 0.5244470856538797,
+      "learning_rate": 4.035824359684253e-05,
+      "loss": 0.5559,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3416511194029851,
+      "grad_norm": 0.46218246493450177,
+      "learning_rate": 4.032262008116559e-05,
+      "loss": 0.5496,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3422341417910448,
+      "grad_norm": 0.5460308556475504,
+      "learning_rate": 4.02869488981438e-05,
+      "loss": 0.5726,
+      "step": 2935
+    },
+    {
+      "epoch": 0.34281716417910446,
+      "grad_norm": 0.4623737344709334,
+      "learning_rate": 4.025123018038271e-05,
+      "loss": 0.5422,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3434001865671642,
+      "grad_norm": 0.5813565724181794,
+      "learning_rate": 4.0215464060664564e-05,
+      "loss": 0.5822,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3439832089552239,
+      "grad_norm": 0.49376149527368923,
+      "learning_rate": 4.017965067194783e-05,
+      "loss": 0.5821,
+      "step": 2950
+    },
+    {
+      "epoch": 0.34456623134328357,
+      "grad_norm": 0.5169325234827248,
+      "learning_rate": 4.0143790147366724e-05,
+      "loss": 0.5482,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3451492537313433,
+      "grad_norm": 0.515737079500457,
+      "learning_rate": 4.010788262023064e-05,
+      "loss": 0.5709,
+      "step": 2960
+    },
+    {
+      "epoch": 0.345732276119403,
+      "grad_norm": 0.5087208722494978,
+      "learning_rate": 4.007192822402372e-05,
+      "loss": 0.5565,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3463152985074627,
+      "grad_norm": 0.4796963642209258,
+      "learning_rate": 4.003592709240438e-05,
+      "loss": 0.5494,
+      "step": 2970
+    },
+    {
+      "epoch": 0.34689832089552236,
+      "grad_norm": 0.49755643832004437,
+      "learning_rate": 3.9999879359204676e-05,
+      "loss": 0.5382,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3474813432835821,
+      "grad_norm": 0.4690304314754959,
+      "learning_rate": 3.996378515843001e-05,
+      "loss": 0.5334,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3480643656716418,
+      "grad_norm": 0.4772189221465707,
+      "learning_rate": 3.9927644624258445e-05,
+      "loss": 0.5902,
+      "step": 2985
+    },
+    {
+      "epoch": 0.34864738805970147,
+      "grad_norm": 0.46404845790530064,
+      "learning_rate": 3.989145789104033e-05,
+      "loss": 0.5431,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3492304104477612,
+      "grad_norm": 0.5340362988246943,
+      "learning_rate": 3.985522509329775e-05,
+      "loss": 0.5486,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3498134328358209,
+      "grad_norm": 0.510771920504794,
+      "learning_rate": 3.9818946365724004e-05,
+      "loss": 0.5401,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3503964552238806,
+      "grad_norm": 0.49718158294490394,
+      "learning_rate": 3.978262184318317e-05,
+      "loss": 0.5626,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3509794776119403,
+      "grad_norm": 0.5158633729934479,
+      "learning_rate": 3.974625166070953e-05,
+      "loss": 0.5364,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3515625,
+      "grad_norm": 0.5204824836614376,
+      "learning_rate": 3.970983595350714e-05,
+      "loss": 0.5545,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3521455223880597,
+      "grad_norm": 0.506085761814899,
+      "learning_rate": 3.967337485694929e-05,
+      "loss": 0.5492,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3527285447761194,
+      "grad_norm": 0.5167191975821872,
+      "learning_rate": 3.963686850657795e-05,
+      "loss": 0.5326,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3533115671641791,
+      "grad_norm": 0.49528479007831655,
+      "learning_rate": 3.9600317038103385e-05,
+      "loss": 0.548,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3538945895522388,
+      "grad_norm": 0.514405428467145,
+      "learning_rate": 3.956372058740354e-05,
+      "loss": 0.5708,
+      "step": 3035
+    },
+    {
+      "epoch": 0.35447761194029853,
+      "grad_norm": 0.4760653718744451,
+      "learning_rate": 3.952707929052359e-05,
+      "loss": 0.5385,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3550606343283582,
+      "grad_norm": 0.5086645138695323,
+      "learning_rate": 3.9490393283675445e-05,
+      "loss": 0.5425,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3556436567164179,
+      "grad_norm": 0.48138382467030305,
+      "learning_rate": 3.9453662703237186e-05,
+      "loss": 0.5599,
+      "step": 3050
+    },
+    {
+      "epoch": 0.35622667910447764,
+      "grad_norm": 0.594129469541421,
+      "learning_rate": 3.941688768575261e-05,
+      "loss": 0.558,
+      "step": 3055
+    },
+    {
+      "epoch": 0.3568097014925373,
+      "grad_norm": 0.4896479074572315,
+      "learning_rate": 3.938006836793073e-05,
+      "loss": 0.5399,
+      "step": 3060
+    },
+    {
+      "epoch": 0.357392723880597,
+      "grad_norm": 0.5430860896504877,
+      "learning_rate": 3.934320488664519e-05,
+      "loss": 0.5614,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3579757462686567,
+      "grad_norm": 0.5171549846963465,
+      "learning_rate": 3.9306297378933855e-05,
+      "loss": 0.5338,
+      "step": 3070
+    },
+    {
+      "epoch": 0.35855876865671643,
+      "grad_norm": 0.5030297563055794,
+      "learning_rate": 3.926934598199824e-05,
+      "loss": 0.5671,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3591417910447761,
+      "grad_norm": 0.4738604088361965,
+      "learning_rate": 3.923235083320301e-05,
+      "loss": 0.5204,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3597248134328358,
+      "grad_norm": 0.5081725848955316,
+      "learning_rate": 3.919531207007548e-05,
+      "loss": 0.5382,
+      "step": 3085
+    },
+    {
+      "epoch": 0.36030783582089554,
+      "grad_norm": 0.5336505974453601,
+      "learning_rate": 3.915822983030512e-05,
+      "loss": 0.5498,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3608908582089552,
+      "grad_norm": 0.45862890034907244,
+      "learning_rate": 3.912110425174296e-05,
+      "loss": 0.5478,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3614738805970149,
+      "grad_norm": 0.5692817255847623,
+      "learning_rate": 3.9083935472401214e-05,
+      "loss": 0.5511,
+      "step": 3100
+    },
+    {
+      "epoch": 0.36205690298507465,
+      "grad_norm": 0.5588630515105258,
+      "learning_rate": 3.904672363045265e-05,
+      "loss": 0.5713,
+      "step": 3105
+    },
+    {
+      "epoch": 0.36263992537313433,
+      "grad_norm": 0.7882164929530883,
+      "learning_rate": 3.900946886423012e-05,
+      "loss": 0.5967,
+      "step": 3110
+    },
+    {
+      "epoch": 0.363222947761194,
+      "grad_norm": 0.44828952211943274,
+      "learning_rate": 3.897217131222606e-05,
+      "loss": 0.5117,
+      "step": 3115
+    },
+    {
+      "epoch": 0.36380597014925375,
+      "grad_norm": 0.492899081711114,
+      "learning_rate": 3.893483111309196e-05,
+      "loss": 0.5381,
+      "step": 3120
+    },
+    {
+      "epoch": 0.36438899253731344,
+      "grad_norm": 0.46780156460670047,
+      "learning_rate": 3.889744840563781e-05,
+      "loss": 0.5561,
+      "step": 3125
+    },
+    {
+      "epoch": 0.3649720149253731,
+      "grad_norm": 0.5062858013949219,
+      "learning_rate": 3.886002332883169e-05,
+      "loss": 0.5501,
+      "step": 3130
+    },
+    {
+      "epoch": 0.36555503731343286,
+      "grad_norm": 0.4693101482730097,
+      "learning_rate": 3.8822556021799114e-05,
+      "loss": 0.5193,
+      "step": 3135
+    },
+    {
+      "epoch": 0.36613805970149255,
+      "grad_norm": 0.5060844807583814,
+      "learning_rate": 3.878504662382264e-05,
+      "loss": 0.5532,
+      "step": 3140
+    },
+    {
+      "epoch": 0.36672108208955223,
+      "grad_norm": 0.5201363699624885,
+      "learning_rate": 3.8747495274341274e-05,
+      "loss": 0.5845,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3673041044776119,
+      "grad_norm": 0.5073064897043211,
+      "learning_rate": 3.870990211294997e-05,
+      "loss": 0.5444,
+      "step": 3150
+    },
+    {
+      "epoch": 0.36788712686567165,
+      "grad_norm": 0.5344016844623599,
+      "learning_rate": 3.867226727939912e-05,
+      "loss": 0.5606,
+      "step": 3155
+    },
+    {
+      "epoch": 0.36847014925373134,
+      "grad_norm": 0.5139307938716214,
+      "learning_rate": 3.863459091359401e-05,
+      "loss": 0.5882,
+      "step": 3160
+    },
+    {
+      "epoch": 0.369053171641791,
+      "grad_norm": 0.4425759816838131,
+      "learning_rate": 3.8596873155594385e-05,
+      "loss": 0.5202,
+      "step": 3165
+    },
+    {
+      "epoch": 0.36963619402985076,
+      "grad_norm": 0.5742393732533151,
+      "learning_rate": 3.855911414561378e-05,
+      "loss": 0.5662,
+      "step": 3170
+    },
+    {
+      "epoch": 0.37021921641791045,
+      "grad_norm": 0.44175421568947043,
+      "learning_rate": 3.852131402401914e-05,
+      "loss": 0.4854,
+      "step": 3175
+    },
+    {
+      "epoch": 0.37080223880597013,
+      "grad_norm": 0.4984055964239209,
+      "learning_rate": 3.848347293133021e-05,
+      "loss": 0.573,
+      "step": 3180
+    },
+    {
+      "epoch": 0.37138526119402987,
+      "grad_norm": 0.49757126890693293,
+      "learning_rate": 3.844559100821906e-05,
+      "loss": 0.5549,
+      "step": 3185
+    },
+    {
+      "epoch": 0.37196828358208955,
+      "grad_norm": 0.57357388001981,
+      "learning_rate": 3.8407668395509526e-05,
+      "loss": 0.5574,
+      "step": 3190
+    },
+    {
+      "epoch": 0.37255130597014924,
+      "grad_norm": 0.5019623696079913,
+      "learning_rate": 3.8369705234176726e-05,
+      "loss": 0.5894,
+      "step": 3195
+    },
+    {
+      "epoch": 0.373134328358209,
+      "grad_norm": 0.4858155386150297,
+      "learning_rate": 3.8331701665346495e-05,
+      "loss": 0.5383,
+      "step": 3200
+    },
+    {
+      "epoch": 0.37371735074626866,
+      "grad_norm": 0.5624113035875654,
+      "learning_rate": 3.829365783029492e-05,
+      "loss": 0.5585,
+      "step": 3205
+    },
+    {
+      "epoch": 0.37430037313432835,
+      "grad_norm": 0.5052807550374429,
+      "learning_rate": 3.8255573870447716e-05,
+      "loss": 0.5439,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3748833955223881,
+      "grad_norm": 0.5085388458884429,
+      "learning_rate": 3.82174499273798e-05,
+      "loss": 0.565,
+      "step": 3215
+    },
+    {
+      "epoch": 0.37546641791044777,
+      "grad_norm": 0.4989534046655282,
+      "learning_rate": 3.817928614281471e-05,
+      "loss": 0.5324,
+      "step": 3220
+    },
+    {
+      "epoch": 0.37604944029850745,
+      "grad_norm": 0.8403204163889687,
+      "learning_rate": 3.8141082658624106e-05,
+      "loss": 0.5626,
+      "step": 3225
+    },
+    {
+      "epoch": 0.37663246268656714,
+      "grad_norm": 0.5559098643501764,
+      "learning_rate": 3.810283961682722e-05,
+      "loss": 0.5583,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3772154850746269,
+      "grad_norm": 0.47644518540184416,
+      "learning_rate": 3.806455715959032e-05,
+      "loss": 0.5233,
+      "step": 3235
+    },
+    {
+      "epoch": 0.37779850746268656,
+      "grad_norm": 0.4930068420607561,
+      "learning_rate": 3.8026235429226236e-05,
+      "loss": 0.5642,
+      "step": 3240
+    },
+    {
+      "epoch": 0.37838152985074625,
+      "grad_norm": 0.5290426787630793,
+      "learning_rate": 3.798787456819377e-05,
+      "loss": 0.5234,
+      "step": 3245
+    },
+    {
+      "epoch": 0.378964552238806,
+      "grad_norm": 0.4868258441116788,
+      "learning_rate": 3.794947471909719e-05,
+      "loss": 0.5395,
+      "step": 3250
+    },
+    {
+      "epoch": 0.37954757462686567,
+      "grad_norm": 0.47174968576411624,
+      "learning_rate": 3.791103602468569e-05,
+      "loss": 0.5395,
+      "step": 3255
+    },
+    {
+      "epoch": 0.38013059701492535,
+      "grad_norm": 0.5363107908185231,
+      "learning_rate": 3.7872558627852905e-05,
+      "loss": 0.5557,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3807136194029851,
+      "grad_norm": 0.49188998699814623,
+      "learning_rate": 3.78340426716363e-05,
+      "loss": 0.5268,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3812966417910448,
+      "grad_norm": 0.48927529818476645,
+      "learning_rate": 3.779548829921673e-05,
+      "loss": 0.5387,
+      "step": 3270
+    },
+    {
+      "epoch": 0.38187966417910446,
+      "grad_norm": 0.47364382567752494,
+      "learning_rate": 3.775689565391781e-05,
+      "loss": 0.5129,
+      "step": 3275
+    },
+    {
+      "epoch": 0.3824626865671642,
+      "grad_norm": 0.5126016691334612,
+      "learning_rate": 3.771826487920546e-05,
+      "loss": 0.5467,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3830457089552239,
+      "grad_norm": 0.5430085481644645,
+      "learning_rate": 3.767959611868734e-05,
+      "loss": 0.5536,
+      "step": 3285
+    },
+    {
+      "epoch": 0.38362873134328357,
+      "grad_norm": 0.4763687695806715,
+      "learning_rate": 3.764088951611233e-05,
+      "loss": 0.5071,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3842117537313433,
+      "grad_norm": 0.47538986078328227,
+      "learning_rate": 3.7602145215369965e-05,
+      "loss": 0.5431,
+      "step": 3295
+    },
+    {
+      "epoch": 0.384794776119403,
+      "grad_norm": 0.44993211618209894,
+      "learning_rate": 3.756336336048994e-05,
+      "loss": 0.5608,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3853777985074627,
+      "grad_norm": 0.4857510707261703,
+      "learning_rate": 3.752454409564152e-05,
+      "loss": 0.5307,
+      "step": 3305
+    },
+    {
+      "epoch": 0.38596082089552236,
+      "grad_norm": 0.49584291918089757,
+      "learning_rate": 3.74856875651331e-05,
+      "loss": 0.6047,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3865438432835821,
+      "grad_norm": 0.5136523336548113,
+      "learning_rate": 3.744679391341157e-05,
+      "loss": 0.5516,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3871268656716418,
+      "grad_norm": 0.5181136974711912,
+      "learning_rate": 3.740786328506179e-05,
+      "loss": 0.5439,
+      "step": 3320
+    },
+    {
+      "epoch": 0.38770988805970147,
+      "grad_norm": 0.5089785637081085,
+      "learning_rate": 3.7368895824806146e-05,
+      "loss": 0.5585,
+      "step": 3325
+    },
+    {
+      "epoch": 0.3882929104477612,
+      "grad_norm": 0.49682305539963434,
+      "learning_rate": 3.732989167750388e-05,
+      "loss": 0.562,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3888759328358209,
+      "grad_norm": 0.5243718121493276,
+      "learning_rate": 3.7290850988150644e-05,
+      "loss": 0.5851,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3894589552238806,
+      "grad_norm": 0.5098618228761496,
+      "learning_rate": 3.7251773901877945e-05,
+      "loss": 0.5191,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3900419776119403,
+      "grad_norm": 0.47672882853039394,
+      "learning_rate": 3.721266056395257e-05,
+      "loss": 0.5284,
+      "step": 3345
+    },
+    {
+      "epoch": 0.390625,
+      "grad_norm": 0.4822727732005445,
+      "learning_rate": 3.7173511119776086e-05,
+      "loss": 0.5794,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3912080223880597,
+      "grad_norm": 0.5222854783235708,
+      "learning_rate": 3.713432571488427e-05,
+      "loss": 0.5391,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3917910447761194,
+      "grad_norm": 7.9962975681807595,
+      "learning_rate": 3.70951044949466e-05,
+      "loss": 0.5534,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3923740671641791,
+      "grad_norm": 0.4764348433947033,
+      "learning_rate": 3.705584760576566e-05,
+      "loss": 0.5452,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3929570895522388,
+      "grad_norm": 0.49080831286781906,
+      "learning_rate": 3.7016555193276667e-05,
+      "loss": 0.5746,
+      "step": 3370
+    },
+    {
+      "epoch": 0.39354011194029853,
+      "grad_norm": 0.5101400424865905,
+      "learning_rate": 3.697722740354688e-05,
+      "loss": 0.5729,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3941231343283582,
+      "grad_norm": 0.4479587001588852,
+      "learning_rate": 3.6937864382775076e-05,
+      "loss": 0.5447,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3947061567164179,
+      "grad_norm": 0.48466656620240023,
+      "learning_rate": 3.689846627729098e-05,
+      "loss": 0.5327,
+      "step": 3385
+    },
+    {
+      "epoch": 0.39528917910447764,
+      "grad_norm": 0.5157423888893855,
+      "learning_rate": 3.685903323355477e-05,
+      "loss": 0.5381,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3958722014925373,
+      "grad_norm": 0.4820557574738787,
+      "learning_rate": 3.68195653981565e-05,
+      "loss": 0.5183,
+      "step": 3395
+    },
+    {
+      "epoch": 0.396455223880597,
+      "grad_norm": 0.48067690178778916,
+      "learning_rate": 3.678006291781555e-05,
+      "loss": 0.6121,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3970382462686567,
+      "grad_norm": 0.48140274545401546,
+      "learning_rate": 3.6740525939380084e-05,
+      "loss": 0.5763,
+      "step": 3405
+    },
+    {
+      "epoch": 0.39762126865671643,
+      "grad_norm": 0.4821460133421974,
+      "learning_rate": 3.6700954609826535e-05,
+      "loss": 0.5498,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3982042910447761,
+      "grad_norm": 0.5062442615898354,
+      "learning_rate": 3.6661349076259015e-05,
+      "loss": 0.5301,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3987873134328358,
+      "grad_norm": 0.4894842886999361,
+      "learning_rate": 3.662170948590879e-05,
+      "loss": 0.5647,
+      "step": 3420
+    },
+    {
+      "epoch": 0.39937033582089554,
+      "grad_norm": 0.4635809485827499,
+      "learning_rate": 3.658203598613375e-05,
+      "loss": 0.5658,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3999533582089552,
+      "grad_norm": 0.4756142318658808,
+      "learning_rate": 3.65423287244178e-05,
+      "loss": 0.5434,
+      "step": 3430
+    },
+    {
+      "epoch": 0.4005363805970149,
+      "grad_norm": 0.4768793799786962,
+      "learning_rate": 3.6502587848370395e-05,
+      "loss": 0.5129,
+      "step": 3435
+    },
+    {
+      "epoch": 0.40111940298507465,
+      "grad_norm": 0.49735047427469553,
+      "learning_rate": 3.6462813505725915e-05,
+      "loss": 0.5693,
+      "step": 3440
+    },
+    {
+      "epoch": 0.40170242537313433,
+      "grad_norm": 0.46768242382881264,
+      "learning_rate": 3.642300584434319e-05,
+      "loss": 0.5424,
+      "step": 3445
+    },
+    {
+      "epoch": 0.402285447761194,
+      "grad_norm": 0.48929227885921284,
+      "learning_rate": 3.638316501220487e-05,
+      "loss": 0.5613,
+      "step": 3450
+    },
+    {
+      "epoch": 0.40286847014925375,
+      "grad_norm": 0.49035092965063865,
+      "learning_rate": 3.6343291157416937e-05,
+      "loss": 0.5747,
+      "step": 3455
+    },
+    {
+      "epoch": 0.40345149253731344,
+      "grad_norm": 0.4920873718095396,
+      "learning_rate": 3.630338442820814e-05,
+      "loss": 0.5621,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4040345149253731,
+      "grad_norm": 0.5721821448291733,
+      "learning_rate": 3.6263444972929395e-05,
+      "loss": 0.5642,
+      "step": 3465
+    },
+    {
+      "epoch": 0.40461753731343286,
+      "grad_norm": 0.4900664650078509,
+      "learning_rate": 3.622347294005334e-05,
+      "loss": 0.5637,
+      "step": 3470
+    },
+    {
+      "epoch": 0.40520055970149255,
+      "grad_norm": 0.4784540540109224,
+      "learning_rate": 3.618346847817366e-05,
+      "loss": 0.5166,
+      "step": 3475
+    },
+    {
+      "epoch": 0.40578358208955223,
+      "grad_norm": 0.5033729998350225,
+      "learning_rate": 3.6143431736004636e-05,
+      "loss": 0.5604,
+      "step": 3480
+    },
+    {
+      "epoch": 0.4063666044776119,
+      "grad_norm": 0.47667240088779256,
+      "learning_rate": 3.610336286238051e-05,
+      "loss": 0.5492,
+      "step": 3485
+    },
+    {
+      "epoch": 0.40694962686567165,
+      "grad_norm": 0.4874513651352348,
+      "learning_rate": 3.6063262006255006e-05,
+      "loss": 0.5563,
+      "step": 3490
+    },
+    {
+      "epoch": 0.40753264925373134,
+      "grad_norm": 0.523091113003475,
+      "learning_rate": 3.602312931670073e-05,
+      "loss": 0.5488,
+      "step": 3495
+    },
+    {
+      "epoch": 0.408115671641791,
+      "grad_norm": 0.5531785395812051,
+      "learning_rate": 3.59829649429086e-05,
+      "loss": 0.6013,
+      "step": 3500
+    },
+    {
+      "epoch": 0.40869869402985076,
+      "grad_norm": 0.47185275980717517,
+      "learning_rate": 3.5942769034187354e-05,
+      "loss": 0.536,
+      "step": 3505
+    },
+    {
+      "epoch": 0.40928171641791045,
+      "grad_norm": 0.46288526087001597,
+      "learning_rate": 3.590254173996295e-05,
+      "loss": 0.5169,
+      "step": 3510
+    },
+    {
+      "epoch": 0.40986473880597013,
+      "grad_norm": 0.48524872738357405,
+      "learning_rate": 3.586228320977801e-05,
+      "loss": 0.5271,
+      "step": 3515
+    },
+    {
+      "epoch": 0.41044776119402987,
+      "grad_norm": 0.5450240726382458,
+      "learning_rate": 3.582199359329129e-05,
+      "loss": 0.5427,
+      "step": 3520
+    },
+    {
+      "epoch": 0.41103078358208955,
+      "grad_norm": 0.48693873450848696,
+      "learning_rate": 3.5781673040277084e-05,
+      "loss": 0.5088,
+      "step": 3525
+    },
+    {
+      "epoch": 0.41161380597014924,
+      "grad_norm": 0.46864955154362703,
+      "learning_rate": 3.5741321700624726e-05,
+      "loss": 0.5174,
+      "step": 3530
+    },
+    {
+      "epoch": 0.412196828358209,
+      "grad_norm": 0.46728302060083887,
+      "learning_rate": 3.570093972433794e-05,
+      "loss": 0.5687,
+      "step": 3535
+    },
+    {
+      "epoch": 0.41277985074626866,
+      "grad_norm": 0.4634797945645771,
+      "learning_rate": 3.56605272615344e-05,
+      "loss": 0.528,
+      "step": 3540
+    },
+    {
+      "epoch": 0.41336287313432835,
+      "grad_norm": 0.5238785746405048,
+      "learning_rate": 3.562008446244509e-05,
+      "loss": 0.5375,
+      "step": 3545
+    },
+    {
+      "epoch": 0.4139458955223881,
+      "grad_norm": 0.5187794803590449,
+      "learning_rate": 3.557961147741376e-05,
+      "loss": 0.5985,
+      "step": 3550
+    },
+    {
+      "epoch": 0.41452891791044777,
+      "grad_norm": 0.4379680762591855,
+      "learning_rate": 3.553910845689638e-05,
+      "loss": 0.5554,
+      "step": 3555
+    },
+    {
+      "epoch": 0.41511194029850745,
+      "grad_norm": 0.46577740202461215,
+      "learning_rate": 3.549857555146056e-05,
+      "loss": 0.5472,
+      "step": 3560
+    },
+    {
+      "epoch": 0.41569496268656714,
+      "grad_norm": 0.4789932179858049,
+      "learning_rate": 3.5458012911785036e-05,
+      "loss": 0.5451,
+      "step": 3565
+    },
+    {
+      "epoch": 0.4162779850746269,
+      "grad_norm": 0.49654652817675987,
+      "learning_rate": 3.541742068865907e-05,
+      "loss": 0.5513,
+      "step": 3570
+    },
+    {
+      "epoch": 0.41686100746268656,
+      "grad_norm": 0.4692894243690009,
+      "learning_rate": 3.537679903298187e-05,
+      "loss": 0.5185,
+      "step": 3575
+    },
+    {
+      "epoch": 0.41744402985074625,
+      "grad_norm": 0.47941339153241536,
+      "learning_rate": 3.53361480957621e-05,
+      "loss": 0.5692,
+      "step": 3580
+    },
+    {
+      "epoch": 0.418027052238806,
+      "grad_norm": 0.46544865766435817,
+      "learning_rate": 3.529546802811725e-05,
+      "loss": 0.5405,
+      "step": 3585
+    },
+    {
+      "epoch": 0.41861007462686567,
+      "grad_norm": 0.46403588392065775,
+      "learning_rate": 3.5254758981273106e-05,
+      "loss": 0.5437,
+      "step": 3590
+    },
+    {
+      "epoch": 0.41919309701492535,
+      "grad_norm": 0.5179906749611947,
+      "learning_rate": 3.521402110656318e-05,
+      "loss": 0.5593,
+      "step": 3595
+    },
+    {
+      "epoch": 0.4197761194029851,
+      "grad_norm": 0.4781025455029735,
+      "learning_rate": 3.517325455542815e-05,
+      "loss": 0.5498,
+      "step": 3600
+    },
+    {
+      "epoch": 0.4203591417910448,
+      "grad_norm": 0.4934656687592932,
+      "learning_rate": 3.513245947941531e-05,
+      "loss": 0.5215,
+      "step": 3605
+    },
+    {
+      "epoch": 0.42094216417910446,
+      "grad_norm": 0.5326144345367541,
+      "learning_rate": 3.5091636030177995e-05,
+      "loss": 0.535,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4215251865671642,
+      "grad_norm": 0.47801066829681754,
+      "learning_rate": 3.505078435947498e-05,
+      "loss": 0.5137,
+      "step": 3615
+    },
+    {
+      "epoch": 0.4221082089552239,
+      "grad_norm": 0.5286215614880156,
+      "learning_rate": 3.500990461916998e-05,
+      "loss": 0.5733,
+      "step": 3620
+    },
+    {
+      "epoch": 0.42269123134328357,
+      "grad_norm": 0.5073258841460547,
+      "learning_rate": 3.496899696123107e-05,
+      "loss": 0.5746,
+      "step": 3625
+    },
+    {
+      "epoch": 0.4232742537313433,
+      "grad_norm": 0.48126213143188057,
+      "learning_rate": 3.492806153773007e-05,
+      "loss": 0.559,
+      "step": 3630
+    },
+    {
+      "epoch": 0.423857276119403,
+      "grad_norm": 0.43727021104667335,
+      "learning_rate": 3.488709850084206e-05,
+      "loss": 0.4998,
+      "step": 3635
+    },
+    {
+      "epoch": 0.4244402985074627,
+      "grad_norm": 0.4897298712066961,
+      "learning_rate": 3.484610800284473e-05,
+      "loss": 0.5463,
+      "step": 3640
+    },
+    {
+      "epoch": 0.42502332089552236,
+      "grad_norm": 0.523356089677995,
+      "learning_rate": 3.480509019611788e-05,
+      "loss": 0.5659,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4256063432835821,
+      "grad_norm": 0.5277720573146358,
+      "learning_rate": 3.476404523314282e-05,
+      "loss": 0.5241,
+      "step": 3650
+    },
+    {
+      "epoch": 0.4261893656716418,
+      "grad_norm": 0.526635621277189,
+      "learning_rate": 3.472297326650183e-05,
+      "loss": 0.5543,
+      "step": 3655
+    },
+    {
+      "epoch": 0.42677238805970147,
+      "grad_norm": 0.45989723897065593,
+      "learning_rate": 3.468187444887754e-05,
+      "loss": 0.4939,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4273554104477612,
+      "grad_norm": 0.4622944623408051,
+      "learning_rate": 3.464074893305242e-05,
+      "loss": 0.5297,
+      "step": 3665
+    },
+    {
+      "epoch": 0.4279384328358209,
+      "grad_norm": 0.4783006627278798,
+      "learning_rate": 3.45995968719082e-05,
+      "loss": 0.5255,
+      "step": 3670
+    },
+    {
+      "epoch": 0.4285214552238806,
+      "grad_norm": 0.5059594374936409,
+      "learning_rate": 3.455841841842524e-05,
+      "loss": 0.5773,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4291044776119403,
+      "grad_norm": 0.44346639279558187,
+      "learning_rate": 3.4517213725682085e-05,
+      "loss": 0.5196,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4296875,
+      "grad_norm": 0.4742923141132099,
+      "learning_rate": 3.447598294685476e-05,
+      "loss": 0.5211,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4302705223880597,
+      "grad_norm": 0.5096994312042691,
+      "learning_rate": 3.443472623521631e-05,
+      "loss": 0.5576,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4308535447761194,
+      "grad_norm": 0.4757579513441239,
+      "learning_rate": 3.4393443744136136e-05,
+      "loss": 0.5342,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4314365671641791,
+      "grad_norm": 0.4525723387319913,
+      "learning_rate": 3.435213562707953e-05,
+      "loss": 0.521,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4320195895522388,
+      "grad_norm": 0.4655324458390628,
+      "learning_rate": 3.431080203760699e-05,
+      "loss": 0.5143,
+      "step": 3705
+    },
+    {
+      "epoch": 0.43260261194029853,
+      "grad_norm": 0.46493164741995624,
+      "learning_rate": 3.426944312937376e-05,
+      "loss": 0.5448,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4331856343283582,
+      "grad_norm": 0.4705834455942846,
+      "learning_rate": 3.422805905612914e-05,
+      "loss": 0.5132,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4337686567164179,
+      "grad_norm": 0.44735510242287085,
+      "learning_rate": 3.4186649971716044e-05,
+      "loss": 0.5078,
+      "step": 3720
+    },
+    {
+      "epoch": 0.43435167910447764,
+      "grad_norm": 0.4248977052828472,
+      "learning_rate": 3.4145216030070344e-05,
+      "loss": 0.5224,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4349347014925373,
+      "grad_norm": 0.48361810889787654,
+      "learning_rate": 3.410375738522028e-05,
+      "loss": 0.5696,
+      "step": 3730
+    },
+    {
+      "epoch": 0.435517723880597,
+      "grad_norm": 0.5169029233248057,
+      "learning_rate": 3.406227419128596e-05,
+      "loss": 0.548,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4361007462686567,
+      "grad_norm": 0.47352743387761187,
+      "learning_rate": 3.402076660247878e-05,
+      "loss": 0.5735,
+      "step": 3740
+    },
+    {
+      "epoch": 0.43668376865671643,
+      "grad_norm": 0.4598793206254637,
+      "learning_rate": 3.397923477310074e-05,
+      "loss": 0.5286,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4372667910447761,
+      "grad_norm": 0.7329855818387587,
+      "learning_rate": 3.393767885754405e-05,
+      "loss": 0.5274,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4378498134328358,
+      "grad_norm": 0.5374277209323949,
+      "learning_rate": 3.389609901029038e-05,
+      "loss": 0.569,
+      "step": 3755
+    },
+    {
+      "epoch": 0.43843283582089554,
+      "grad_norm": 0.4851132196571775,
+      "learning_rate": 3.38544953859104e-05,
+      "loss": 0.5658,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4390158582089552,
+      "grad_norm": 0.4588569006808473,
+      "learning_rate": 3.381286813906317e-05,
+      "loss": 0.5463,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4395988805970149,
+      "grad_norm": 0.49195337073823053,
+      "learning_rate": 3.3771217424495555e-05,
+      "loss": 0.5855,
+      "step": 3770
+    },
+    {
+      "epoch": 0.44018190298507465,
+      "grad_norm": 0.47991606249122304,
+      "learning_rate": 3.372954339704167e-05,
+      "loss": 0.5496,
+      "step": 3775
+    },
+    {
+      "epoch": 0.44076492537313433,
+      "grad_norm": 0.4690325622228386,
+      "learning_rate": 3.368784621162229e-05,
+      "loss": 0.5647,
+      "step": 3780
+    },
+    {
+      "epoch": 0.441347947761194,
+      "grad_norm": 0.5399827946397292,
+      "learning_rate": 3.364612602324429e-05,
+      "loss": 0.554,
+      "step": 3785
+    },
+    {
+      "epoch": 0.44193097014925375,
+      "grad_norm": 0.44956816879532197,
+      "learning_rate": 3.3604382987000016e-05,
+      "loss": 0.5033,
+      "step": 3790
+    },
+    {
+      "epoch": 0.44251399253731344,
+      "grad_norm": 0.49304141117786027,
+      "learning_rate": 3.356261725806681e-05,
+      "loss": 0.5651,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4430970149253731,
+      "grad_norm": 0.4984710533841555,
+      "learning_rate": 3.352082899170631e-05,
+      "loss": 0.5238,
+      "step": 3800
+    },
+    {
+      "epoch": 0.44368003731343286,
+      "grad_norm": 0.5007565543332921,
+      "learning_rate": 3.3479018343264e-05,
+      "loss": 0.5519,
+      "step": 3805
+    },
+    {
+      "epoch": 0.44426305970149255,
+      "grad_norm": 0.5011462730265147,
+      "learning_rate": 3.343718546816852e-05,
+      "loss": 0.5523,
+      "step": 3810
+    },
+    {
+      "epoch": 0.44484608208955223,
+      "grad_norm": 0.49568146210324576,
+      "learning_rate": 3.339533052193114e-05,
+      "loss": 0.5132,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4454291044776119,
+      "grad_norm": 0.45124156599655735,
+      "learning_rate": 3.335345366014522e-05,
+      "loss": 0.55,
+      "step": 3820
+    },
+    {
+      "epoch": 0.44601212686567165,
+      "grad_norm": 0.47223468640192207,
+      "learning_rate": 3.331155503848553e-05,
+      "loss": 0.5438,
+      "step": 3825
+    },
+    {
+      "epoch": 0.44659514925373134,
+      "grad_norm": 0.5287373716087295,
+      "learning_rate": 3.326963481270778e-05,
+      "loss": 0.5411,
+      "step": 3830
+    },
+    {
+      "epoch": 0.447178171641791,
+      "grad_norm": 0.44674599573765417,
+      "learning_rate": 3.322769313864796e-05,
+      "loss": 0.5363,
+      "step": 3835
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.4434254783508246,
+      "learning_rate": 3.3185730172221814e-05,
+      "loss": 0.537,
+      "step": 3840
+    },
+    {
+      "epoch": 0.44834421641791045,
+      "grad_norm": 0.5046794647102569,
+      "learning_rate": 3.3143746069424215e-05,
+      "loss": 0.5582,
+      "step": 3845
+    },
+    {
+      "epoch": 0.44892723880597013,
+      "grad_norm": 0.489005919425733,
+      "learning_rate": 3.310174098632865e-05,
+      "loss": 0.5549,
+      "step": 3850
+    },
+    {
+      "epoch": 0.44951026119402987,
+      "grad_norm": 0.49266732159431287,
+      "learning_rate": 3.305971507908655e-05,
+      "loss": 0.5313,
+      "step": 3855
+    },
+    {
+      "epoch": 0.45009328358208955,
+      "grad_norm": 0.5315237555379242,
+      "learning_rate": 3.301766850392681e-05,
+      "loss": 0.5581,
+      "step": 3860
+    },
+    {
+      "epoch": 0.45067630597014924,
+      "grad_norm": 0.46700902921771087,
+      "learning_rate": 3.29756014171551e-05,
+      "loss": 0.5371,
+      "step": 3865
+    },
+    {
+      "epoch": 0.451259328358209,
+      "grad_norm": 0.4395133061533244,
+      "learning_rate": 3.2933513975153384e-05,
+      "loss": 0.5399,
+      "step": 3870
+    },
+    {
+      "epoch": 0.45184235074626866,
+      "grad_norm": 0.485910728377908,
+      "learning_rate": 3.2891406334379285e-05,
+      "loss": 0.525,
+      "step": 3875
+    },
+    {
+      "epoch": 0.45242537313432835,
+      "grad_norm": 0.46551066433925664,
+      "learning_rate": 3.284927865136551e-05,
+      "loss": 0.4913,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4530083955223881,
+      "grad_norm": 0.4792517888689515,
+      "learning_rate": 3.280713108271926e-05,
+      "loss": 0.5237,
+      "step": 3885
+    },
+    {
+      "epoch": 0.45359141791044777,
+      "grad_norm": 0.44580552095045006,
+      "learning_rate": 3.276496378512168e-05,
+      "loss": 0.5716,
+      "step": 3890
+    },
+    {
+      "epoch": 0.45417444029850745,
+      "grad_norm": 0.46092489328912967,
+      "learning_rate": 3.272277691532725e-05,
+      "loss": 0.5402,
+      "step": 3895
+    },
+    {
+      "epoch": 0.45475746268656714,
+      "grad_norm": 0.4413897932753195,
+      "learning_rate": 3.268057063016319e-05,
+      "loss": 0.5305,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4553404850746269,
+      "grad_norm": 0.4782676785091174,
+      "learning_rate": 3.263834508652894e-05,
+      "loss": 0.5946,
+      "step": 3905
+    },
+    {
+      "epoch": 0.45592350746268656,
+      "grad_norm": 0.5238873675452039,
+      "learning_rate": 3.259610044139548e-05,
+      "loss": 0.5478,
+      "step": 3910
+    },
+    {
+      "epoch": 0.45650652985074625,
+      "grad_norm": 0.5182499166196407,
+      "learning_rate": 3.255383685180484e-05,
+      "loss": 0.5454,
+      "step": 3915
+    },
+    {
+      "epoch": 0.457089552238806,
+      "grad_norm": 0.47638810180727925,
+      "learning_rate": 3.251155447486945e-05,
+      "loss": 0.5446,
+      "step": 3920
+    },
+    {
+      "epoch": 0.45767257462686567,
+      "grad_norm": 0.4523229716036492,
+      "learning_rate": 3.246925346777158e-05,
+      "loss": 0.522,
+      "step": 3925
+    },
+    {
+      "epoch": 0.45825559701492535,
+      "grad_norm": 0.46909930899902835,
+      "learning_rate": 3.2426933987762785e-05,
+      "loss": 0.5393,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4588386194029851,
+      "grad_norm": 0.43989917162010367,
+      "learning_rate": 3.238459619216326e-05,
+      "loss": 0.5211,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4594216417910448,
+      "grad_norm": 0.48026394452289217,
+      "learning_rate": 3.23422402383613e-05,
+      "loss": 0.5547,
+      "step": 3940
+    },
+    {
+      "epoch": 0.46000466417910446,
+      "grad_norm": 0.48185947326085543,
+      "learning_rate": 3.22998662838127e-05,
+      "loss": 0.5041,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4605876865671642,
+      "grad_norm": 0.4734018707747314,
+      "learning_rate": 3.2257474486040166e-05,
+      "loss": 0.5038,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4611707089552239,
+      "grad_norm": 0.5075617095429634,
+      "learning_rate": 3.221506500263276e-05,
+      "loss": 0.5447,
+      "step": 3955
+    },
+    {
+      "epoch": 0.46175373134328357,
+      "grad_norm": 0.5046083432543891,
+      "learning_rate": 3.217263799124527e-05,
+      "loss": 0.5772,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4623367537313433,
+      "grad_norm": 0.47069172866304554,
+      "learning_rate": 3.213019360959762e-05,
+      "loss": 0.5341,
+      "step": 3965
+    },
+    {
+      "epoch": 0.462919776119403,
+      "grad_norm": 0.47866635232292687,
+      "learning_rate": 3.2087732015474366e-05,
+      "loss": 0.5208,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4635027985074627,
+      "grad_norm": 0.4516545061013375,
+      "learning_rate": 3.204525336672399e-05,
+      "loss": 0.5382,
+      "step": 3975
+    },
+    {
+      "epoch": 0.46408582089552236,
+      "grad_norm": 0.4987443516763455,
+      "learning_rate": 3.200275782125842e-05,
+      "loss": 0.5319,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4646688432835821,
+      "grad_norm": 0.5260827157599703,
+      "learning_rate": 3.196024553705235e-05,
+      "loss": 0.5355,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4652518656716418,
+      "grad_norm": 0.4668030990105745,
+      "learning_rate": 3.1917716672142746e-05,
+      "loss": 0.5057,
+      "step": 3990
+    },
+    {
+      "epoch": 0.46583488805970147,
+      "grad_norm": 0.4426983151393079,
+      "learning_rate": 3.187517138462819e-05,
+      "loss": 0.5254,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4664179104477612,
+      "grad_norm": 0.4925988353031197,
+      "learning_rate": 3.1832609832668314e-05,
+      "loss": 0.5422,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4670009328358209,
+      "grad_norm": 0.4560283471593519,
+      "learning_rate": 3.179003217448321e-05,
+      "loss": 0.5013,
+      "step": 4005
+    },
+    {
+      "epoch": 0.4675839552238806,
+      "grad_norm": 0.5244233872415475,
+      "learning_rate": 3.1747438568352844e-05,
+      "loss": 0.5736,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4681669776119403,
+      "grad_norm": 0.5512150192012447,
+      "learning_rate": 3.170482917261648e-05,
+      "loss": 0.5688,
+      "step": 4015
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 0.5009381195241619,
+      "learning_rate": 3.166220414567206e-05,
+      "loss": 0.5468,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4693330223880597,
+      "grad_norm": 0.47235216856933904,
+      "learning_rate": 3.161956364597566e-05,
+      "loss": 0.5318,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4699160447761194,
+      "grad_norm": 0.5634765857129355,
+      "learning_rate": 3.1576907832040855e-05,
+      "loss": 0.5188,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4704990671641791,
+      "grad_norm": 0.48072091370811076,
+      "learning_rate": 3.153423686243813e-05,
+      "loss": 0.5313,
+      "step": 4035
+    },
+    {
+      "epoch": 0.4710820895522388,
+      "grad_norm": 0.5120155067954245,
+      "learning_rate": 3.149155089579437e-05,
+      "loss": 0.5572,
+      "step": 4040
+    },
+    {
+      "epoch": 0.47166511194029853,
+      "grad_norm": 0.4812745114629162,
+      "learning_rate": 3.144885009079215e-05,
+      "loss": 0.5578,
+      "step": 4045
+    },
+    {
+      "epoch": 0.4722481343283582,
+      "grad_norm": 0.5058486137109877,
+      "learning_rate": 3.140613460616924e-05,
+      "loss": 0.5199,
+      "step": 4050
+    },
+    {
+      "epoch": 0.4728311567164179,
+      "grad_norm": 0.48002106325579913,
+      "learning_rate": 3.1363404600717965e-05,
+      "loss": 0.5659,
+      "step": 4055
+    },
+    {
+      "epoch": 0.47341417910447764,
+      "grad_norm": 0.48245091505853244,
+      "learning_rate": 3.132066023328465e-05,
+      "loss": 0.533,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4739972014925373,
+      "grad_norm": 0.47536248109948226,
+      "learning_rate": 3.1277901662768983e-05,
+      "loss": 0.5433,
+      "step": 4065
+    },
+    {
+      "epoch": 0.474580223880597,
+      "grad_norm": 0.47200232932636466,
+      "learning_rate": 3.123512904812347e-05,
+      "loss": 0.5322,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4751632462686567,
+      "grad_norm": 0.44740442637527467,
+      "learning_rate": 3.119234254835282e-05,
+      "loss": 0.5107,
+      "step": 4075
+    },
+    {
+      "epoch": 0.47574626865671643,
+      "grad_norm": 0.46154208469718466,
+      "learning_rate": 3.114954232251336e-05,
+      "loss": 0.527,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4763292910447761,
+      "grad_norm": 0.5341929895695878,
+      "learning_rate": 3.110672852971243e-05,
+      "loss": 0.5689,
+      "step": 4085
+    },
+    {
+      "epoch": 0.4769123134328358,
+      "grad_norm": 0.4652305557977976,
+      "learning_rate": 3.1063901329107843e-05,
+      "loss": 0.5167,
+      "step": 4090
+    },
+    {
+      "epoch": 0.47749533582089554,
+      "grad_norm": 0.44417381469876005,
+      "learning_rate": 3.10210608799072e-05,
+      "loss": 0.5413,
+      "step": 4095
+    },
+    {
+      "epoch": 0.4780783582089552,
+      "grad_norm": 0.4507749342993046,
+      "learning_rate": 3.097820734136739e-05,
+      "loss": 0.5282,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4786613805970149,
+      "grad_norm": 0.48297525105833106,
+      "learning_rate": 3.093534087279397e-05,
+      "loss": 0.5347,
+      "step": 4105
+    },
+    {
+      "epoch": 0.47924440298507465,
+      "grad_norm": 0.5100316904667305,
+      "learning_rate": 3.089246163354051e-05,
+      "loss": 0.5395,
+      "step": 4110
+    },
+    {
+      "epoch": 0.47982742537313433,
+      "grad_norm": 0.4995786265506999,
+      "learning_rate": 3.084956978300812e-05,
+      "loss": 0.5311,
+      "step": 4115
+    },
+    {
+      "epoch": 0.480410447761194,
+      "grad_norm": 0.4471832602242875,
+      "learning_rate": 3.080666548064475e-05,
+      "loss": 0.5193,
+      "step": 4120
+    },
+    {
+      "epoch": 0.48099347014925375,
+      "grad_norm": 0.46140397088213214,
+      "learning_rate": 3.076374888594464e-05,
+      "loss": 0.5345,
+      "step": 4125
+    },
+    {
+      "epoch": 0.48157649253731344,
+      "grad_norm": 0.44371711691404125,
+      "learning_rate": 3.0720820158447766e-05,
+      "loss": 0.5072,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4821595149253731,
+      "grad_norm": 0.4561270995882789,
+      "learning_rate": 3.067787945773915e-05,
+      "loss": 0.5181,
+      "step": 4135
+    },
+    {
+      "epoch": 0.48274253731343286,
+      "grad_norm": 0.4468599316613869,
+      "learning_rate": 3.063492694344835e-05,
+      "loss": 0.5286,
+      "step": 4140
+    },
+    {
+      "epoch": 0.48332555970149255,
+      "grad_norm": 0.4921853335929309,
+      "learning_rate": 3.059196277524886e-05,
+      "loss": 0.5075,
+      "step": 4145
+    },
+    {
+      "epoch": 0.48390858208955223,
+      "grad_norm": 0.47049020704612843,
+      "learning_rate": 3.054898711285747e-05,
+      "loss": 0.5475,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4844916044776119,
+      "grad_norm": 1.0060177708740357,
+      "learning_rate": 3.05060001160337e-05,
+      "loss": 0.5293,
+      "step": 4155
+    },
+    {
+      "epoch": 0.48507462686567165,
+      "grad_norm": 0.48784465637028546,
+      "learning_rate": 3.046300194457923e-05,
+      "loss": 0.534,
+      "step": 4160
+    },
+    {
+      "epoch": 0.48565764925373134,
+      "grad_norm": 0.4518879881598059,
+      "learning_rate": 3.0419992758337235e-05,
+      "loss": 0.5209,
+      "step": 4165
+    },
+    {
+      "epoch": 0.486240671641791,
+      "grad_norm": 0.4452974474644216,
+      "learning_rate": 3.0376972717191894e-05,
+      "loss": 0.5245,
+      "step": 4170
+    },
+    {
+      "epoch": 0.48682369402985076,
+      "grad_norm": 0.44833371464722693,
+      "learning_rate": 3.0333941981067688e-05,
+      "loss": 0.5156,
+      "step": 4175
+    },
+    {
+      "epoch": 0.48740671641791045,
+      "grad_norm": 0.44934997583495073,
+      "learning_rate": 3.029090070992889e-05,
+      "loss": 0.5144,
+      "step": 4180
+    },
+    {
+      "epoch": 0.48798973880597013,
+      "grad_norm": 0.47274610781733095,
+      "learning_rate": 3.0247849063778917e-05,
+      "loss": 0.5317,
+      "step": 4185
+    },
+    {
+      "epoch": 0.48857276119402987,
+      "grad_norm": 0.47590787280797764,
+      "learning_rate": 3.020478720265977e-05,
+      "loss": 0.51,
+      "step": 4190
+    },
+    {
+      "epoch": 0.48915578358208955,
+      "grad_norm": 0.44563102351746864,
+      "learning_rate": 3.01617152866514e-05,
+      "loss": 0.5301,
+      "step": 4195
+    },
+    {
+      "epoch": 0.48973880597014924,
+      "grad_norm": 0.4916861483726837,
+      "learning_rate": 3.0118633475871167e-05,
+      "loss": 0.5772,
+      "step": 4200
+    },
+    {
+      "epoch": 0.490321828358209,
+      "grad_norm": 0.4684037067244426,
+      "learning_rate": 3.0075541930473183e-05,
+      "loss": 0.4969,
+      "step": 4205
+    },
+    {
+      "epoch": 0.49090485074626866,
+      "grad_norm": 0.4797812232043466,
+      "learning_rate": 3.0032440810647783e-05,
+      "loss": 0.5038,
+      "step": 4210
+    },
+    {
+      "epoch": 0.49148787313432835,
+      "grad_norm": 0.4525162593921102,
+      "learning_rate": 2.998933027662086e-05,
+      "loss": 0.5266,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4920708955223881,
+      "grad_norm": 0.4822829175013609,
+      "learning_rate": 2.9946210488653316e-05,
+      "loss": 0.5243,
+      "step": 4220
+    },
+    {
+      "epoch": 0.49265391791044777,
+      "grad_norm": 0.45841709338227005,
+      "learning_rate": 2.9903081607040473e-05,
+      "loss": 0.4948,
+      "step": 4225
+    },
+    {
+      "epoch": 0.49323694029850745,
+      "grad_norm": 0.4465507453121582,
+      "learning_rate": 2.9859943792111422e-05,
+      "loss": 0.5228,
+      "step": 4230
+    },
+    {
+      "epoch": 0.49381996268656714,
+      "grad_norm": 0.613185904602815,
+      "learning_rate": 2.9816797204228497e-05,
+      "loss": 0.5345,
+      "step": 4235
+    },
+    {
+      "epoch": 0.4944029850746269,
+      "grad_norm": 0.49396745715444307,
+      "learning_rate": 2.9773642003786627e-05,
+      "loss": 0.5412,
+      "step": 4240
+    },
+    {
+      "epoch": 0.49498600746268656,
+      "grad_norm": 0.5239548031082005,
+      "learning_rate": 2.9730478351212754e-05,
+      "loss": 0.5821,
+      "step": 4245
+    },
+    {
+      "epoch": 0.49556902985074625,
+      "grad_norm": 0.4221114790983556,
+      "learning_rate": 2.968730640696526e-05,
+      "loss": 0.5199,
+      "step": 4250
+    },
+    {
+      "epoch": 0.496152052238806,
+      "grad_norm": 0.480540611050451,
+      "learning_rate": 2.9644126331533328e-05,
+      "loss": 0.5149,
+      "step": 4255
+    },
+    {
+      "epoch": 0.49673507462686567,
+      "grad_norm": 0.4972848047633413,
+      "learning_rate": 2.9600938285436385e-05,
+      "loss": 0.5155,
+      "step": 4260
+    },
+    {
+      "epoch": 0.49731809701492535,
+      "grad_norm": 0.5100758433643319,
+      "learning_rate": 2.9557742429223478e-05,
+      "loss": 0.5495,
+      "step": 4265
+    },
+    {
+      "epoch": 0.4979011194029851,
+      "grad_norm": 0.46169645444430685,
+      "learning_rate": 2.95145389234727e-05,
+      "loss": 0.5432,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4984841417910448,
+      "grad_norm": 0.478683421695544,
+      "learning_rate": 2.947132792879056e-05,
+      "loss": 0.5622,
+      "step": 4275
+    },
+    {
+      "epoch": 0.49906716417910446,
+      "grad_norm": 0.4937574145856353,
+      "learning_rate": 2.9428109605811427e-05,
+      "loss": 0.5292,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4996501865671642,
+      "grad_norm": 0.4482984531417574,
+      "learning_rate": 2.9384884115196898e-05,
+      "loss": 0.5115,
+      "step": 4285
+    },
+    {
+      "epoch": 0.5002332089552238,
+      "grad_norm": 0.4581506836353169,
+      "learning_rate": 2.9341651617635236e-05,
+      "loss": 0.5369,
+      "step": 4290
+    },
+    {
+      "epoch": 0.5008162313432836,
+      "grad_norm": 0.4636121196389006,
+      "learning_rate": 2.929841227384072e-05,
+      "loss": 0.563,
+      "step": 4295
+    },
+    {
+      "epoch": 0.5013992537313433,
+      "grad_norm": 0.49303476266346524,
+      "learning_rate": 2.925516624455311e-05,
+      "loss": 0.5323,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5019822761194029,
+      "grad_norm": 0.4792765198500162,
+      "learning_rate": 2.9211913690537003e-05,
+      "loss": 0.5238,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5025652985074627,
+      "grad_norm": 0.47595384636734767,
+      "learning_rate": 2.9168654772581257e-05,
+      "loss": 0.5487,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5031483208955224,
+      "grad_norm": 0.45475337905445856,
+      "learning_rate": 2.9125389651498374e-05,
+      "loss": 0.5551,
+      "step": 4315
+    },
+    {
+      "epoch": 0.503731343283582,
+      "grad_norm": 0.45368396432436275,
+      "learning_rate": 2.908211848812394e-05,
+      "loss": 0.5096,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5043143656716418,
+      "grad_norm": 0.4720732964199861,
+      "learning_rate": 2.903884144331598e-05,
+      "loss": 0.5375,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5048973880597015,
+      "grad_norm": 0.4609507788137596,
+      "learning_rate": 2.89955586779544e-05,
+      "loss": 0.5193,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5054804104477612,
+      "grad_norm": 0.4437193884044816,
+      "learning_rate": 2.8952270352940362e-05,
+      "loss": 0.538,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5060634328358209,
+      "grad_norm": 0.4686721143893824,
+      "learning_rate": 2.89089766291957e-05,
+      "loss": 0.5359,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5066464552238806,
+      "grad_norm": 0.48522442554794026,
+      "learning_rate": 2.886567766766231e-05,
+      "loss": 0.5367,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5072294776119403,
+      "grad_norm": 0.4682334708596764,
+      "learning_rate": 2.8822373629301573e-05,
+      "loss": 0.4979,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5078125,
+      "grad_norm": 0.4461441009977717,
+      "learning_rate": 2.8779064675093724e-05,
+      "loss": 0.5473,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5083955223880597,
+      "grad_norm": 0.523425795341376,
+      "learning_rate": 2.8735750966037295e-05,
+      "loss": 0.592,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5089785447761194,
+      "grad_norm": 0.41208414865400855,
+      "learning_rate": 2.869243266314847e-05,
+      "loss": 0.5146,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5095615671641791,
+      "grad_norm": 0.45512638703063274,
+      "learning_rate": 2.8649109927460533e-05,
+      "loss": 0.517,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5101445895522388,
+      "grad_norm": 0.449015485615699,
+      "learning_rate": 2.8605782920023227e-05,
+      "loss": 0.5235,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5107276119402985,
+      "grad_norm": 0.49156178005481366,
+      "learning_rate": 2.8562451801902197e-05,
+      "loss": 0.5308,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5113106343283582,
+      "grad_norm": 0.4348710846671963,
+      "learning_rate": 2.8519116734178336e-05,
+      "loss": 0.5387,
+      "step": 4385
+    },
+    {
+      "epoch": 0.511893656716418,
+      "grad_norm": 0.4902492973827986,
+      "learning_rate": 2.8475777877947264e-05,
+      "loss": 0.5417,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5124766791044776,
+      "grad_norm": 0.6547931036800456,
+      "learning_rate": 2.843243539431863e-05,
+      "loss": 0.5444,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5130597014925373,
+      "grad_norm": 0.4876654982418252,
+      "learning_rate": 2.838908944441562e-05,
+      "loss": 0.5558,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5136427238805971,
+      "grad_norm": 0.46099868761271035,
+      "learning_rate": 2.834574018937428e-05,
+      "loss": 0.5605,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5142257462686567,
+      "grad_norm": 0.4545024437386275,
+      "learning_rate": 2.8302387790342943e-05,
+      "loss": 0.5175,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5148087686567164,
+      "grad_norm": 0.4336617180340361,
+      "learning_rate": 2.8259032408481635e-05,
+      "loss": 0.5038,
+      "step": 4415
+    },
+    {
+      "epoch": 0.5153917910447762,
+      "grad_norm": 0.47838755970386043,
+      "learning_rate": 2.8215674204961462e-05,
+      "loss": 0.5594,
+      "step": 4420
+    },
+    {
+      "epoch": 0.5159748134328358,
+      "grad_norm": 0.5142813582002242,
+      "learning_rate": 2.817231334096403e-05,
+      "loss": 0.5177,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5165578358208955,
+      "grad_norm": 0.44541441218084665,
+      "learning_rate": 2.812894997768083e-05,
+      "loss": 0.5465,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5171408582089553,
+      "grad_norm": 0.4560004171104499,
+      "learning_rate": 2.8085584276312644e-05,
+      "loss": 0.5201,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5177238805970149,
+      "grad_norm": 0.47964290891517664,
+      "learning_rate": 2.8042216398068942e-05,
+      "loss": 0.5247,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5183069029850746,
+      "grad_norm": 0.8168752970003079,
+      "learning_rate": 2.7998846504167308e-05,
+      "loss": 0.5277,
+      "step": 4445
+    },
+    {
+      "epoch": 0.5188899253731343,
+      "grad_norm": 2.544656980904975,
+      "learning_rate": 2.7955474755832784e-05,
+      "loss": 0.505,
+      "step": 4450
+    },
+    {
+      "epoch": 0.519472947761194,
+      "grad_norm": 0.4659395065613913,
+      "learning_rate": 2.7912101314297327e-05,
+      "loss": 0.5269,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5200559701492538,
+      "grad_norm": 2.6969407750363943,
+      "learning_rate": 2.7868726340799184e-05,
+      "loss": 0.5306,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5206389925373134,
+      "grad_norm": 0.5270843633040919,
+      "learning_rate": 2.7825349996582313e-05,
+      "loss": 0.5711,
+      "step": 4465
+    },
+    {
+      "epoch": 0.5212220149253731,
+      "grad_norm": 0.5091777799722457,
+      "learning_rate": 2.7781972442895726e-05,
+      "loss": 0.5365,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5218050373134329,
+      "grad_norm": 1.3533639882343287,
+      "learning_rate": 2.7738593840992975e-05,
+      "loss": 0.5539,
+      "step": 4475
+    },
+    {
+      "epoch": 0.5223880597014925,
+      "grad_norm": 0.4547594298673975,
+      "learning_rate": 2.769521435213149e-05,
+      "loss": 0.544,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5229710820895522,
+      "grad_norm": 0.5510873748275549,
+      "learning_rate": 2.7651834137572003e-05,
+      "loss": 0.5353,
+      "step": 4485
+    },
+    {
+      "epoch": 0.523554104477612,
+      "grad_norm": 0.4328373857127542,
+      "learning_rate": 2.760845335857793e-05,
+      "loss": 0.5132,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5241371268656716,
+      "grad_norm": 0.45053079578268007,
+      "learning_rate": 2.7565072176414803e-05,
+      "loss": 0.5231,
+      "step": 4495
+    },
+    {
+      "epoch": 0.5247201492537313,
+      "grad_norm": 0.49741815715434656,
+      "learning_rate": 2.7521690752349643e-05,
+      "loss": 0.5281,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5253031716417911,
+      "grad_norm": 0.4607291755452952,
+      "learning_rate": 2.7478309247650362e-05,
+      "loss": 0.5558,
+      "step": 4505
+    },
+    {
+      "epoch": 0.5258861940298507,
+      "grad_norm": 0.5194917261722904,
+      "learning_rate": 2.7434927823585206e-05,
+      "loss": 0.539,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5264692164179104,
+      "grad_norm": 0.45463126850370955,
+      "learning_rate": 2.739154664142208e-05,
+      "loss": 0.5215,
+      "step": 4515
+    },
+    {
+      "epoch": 0.5270522388059702,
+      "grad_norm": 0.4892107161079775,
+      "learning_rate": 2.7348165862428e-05,
+      "loss": 0.4891,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5276352611940298,
+      "grad_norm": 0.5045457990411999,
+      "learning_rate": 2.7304785647868507e-05,
+      "loss": 0.5416,
+      "step": 4525
+    },
+    {
+      "epoch": 0.5282182835820896,
+      "grad_norm": 0.44346437150982526,
+      "learning_rate": 2.726140615900703e-05,
+      "loss": 0.5048,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5288013059701493,
+      "grad_norm": 0.48666851854991494,
+      "learning_rate": 2.7218027557104286e-05,
+      "loss": 0.5034,
+      "step": 4535
+    },
+    {
+      "epoch": 0.5293843283582089,
+      "grad_norm": 0.4611162809005146,
+      "learning_rate": 2.7174650003417696e-05,
+      "loss": 0.5142,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5299673507462687,
+      "grad_norm": 0.45514376590541833,
+      "learning_rate": 2.7131273659200818e-05,
+      "loss": 0.5162,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5305503731343284,
+      "grad_norm": 0.48960501435122616,
+      "learning_rate": 2.7087898685702685e-05,
+      "loss": 0.5069,
+      "step": 4550
+    },
+    {
+      "epoch": 0.531133395522388,
+      "grad_norm": 0.4797862989191752,
+      "learning_rate": 2.704452524416722e-05,
+      "loss": 0.531,
+      "step": 4555
+    },
+    {
+      "epoch": 0.5317164179104478,
+      "grad_norm": 0.5893011284468495,
+      "learning_rate": 2.7001153495832697e-05,
+      "loss": 0.5061,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5322994402985075,
+      "grad_norm": 0.4461810976300674,
+      "learning_rate": 2.6957783601931063e-05,
+      "loss": 0.5203,
+      "step": 4565
+    },
+    {
+      "epoch": 0.5328824626865671,
+      "grad_norm": 0.49982880523004874,
+      "learning_rate": 2.691441572368737e-05,
+      "loss": 0.5284,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5334654850746269,
+      "grad_norm": 0.8194747676964631,
+      "learning_rate": 2.6871050022319177e-05,
+      "loss": 0.5642,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5340485074626866,
+      "grad_norm": 0.45941020004877126,
+      "learning_rate": 2.6827686659035983e-05,
+      "loss": 0.5444,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5346315298507462,
+      "grad_norm": 0.7401627959971693,
+      "learning_rate": 2.678432579503855e-05,
+      "loss": 0.5138,
+      "step": 4585
+    },
+    {
+      "epoch": 0.535214552238806,
+      "grad_norm": 0.49796442033400656,
+      "learning_rate": 2.6740967591518374e-05,
+      "loss": 0.5264,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5357975746268657,
+      "grad_norm": 1.3853714172153995,
+      "learning_rate": 2.6697612209657063e-05,
+      "loss": 0.5375,
+      "step": 4595
+    },
+    {
+      "epoch": 0.5363805970149254,
+      "grad_norm": 0.4543560797621967,
+      "learning_rate": 2.665425981062573e-05,
+      "loss": 0.5155,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5369636194029851,
+      "grad_norm": 0.484187959090983,
+      "learning_rate": 2.6610910555584384e-05,
+      "loss": 0.5478,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5375466417910447,
+      "grad_norm": 0.49626488411866204,
+      "learning_rate": 2.6567564605681376e-05,
+      "loss": 0.5352,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5381296641791045,
+      "grad_norm": 0.514181721385795,
+      "learning_rate": 2.652422212205275e-05,
+      "loss": 0.5387,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5387126865671642,
+      "grad_norm": 0.5832049097179766,
+      "learning_rate": 2.6480883265821673e-05,
+      "loss": 0.5355,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5392957089552238,
+      "grad_norm": 1.5120035406279293,
+      "learning_rate": 2.643754819809781e-05,
+      "loss": 0.5144,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5398787313432836,
+      "grad_norm": 0.49157920203901545,
+      "learning_rate": 2.639421707997678e-05,
+      "loss": 0.5176,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5404617537313433,
+      "grad_norm": 0.52902066701415,
+      "learning_rate": 2.6350890072539476e-05,
+      "loss": 0.5635,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5410447761194029,
+      "grad_norm": 0.4637572082859708,
+      "learning_rate": 2.630756733685153e-05,
+      "loss": 0.5346,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5416277985074627,
+      "grad_norm": 0.47595145214824397,
+      "learning_rate": 2.6264249033962713e-05,
+      "loss": 0.5249,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5422108208955224,
+      "grad_norm": 1.6764787888943493,
+      "learning_rate": 2.622093532490628e-05,
+      "loss": 0.5064,
+      "step": 4650
+    },
+    {
+      "epoch": 0.542793843283582,
+      "grad_norm": 0.5788528459186878,
+      "learning_rate": 2.6177626370698443e-05,
+      "loss": 0.5051,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5433768656716418,
+      "grad_norm": 0.48162500018331245,
+      "learning_rate": 2.6134322332337695e-05,
+      "loss": 0.5043,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5439598880597015,
+      "grad_norm": 0.4737951941511267,
+      "learning_rate": 2.6091023370804307e-05,
+      "loss": 0.5142,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5445429104477612,
+      "grad_norm": 0.5575644816288676,
+      "learning_rate": 2.604772964705965e-05,
+      "loss": 0.5505,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5451259328358209,
+      "grad_norm": 0.4758372366326835,
+      "learning_rate": 2.6004441322045603e-05,
+      "loss": 0.5373,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5457089552238806,
+      "grad_norm": 0.508958909313539,
+      "learning_rate": 2.596115855668403e-05,
+      "loss": 0.5353,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5462919776119403,
+      "grad_norm": 0.4683786155826753,
+      "learning_rate": 2.5917881511876073e-05,
+      "loss": 0.4946,
+      "step": 4685
+    },
+    {
+      "epoch": 0.546875,
+      "grad_norm": 0.5456678942382134,
+      "learning_rate": 2.5874610348501632e-05,
+      "loss": 0.5371,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5474580223880597,
+      "grad_norm": 0.4552682312653168,
+      "learning_rate": 2.5831345227418752e-05,
+      "loss": 0.5605,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5480410447761194,
+      "grad_norm": 0.4776001785979188,
+      "learning_rate": 2.5788086309463006e-05,
+      "loss": 0.5005,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5486240671641791,
+      "grad_norm": 0.5453229696717531,
+      "learning_rate": 2.57448337554469e-05,
+      "loss": 0.5194,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5492070895522388,
+      "grad_norm": 0.5387817919914101,
+      "learning_rate": 2.570158772615928e-05,
+      "loss": 0.5333,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5497901119402985,
+      "grad_norm": 0.48603015940285904,
+      "learning_rate": 2.5658348382364773e-05,
+      "loss": 0.5117,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5503731343283582,
+      "grad_norm": 0.4789950555392135,
+      "learning_rate": 2.5615115884803108e-05,
+      "loss": 0.5168,
+      "step": 4720
+    },
+    {
+      "epoch": 0.550956156716418,
+      "grad_norm": 0.4593000342095894,
+      "learning_rate": 2.557189039418858e-05,
+      "loss": 0.5238,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5515391791044776,
+      "grad_norm": 0.4671404581672369,
+      "learning_rate": 2.552867207120945e-05,
+      "loss": 0.5336,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5521222014925373,
+      "grad_norm": 0.4867006683126969,
+      "learning_rate": 2.5485461076527308e-05,
+      "loss": 0.5432,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5527052238805971,
+      "grad_norm": 0.4603754841175435,
+      "learning_rate": 2.5442257570776527e-05,
+      "loss": 0.5464,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5532882462686567,
+      "grad_norm": 0.5322095390809611,
+      "learning_rate": 2.539906171456362e-05,
+      "loss": 0.5102,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5538712686567164,
+      "grad_norm": 0.4501063054653902,
+      "learning_rate": 2.5355873668466677e-05,
+      "loss": 0.526,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5544542910447762,
+      "grad_norm": 0.4645529985985766,
+      "learning_rate": 2.5312693593034746e-05,
+      "loss": 0.5035,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5550373134328358,
+      "grad_norm": 0.4764710401107973,
+      "learning_rate": 2.5269521648787247e-05,
+      "loss": 0.5458,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5556203358208955,
+      "grad_norm": 0.4534954950893274,
+      "learning_rate": 2.5226357996213378e-05,
+      "loss": 0.5199,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5562033582089553,
+      "grad_norm": 0.4810706685920549,
+      "learning_rate": 2.518320279577151e-05,
+      "loss": 0.5263,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5567863805970149,
+      "grad_norm": 0.4690472358261824,
+      "learning_rate": 2.514005620788858e-05,
+      "loss": 0.5245,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5573694029850746,
+      "grad_norm": 0.46880508544260313,
+      "learning_rate": 2.5096918392959532e-05,
+      "loss": 0.5232,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5579524253731343,
+      "grad_norm": 0.513925549118589,
+      "learning_rate": 2.5053789511346693e-05,
+      "loss": 0.5157,
+      "step": 4785
+    },
+    {
+      "epoch": 0.558535447761194,
+      "grad_norm": 0.5043455996761089,
+      "learning_rate": 2.5010669723379154e-05,
+      "loss": 0.5128,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5591184701492538,
+      "grad_norm": 0.5924109641064016,
+      "learning_rate": 2.4967559189352226e-05,
+      "loss": 0.4921,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5597014925373134,
+      "grad_norm": 0.47567528373039897,
+      "learning_rate": 2.492445806952682e-05,
+      "loss": 0.5406,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5602845149253731,
+      "grad_norm": 0.4707447199114358,
+      "learning_rate": 2.4881366524128845e-05,
+      "loss": 0.5325,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5608675373134329,
+      "grad_norm": 0.444463840534227,
+      "learning_rate": 2.4838284713348602e-05,
+      "loss": 0.4945,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5614505597014925,
+      "grad_norm": 0.51359075031383,
+      "learning_rate": 2.479521279734024e-05,
+      "loss": 0.4986,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5620335820895522,
+      "grad_norm": 0.508318043421021,
+      "learning_rate": 2.475215093622109e-05,
+      "loss": 0.5362,
+      "step": 4820
+    },
+    {
+      "epoch": 0.562616604477612,
+      "grad_norm": 0.4798125508084004,
+      "learning_rate": 2.4709099290071126e-05,
+      "loss": 0.498,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5631996268656716,
+      "grad_norm": 0.46383793437769893,
+      "learning_rate": 2.4666058018932314e-05,
+      "loss": 0.5254,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5637826492537313,
+      "grad_norm": 0.4825804953446005,
+      "learning_rate": 2.4623027282808114e-05,
+      "loss": 0.515,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5643656716417911,
+      "grad_norm": 0.49356516542621953,
+      "learning_rate": 2.4580007241662773e-05,
+      "loss": 0.4913,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5649486940298507,
+      "grad_norm": 0.46504399968658483,
+      "learning_rate": 2.4536998055420783e-05,
+      "loss": 0.5055,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5655317164179104,
+      "grad_norm": 0.5178266867135491,
+      "learning_rate": 2.4493999883966308e-05,
+      "loss": 0.5338,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5661147388059702,
+      "grad_norm": 0.443773396289914,
+      "learning_rate": 2.445101288714254e-05,
+      "loss": 0.527,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5666977611940298,
+      "grad_norm": 0.4361179409054666,
+      "learning_rate": 2.440803722475114e-05,
+      "loss": 0.5115,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5672807835820896,
+      "grad_norm": 0.4957031944579591,
+      "learning_rate": 2.436507305655165e-05,
+      "loss": 0.5224,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5678638059701493,
+      "grad_norm": 0.5179909459677171,
+      "learning_rate": 2.4322120542260864e-05,
+      "loss": 0.5026,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5684468283582089,
+      "grad_norm": 0.45931411628609564,
+      "learning_rate": 2.4279179841552246e-05,
+      "loss": 0.5492,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5690298507462687,
+      "grad_norm": 0.4718292947061309,
+      "learning_rate": 2.4236251114055358e-05,
+      "loss": 0.5485,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5696128731343284,
+      "grad_norm": 0.4509426009796579,
+      "learning_rate": 2.419333451935526e-05,
+      "loss": 0.5262,
+      "step": 4885
+    },
+    {
+      "epoch": 0.570195895522388,
+      "grad_norm": 0.520970250021266,
+      "learning_rate": 2.4150430216991888e-05,
+      "loss": 0.54,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5707789179104478,
+      "grad_norm": 0.48125165415242044,
+      "learning_rate": 2.4107538366459494e-05,
+      "loss": 0.5535,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5713619402985075,
+      "grad_norm": 0.49924612989919914,
+      "learning_rate": 2.406465912720604e-05,
+      "loss": 0.5378,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5719449626865671,
+      "grad_norm": 0.4618022420362027,
+      "learning_rate": 2.4021792658632612e-05,
+      "loss": 0.5385,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5725279850746269,
+      "grad_norm": 0.43524447366727087,
+      "learning_rate": 2.3978939120092814e-05,
+      "loss": 0.5275,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5731110074626866,
+      "grad_norm": 0.5235831961960491,
+      "learning_rate": 2.3936098670892165e-05,
+      "loss": 0.5171,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5736940298507462,
+      "grad_norm": 0.47253982117527504,
+      "learning_rate": 2.389327147028757e-05,
+      "loss": 0.5149,
+      "step": 4920
+    },
+    {
+      "epoch": 0.574277052238806,
+      "grad_norm": 0.536733235694404,
+      "learning_rate": 2.3850457677486655e-05,
+      "loss": 0.5385,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5748600746268657,
+      "grad_norm": 0.49487264913586265,
+      "learning_rate": 2.380765745164718e-05,
+      "loss": 0.5453,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5754430970149254,
+      "grad_norm": 0.4782990502193716,
+      "learning_rate": 2.376487095187654e-05,
+      "loss": 0.496,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5760261194029851,
+      "grad_norm": 1.5309625007086856,
+      "learning_rate": 2.3722098337231025e-05,
+      "loss": 0.507,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5766091417910447,
+      "grad_norm": 0.4155961423348815,
+      "learning_rate": 2.3679339766715358e-05,
+      "loss": 0.5219,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5771921641791045,
+      "grad_norm": 0.4914382394629777,
+      "learning_rate": 2.363659539928204e-05,
+      "loss": 0.5107,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5777751865671642,
+      "grad_norm": 0.4984691122251259,
+      "learning_rate": 2.3593865393830766e-05,
+      "loss": 0.5007,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5783582089552238,
+      "grad_norm": 0.504442137117383,
+      "learning_rate": 2.355114990920786e-05,
+      "loss": 0.5103,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5789412313432836,
+      "grad_norm": 0.774789963927665,
+      "learning_rate": 2.3508449104205636e-05,
+      "loss": 0.516,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5795242537313433,
+      "grad_norm": 0.49669070623006295,
+      "learning_rate": 2.3465763137561875e-05,
+      "loss": 0.4984,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5801072761194029,
+      "grad_norm": 0.4915699683156642,
+      "learning_rate": 2.342309216795916e-05,
+      "loss": 0.5237,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5806902985074627,
+      "grad_norm": 0.7919278813140391,
+      "learning_rate": 2.3380436354024338e-05,
+      "loss": 0.5327,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5812733208955224,
+      "grad_norm": 0.513367233603117,
+      "learning_rate": 2.333779585432794e-05,
+      "loss": 0.545,
+      "step": 4985
+    },
+    {
+      "epoch": 0.581856343283582,
+      "grad_norm": 0.5185055850159029,
+      "learning_rate": 2.329517082738353e-05,
+      "loss": 0.5671,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5824393656716418,
+      "grad_norm": 0.4535336329617379,
+      "learning_rate": 2.3252561431647158e-05,
+      "loss": 0.5072,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5830223880597015,
+      "grad_norm": 0.4214639260752155,
+      "learning_rate": 2.32099678255168e-05,
+      "loss": 0.4765,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5836054104477612,
+      "grad_norm": 0.4292230083522436,
+      "learning_rate": 2.316739016733169e-05,
+      "loss": 0.535,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5841884328358209,
+      "grad_norm": 0.4500847118246923,
+      "learning_rate": 2.3124828615371817e-05,
+      "loss": 0.5579,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5847714552238806,
+      "grad_norm": 0.4441260617873408,
+      "learning_rate": 2.3082283327857253e-05,
+      "loss": 0.502,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5853544776119403,
+      "grad_norm": 0.4929193442519856,
+      "learning_rate": 2.3039754462947653e-05,
+      "loss": 0.5154,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5859375,
+      "grad_norm": 0.4604596804207028,
+      "learning_rate": 2.2997242178741596e-05,
+      "loss": 0.5173,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5865205223880597,
+      "grad_norm": 0.4671301724369313,
+      "learning_rate": 2.2954746633276016e-05,
+      "loss": 0.5022,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5871035447761194,
+      "grad_norm": 0.4666266550085784,
+      "learning_rate": 2.2912267984525643e-05,
+      "loss": 0.5263,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5876865671641791,
+      "grad_norm": 0.5663717691966537,
+      "learning_rate": 2.2869806390402384e-05,
+      "loss": 0.5126,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5882695895522388,
+      "grad_norm": 0.4709782228008594,
+      "learning_rate": 2.2827362008754743e-05,
+      "loss": 0.5374,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5888526119402985,
+      "grad_norm": 0.523861503593284,
+      "learning_rate": 2.278493499736724e-05,
+      "loss": 0.5313,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5894356343283582,
+      "grad_norm": 0.4797723190933745,
+      "learning_rate": 2.2742525513959832e-05,
+      "loss": 0.5382,
+      "step": 5055
+    },
+    {
+      "epoch": 0.590018656716418,
+      "grad_norm": 0.43860453059040094,
+      "learning_rate": 2.2700133716187316e-05,
+      "loss": 0.5265,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5906016791044776,
+      "grad_norm": 0.4740764262851301,
+      "learning_rate": 2.2657759761638707e-05,
+      "loss": 0.5148,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5911847014925373,
+      "grad_norm": 0.48957608183212625,
+      "learning_rate": 2.261540380783675e-05,
+      "loss": 0.4952,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5917677238805971,
+      "grad_norm": 0.5162045189958194,
+      "learning_rate": 2.257306601223722e-05,
+      "loss": 0.5331,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5923507462686567,
+      "grad_norm": 0.5728273214706431,
+      "learning_rate": 2.2530746532228413e-05,
+      "loss": 0.5143,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5929337686567164,
+      "grad_norm": 0.44294658485783917,
+      "learning_rate": 2.2488445525130557e-05,
+      "loss": 0.5385,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5935167910447762,
+      "grad_norm": 0.4728473251887733,
+      "learning_rate": 2.2446163148195164e-05,
+      "loss": 0.5249,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5940998134328358,
+      "grad_norm": 0.42056687179230184,
+      "learning_rate": 2.2403899558604525e-05,
+      "loss": 0.4896,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5946828358208955,
+      "grad_norm": 0.42312832455359684,
+      "learning_rate": 2.2361654913471065e-05,
+      "loss": 0.4831,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5952658582089553,
+      "grad_norm": 0.4310382361089622,
+      "learning_rate": 2.2319429369836815e-05,
+      "loss": 0.5038,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5958488805970149,
+      "grad_norm": 0.43349085099461926,
+      "learning_rate": 2.2277223084672765e-05,
+      "loss": 0.5332,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5964319029850746,
+      "grad_norm": 0.4358686772263049,
+      "learning_rate": 2.2235036214878325e-05,
+      "loss": 0.4716,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.4998061706788218,
+      "learning_rate": 2.2192868917280745e-05,
+      "loss": 0.5277,
+      "step": 5120
+    },
+    {
+      "epoch": 0.597597947761194,
+      "grad_norm": 0.4569209342643407,
+      "learning_rate": 2.21507213486345e-05,
+      "loss": 0.4995,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5981809701492538,
+      "grad_norm": 0.45297207736047584,
+      "learning_rate": 2.2108593665620724e-05,
+      "loss": 0.5254,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5987639925373134,
+      "grad_norm": 0.49176995249756655,
+      "learning_rate": 2.2066486024846615e-05,
+      "loss": 0.5312,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5993470149253731,
+      "grad_norm": 0.45352488174846867,
+      "learning_rate": 2.2024398582844906e-05,
+      "loss": 0.5183,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5999300373134329,
+      "grad_norm": 0.4686989486366428,
+      "learning_rate": 2.19823314960732e-05,
+      "loss": 0.5337,
+      "step": 5145
+    },
+    {
+      "epoch": 0.6005130597014925,
+      "grad_norm": 0.4989471659468385,
+      "learning_rate": 2.1940284920913445e-05,
+      "loss": 0.513,
+      "step": 5150
+    },
+    {
+      "epoch": 0.6010960820895522,
+      "grad_norm": 0.7737032679878307,
+      "learning_rate": 2.1898259013671357e-05,
+      "loss": 0.4948,
+      "step": 5155
+    },
+    {
+      "epoch": 0.601679104477612,
+      "grad_norm": 0.6456584381963478,
+      "learning_rate": 2.1856253930575787e-05,
+      "loss": 0.5506,
+      "step": 5160
+    },
+    {
+      "epoch": 0.6022621268656716,
+      "grad_norm": 0.49420633212169496,
+      "learning_rate": 2.18142698277782e-05,
+      "loss": 0.534,
+      "step": 5165
+    },
+    {
+      "epoch": 0.6028451492537313,
+      "grad_norm": 0.43079700140392413,
+      "learning_rate": 2.1772306861352044e-05,
+      "loss": 0.4994,
+      "step": 5170
+    },
+    {
+      "epoch": 0.6034281716417911,
+      "grad_norm": 0.4477340131354866,
+      "learning_rate": 2.1730365187292228e-05,
+      "loss": 0.4895,
+      "step": 5175
+    },
+    {
+      "epoch": 0.6040111940298507,
+      "grad_norm": 0.45974056791460993,
+      "learning_rate": 2.1688444961514476e-05,
+      "loss": 0.4963,
+      "step": 5180
+    },
+    {
+      "epoch": 0.6045942164179104,
+      "grad_norm": 0.4463763393224647,
+      "learning_rate": 2.1646546339854788e-05,
+      "loss": 0.5139,
+      "step": 5185
+    },
+    {
+      "epoch": 0.6051772388059702,
+      "grad_norm": 0.5136717833408538,
+      "learning_rate": 2.1604669478068863e-05,
+      "loss": 0.5195,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6057602611940298,
+      "grad_norm": 0.4554979479736964,
+      "learning_rate": 2.1562814531831487e-05,
+      "loss": 0.5062,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6063432835820896,
+      "grad_norm": 0.45967404846816146,
+      "learning_rate": 2.1520981656736e-05,
+      "loss": 0.4984,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6069263059701493,
+      "grad_norm": 0.44666913701763594,
+      "learning_rate": 2.1479171008293686e-05,
+      "loss": 0.4939,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6075093283582089,
+      "grad_norm": 1.738573008880425,
+      "learning_rate": 2.1437382741933204e-05,
+      "loss": 0.542,
+      "step": 5210
+    },
+    {
+      "epoch": 0.6080923507462687,
+      "grad_norm": 0.45746766067753797,
+      "learning_rate": 2.1395617012999993e-05,
+      "loss": 0.5014,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6086753731343284,
+      "grad_norm": 0.4542305089948559,
+      "learning_rate": 2.1353873976755716e-05,
+      "loss": 0.4852,
+      "step": 5220
+    },
+    {
+      "epoch": 0.609258395522388,
+      "grad_norm": 0.4734086638694241,
+      "learning_rate": 2.131215378837771e-05,
+      "loss": 0.5218,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6098414179104478,
+      "grad_norm": 0.4738565708182124,
+      "learning_rate": 2.1270456602958332e-05,
+      "loss": 0.5322,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6104244402985075,
+      "grad_norm": 0.49359869399689404,
+      "learning_rate": 2.1228782575504447e-05,
+      "loss": 0.537,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6110074626865671,
+      "grad_norm": 0.47548423763939335,
+      "learning_rate": 2.1187131860936845e-05,
+      "loss": 0.5183,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6115904850746269,
+      "grad_norm": 0.6813270131879138,
+      "learning_rate": 2.114550461408961e-05,
+      "loss": 0.5097,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6121735074626866,
+      "grad_norm": 0.4858514966177657,
+      "learning_rate": 2.1103900989709623e-05,
+      "loss": 0.5267,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6127565298507462,
+      "grad_norm": 0.46679943200730223,
+      "learning_rate": 2.1062321142455953e-05,
+      "loss": 0.5187,
+      "step": 5255
+    },
+    {
+      "epoch": 0.613339552238806,
+      "grad_norm": 0.43680321948998807,
+      "learning_rate": 2.1020765226899257e-05,
+      "loss": 0.5072,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6139225746268657,
+      "grad_norm": 0.48765723203673156,
+      "learning_rate": 2.0979233397521237e-05,
+      "loss": 0.5172,
+      "step": 5265
+    },
+    {
+      "epoch": 0.6145055970149254,
+      "grad_norm": 0.5421523270806327,
+      "learning_rate": 2.0937725808714037e-05,
+      "loss": 0.5173,
+      "step": 5270
+    },
+    {
+      "epoch": 0.6150886194029851,
+      "grad_norm": 0.4821842656660482,
+      "learning_rate": 2.089624261477973e-05,
+      "loss": 0.522,
+      "step": 5275
+    },
+    {
+      "epoch": 0.6156716417910447,
+      "grad_norm": 0.473993764265411,
+      "learning_rate": 2.0854783969929668e-05,
+      "loss": 0.5394,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6162546641791045,
+      "grad_norm": 0.4419301919043876,
+      "learning_rate": 2.0813350028283958e-05,
+      "loss": 0.5152,
+      "step": 5285
+    },
+    {
+      "epoch": 0.6168376865671642,
+      "grad_norm": 0.4672525694912507,
+      "learning_rate": 2.0771940943870866e-05,
+      "loss": 0.5195,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6174207089552238,
+      "grad_norm": 0.4420478828415689,
+      "learning_rate": 2.073055687062625e-05,
+      "loss": 0.5131,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6180037313432836,
+      "grad_norm": 0.4626362183961214,
+      "learning_rate": 2.0689197962393007e-05,
+      "loss": 0.5162,
+      "step": 5300
+    },
+    {
+      "epoch": 0.6185867537313433,
+      "grad_norm": 0.49784203270276733,
+      "learning_rate": 2.0647864372920472e-05,
+      "loss": 0.5279,
+      "step": 5305
+    },
+    {
+      "epoch": 0.6191697761194029,
+      "grad_norm": 0.454384681268438,
+      "learning_rate": 2.0606556255863862e-05,
+      "loss": 0.5193,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6197527985074627,
+      "grad_norm": 0.549168830900785,
+      "learning_rate": 2.05652737647837e-05,
+      "loss": 0.508,
+      "step": 5315
+    },
+    {
+      "epoch": 0.6203358208955224,
+      "grad_norm": 0.4790027663801494,
+      "learning_rate": 2.0524017053145238e-05,
+      "loss": 0.5009,
+      "step": 5320
+    },
+    {
+      "epoch": 0.620918843283582,
+      "grad_norm": 0.4858133125717372,
+      "learning_rate": 2.0482786274317923e-05,
+      "loss": 0.5257,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6215018656716418,
+      "grad_norm": 0.4686160974815818,
+      "learning_rate": 2.0441581581574765e-05,
+      "loss": 0.5006,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6220848880597015,
+      "grad_norm": 0.4623666153609126,
+      "learning_rate": 2.0400403128091812e-05,
+      "loss": 0.5169,
+      "step": 5335
+    },
+    {
+      "epoch": 0.6226679104477612,
+      "grad_norm": 0.45877847406013955,
+      "learning_rate": 2.0359251066947583e-05,
+      "loss": 0.5334,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6232509328358209,
+      "grad_norm": 0.4835869467134031,
+      "learning_rate": 2.0318125551122468e-05,
+      "loss": 0.4976,
+      "step": 5345
+    },
+    {
+      "epoch": 0.6238339552238806,
+      "grad_norm": 0.4575233875191603,
+      "learning_rate": 2.027702673349818e-05,
+      "loss": 0.5078,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6244169776119403,
+      "grad_norm": 0.42998314282841227,
+      "learning_rate": 2.023595476685718e-05,
+      "loss": 0.5217,
+      "step": 5355
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.40746461264097383,
+      "learning_rate": 2.0194909803882128e-05,
+      "loss": 0.4901,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6255830223880597,
+      "grad_norm": 0.4597436174722755,
+      "learning_rate": 2.0153891997155282e-05,
+      "loss": 0.5494,
+      "step": 5365
+    },
+    {
+      "epoch": 0.6261660447761194,
+      "grad_norm": 0.47770020364059385,
+      "learning_rate": 2.011290149915795e-05,
+      "loss": 0.5137,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6267490671641791,
+      "grad_norm": 0.4282249414588399,
+      "learning_rate": 2.0071938462269936e-05,
+      "loss": 0.4793,
+      "step": 5375
+    },
+    {
+      "epoch": 0.6273320895522388,
+      "grad_norm": 0.4509595314631354,
+      "learning_rate": 2.0031003038768942e-05,
+      "loss": 0.498,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6279151119402985,
+      "grad_norm": 0.47714728709347026,
+      "learning_rate": 1.999009538083003e-05,
+      "loss": 0.545,
+      "step": 5385
+    },
+    {
+      "epoch": 0.6284981343283582,
+      "grad_norm": 0.47392502530376346,
+      "learning_rate": 1.994921564052503e-05,
+      "loss": 0.5133,
+      "step": 5390
+    },
+    {
+      "epoch": 0.629081156716418,
+      "grad_norm": 0.44759229419435465,
+      "learning_rate": 1.990836396982202e-05,
+      "loss": 0.5262,
+      "step": 5395
+    },
+    {
+      "epoch": 0.6296641791044776,
+      "grad_norm": 0.4912367043364836,
+      "learning_rate": 1.9867540520584693e-05,
+      "loss": 0.5346,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6302472014925373,
+      "grad_norm": 0.4501000804982757,
+      "learning_rate": 1.9826745444571853e-05,
+      "loss": 0.5019,
+      "step": 5405
+    },
+    {
+      "epoch": 0.6308302238805971,
+      "grad_norm": 0.48836277098183645,
+      "learning_rate": 1.978597889343683e-05,
+      "loss": 0.5289,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6314132462686567,
+      "grad_norm": 0.4625223996087579,
+      "learning_rate": 1.97452410187269e-05,
+      "loss": 0.4958,
+      "step": 5415
+    },
+    {
+      "epoch": 0.6319962686567164,
+      "grad_norm": 0.48092056548731926,
+      "learning_rate": 1.970453197188275e-05,
+      "loss": 0.5015,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6325792910447762,
+      "grad_norm": 0.41113077299689604,
+      "learning_rate": 1.9663851904237903e-05,
+      "loss": 0.4989,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6331623134328358,
+      "grad_norm": 0.43567184997194336,
+      "learning_rate": 1.9623200967018134e-05,
+      "loss": 0.5066,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6337453358208955,
+      "grad_norm": 0.480598507617135,
+      "learning_rate": 1.9582579311340943e-05,
+      "loss": 0.5476,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6343283582089553,
+      "grad_norm": 0.4632420890807915,
+      "learning_rate": 1.9541987088214963e-05,
+      "loss": 0.514,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6349113805970149,
+      "grad_norm": 0.461886176881567,
+      "learning_rate": 1.9501424448539445e-05,
+      "loss": 0.4993,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6354944029850746,
+      "grad_norm": 0.4863359836422005,
+      "learning_rate": 1.946089154310364e-05,
+      "loss": 0.5349,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6360774253731343,
+      "grad_norm": 0.44233037285547316,
+      "learning_rate": 1.9420388522586242e-05,
+      "loss": 0.5036,
+      "step": 5455
+    },
+    {
+      "epoch": 0.636660447761194,
+      "grad_norm": 0.4399036507814764,
+      "learning_rate": 1.937991553755491e-05,
+      "loss": 0.5198,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6372434701492538,
+      "grad_norm": 0.4644842288483705,
+      "learning_rate": 1.9339472738465604e-05,
+      "loss": 0.468,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6378264925373134,
+      "grad_norm": 0.5021801630911482,
+      "learning_rate": 1.929906027566207e-05,
+      "loss": 0.5046,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6384095149253731,
+      "grad_norm": 0.4906231561063412,
+      "learning_rate": 1.9258678299375287e-05,
+      "loss": 0.5637,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6389925373134329,
+      "grad_norm": 0.4411840007547938,
+      "learning_rate": 1.9218326959722915e-05,
+      "loss": 0.4941,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6395755597014925,
+      "grad_norm": 0.44492116681266264,
+      "learning_rate": 1.9178006406708716e-05,
+      "loss": 0.4999,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6401585820895522,
+      "grad_norm": 0.45899402350391216,
+      "learning_rate": 1.913771679022199e-05,
+      "loss": 0.5143,
+      "step": 5490
+    },
+    {
+      "epoch": 0.640741604477612,
+      "grad_norm": 0.5002130360013761,
+      "learning_rate": 1.9097458260037055e-05,
+      "loss": 0.5223,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6413246268656716,
+      "grad_norm": 0.4445252627269499,
+      "learning_rate": 1.9057230965812652e-05,
+      "loss": 0.5277,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6419076492537313,
+      "grad_norm": 0.4327961966702548,
+      "learning_rate": 1.901703505709141e-05,
+      "loss": 0.4689,
+      "step": 5505
+    },
+    {
+      "epoch": 0.6424906716417911,
+      "grad_norm": 0.47514425543756855,
+      "learning_rate": 1.897687068329928e-05,
+      "loss": 0.499,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6430736940298507,
+      "grad_norm": 0.4544747659785093,
+      "learning_rate": 1.8936737993744996e-05,
+      "loss": 0.5026,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6436567164179104,
+      "grad_norm": 0.45829680386820026,
+      "learning_rate": 1.8896637137619495e-05,
+      "loss": 0.5066,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6442397388059702,
+      "grad_norm": 0.42928587523633865,
+      "learning_rate": 1.8856568263995373e-05,
+      "loss": 0.4651,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6448227611940298,
+      "grad_norm": 0.4785710795086254,
+      "learning_rate": 1.8816531521826346e-05,
+      "loss": 0.5118,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6454057835820896,
+      "grad_norm": 0.44984484126810564,
+      "learning_rate": 1.8776527059946676e-05,
+      "loss": 0.4979,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6459888059701493,
+      "grad_norm": 0.5084605177342773,
+      "learning_rate": 1.8736555027070607e-05,
+      "loss": 0.513,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6465718283582089,
+      "grad_norm": 0.4587926020250396,
+      "learning_rate": 1.8696615571791876e-05,
+      "loss": 0.5056,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6471548507462687,
+      "grad_norm": 0.5265935639558714,
+      "learning_rate": 1.865670884258307e-05,
+      "loss": 0.5328,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6477378731343284,
+      "grad_norm": 0.48393501644987547,
+      "learning_rate": 1.861683498779514e-05,
+      "loss": 0.5409,
+      "step": 5555
+    },
+    {
+      "epoch": 0.648320895522388,
+      "grad_norm": 0.42845129776042723,
+      "learning_rate": 1.8576994155656814e-05,
+      "loss": 0.5036,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6489039179104478,
+      "grad_norm": 0.45750371585274724,
+      "learning_rate": 1.853718649427409e-05,
+      "loss": 0.5458,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6494869402985075,
+      "grad_norm": 0.5071290594881412,
+      "learning_rate": 1.8497412151629617e-05,
+      "loss": 0.5574,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6500699626865671,
+      "grad_norm": 0.4193141376087725,
+      "learning_rate": 1.8457671275582202e-05,
+      "loss": 0.501,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6506529850746269,
+      "grad_norm": 0.5097091316108154,
+      "learning_rate": 1.841796401386626e-05,
+      "loss": 0.5424,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6512360074626866,
+      "grad_norm": 0.4574079642887348,
+      "learning_rate": 1.8378290514091214e-05,
+      "loss": 0.5143,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6518190298507462,
+      "grad_norm": 0.48232820582938274,
+      "learning_rate": 1.8338650923740984e-05,
+      "loss": 0.5068,
+      "step": 5590
+    },
+    {
+      "epoch": 0.652402052238806,
+      "grad_norm": 0.4295823610973414,
+      "learning_rate": 1.829904539017347e-05,
+      "loss": 0.5076,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6529850746268657,
+      "grad_norm": 0.4893118320191642,
+      "learning_rate": 1.8259474060619925e-05,
+      "loss": 0.5219,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6535680970149254,
+      "grad_norm": 0.44539776731173597,
+      "learning_rate": 1.8219937082184462e-05,
+      "loss": 0.4935,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6541511194029851,
+      "grad_norm": 0.5129311723129513,
+      "learning_rate": 1.8180434601843505e-05,
+      "loss": 0.5179,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6547341417910447,
+      "grad_norm": 0.4864225115581459,
+      "learning_rate": 1.8140966766445235e-05,
+      "loss": 0.4969,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6553171641791045,
+      "grad_norm": 0.47007224857589525,
+      "learning_rate": 1.8101533722709036e-05,
+      "loss": 0.5005,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6559001865671642,
+      "grad_norm": 0.4597974424746557,
+      "learning_rate": 1.8062135617224933e-05,
+      "loss": 0.5294,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6564832089552238,
+      "grad_norm": 0.49917347832714665,
+      "learning_rate": 1.802277259645313e-05,
+      "loss": 0.5261,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6570662313432836,
+      "grad_norm": 0.46525260879023533,
+      "learning_rate": 1.798344480672334e-05,
+      "loss": 0.5119,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6576492537313433,
+      "grad_norm": 0.49319911982873976,
+      "learning_rate": 1.7944152394234354e-05,
+      "loss": 0.4884,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6582322761194029,
+      "grad_norm": 0.5132851806324479,
+      "learning_rate": 1.7904895505053405e-05,
+      "loss": 0.5158,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6588152985074627,
+      "grad_norm": 0.47569949694794605,
+      "learning_rate": 1.7865674285115735e-05,
+      "loss": 0.5121,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6593983208955224,
+      "grad_norm": 0.44552199263042797,
+      "learning_rate": 1.7826488880223913e-05,
+      "loss": 0.4761,
+      "step": 5655
+    },
+    {
+      "epoch": 0.659981343283582,
+      "grad_norm": 0.47720149730201705,
+      "learning_rate": 1.778733943604743e-05,
+      "loss": 0.5124,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6605643656716418,
+      "grad_norm": 0.4432509370944976,
+      "learning_rate": 1.774822609812205e-05,
+      "loss": 0.4711,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6611473880597015,
+      "grad_norm": 0.4450343141290362,
+      "learning_rate": 1.7709149011849364e-05,
+      "loss": 0.4761,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6617304104477612,
+      "grad_norm": 0.4797782997827445,
+      "learning_rate": 1.767010832249613e-05,
+      "loss": 0.5175,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6623134328358209,
+      "grad_norm": 0.47463931278850086,
+      "learning_rate": 1.7631104175193863e-05,
+      "loss": 0.5208,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6628964552238806,
+      "grad_norm": 0.49271476148028914,
+      "learning_rate": 1.7592136714938206e-05,
+      "loss": 0.522,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6634794776119403,
+      "grad_norm": 0.4366587132398326,
+      "learning_rate": 1.755320608658844e-05,
+      "loss": 0.4709,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6640625,
+      "grad_norm": 0.44579614307494786,
+      "learning_rate": 1.7514312434866904e-05,
+      "loss": 0.5267,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6646455223880597,
+      "grad_norm": 0.4487339036982587,
+      "learning_rate": 1.747545590435848e-05,
+      "loss": 0.4991,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6652285447761194,
+      "grad_norm": 0.4383027709136693,
+      "learning_rate": 1.7436636639510082e-05,
+      "loss": 0.5141,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6658115671641791,
+      "grad_norm": 0.4458649709251083,
+      "learning_rate": 1.739785478463004e-05,
+      "loss": 0.4921,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6663945895522388,
+      "grad_norm": 0.49612990301681104,
+      "learning_rate": 1.735911048388768e-05,
+      "loss": 0.5081,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6669776119402985,
+      "grad_norm": 0.4717384137501139,
+      "learning_rate": 1.7320403881312665e-05,
+      "loss": 0.4909,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6675606343283582,
+      "grad_norm": 0.5029540121644192,
+      "learning_rate": 1.7281735120794555e-05,
+      "loss": 0.5439,
+      "step": 5725
+    },
+    {
+      "epoch": 0.668143656716418,
+      "grad_norm": 0.4643667414061285,
+      "learning_rate": 1.7243104346082194e-05,
+      "loss": 0.4794,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6687266791044776,
+      "grad_norm": 0.47417180490760813,
+      "learning_rate": 1.720451170078328e-05,
+      "loss": 0.4996,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6693097014925373,
+      "grad_norm": 0.4814261614829366,
+      "learning_rate": 1.7165957328363703e-05,
+      "loss": 0.5027,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6698927238805971,
+      "grad_norm": 0.45627638279169563,
+      "learning_rate": 1.71274413721471e-05,
+      "loss": 0.513,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6704757462686567,
+      "grad_norm": 0.4334633223796316,
+      "learning_rate": 1.708896397531431e-05,
+      "loss": 0.4659,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6710587686567164,
+      "grad_norm": 0.4579206868255898,
+      "learning_rate": 1.7050525280902824e-05,
+      "loss": 0.5158,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6716417910447762,
+      "grad_norm": 0.4386284661436938,
+      "learning_rate": 1.701212543180623e-05,
+      "loss": 0.5206,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6722248134328358,
+      "grad_norm": 0.45478097582209165,
+      "learning_rate": 1.6973764570773766e-05,
+      "loss": 0.4958,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6728078358208955,
+      "grad_norm": 0.49448818817184126,
+      "learning_rate": 1.693544284040968e-05,
+      "loss": 0.532,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6733908582089553,
+      "grad_norm": 0.44392640275278356,
+      "learning_rate": 1.6897160383172794e-05,
+      "loss": 0.5102,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6739738805970149,
+      "grad_norm": 0.4681803651785721,
+      "learning_rate": 1.6858917341375893e-05,
+      "loss": 0.5033,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6745569029850746,
+      "grad_norm": 0.5012783536207814,
+      "learning_rate": 1.6820713857185296e-05,
+      "loss": 0.5261,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6751399253731343,
+      "grad_norm": 0.5021117339189587,
+      "learning_rate": 1.6782550072620208e-05,
+      "loss": 0.4987,
+      "step": 5790
+    },
+    {
+      "epoch": 0.675722947761194,
+      "grad_norm": 0.48367815094839645,
+      "learning_rate": 1.674442612955229e-05,
+      "loss": 0.5231,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6763059701492538,
+      "grad_norm": 0.42292867586088734,
+      "learning_rate": 1.6706342169705092e-05,
+      "loss": 0.4833,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6768889925373134,
+      "grad_norm": 0.5498079373194198,
+      "learning_rate": 1.6668298334653504e-05,
+      "loss": 0.5303,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6774720149253731,
+      "grad_norm": 0.45613249161600566,
+      "learning_rate": 1.663029476582328e-05,
+      "loss": 0.4946,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6780550373134329,
+      "grad_norm": 0.4300584687753966,
+      "learning_rate": 1.659233160449048e-05,
+      "loss": 0.4835,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6786380597014925,
+      "grad_norm": 0.47672064051122565,
+      "learning_rate": 1.6554408991780958e-05,
+      "loss": 0.5239,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6792210820895522,
+      "grad_norm": 0.4363678330399839,
+      "learning_rate": 1.65165270686698e-05,
+      "loss": 0.5095,
+      "step": 5825
+    },
+    {
+      "epoch": 0.679804104477612,
+      "grad_norm": 0.4386976710048956,
+      "learning_rate": 1.6478685975980867e-05,
+      "loss": 0.5037,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6803871268656716,
+      "grad_norm": 0.4713692579068845,
+      "learning_rate": 1.6440885854386223e-05,
+      "loss": 0.4919,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6809701492537313,
+      "grad_norm": 0.4590294577914819,
+      "learning_rate": 1.6403126844405627e-05,
+      "loss": 0.5001,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6815531716417911,
+      "grad_norm": 0.48010469655748705,
+      "learning_rate": 1.6365409086405982e-05,
+      "loss": 0.523,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6821361940298507,
+      "grad_norm": 0.4572102553531434,
+      "learning_rate": 1.6327732720600893e-05,
+      "loss": 0.4981,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6827192164179104,
+      "grad_norm": 0.42279317577956194,
+      "learning_rate": 1.6290097887050037e-05,
+      "loss": 0.4824,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6833022388059702,
+      "grad_norm": 0.4672338547985673,
+      "learning_rate": 1.6252504725658738e-05,
+      "loss": 0.5276,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6838852611940298,
+      "grad_norm": 0.5057683989918992,
+      "learning_rate": 1.6214953376177355e-05,
+      "loss": 0.5108,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6844682835820896,
+      "grad_norm": 0.4285897123749069,
+      "learning_rate": 1.617744397820089e-05,
+      "loss": 0.4943,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6850513059701493,
+      "grad_norm": 0.47280662245843785,
+      "learning_rate": 1.613997667116832e-05,
+      "loss": 0.5127,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6856343283582089,
+      "grad_norm": 0.467214604689162,
+      "learning_rate": 1.610255159436219e-05,
+      "loss": 0.4953,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6862173507462687,
+      "grad_norm": 0.4452219391184069,
+      "learning_rate": 1.6065168886908046e-05,
+      "loss": 0.4957,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6868003731343284,
+      "grad_norm": 0.44301870809154087,
+      "learning_rate": 1.6027828687773947e-05,
+      "loss": 0.4975,
+      "step": 5890
+    },
+    {
+      "epoch": 0.687383395522388,
+      "grad_norm": 0.4401277833271277,
+      "learning_rate": 1.5990531135769885e-05,
+      "loss": 0.5195,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6879664179104478,
+      "grad_norm": 2.0394489198103662,
+      "learning_rate": 1.5953276369547356e-05,
+      "loss": 0.4956,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6885494402985075,
+      "grad_norm": 0.46836588645884025,
+      "learning_rate": 1.591606452759879e-05,
+      "loss": 0.5327,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6891324626865671,
+      "grad_norm": 0.44598757398103644,
+      "learning_rate": 1.587889574825705e-05,
+      "loss": 0.4764,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6897154850746269,
+      "grad_norm": 0.49742861746687556,
+      "learning_rate": 1.5841770169694895e-05,
+      "loss": 0.5111,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6902985074626866,
+      "grad_norm": 0.47331933429323847,
+      "learning_rate": 1.5804687929924522e-05,
+      "loss": 0.5047,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6908815298507462,
+      "grad_norm": 0.4652913554765252,
+      "learning_rate": 1.5767649166796995e-05,
+      "loss": 0.5038,
+      "step": 5925
+    },
+    {
+      "epoch": 0.691464552238806,
+      "grad_norm": 0.453898586144349,
+      "learning_rate": 1.573065401800176e-05,
+      "loss": 0.4755,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6920475746268657,
+      "grad_norm": 0.43523535343423403,
+      "learning_rate": 1.569370262106615e-05,
+      "loss": 0.553,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6926305970149254,
+      "grad_norm": 0.46238104794971113,
+      "learning_rate": 1.5656795113354816e-05,
+      "loss": 0.4496,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6932136194029851,
+      "grad_norm": 0.4612926915904917,
+      "learning_rate": 1.5619931632069284e-05,
+      "loss": 0.5039,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6937966417910447,
+      "grad_norm": 0.46826494703709465,
+      "learning_rate": 1.5583112314247386e-05,
+      "loss": 0.5022,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6943796641791045,
+      "grad_norm": 0.47987543937588256,
+      "learning_rate": 1.5546337296762826e-05,
+      "loss": 0.5097,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6949626865671642,
+      "grad_norm": 0.43334758409327906,
+      "learning_rate": 1.5509606716324563e-05,
+      "loss": 0.4943,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6955457089552238,
+      "grad_norm": 0.4681763383562353,
+      "learning_rate": 1.547292070947641e-05,
+      "loss": 0.5063,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6961287313432836,
+      "grad_norm": 0.4924802920273831,
+      "learning_rate": 1.5436279412596466e-05,
+      "loss": 0.5047,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6967117537313433,
+      "grad_norm": 0.5347314807147506,
+      "learning_rate": 1.5399682961896627e-05,
+      "loss": 0.5254,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6972947761194029,
+      "grad_norm": 0.46161765208167127,
+      "learning_rate": 1.5363131493422045e-05,
+      "loss": 0.5527,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6978777985074627,
+      "grad_norm": 0.481592995924184,
+      "learning_rate": 1.5326625143050717e-05,
+      "loss": 0.5093,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6984608208955224,
+      "grad_norm": 0.43698502229598946,
+      "learning_rate": 1.5290164046492855e-05,
+      "loss": 0.5113,
+      "step": 5990
+    },
+    {
+      "epoch": 0.699043843283582,
+      "grad_norm": 0.45737599000837936,
+      "learning_rate": 1.5253748339290478e-05,
+      "loss": 0.5072,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6996268656716418,
+      "grad_norm": 0.4632305738785457,
+      "learning_rate": 1.5217378156816836e-05,
+      "loss": 0.5092,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7002098880597015,
+      "grad_norm": 0.5658113543744521,
+      "learning_rate": 1.5181053634276005e-05,
+      "loss": 0.4714,
+      "step": 6005
+    },
+    {
+      "epoch": 0.7007929104477612,
+      "grad_norm": 0.5024081912461719,
+      "learning_rate": 1.5144774906702261e-05,
+      "loss": 0.5587,
+      "step": 6010
+    },
+    {
+      "epoch": 0.7013759328358209,
+      "grad_norm": 0.4346747521893763,
+      "learning_rate": 1.5108542108959666e-05,
+      "loss": 0.4874,
+      "step": 6015
+    },
+    {
+      "epoch": 0.7019589552238806,
+      "grad_norm": 0.48751850242768197,
+      "learning_rate": 1.5072355375741564e-05,
+      "loss": 0.5152,
+      "step": 6020
+    },
+    {
+      "epoch": 0.7025419776119403,
+      "grad_norm": 0.4679030016933106,
+      "learning_rate": 1.5036214841570002e-05,
+      "loss": 0.5177,
+      "step": 6025
+    },
+    {
+      "epoch": 0.703125,
+      "grad_norm": 0.43192813973658906,
+      "learning_rate": 1.500012064079533e-05,
+      "loss": 0.5381,
+      "step": 6030
+    },
+    {
+      "epoch": 0.7037080223880597,
+      "grad_norm": 0.47334729350720534,
+      "learning_rate": 1.4964072907595633e-05,
+      "loss": 0.5428,
+      "step": 6035
+    },
+    {
+      "epoch": 0.7042910447761194,
+      "grad_norm": 0.4625335667175621,
+      "learning_rate": 1.4928071775976283e-05,
+      "loss": 0.4965,
+      "step": 6040
+    },
+    {
+      "epoch": 0.7048740671641791,
+      "grad_norm": 0.4837863609737646,
+      "learning_rate": 1.489211737976937e-05,
+      "loss": 0.5201,
+      "step": 6045
+    },
+    {
+      "epoch": 0.7054570895522388,
+      "grad_norm": 0.45448831055589556,
+      "learning_rate": 1.4856209852633282e-05,
+      "loss": 0.4988,
+      "step": 6050
+    },
+    {
+      "epoch": 0.7060401119402985,
+      "grad_norm": 0.4470633313561567,
+      "learning_rate": 1.482034932805217e-05,
+      "loss": 0.4923,
+      "step": 6055
+    },
+    {
+      "epoch": 0.7066231343283582,
+      "grad_norm": 0.46146957236249214,
+      "learning_rate": 1.478453593933545e-05,
+      "loss": 0.4966,
+      "step": 6060
+    },
+    {
+      "epoch": 0.707206156716418,
+      "grad_norm": 0.44846726296477046,
+      "learning_rate": 1.4748769819617291e-05,
+      "loss": 0.5121,
+      "step": 6065
+    },
+    {
+      "epoch": 0.7077891791044776,
+      "grad_norm": 0.445769696660708,
+      "learning_rate": 1.47130511018562e-05,
+      "loss": 0.49,
+      "step": 6070
+    },
+    {
+      "epoch": 0.7083722014925373,
+      "grad_norm": 0.44424913913430536,
+      "learning_rate": 1.4677379918834408e-05,
+      "loss": 0.5154,
+      "step": 6075
+    },
+    {
+      "epoch": 0.7089552238805971,
+      "grad_norm": 0.5025893980492081,
+      "learning_rate": 1.464175640315748e-05,
+      "loss": 0.4886,
+      "step": 6080
+    },
+    {
+      "epoch": 0.7095382462686567,
+      "grad_norm": 0.4287175183865977,
+      "learning_rate": 1.460618068725374e-05,
+      "loss": 0.4943,
+      "step": 6085
+    },
+    {
+      "epoch": 0.7101212686567164,
+      "grad_norm": 0.46657717095673207,
+      "learning_rate": 1.4570652903373877e-05,
+      "loss": 0.4866,
+      "step": 6090
+    },
+    {
+      "epoch": 0.7107042910447762,
+      "grad_norm": 0.44658175711235154,
+      "learning_rate": 1.453517318359034e-05,
+      "loss": 0.4952,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7112873134328358,
+      "grad_norm": 0.5109217950300224,
+      "learning_rate": 1.4499741659796927e-05,
+      "loss": 0.5495,
+      "step": 6100
+    },
+    {
+      "epoch": 0.7118703358208955,
+      "grad_norm": 0.439976204898742,
+      "learning_rate": 1.4464358463708277e-05,
+      "loss": 0.4947,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7124533582089553,
+      "grad_norm": 0.442606078185802,
+      "learning_rate": 1.442902372685937e-05,
+      "loss": 0.5125,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7130363805970149,
+      "grad_norm": 0.5113814435199087,
+      "learning_rate": 1.4393737580605019e-05,
+      "loss": 0.5397,
+      "step": 6115
+    },
+    {
+      "epoch": 0.7136194029850746,
+      "grad_norm": 0.4292248926969336,
+      "learning_rate": 1.435850015611943e-05,
+      "loss": 0.5038,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7142024253731343,
+      "grad_norm": 0.4436015934650438,
+      "learning_rate": 1.432331158439568e-05,
+      "loss": 0.5015,
+      "step": 6125
+    },
+    {
+      "epoch": 0.714785447761194,
+      "grad_norm": 0.4679393946330053,
+      "learning_rate": 1.4288171996245247e-05,
+      "loss": 0.4961,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7153684701492538,
+      "grad_norm": 0.46822261278058663,
+      "learning_rate": 1.425308152229749e-05,
+      "loss": 0.5565,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7159514925373134,
+      "grad_norm": 0.48866887191385927,
+      "learning_rate": 1.4218040292999221e-05,
+      "loss": 0.531,
+      "step": 6140
+    },
+    {
+      "epoch": 0.7165345149253731,
+      "grad_norm": 0.4487435192453521,
+      "learning_rate": 1.4183048438614166e-05,
+      "loss": 0.4994,
+      "step": 6145
+    },
+    {
+      "epoch": 0.7171175373134329,
+      "grad_norm": 0.5036947554976395,
+      "learning_rate": 1.4148106089222513e-05,
+      "loss": 0.519,
+      "step": 6150
+    },
+    {
+      "epoch": 0.7177005597014925,
+      "grad_norm": 0.49765078956448927,
+      "learning_rate": 1.4113213374720425e-05,
+      "loss": 0.4825,
+      "step": 6155
+    },
+    {
+      "epoch": 0.7182835820895522,
+      "grad_norm": 0.49796549587751004,
+      "learning_rate": 1.4078370424819515e-05,
+      "loss": 0.5142,
+      "step": 6160
+    },
+    {
+      "epoch": 0.718866604477612,
+      "grad_norm": 0.46661776742591504,
+      "learning_rate": 1.404357736904645e-05,
+      "loss": 0.5121,
+      "step": 6165
+    },
+    {
+      "epoch": 0.7194496268656716,
+      "grad_norm": 0.4206361496368435,
+      "learning_rate": 1.4008834336742366e-05,
+      "loss": 0.5043,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7200326492537313,
+      "grad_norm": 0.49140533816343374,
+      "learning_rate": 1.3974141457062498e-05,
+      "loss": 0.5465,
+      "step": 6175
+    },
+    {
+      "epoch": 0.7206156716417911,
+      "grad_norm": 0.4730841055221398,
+      "learning_rate": 1.3939498858975584e-05,
+      "loss": 0.496,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7211986940298507,
+      "grad_norm": 0.5165044754977881,
+      "learning_rate": 1.390490667126348e-05,
+      "loss": 0.5699,
+      "step": 6185
+    },
+    {
+      "epoch": 0.7217817164179104,
+      "grad_norm": 0.4574833927556042,
+      "learning_rate": 1.3870365022520627e-05,
+      "loss": 0.494,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7223647388059702,
+      "grad_norm": 0.4797480787975331,
+      "learning_rate": 1.3835874041153607e-05,
+      "loss": 0.5,
+      "step": 6195
+    },
+    {
+      "epoch": 0.7229477611940298,
+      "grad_norm": 0.5266171413115511,
+      "learning_rate": 1.380143385538063e-05,
+      "loss": 0.5248,
+      "step": 6200
+    },
+    {
+      "epoch": 0.7235307835820896,
+      "grad_norm": 0.4787651159553734,
+      "learning_rate": 1.3767044593231082e-05,
+      "loss": 0.4999,
+      "step": 6205
+    },
+    {
+      "epoch": 0.7241138059701493,
+      "grad_norm": 0.42275543868072146,
+      "learning_rate": 1.3732706382545054e-05,
+      "loss": 0.4731,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7246968283582089,
+      "grad_norm": 0.48666619671292066,
+      "learning_rate": 1.3698419350972851e-05,
+      "loss": 0.5242,
+      "step": 6215
+    },
+    {
+      "epoch": 0.7252798507462687,
+      "grad_norm": 0.41898376072456794,
+      "learning_rate": 1.3664183625974503e-05,
+      "loss": 0.5104,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7258628731343284,
+      "grad_norm": 0.4929104768550004,
+      "learning_rate": 1.362999933481935e-05,
+      "loss": 0.5206,
+      "step": 6225
+    },
+    {
+      "epoch": 0.726445895522388,
+      "grad_norm": 0.4308013069303933,
+      "learning_rate": 1.3595866604585492e-05,
+      "loss": 0.5074,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7270289179104478,
+      "grad_norm": 0.4421936343319682,
+      "learning_rate": 1.3561785562159374e-05,
+      "loss": 0.4808,
+      "step": 6235
+    },
+    {
+      "epoch": 0.7276119402985075,
+      "grad_norm": 0.4980669606106666,
+      "learning_rate": 1.3527756334235288e-05,
+      "loss": 0.4746,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7281949626865671,
+      "grad_norm": 0.4364859749653744,
+      "learning_rate": 1.3493779047314925e-05,
+      "loss": 0.4967,
+      "step": 6245
+    },
+    {
+      "epoch": 0.7287779850746269,
+      "grad_norm": 0.4424531855538569,
+      "learning_rate": 1.3459853827706853e-05,
+      "loss": 0.4962,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7293610074626866,
+      "grad_norm": 0.5043638790699465,
+      "learning_rate": 1.3425980801526118e-05,
+      "loss": 0.5095,
+      "step": 6255
+    },
+    {
+      "epoch": 0.7299440298507462,
+      "grad_norm": 0.464384451619953,
+      "learning_rate": 1.3392160094693724e-05,
+      "loss": 0.5008,
+      "step": 6260
+    },
+    {
+      "epoch": 0.730527052238806,
+      "grad_norm": 0.44362217621259836,
+      "learning_rate": 1.3358391832936174e-05,
+      "loss": 0.4965,
+      "step": 6265
+    },
+    {
+      "epoch": 0.7311100746268657,
+      "grad_norm": 0.4092013778366071,
+      "learning_rate": 1.3324676141785029e-05,
+      "loss": 0.5133,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7316930970149254,
+      "grad_norm": 0.4350575485273892,
+      "learning_rate": 1.3291013146576403e-05,
+      "loss": 0.5128,
+      "step": 6275
+    },
+    {
+      "epoch": 0.7322761194029851,
+      "grad_norm": 0.4744197561421012,
+      "learning_rate": 1.3257402972450539e-05,
+      "loss": 0.4784,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7328591417910447,
+      "grad_norm": 0.4633376651924879,
+      "learning_rate": 1.3223845744351287e-05,
+      "loss": 0.475,
+      "step": 6285
+    },
+    {
+      "epoch": 0.7334421641791045,
+      "grad_norm": 0.47064248305550177,
+      "learning_rate": 1.3190341587025698e-05,
+      "loss": 0.5147,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7340251865671642,
+      "grad_norm": 0.4745007087563231,
+      "learning_rate": 1.3156890625023532e-05,
+      "loss": 0.5131,
+      "step": 6295
+    },
+    {
+      "epoch": 0.7346082089552238,
+      "grad_norm": 0.47910451161364465,
+      "learning_rate": 1.3123492982696806e-05,
+      "loss": 0.5125,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7351912313432836,
+      "grad_norm": 0.45150153399392984,
+      "learning_rate": 1.3090148784199288e-05,
+      "loss": 0.5195,
+      "step": 6305
+    },
+    {
+      "epoch": 0.7357742537313433,
+      "grad_norm": 0.4666852788990984,
+      "learning_rate": 1.305685815348613e-05,
+      "loss": 0.5348,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7363572761194029,
+      "grad_norm": 0.43969109543694684,
+      "learning_rate": 1.3023621214313289e-05,
+      "loss": 0.4887,
+      "step": 6315
+    },
+    {
+      "epoch": 0.7369402985074627,
+      "grad_norm": 0.4520730263820748,
+      "learning_rate": 1.2990438090237167e-05,
+      "loss": 0.5174,
+      "step": 6320
+    },
+    {
+      "epoch": 0.7375233208955224,
+      "grad_norm": 0.4531016913171499,
+      "learning_rate": 1.2957308904614099e-05,
+      "loss": 0.513,
+      "step": 6325
+    },
+    {
+      "epoch": 0.738106343283582,
+      "grad_norm": 0.43990578576591227,
+      "learning_rate": 1.2924233780599915e-05,
+      "loss": 0.469,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7386893656716418,
+      "grad_norm": 0.44084700165618407,
+      "learning_rate": 1.2891212841149447e-05,
+      "loss": 0.4997,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7392723880597015,
+      "grad_norm": 0.49100928132633287,
+      "learning_rate": 1.2858246209016128e-05,
+      "loss": 0.5187,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7398554104477612,
+      "grad_norm": 1.5928599989218044,
+      "learning_rate": 1.2825334006751493e-05,
+      "loss": 0.4954,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7404384328358209,
+      "grad_norm": 0.4278160829593741,
+      "learning_rate": 1.2792476356704759e-05,
+      "loss": 0.462,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7410214552238806,
+      "grad_norm": 0.45173765932520604,
+      "learning_rate": 1.2759673381022305e-05,
+      "loss": 0.5198,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7416044776119403,
+      "grad_norm": 0.46060933457902786,
+      "learning_rate": 1.27269252016473e-05,
+      "loss": 0.501,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7421875,
+      "grad_norm": 0.5194412989177508,
+      "learning_rate": 1.2694231940319192e-05,
+      "loss": 0.4862,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7427705223880597,
+      "grad_norm": 0.46637710292865725,
+      "learning_rate": 1.2661593718573294e-05,
+      "loss": 0.4913,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7433535447761194,
+      "grad_norm": 0.49083348251147924,
+      "learning_rate": 1.2629010657740275e-05,
+      "loss": 0.5073,
+      "step": 6375
+    },
+    {
+      "epoch": 0.7439365671641791,
+      "grad_norm": 0.476732025190355,
+      "learning_rate": 1.2596482878945787e-05,
+      "loss": 0.5343,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7445195895522388,
+      "grad_norm": 0.43400279583829604,
+      "learning_rate": 1.2564010503109952e-05,
+      "loss": 0.4952,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7451026119402985,
+      "grad_norm": 0.4474931990549133,
+      "learning_rate": 1.2531593650946932e-05,
+      "loss": 0.4966,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7456856343283582,
+      "grad_norm": 0.4746293119307018,
+      "learning_rate": 1.2499232442964506e-05,
+      "loss": 0.5241,
+      "step": 6395
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 0.4596957874880727,
+      "learning_rate": 1.2466926999463575e-05,
+      "loss": 0.4931,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7468516791044776,
+      "grad_norm": 0.48590988974018173,
+      "learning_rate": 1.2434677440537745e-05,
+      "loss": 0.5498,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7474347014925373,
+      "grad_norm": 0.43144811701119584,
+      "learning_rate": 1.2402483886072883e-05,
+      "loss": 0.4673,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7480177238805971,
+      "grad_norm": 0.4694370814956435,
+      "learning_rate": 1.237034645574666e-05,
+      "loss": 0.5173,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7486007462686567,
+      "grad_norm": 0.5215126950510659,
+      "learning_rate": 1.233826526902809e-05,
+      "loss": 0.5413,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7491837686567164,
+      "grad_norm": 0.4667853644040054,
+      "learning_rate": 1.230624044517713e-05,
+      "loss": 0.5206,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7497667910447762,
+      "grad_norm": 0.49822013134556736,
+      "learning_rate": 1.2274272103244201e-05,
+      "loss": 0.5264,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7503498134328358,
+      "grad_norm": 0.4924144531446238,
+      "learning_rate": 1.2242360362069763e-05,
+      "loss": 0.4843,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7509328358208955,
+      "grad_norm": 0.47214741828129897,
+      "learning_rate": 1.2210505340283838e-05,
+      "loss": 0.529,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7515158582089553,
+      "grad_norm": 0.4726759175626485,
+      "learning_rate": 1.2178707156305644e-05,
+      "loss": 0.4993,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7520988805970149,
+      "grad_norm": 0.47497735759010773,
+      "learning_rate": 1.2146965928343062e-05,
+      "loss": 0.4923,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7526819029850746,
+      "grad_norm": 0.5282112383115118,
+      "learning_rate": 1.2115281774392278e-05,
+      "loss": 0.5043,
+      "step": 6455
+    },
+    {
+      "epoch": 0.7532649253731343,
+      "grad_norm": 0.4289024052469388,
+      "learning_rate": 1.208365481223727e-05,
+      "loss": 0.4915,
+      "step": 6460
+    },
+    {
+      "epoch": 0.753847947761194,
+      "grad_norm": 0.4527587796867484,
+      "learning_rate": 1.2052085159449455e-05,
+      "loss": 0.491,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7544309701492538,
+      "grad_norm": 0.46413997296204734,
+      "learning_rate": 1.202057293338717e-05,
+      "loss": 0.5207,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7550139925373134,
+      "grad_norm": 0.47431690969030404,
+      "learning_rate": 1.1989118251195284e-05,
+      "loss": 0.4807,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7555970149253731,
+      "grad_norm": 0.6124168305417785,
+      "learning_rate": 1.1957721229804761e-05,
+      "loss": 0.4909,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7561800373134329,
+      "grad_norm": 0.4612146149691418,
+      "learning_rate": 1.1926381985932186e-05,
+      "loss": 0.4912,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7567630597014925,
+      "grad_norm": 0.4602032665320846,
+      "learning_rate": 1.1895100636079387e-05,
+      "loss": 0.5287,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7573460820895522,
+      "grad_norm": 0.44850719377837156,
+      "learning_rate": 1.186387729653296e-05,
+      "loss": 0.5147,
+      "step": 6495
+    },
+    {
+      "epoch": 0.757929104477612,
+      "grad_norm": 0.43547527328367247,
+      "learning_rate": 1.1832712083363865e-05,
+      "loss": 0.4774,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7585121268656716,
+      "grad_norm": 0.48648908094016996,
+      "learning_rate": 1.1801605112426953e-05,
+      "loss": 0.4948,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7590951492537313,
+      "grad_norm": 0.4362070385676467,
+      "learning_rate": 1.1770556499360593e-05,
+      "loss": 0.4768,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7596781716417911,
+      "grad_norm": 0.4594448702342764,
+      "learning_rate": 1.1739566359586195e-05,
+      "loss": 0.5192,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7602611940298507,
+      "grad_norm": 0.4895001977535846,
+      "learning_rate": 1.170863480830781e-05,
+      "loss": 0.5029,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7608442164179104,
+      "grad_norm": 0.4811285447973769,
+      "learning_rate": 1.167776196051166e-05,
+      "loss": 0.5089,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7614272388059702,
+      "grad_norm": 0.4169504765893255,
+      "learning_rate": 1.1646947930965795e-05,
+      "loss": 0.4477,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7620102611940298,
+      "grad_norm": 0.4342791005744433,
+      "learning_rate": 1.1616192834219553e-05,
+      "loss": 0.4957,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7625932835820896,
+      "grad_norm": 0.500766952555154,
+      "learning_rate": 1.1585496784603234e-05,
+      "loss": 0.474,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7631763059701493,
+      "grad_norm": 0.459834987252376,
+      "learning_rate": 1.15548598962276e-05,
+      "loss": 0.4653,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7637593283582089,
+      "grad_norm": 0.4634667523018127,
+      "learning_rate": 1.1524282282983526e-05,
+      "loss": 0.4952,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7643423507462687,
+      "grad_norm": 0.48306231366385766,
+      "learning_rate": 1.1493764058541493e-05,
+      "loss": 0.5092,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7649253731343284,
+      "grad_norm": 0.4412886363099704,
+      "learning_rate": 1.1463305336351233e-05,
+      "loss": 0.4836,
+      "step": 6560
+    },
+    {
+      "epoch": 0.765508395522388,
+      "grad_norm": 0.4594326939876668,
+      "learning_rate": 1.143290622964128e-05,
+      "loss": 0.4858,
+      "step": 6565
+    },
+    {
+      "epoch": 0.7660914179104478,
+      "grad_norm": 0.495560886941702,
+      "learning_rate": 1.1402566851418545e-05,
+      "loss": 0.484,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7666744402985075,
+      "grad_norm": 0.44698231009276274,
+      "learning_rate": 1.1372287314467896e-05,
+      "loss": 0.4938,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7672574626865671,
+      "grad_norm": 0.48489845254460334,
+      "learning_rate": 1.1342067731351754e-05,
+      "loss": 0.5349,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7678404850746269,
+      "grad_norm": 0.3983706507650223,
+      "learning_rate": 1.1311908214409666e-05,
+      "loss": 0.4916,
+      "step": 6585
+    },
+    {
+      "epoch": 0.7684235074626866,
+      "grad_norm": 0.4829203676168014,
+      "learning_rate": 1.128180887575789e-05,
+      "loss": 0.5538,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7690065298507462,
+      "grad_norm": 0.5181252743781259,
+      "learning_rate": 1.1251769827288953e-05,
+      "loss": 0.5103,
+      "step": 6595
+    },
+    {
+      "epoch": 0.769589552238806,
+      "grad_norm": 0.45245249797253495,
+      "learning_rate": 1.122179118067128e-05,
+      "loss": 0.4877,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7701725746268657,
+      "grad_norm": 0.45213952635934074,
+      "learning_rate": 1.1191873047348743e-05,
+      "loss": 0.4699,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7707555970149254,
+      "grad_norm": 0.4384946771009236,
+      "learning_rate": 1.1162015538540268e-05,
+      "loss": 0.5199,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7713386194029851,
+      "grad_norm": 1.4634272550071448,
+      "learning_rate": 1.1132218765239417e-05,
+      "loss": 0.5111,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7719216417910447,
+      "grad_norm": 0.4528179441810264,
+      "learning_rate": 1.1102482838213945e-05,
+      "loss": 0.5095,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7725046641791045,
+      "grad_norm": 0.48790948502665343,
+      "learning_rate": 1.1072807868005438e-05,
+      "loss": 0.5375,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7730876865671642,
+      "grad_norm": 0.43356020787990573,
+      "learning_rate": 1.104319396492888e-05,
+      "loss": 0.494,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7736707089552238,
+      "grad_norm": 0.46568426433428856,
+      "learning_rate": 1.1013641239072233e-05,
+      "loss": 0.5089,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7742537313432836,
+      "grad_norm": 0.5058210347479575,
+      "learning_rate": 1.098414980029603e-05,
+      "loss": 0.5278,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7748367537313433,
+      "grad_norm": 0.46834473900137596,
+      "learning_rate": 1.0954719758232983e-05,
+      "loss": 0.5183,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7754197761194029,
+      "grad_norm": 0.4485355699073728,
+      "learning_rate": 1.092535122228757e-05,
+      "loss": 0.5067,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7760027985074627,
+      "grad_norm": 0.44029418480334825,
+      "learning_rate": 1.0896044301635616e-05,
+      "loss": 0.4902,
+      "step": 6655
+    },
+    {
+      "epoch": 0.7765858208955224,
+      "grad_norm": 0.46043398790080686,
+      "learning_rate": 1.0866799105223877e-05,
+      "loss": 0.4752,
+      "step": 6660
+    },
+    {
+      "epoch": 0.777168843283582,
+      "grad_norm": 0.45411148412679725,
+      "learning_rate": 1.0837615741769695e-05,
+      "loss": 0.5027,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7777518656716418,
+      "grad_norm": 0.46002535963451235,
+      "learning_rate": 1.0808494319760511e-05,
+      "loss": 0.4818,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7783348880597015,
+      "grad_norm": 0.4613160618534238,
+      "learning_rate": 1.0779434947453531e-05,
+      "loss": 0.5305,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7789179104477612,
+      "grad_norm": 0.5192961001346327,
+      "learning_rate": 1.0750437732875265e-05,
+      "loss": 0.4909,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7795009328358209,
+      "grad_norm": 0.47703840671844866,
+      "learning_rate": 1.0721502783821194e-05,
+      "loss": 0.5433,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7800839552238806,
+      "grad_norm": 0.4027202848278968,
+      "learning_rate": 1.0692630207855296e-05,
+      "loss": 0.4795,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7806669776119403,
+      "grad_norm": 0.5254895186437003,
+      "learning_rate": 1.0663820112309695e-05,
+      "loss": 0.5234,
+      "step": 6695
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.4714359353636831,
+      "learning_rate": 1.0635072604284254e-05,
+      "loss": 0.4837,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7818330223880597,
+      "grad_norm": 0.4409933416168129,
+      "learning_rate": 1.0606387790646154e-05,
+      "loss": 0.5124,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7824160447761194,
+      "grad_norm": 0.47666045358791953,
+      "learning_rate": 1.0577765778029525e-05,
+      "loss": 0.4762,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7829990671641791,
+      "grad_norm": 0.48287842316913426,
+      "learning_rate": 1.0549206672835033e-05,
+      "loss": 0.4879,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7835820895522388,
+      "grad_norm": 0.4494356289008418,
+      "learning_rate": 1.0520710581229507e-05,
+      "loss": 0.4816,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7841651119402985,
+      "grad_norm": 0.46052908698109446,
+      "learning_rate": 1.049227760914549e-05,
+      "loss": 0.4516,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7847481343283582,
+      "grad_norm": 0.4567314964584747,
+      "learning_rate": 1.0463907862280913e-05,
+      "loss": 0.4871,
+      "step": 6730
+    },
+    {
+      "epoch": 0.785331156716418,
+      "grad_norm": 0.44868088537045625,
+      "learning_rate": 1.043560144609866e-05,
+      "loss": 0.4955,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7859141791044776,
+      "grad_norm": 0.4530134493607295,
+      "learning_rate": 1.0407358465826198e-05,
+      "loss": 0.5199,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7864972014925373,
+      "grad_norm": 0.4463551169201282,
+      "learning_rate": 1.0379179026455136e-05,
+      "loss": 0.4913,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7870802238805971,
+      "grad_norm": 0.47099610185569996,
+      "learning_rate": 1.0351063232740937e-05,
+      "loss": 0.5332,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7876632462686567,
+      "grad_norm": 0.4402447824147342,
+      "learning_rate": 1.0323011189202408e-05,
+      "loss": 0.5053,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7882462686567164,
+      "grad_norm": 0.45593057954899197,
+      "learning_rate": 1.0295023000121404e-05,
+      "loss": 0.474,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7888292910447762,
+      "grad_norm": 0.4374300903516831,
+      "learning_rate": 1.0267098769542368e-05,
+      "loss": 0.5427,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7894123134328358,
+      "grad_norm": 0.4216054919323253,
+      "learning_rate": 1.0239238601272036e-05,
+      "loss": 0.4862,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7899953358208955,
+      "grad_norm": 0.43838979602724326,
+      "learning_rate": 1.0211442598878936e-05,
+      "loss": 0.4697,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7905783582089553,
+      "grad_norm": 0.41927389582266983,
+      "learning_rate": 1.0183710865693105e-05,
+      "loss": 0.4731,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7911613805970149,
+      "grad_norm": 0.4618407627786652,
+      "learning_rate": 1.0156043504805648e-05,
+      "loss": 0.4946,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7917444029850746,
+      "grad_norm": 0.49476862297527385,
+      "learning_rate": 1.0128440619068379e-05,
+      "loss": 0.5218,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7923274253731343,
+      "grad_norm": 0.4878473893991779,
+      "learning_rate": 1.0100902311093405e-05,
+      "loss": 0.5127,
+      "step": 6795
+    },
+    {
+      "epoch": 0.792910447761194,
+      "grad_norm": 0.44555429187543133,
+      "learning_rate": 1.0073428683252788e-05,
+      "loss": 0.4893,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7934934701492538,
+      "grad_norm": 0.44977845000528854,
+      "learning_rate": 1.0046019837678153e-05,
+      "loss": 0.4687,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7940764925373134,
+      "grad_norm": 0.447310487987635,
+      "learning_rate": 1.001867587626029e-05,
+      "loss": 0.484,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7946595149253731,
+      "grad_norm": 0.48144807868357464,
+      "learning_rate": 9.991396900648774e-06,
+      "loss": 0.5395,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7952425373134329,
+      "grad_norm": 0.46182249334458747,
+      "learning_rate": 9.964183012251619e-06,
+      "loss": 0.4914,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7958255597014925,
+      "grad_norm": 0.4389736266429193,
+      "learning_rate": 9.937034312234872e-06,
+      "loss": 0.4966,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7964085820895522,
+      "grad_norm": 0.4683635632125285,
+      "learning_rate": 9.90995090152225e-06,
+      "loss": 0.4872,
+      "step": 6830
+    },
+    {
+      "epoch": 0.796991604477612,
+      "grad_norm": 0.449577983637824,
+      "learning_rate": 9.88293288079476e-06,
+      "loss": 0.4775,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7975746268656716,
+      "grad_norm": 0.4679257203680657,
+      "learning_rate": 9.855980350490315e-06,
+      "loss": 0.4628,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7981576492537313,
+      "grad_norm": 0.4910686168202571,
+      "learning_rate": 9.82909341080339e-06,
+      "loss": 0.516,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7987406716417911,
+      "grad_norm": 0.4328561795827866,
+      "learning_rate": 9.802272161684601e-06,
+      "loss": 0.5062,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7993236940298507,
+      "grad_norm": 0.48810830150277185,
+      "learning_rate": 9.775516702840411e-06,
+      "loss": 0.542,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7999067164179104,
+      "grad_norm": 0.4326359263896414,
+      "learning_rate": 9.748827133732665e-06,
+      "loss": 0.491,
+      "step": 6860
+    },
+    {
+      "epoch": 0.8004897388059702,
+      "grad_norm": 0.45895357662154007,
+      "learning_rate": 9.722203553578288e-06,
+      "loss": 0.5017,
+      "step": 6865
+    },
+    {
+      "epoch": 0.8010727611940298,
+      "grad_norm": 0.4201551976753478,
+      "learning_rate": 9.695646061348892e-06,
+      "loss": 0.5233,
+      "step": 6870
+    },
+    {
+      "epoch": 0.8016557835820896,
+      "grad_norm": 0.4261865295483104,
+      "learning_rate": 9.669154755770415e-06,
+      "loss": 0.4615,
+      "step": 6875
+    },
+    {
+      "epoch": 0.8022388059701493,
+      "grad_norm": 0.7140092170328386,
+      "learning_rate": 9.642729735322733e-06,
+      "loss": 0.5097,
+      "step": 6880
+    },
+    {
+      "epoch": 0.8028218283582089,
+      "grad_norm": 0.4990283440949626,
+      "learning_rate": 9.616371098239346e-06,
+      "loss": 0.5716,
+      "step": 6885
+    },
+    {
+      "epoch": 0.8034048507462687,
+      "grad_norm": 0.4481559993333328,
+      "learning_rate": 9.590078942506933e-06,
+      "loss": 0.4856,
+      "step": 6890
+    },
+    {
+      "epoch": 0.8039878731343284,
+      "grad_norm": 0.4607158596942868,
+      "learning_rate": 9.56385336586507e-06,
+      "loss": 0.5107,
+      "step": 6895
+    },
+    {
+      "epoch": 0.804570895522388,
+      "grad_norm": 0.476387734456041,
+      "learning_rate": 9.537694465805797e-06,
+      "loss": 0.4998,
+      "step": 6900
+    },
+    {
+      "epoch": 0.8051539179104478,
+      "grad_norm": 0.46770623541362477,
+      "learning_rate": 9.511602339573324e-06,
+      "loss": 0.4998,
+      "step": 6905
+    },
+    {
+      "epoch": 0.8057369402985075,
+      "grad_norm": 0.5086514392993742,
+      "learning_rate": 9.485577084163604e-06,
+      "loss": 0.5105,
+      "step": 6910
+    },
+    {
+      "epoch": 0.8063199626865671,
+      "grad_norm": 0.5077908007862963,
+      "learning_rate": 9.45961879632401e-06,
+      "loss": 0.5116,
+      "step": 6915
+    },
+    {
+      "epoch": 0.8069029850746269,
+      "grad_norm": 0.5463748621976762,
+      "learning_rate": 9.43372757255297e-06,
+      "loss": 0.4687,
+      "step": 6920
+    },
+    {
+      "epoch": 0.8074860074626866,
+      "grad_norm": 0.4191317799403366,
+      "learning_rate": 9.40790350909961e-06,
+      "loss": 0.498,
+      "step": 6925
+    },
+    {
+      "epoch": 0.8080690298507462,
+      "grad_norm": 0.4865303702117012,
+      "learning_rate": 9.382146701963373e-06,
+      "loss": 0.456,
+      "step": 6930
+    },
+    {
+      "epoch": 0.808652052238806,
+      "grad_norm": 0.42674988798148156,
+      "learning_rate": 9.356457246893695e-06,
+      "loss": 0.5227,
+      "step": 6935
+    },
+    {
+      "epoch": 0.8092350746268657,
+      "grad_norm": 0.46859664111243005,
+      "learning_rate": 9.330835239389645e-06,
+      "loss": 0.5018,
+      "step": 6940
+    },
+    {
+      "epoch": 0.8098180970149254,
+      "grad_norm": 0.5581605519629456,
+      "learning_rate": 9.305280774699531e-06,
+      "loss": 0.4893,
+      "step": 6945
+    },
+    {
+      "epoch": 0.8104011194029851,
+      "grad_norm": 0.4648844812652044,
+      "learning_rate": 9.279793947820596e-06,
+      "loss": 0.5034,
+      "step": 6950
+    },
+    {
+      "epoch": 0.8109841417910447,
+      "grad_norm": 0.4348511386801257,
+      "learning_rate": 9.254374853498636e-06,
+      "loss": 0.476,
+      "step": 6955
+    },
+    {
+      "epoch": 0.8115671641791045,
+      "grad_norm": 0.4471748979975722,
+      "learning_rate": 9.229023586227666e-06,
+      "loss": 0.4746,
+      "step": 6960
+    },
+    {
+      "epoch": 0.8121501865671642,
+      "grad_norm": 0.49028467746657334,
+      "learning_rate": 9.203740240249527e-06,
+      "loss": 0.5075,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8127332089552238,
+      "grad_norm": 0.45156139381243476,
+      "learning_rate": 9.178524909553617e-06,
+      "loss": 0.4997,
+      "step": 6970
+    },
+    {
+      "epoch": 0.8133162313432836,
+      "grad_norm": 0.4580984480401016,
+      "learning_rate": 9.153377687876439e-06,
+      "loss": 0.5098,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8138992537313433,
+      "grad_norm": 0.4417064208682211,
+      "learning_rate": 9.128298668701341e-06,
+      "loss": 0.4866,
+      "step": 6980
+    },
+    {
+      "epoch": 0.8144822761194029,
+      "grad_norm": 0.4396784431419953,
+      "learning_rate": 9.103287945258104e-06,
+      "loss": 0.4843,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8150652985074627,
+      "grad_norm": 0.42781740173572586,
+      "learning_rate": 9.078345610522662e-06,
+      "loss": 0.4864,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8156483208955224,
+      "grad_norm": 0.4706194313899779,
+      "learning_rate": 9.053471757216675e-06,
+      "loss": 0.4829,
+      "step": 6995
+    },
+    {
+      "epoch": 0.816231343283582,
+      "grad_norm": 0.4373005186110652,
+      "learning_rate": 9.028666477807253e-06,
+      "loss": 0.4946,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8168143656716418,
+      "grad_norm": 0.45822848787961307,
+      "learning_rate": 9.003929864506583e-06,
+      "loss": 0.4747,
+      "step": 7005
+    },
+    {
+      "epoch": 0.8173973880597015,
+      "grad_norm": 0.4376578697732612,
+      "learning_rate": 8.979262009271589e-06,
+      "loss": 0.4982,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8179804104477612,
+      "grad_norm": 0.50115571632827,
+      "learning_rate": 8.954663003803579e-06,
+      "loss": 0.4934,
+      "step": 7015
+    },
+    {
+      "epoch": 0.8185634328358209,
+      "grad_norm": 0.44219752315988364,
+      "learning_rate": 8.930132939547932e-06,
+      "loss": 0.4663,
+      "step": 7020
+    },
+    {
+      "epoch": 0.8191464552238806,
+      "grad_norm": 0.440108685069607,
+      "learning_rate": 8.905671907693738e-06,
+      "loss": 0.4856,
+      "step": 7025
+    },
+    {
+      "epoch": 0.8197294776119403,
+      "grad_norm": 0.5246570100006753,
+      "learning_rate": 8.881279999173466e-06,
+      "loss": 0.5021,
+      "step": 7030
+    },
+    {
+      "epoch": 0.8203125,
+      "grad_norm": 0.4319239074767847,
+      "learning_rate": 8.856957304662602e-06,
+      "loss": 0.5123,
+      "step": 7035
+    },
+    {
+      "epoch": 0.8208955223880597,
+      "grad_norm": 0.4326829341676435,
+      "learning_rate": 8.832703914579363e-06,
+      "loss": 0.5021,
+      "step": 7040
+    },
+    {
+      "epoch": 0.8214785447761194,
+      "grad_norm": 0.4380817083829351,
+      "learning_rate": 8.80851991908431e-06,
+      "loss": 0.5044,
+      "step": 7045
+    },
+    {
+      "epoch": 0.8220615671641791,
+      "grad_norm": 0.4637369898587162,
+      "learning_rate": 8.784405408080046e-06,
+      "loss": 0.4852,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8226445895522388,
+      "grad_norm": 0.4796910879347511,
+      "learning_rate": 8.760360471210865e-06,
+      "loss": 0.5338,
+      "step": 7055
+    },
+    {
+      "epoch": 0.8232276119402985,
+      "grad_norm": 0.4135280946993472,
+      "learning_rate": 8.736385197862415e-06,
+      "loss": 0.4824,
+      "step": 7060
+    },
+    {
+      "epoch": 0.8238106343283582,
+      "grad_norm": 0.4287858690307368,
+      "learning_rate": 8.712479677161388e-06,
+      "loss": 0.4869,
+      "step": 7065
+    },
+    {
+      "epoch": 0.824393656716418,
+      "grad_norm": 0.4714198385972294,
+      "learning_rate": 8.688643997975156e-06,
+      "loss": 0.5143,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8249766791044776,
+      "grad_norm": 0.44395978208148723,
+      "learning_rate": 8.66487824891149e-06,
+      "loss": 0.4922,
+      "step": 7075
+    },
+    {
+      "epoch": 0.8255597014925373,
+      "grad_norm": 0.508754740613158,
+      "learning_rate": 8.641182518318162e-06,
+      "loss": 0.5184,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8261427238805971,
+      "grad_norm": 0.4026422778012762,
+      "learning_rate": 8.617556894282683e-06,
+      "loss": 0.4886,
+      "step": 7085
+    },
+    {
+      "epoch": 0.8267257462686567,
+      "grad_norm": 0.47446388322469524,
+      "learning_rate": 8.594001464631938e-06,
+      "loss": 0.5189,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8273087686567164,
+      "grad_norm": 0.47024774444454565,
+      "learning_rate": 8.570516316931869e-06,
+      "loss": 0.5266,
+      "step": 7095
+    },
+    {
+      "epoch": 0.8278917910447762,
+      "grad_norm": 0.44748825048810753,
+      "learning_rate": 8.547101538487136e-06,
+      "loss": 0.4845,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8284748134328358,
+      "grad_norm": 0.4806024012846129,
+      "learning_rate": 8.52375721634083e-06,
+      "loss": 0.5031,
+      "step": 7105
+    },
+    {
+      "epoch": 0.8290578358208955,
+      "grad_norm": 0.5413375895981546,
+      "learning_rate": 8.5004834372741e-06,
+      "loss": 0.5106,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8296408582089553,
+      "grad_norm": 0.4337024823163042,
+      "learning_rate": 8.477280287805883e-06,
+      "loss": 0.4954,
+      "step": 7115
+    },
+    {
+      "epoch": 0.8302238805970149,
+      "grad_norm": 0.4344528368728774,
+      "learning_rate": 8.454147854192515e-06,
+      "loss": 0.5021,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8308069029850746,
+      "grad_norm": 0.45520925847409455,
+      "learning_rate": 8.4310862224275e-06,
+      "loss": 0.4828,
+      "step": 7125
+    },
+    {
+      "epoch": 0.8313899253731343,
+      "grad_norm": 0.4734386762743593,
+      "learning_rate": 8.408095478241099e-06,
+      "loss": 0.5243,
+      "step": 7130
+    },
+    {
+      "epoch": 0.831972947761194,
+      "grad_norm": 0.4717328221940539,
+      "learning_rate": 8.385175707100064e-06,
+      "loss": 0.4907,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8325559701492538,
+      "grad_norm": 0.41240208092851716,
+      "learning_rate": 8.36232699420732e-06,
+      "loss": 0.5031,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8331389925373134,
+      "grad_norm": 0.41514081930390884,
+      "learning_rate": 8.33954942450163e-06,
+      "loss": 0.4896,
+      "step": 7145
+    },
+    {
+      "epoch": 0.8337220149253731,
+      "grad_norm": 0.4675941109997365,
+      "learning_rate": 8.316843082657277e-06,
+      "loss": 0.5009,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8343050373134329,
+      "grad_norm": 0.4427930397053809,
+      "learning_rate": 8.294208053083771e-06,
+      "loss": 0.511,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8348880597014925,
+      "grad_norm": 0.45949248005156934,
+      "learning_rate": 8.271644419925526e-06,
+      "loss": 0.4719,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8354710820895522,
+      "grad_norm": 0.471413986453529,
+      "learning_rate": 8.249152267061524e-06,
+      "loss": 0.4994,
+      "step": 7165
+    },
+    {
+      "epoch": 0.836054104477612,
+      "grad_norm": 0.44813478705837095,
+      "learning_rate": 8.226731678105045e-06,
+      "loss": 0.4804,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8366371268656716,
+      "grad_norm": 0.4426301509227603,
+      "learning_rate": 8.20438273640332e-06,
+      "loss": 0.5143,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8372201492537313,
+      "grad_norm": 0.4465652598526728,
+      "learning_rate": 8.18210552503725e-06,
+      "loss": 0.4871,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8378031716417911,
+      "grad_norm": 0.47244655194156704,
+      "learning_rate": 8.159900126821062e-06,
+      "loss": 0.5003,
+      "step": 7185
+    },
+    {
+      "epoch": 0.8383861940298507,
+      "grad_norm": 0.4471968512251493,
+      "learning_rate": 8.137766624302036e-06,
+      "loss": 0.5142,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8389692164179104,
+      "grad_norm": 0.4709917176706898,
+      "learning_rate": 8.115705099760184e-06,
+      "loss": 0.5195,
+      "step": 7195
+    },
+    {
+      "epoch": 0.8395522388059702,
+      "grad_norm": 0.4380879394783618,
+      "learning_rate": 8.093715635207948e-06,
+      "loss": 0.4885,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8401352611940298,
+      "grad_norm": 0.41712410093981933,
+      "learning_rate": 8.071798312389863e-06,
+      "loss": 0.5116,
+      "step": 7205
+    },
+    {
+      "epoch": 0.8407182835820896,
+      "grad_norm": 0.4325869434630372,
+      "learning_rate": 8.049953212782329e-06,
+      "loss": 0.4478,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8413013059701493,
+      "grad_norm": 0.4166574600617677,
+      "learning_rate": 8.028180417593215e-06,
+      "loss": 0.4824,
+      "step": 7215
+    },
+    {
+      "epoch": 0.8418843283582089,
+      "grad_norm": 0.43830478937394324,
+      "learning_rate": 8.006480007761628e-06,
+      "loss": 0.4674,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8424673507462687,
+      "grad_norm": 0.4817669338294806,
+      "learning_rate": 7.98485206395758e-06,
+      "loss": 0.5077,
+      "step": 7225
+    },
+    {
+      "epoch": 0.8430503731343284,
+      "grad_norm": 0.4647045998674549,
+      "learning_rate": 7.963296666581703e-06,
+      "loss": 0.4702,
+      "step": 7230
+    },
+    {
+      "epoch": 0.843633395522388,
+      "grad_norm": 0.49174994550827694,
+      "learning_rate": 7.941813895764919e-06,
+      "loss": 0.5212,
+      "step": 7235
+    },
+    {
+      "epoch": 0.8442164179104478,
+      "grad_norm": 0.41757383756848515,
+      "learning_rate": 7.920403831368189e-06,
+      "loss": 0.5016,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8447994402985075,
+      "grad_norm": 0.44263407474750815,
+      "learning_rate": 7.899066552982179e-06,
+      "loss": 0.4994,
+      "step": 7245
+    },
+    {
+      "epoch": 0.8453824626865671,
+      "grad_norm": 0.46533860910272895,
+      "learning_rate": 7.87780213992699e-06,
+      "loss": 0.4973,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8459654850746269,
+      "grad_norm": 0.4396469203604777,
+      "learning_rate": 7.856610671251826e-06,
+      "loss": 0.4962,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8465485074626866,
+      "grad_norm": 0.461412303324744,
+      "learning_rate": 7.835492225734753e-06,
+      "loss": 0.4848,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8471315298507462,
+      "grad_norm": 0.5184880522572166,
+      "learning_rate": 7.81444688188236e-06,
+      "loss": 0.5133,
+      "step": 7265
+    },
+    {
+      "epoch": 0.847714552238806,
+      "grad_norm": 0.47001007740037093,
+      "learning_rate": 7.793474717929495e-06,
+      "loss": 0.4852,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8482975746268657,
+      "grad_norm": 0.47109511656808234,
+      "learning_rate": 7.772575811838948e-06,
+      "loss": 0.4961,
+      "step": 7275
+    },
+    {
+      "epoch": 0.8488805970149254,
+      "grad_norm": 0.44116994424682127,
+      "learning_rate": 7.751750241301192e-06,
+      "loss": 0.4972,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8494636194029851,
+      "grad_norm": 0.4501932519308333,
+      "learning_rate": 7.730998083734083e-06,
+      "loss": 0.4859,
+      "step": 7285
+    },
+    {
+      "epoch": 0.8500466417910447,
+      "grad_norm": 0.48144234778210115,
+      "learning_rate": 7.710319416282543e-06,
+      "loss": 0.4984,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8506296641791045,
+      "grad_norm": 0.456988520365441,
+      "learning_rate": 7.689714315818339e-06,
+      "loss": 0.5232,
+      "step": 7295
+    },
+    {
+      "epoch": 0.8512126865671642,
+      "grad_norm": 0.5275485527755961,
+      "learning_rate": 7.669182858939715e-06,
+      "loss": 0.494,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8517957089552238,
+      "grad_norm": 0.4358686155004703,
+      "learning_rate": 7.648725121971178e-06,
+      "loss": 0.4652,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8523787313432836,
+      "grad_norm": 0.5241681742015615,
+      "learning_rate": 7.628341180963175e-06,
+      "loss": 0.5107,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8529617537313433,
+      "grad_norm": 0.4446315316066114,
+      "learning_rate": 7.608031111691826e-06,
+      "loss": 0.4736,
+      "step": 7315
+    },
+    {
+      "epoch": 0.8535447761194029,
+      "grad_norm": 0.4586742881823037,
+      "learning_rate": 7.587794989658621e-06,
+      "loss": 0.4789,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8541277985074627,
+      "grad_norm": 0.43393817483796693,
+      "learning_rate": 7.567632890090176e-06,
+      "loss": 0.4517,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8547108208955224,
+      "grad_norm": 0.4530263088196386,
+      "learning_rate": 7.5475448879379255e-06,
+      "loss": 0.5204,
+      "step": 7330
+    },
+    {
+      "epoch": 0.855293843283582,
+      "grad_norm": 0.47232925084661864,
+      "learning_rate": 7.527531057877849e-06,
+      "loss": 0.5212,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8558768656716418,
+      "grad_norm": 0.4413601041808957,
+      "learning_rate": 7.507591474310185e-06,
+      "loss": 0.4907,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8564598880597015,
+      "grad_norm": 0.5105056148052758,
+      "learning_rate": 7.487726211359198e-06,
+      "loss": 0.5465,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8570429104477612,
+      "grad_norm": 0.6715647626166893,
+      "learning_rate": 7.46793534287283e-06,
+      "loss": 0.5069,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8576259328358209,
+      "grad_norm": 0.4588809773198785,
+      "learning_rate": 7.448218942422498e-06,
+      "loss": 0.5474,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8582089552238806,
+      "grad_norm": 0.4831576147260091,
+      "learning_rate": 7.428577083302757e-06,
+      "loss": 0.5093,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8587919776119403,
+      "grad_norm": 0.5042721886700091,
+      "learning_rate": 7.409009838531095e-06,
+      "loss": 0.5375,
+      "step": 7365
+    },
+    {
+      "epoch": 0.859375,
+      "grad_norm": 0.46284073380005986,
+      "learning_rate": 7.389517280847598e-06,
+      "loss": 0.5159,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8599580223880597,
+      "grad_norm": 0.4661662360141043,
+      "learning_rate": 7.370099482714715e-06,
+      "loss": 0.5315,
+      "step": 7375
+    },
+    {
+      "epoch": 0.8605410447761194,
+      "grad_norm": 0.4476959939432871,
+      "learning_rate": 7.35075651631699e-06,
+      "loss": 0.4901,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8611240671641791,
+      "grad_norm": 0.44717566109177675,
+      "learning_rate": 7.331488453560767e-06,
+      "loss": 0.4921,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8617070895522388,
+      "grad_norm": 0.4476851149766431,
+      "learning_rate": 7.312295366073952e-06,
+      "loss": 0.4839,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8622901119402985,
+      "grad_norm": 0.4656768131692163,
+      "learning_rate": 7.293177325205734e-06,
+      "loss": 0.5089,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8628731343283582,
+      "grad_norm": 0.47867044879346754,
+      "learning_rate": 7.274134402026321e-06,
+      "loss": 0.5204,
+      "step": 7400
+    },
+    {
+      "epoch": 0.863456156716418,
+      "grad_norm": 0.4495376624711596,
+      "learning_rate": 7.255166667326668e-06,
+      "loss": 0.4864,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8640391791044776,
+      "grad_norm": 0.45540868280070557,
+      "learning_rate": 7.236274191618228e-06,
+      "loss": 0.4969,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8646222014925373,
+      "grad_norm": 0.5275767191951953,
+      "learning_rate": 7.217457045132682e-06,
+      "loss": 0.5086,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8652052238805971,
+      "grad_norm": 0.42863000174753835,
+      "learning_rate": 7.198715297821681e-06,
+      "loss": 0.5033,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8657882462686567,
+      "grad_norm": 0.4834849241582333,
+      "learning_rate": 7.18004901935657e-06,
+      "loss": 0.5057,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8663712686567164,
+      "grad_norm": 0.46483441625060457,
+      "learning_rate": 7.161458279128172e-06,
+      "loss": 0.5058,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8669542910447762,
+      "grad_norm": 0.4699147577232594,
+      "learning_rate": 7.142943146246471e-06,
+      "loss": 0.5052,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8675373134328358,
+      "grad_norm": 0.4728336401428503,
+      "learning_rate": 7.124503689540403e-06,
+      "loss": 0.4945,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8681203358208955,
+      "grad_norm": 0.42756247469245834,
+      "learning_rate": 7.106139977557563e-06,
+      "loss": 0.4868,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8687033582089553,
+      "grad_norm": 0.5508454557117077,
+      "learning_rate": 7.087852078564006e-06,
+      "loss": 0.5078,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8692863805970149,
+      "grad_norm": 0.42365611783227247,
+      "learning_rate": 7.069640060543914e-06,
+      "loss": 0.4795,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8698694029850746,
+      "grad_norm": 0.4619373613002068,
+      "learning_rate": 7.051503991199415e-06,
+      "loss": 0.5093,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8704524253731343,
+      "grad_norm": 0.46430333389633666,
+      "learning_rate": 7.03344393795029e-06,
+      "loss": 0.4628,
+      "step": 7465
+    },
+    {
+      "epoch": 0.871035447761194,
+      "grad_norm": 0.46350593595683315,
+      "learning_rate": 7.0154599679337405e-06,
+      "loss": 0.4966,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8716184701492538,
+      "grad_norm": 0.46997441642678967,
+      "learning_rate": 6.997552148004124e-06,
+      "loss": 0.4619,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8722014925373134,
+      "grad_norm": 0.4940809056382957,
+      "learning_rate": 6.9797205447327236e-06,
+      "loss": 0.5207,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8727845149253731,
+      "grad_norm": 0.42737497657553214,
+      "learning_rate": 6.961965224407487e-06,
+      "loss": 0.4879,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8733675373134329,
+      "grad_norm": 0.4420504460833519,
+      "learning_rate": 6.944286253032789e-06,
+      "loss": 0.4519,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8739505597014925,
+      "grad_norm": 0.4482189949894458,
+      "learning_rate": 6.9266836963291725e-06,
+      "loss": 0.5216,
+      "step": 7495
+    },
+    {
+      "epoch": 0.8745335820895522,
+      "grad_norm": 0.48507550585616305,
+      "learning_rate": 6.90915761973312e-06,
+      "loss": 0.4904,
+      "step": 7500
+    },
+    {
+      "epoch": 0.875116604477612,
+      "grad_norm": 0.4845741351343996,
+      "learning_rate": 6.891708088396803e-06,
+      "loss": 0.5031,
+      "step": 7505
+    },
+    {
+      "epoch": 0.8756996268656716,
+      "grad_norm": 0.45661614078718604,
+      "learning_rate": 6.874335167187844e-06,
+      "loss": 0.4694,
+      "step": 7510
+    },
+    {
+      "epoch": 0.8762826492537313,
+      "grad_norm": 0.4378046368918111,
+      "learning_rate": 6.857038920689068e-06,
+      "loss": 0.473,
+      "step": 7515
+    },
+    {
+      "epoch": 0.8768656716417911,
+      "grad_norm": 0.4351278013857574,
+      "learning_rate": 6.839819413198259e-06,
+      "loss": 0.4686,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8774486940298507,
+      "grad_norm": 0.4598040793979348,
+      "learning_rate": 6.822676708727941e-06,
+      "loss": 0.5058,
+      "step": 7525
+    },
+    {
+      "epoch": 0.8780317164179104,
+      "grad_norm": 0.46734999477037775,
+      "learning_rate": 6.805610871005115e-06,
+      "loss": 0.5142,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8786147388059702,
+      "grad_norm": 0.42358187968622485,
+      "learning_rate": 6.788621963471055e-06,
+      "loss": 0.4656,
+      "step": 7535
+    },
+    {
+      "epoch": 0.8791977611940298,
+      "grad_norm": 0.42943564345062557,
+      "learning_rate": 6.771710049281019e-06,
+      "loss": 0.4866,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8797807835820896,
+      "grad_norm": 0.46090527920896157,
+      "learning_rate": 6.754875191304076e-06,
+      "loss": 0.5283,
+      "step": 7545
+    },
+    {
+      "epoch": 0.8803638059701493,
+      "grad_norm": 0.46574541087337956,
+      "learning_rate": 6.73811745212283e-06,
+      "loss": 0.5072,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8809468283582089,
+      "grad_norm": 0.5402513744522806,
+      "learning_rate": 6.721436894033206e-06,
+      "loss": 0.4474,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8815298507462687,
+      "grad_norm": 0.4887426580865765,
+      "learning_rate": 6.704833579044198e-06,
+      "loss": 0.4945,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8821128731343284,
+      "grad_norm": 0.4720599639216187,
+      "learning_rate": 6.688307568877681e-06,
+      "loss": 0.4757,
+      "step": 7565
+    },
+    {
+      "epoch": 0.882695895522388,
+      "grad_norm": 0.44242282186933635,
+      "learning_rate": 6.6718589249681215e-06,
+      "loss": 0.5141,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8832789179104478,
+      "grad_norm": 0.47443100760410045,
+      "learning_rate": 6.655487708462407e-06,
+      "loss": 0.479,
+      "step": 7575
+    },
+    {
+      "epoch": 0.8838619402985075,
+      "grad_norm": 0.4484432447267727,
+      "learning_rate": 6.639193980219574e-06,
+      "loss": 0.503,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8844449626865671,
+      "grad_norm": 0.4544364445156426,
+      "learning_rate": 6.622977800810626e-06,
+      "loss": 0.4757,
+      "step": 7585
+    },
+    {
+      "epoch": 0.8850279850746269,
+      "grad_norm": 0.470997939385338,
+      "learning_rate": 6.60683923051825e-06,
+      "loss": 0.4791,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8856110074626866,
+      "grad_norm": 0.4213706332803066,
+      "learning_rate": 6.5907783293366525e-06,
+      "loss": 0.468,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8861940298507462,
+      "grad_norm": 0.4352892615766016,
+      "learning_rate": 6.574795156971298e-06,
+      "loss": 0.4843,
+      "step": 7600
+    },
+    {
+      "epoch": 0.886777052238806,
+      "grad_norm": 0.45036633475522897,
+      "learning_rate": 6.5588897728387055e-06,
+      "loss": 0.4705,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8873600746268657,
+      "grad_norm": 0.5077288646140908,
+      "learning_rate": 6.543062236066208e-06,
+      "loss": 0.4791,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8879430970149254,
+      "grad_norm": 0.4513500980150284,
+      "learning_rate": 6.527312605491758e-06,
+      "loss": 0.5178,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8885261194029851,
+      "grad_norm": 0.4274182074103898,
+      "learning_rate": 6.5116409396636935e-06,
+      "loss": 0.4626,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8891091417910447,
+      "grad_norm": 0.4667801241554799,
+      "learning_rate": 6.496047296840513e-06,
+      "loss": 0.5071,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8896921641791045,
+      "grad_norm": 0.4123902058374272,
+      "learning_rate": 6.480531734990686e-06,
+      "loss": 0.4992,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8902751865671642,
+      "grad_norm": 0.47897565830736566,
+      "learning_rate": 6.4650943117924065e-06,
+      "loss": 0.5153,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8908582089552238,
+      "grad_norm": 0.47530139972165525,
+      "learning_rate": 6.449735084633407e-06,
+      "loss": 0.4857,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8914412313432836,
+      "grad_norm": 0.4311055331542708,
+      "learning_rate": 6.4344541106107046e-06,
+      "loss": 0.4877,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8920242537313433,
+      "grad_norm": 0.4668273679980216,
+      "learning_rate": 6.419251446530451e-06,
+      "loss": 0.5164,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8926072761194029,
+      "grad_norm": 0.44140094368731336,
+      "learning_rate": 6.404127148907656e-06,
+      "loss": 0.4784,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8931902985074627,
+      "grad_norm": 0.4544488440695652,
+      "learning_rate": 6.38908127396602e-06,
+      "loss": 0.4988,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8937733208955224,
+      "grad_norm": 0.4557410652922852,
+      "learning_rate": 6.374113877637701e-06,
+      "loss": 0.4987,
+      "step": 7665
+    },
+    {
+      "epoch": 0.894356343283582,
+      "grad_norm": 0.4444789436043567,
+      "learning_rate": 6.359225015563138e-06,
+      "loss": 0.5243,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8949393656716418,
+      "grad_norm": 0.4698237483504685,
+      "learning_rate": 6.3444147430908015e-06,
+      "loss": 0.4691,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.4859998755530479,
+      "learning_rate": 6.329683115277018e-06,
+      "loss": 0.4916,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8961054104477612,
+      "grad_norm": 0.515441325427795,
+      "learning_rate": 6.315030186885763e-06,
+      "loss": 0.5116,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8966884328358209,
+      "grad_norm": 0.5283029015600534,
+      "learning_rate": 6.300456012388446e-06,
+      "loss": 0.4934,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8972714552238806,
+      "grad_norm": 0.4417646004083893,
+      "learning_rate": 6.285960645963708e-06,
+      "loss": 0.5026,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8978544776119403,
+      "grad_norm": 0.4605167879082956,
+      "learning_rate": 6.271544141497232e-06,
+      "loss": 0.4901,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8984375,
+      "grad_norm": 0.4792028373581772,
+      "learning_rate": 6.257206552581541e-06,
+      "loss": 0.5118,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8990205223880597,
+      "grad_norm": 0.4644526939230792,
+      "learning_rate": 6.242947932515786e-06,
+      "loss": 0.5282,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8996035447761194,
+      "grad_norm": 0.5046663803144333,
+      "learning_rate": 6.228768334305555e-06,
+      "loss": 0.4946,
+      "step": 7715
+    },
+    {
+      "epoch": 0.9001865671641791,
+      "grad_norm": 0.43155084823278395,
+      "learning_rate": 6.214667810662682e-06,
+      "loss": 0.4794,
+      "step": 7720
+    },
+    {
+      "epoch": 0.9007695895522388,
+      "grad_norm": 0.4640789279302754,
+      "learning_rate": 6.200646414005046e-06,
+      "loss": 0.5239,
+      "step": 7725
+    },
+    {
+      "epoch": 0.9013526119402985,
+      "grad_norm": 0.45218047321306765,
+      "learning_rate": 6.18670419645637e-06,
+      "loss": 0.5015,
+      "step": 7730
+    },
+    {
+      "epoch": 0.9019356343283582,
+      "grad_norm": 0.4588478078182171,
+      "learning_rate": 6.172841209846046e-06,
+      "loss": 0.4913,
+      "step": 7735
+    },
+    {
+      "epoch": 0.902518656716418,
+      "grad_norm": 0.4433138530597733,
+      "learning_rate": 6.159057505708912e-06,
+      "loss": 0.4594,
+      "step": 7740
+    },
+    {
+      "epoch": 0.9031016791044776,
+      "grad_norm": 0.42245796632484117,
+      "learning_rate": 6.145353135285091e-06,
+      "loss": 0.4945,
+      "step": 7745
+    },
+    {
+      "epoch": 0.9036847014925373,
+      "grad_norm": 0.4756088628902291,
+      "learning_rate": 6.131728149519778e-06,
+      "loss": 0.4932,
+      "step": 7750
+    },
+    {
+      "epoch": 0.9042677238805971,
+      "grad_norm": 0.4597129158786207,
+      "learning_rate": 6.118182599063075e-06,
+      "loss": 0.5354,
+      "step": 7755
+    },
+    {
+      "epoch": 0.9048507462686567,
+      "grad_norm": 0.503389587802853,
+      "learning_rate": 6.104716534269772e-06,
+      "loss": 0.5077,
+      "step": 7760
+    },
+    {
+      "epoch": 0.9054337686567164,
+      "grad_norm": 0.4249365343444884,
+      "learning_rate": 6.091330005199183e-06,
+      "loss": 0.4868,
+      "step": 7765
+    },
+    {
+      "epoch": 0.9060167910447762,
+      "grad_norm": 0.43851878310292347,
+      "learning_rate": 6.078023061614953e-06,
+      "loss": 0.5015,
+      "step": 7770
+    },
+    {
+      "epoch": 0.9065998134328358,
+      "grad_norm": 0.43139132012318054,
+      "learning_rate": 6.064795752984875e-06,
+      "loss": 0.4832,
+      "step": 7775
+    },
+    {
+      "epoch": 0.9071828358208955,
+      "grad_norm": 0.45409545061847195,
+      "learning_rate": 6.0516481284806885e-06,
+      "loss": 0.4794,
+      "step": 7780
+    },
+    {
+      "epoch": 0.9077658582089553,
+      "grad_norm": 0.4411595137434583,
+      "learning_rate": 6.0385802369779385e-06,
+      "loss": 0.5183,
+      "step": 7785
+    },
+    {
+      "epoch": 0.9083488805970149,
+      "grad_norm": 0.45410062982495236,
+      "learning_rate": 6.025592127055741e-06,
+      "loss": 0.4736,
+      "step": 7790
+    },
+    {
+      "epoch": 0.9089319029850746,
+      "grad_norm": 0.42835954681482424,
+      "learning_rate": 6.012683846996645e-06,
+      "loss": 0.4541,
+      "step": 7795
+    },
+    {
+      "epoch": 0.9095149253731343,
+      "grad_norm": 0.42047696751886715,
+      "learning_rate": 5.999855444786425e-06,
+      "loss": 0.4973,
+      "step": 7800
+    },
+    {
+      "epoch": 0.910097947761194,
+      "grad_norm": 0.44914336580430964,
+      "learning_rate": 5.987106968113928e-06,
+      "loss": 0.5061,
+      "step": 7805
+    },
+    {
+      "epoch": 0.9106809701492538,
+      "grad_norm": 0.460206776855078,
+      "learning_rate": 5.974438464370872e-06,
+      "loss": 0.4888,
+      "step": 7810
+    },
+    {
+      "epoch": 0.9112639925373134,
+      "grad_norm": 0.4470862201000093,
+      "learning_rate": 5.961849980651684e-06,
+      "loss": 0.4659,
+      "step": 7815
+    },
+    {
+      "epoch": 0.9118470149253731,
+      "grad_norm": 0.47720200193566115,
+      "learning_rate": 5.949341563753319e-06,
+      "loss": 0.4992,
+      "step": 7820
+    },
+    {
+      "epoch": 0.9124300373134329,
+      "grad_norm": 0.4676169159668866,
+      "learning_rate": 5.936913260175094e-06,
+      "loss": 0.5179,
+      "step": 7825
+    },
+    {
+      "epoch": 0.9130130597014925,
+      "grad_norm": 0.4584704293118159,
+      "learning_rate": 5.924565116118499e-06,
+      "loss": 0.4797,
+      "step": 7830
+    },
+    {
+      "epoch": 0.9135960820895522,
+      "grad_norm": 0.4776754416693919,
+      "learning_rate": 5.9122971774870435e-06,
+      "loss": 0.4678,
+      "step": 7835
+    },
+    {
+      "epoch": 0.914179104477612,
+      "grad_norm": 0.42982811564868484,
+      "learning_rate": 5.900109489886081e-06,
+      "loss": 0.4804,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9147621268656716,
+      "grad_norm": 0.41729830382315986,
+      "learning_rate": 5.8880020986226285e-06,
+      "loss": 0.468,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9153451492537313,
+      "grad_norm": 0.4582133802824904,
+      "learning_rate": 5.875975048705206e-06,
+      "loss": 0.4939,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9159281716417911,
+      "grad_norm": 0.4934433745443256,
+      "learning_rate": 5.864028384843678e-06,
+      "loss": 0.512,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9165111940298507,
+      "grad_norm": 0.4461718822871272,
+      "learning_rate": 5.8521621514490715e-06,
+      "loss": 0.5226,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9170942164179104,
+      "grad_norm": 0.43580261924917685,
+      "learning_rate": 5.8403763926334146e-06,
+      "loss": 0.4712,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9176772388059702,
+      "grad_norm": 0.4289025060075553,
+      "learning_rate": 5.82867115220959e-06,
+      "loss": 0.4857,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9182602611940298,
+      "grad_norm": 0.4471506171646922,
+      "learning_rate": 5.81704647369114e-06,
+      "loss": 0.5177,
+      "step": 7875
+    },
+    {
+      "epoch": 0.9188432835820896,
+      "grad_norm": 0.4349776664841292,
+      "learning_rate": 5.805502400292137e-06,
+      "loss": 0.4925,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9194263059701493,
+      "grad_norm": 0.4524645377436028,
+      "learning_rate": 5.794038974926995e-06,
+      "loss": 0.4807,
+      "step": 7885
+    },
+    {
+      "epoch": 0.9200093283582089,
+      "grad_norm": 0.45618145888853556,
+      "learning_rate": 5.782656240210343e-06,
+      "loss": 0.5078,
+      "step": 7890
+    },
+    {
+      "epoch": 0.9205923507462687,
+      "grad_norm": 0.4316155039975515,
+      "learning_rate": 5.771354238456828e-06,
+      "loss": 0.4795,
+      "step": 7895
+    },
+    {
+      "epoch": 0.9211753731343284,
+      "grad_norm": 0.46008183460748864,
+      "learning_rate": 5.760133011680985e-06,
+      "loss": 0.4788,
+      "step": 7900
+    },
+    {
+      "epoch": 0.921758395522388,
+      "grad_norm": 0.44423836815839907,
+      "learning_rate": 5.748992601597076e-06,
+      "loss": 0.4949,
+      "step": 7905
+    },
+    {
+      "epoch": 0.9223414179104478,
+      "grad_norm": 0.45890602021310334,
+      "learning_rate": 5.737933049618925e-06,
+      "loss": 0.4811,
+      "step": 7910
+    },
+    {
+      "epoch": 0.9229244402985075,
+      "grad_norm": 0.424972332822205,
+      "learning_rate": 5.726954396859773e-06,
+      "loss": 0.4854,
+      "step": 7915
+    },
+    {
+      "epoch": 0.9235074626865671,
+      "grad_norm": 0.42910016132492174,
+      "learning_rate": 5.7160566841321255e-06,
+      "loss": 0.4512,
+      "step": 7920
+    },
+    {
+      "epoch": 0.9240904850746269,
+      "grad_norm": 0.46334797979450737,
+      "learning_rate": 5.705239951947597e-06,
+      "loss": 0.4834,
+      "step": 7925
+    },
+    {
+      "epoch": 0.9246735074626866,
+      "grad_norm": 0.4603535009443525,
+      "learning_rate": 5.694504240516759e-06,
+      "loss": 0.4837,
+      "step": 7930
+    },
+    {
+      "epoch": 0.9252565298507462,
+      "grad_norm": 0.4662682403880509,
+      "learning_rate": 5.683849589748994e-06,
+      "loss": 0.4823,
+      "step": 7935
+    },
+    {
+      "epoch": 0.925839552238806,
+      "grad_norm": 0.44330683887009237,
+      "learning_rate": 5.673276039252347e-06,
+      "loss": 0.5095,
+      "step": 7940
+    },
+    {
+      "epoch": 0.9264225746268657,
+      "grad_norm": 0.43601918566724357,
+      "learning_rate": 5.662783628333379e-06,
+      "loss": 0.5165,
+      "step": 7945
+    },
+    {
+      "epoch": 0.9270055970149254,
+      "grad_norm": 0.41717748530531923,
+      "learning_rate": 5.652372395997015e-06,
+      "loss": 0.4911,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9275886194029851,
+      "grad_norm": 2.401919501282761,
+      "learning_rate": 5.642042380946412e-06,
+      "loss": 0.4687,
+      "step": 7955
+    },
+    {
+      "epoch": 0.9281716417910447,
+      "grad_norm": 0.41624865614532475,
+      "learning_rate": 5.631793621582793e-06,
+      "loss": 0.4945,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9287546641791045,
+      "grad_norm": 0.5068445026903259,
+      "learning_rate": 5.621626156005335e-06,
+      "loss": 0.4786,
+      "step": 7965
+    },
+    {
+      "epoch": 0.9293376865671642,
+      "grad_norm": 0.45630828056750805,
+      "learning_rate": 5.611540022010985e-06,
+      "loss": 0.4951,
+      "step": 7970
+    },
+    {
+      "epoch": 0.9299207089552238,
+      "grad_norm": 0.4956233980595794,
+      "learning_rate": 5.6015352570943755e-06,
+      "loss": 0.4843,
+      "step": 7975
+    },
+    {
+      "epoch": 0.9305037313432836,
+      "grad_norm": 0.4329791152614355,
+      "learning_rate": 5.591611898447632e-06,
+      "loss": 0.4634,
+      "step": 7980
+    },
+    {
+      "epoch": 0.9310867537313433,
+      "grad_norm": 0.4663064517760106,
+      "learning_rate": 5.581769982960261e-06,
+      "loss": 0.5264,
+      "step": 7985
+    },
+    {
+      "epoch": 0.9316697761194029,
+      "grad_norm": 0.4836205261061597,
+      "learning_rate": 5.572009547219013e-06,
+      "loss": 0.5156,
+      "step": 7990
+    },
+    {
+      "epoch": 0.9322527985074627,
+      "grad_norm": 0.4277290525866144,
+      "learning_rate": 5.5623306275077475e-06,
+      "loss": 0.481,
+      "step": 7995
+    },
+    {
+      "epoch": 0.9328358208955224,
+      "grad_norm": 0.43341519970296755,
+      "learning_rate": 5.552733259807276e-06,
+      "loss": 0.498,
+      "step": 8000
+    },
+    {
+      "epoch": 0.933418843283582,
+      "grad_norm": 0.43071532419905234,
+      "learning_rate": 5.543217479795256e-06,
+      "loss": 0.4958,
+      "step": 8005
+    },
+    {
+      "epoch": 0.9340018656716418,
+      "grad_norm": 0.46622528251802137,
+      "learning_rate": 5.533783322846053e-06,
+      "loss": 0.4814,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9345848880597015,
+      "grad_norm": 0.5249920187972582,
+      "learning_rate": 5.524430824030594e-06,
+      "loss": 0.4907,
+      "step": 8015
+    },
+    {
+      "epoch": 0.9351679104477612,
+      "grad_norm": 0.45135919668531826,
+      "learning_rate": 5.515160018116247e-06,
+      "loss": 0.492,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9357509328358209,
+      "grad_norm": 0.4928051727444886,
+      "learning_rate": 5.505970939566699e-06,
+      "loss": 0.5035,
+      "step": 8025
+    },
+    {
+      "epoch": 0.9363339552238806,
+      "grad_norm": 0.44235862805439663,
+      "learning_rate": 5.4968636225418125e-06,
+      "loss": 0.4783,
+      "step": 8030
+    },
+    {
+      "epoch": 0.9369169776119403,
+      "grad_norm": 0.41958300409071425,
+      "learning_rate": 5.487838100897508e-06,
+      "loss": 0.4789,
+      "step": 8035
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.4570459794031607,
+      "learning_rate": 5.478894408185641e-06,
+      "loss": 0.4661,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9380830223880597,
+      "grad_norm": 0.4783472459507973,
+      "learning_rate": 5.470032577653869e-06,
+      "loss": 0.4758,
+      "step": 8045
+    },
+    {
+      "epoch": 0.9386660447761194,
+      "grad_norm": 0.4549507857184879,
+      "learning_rate": 5.4612526422455265e-06,
+      "loss": 0.4961,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9392490671641791,
+      "grad_norm": 0.4650558485006105,
+      "learning_rate": 5.452554634599519e-06,
+      "loss": 0.4891,
+      "step": 8055
+    },
+    {
+      "epoch": 0.9398320895522388,
+      "grad_norm": 0.580533723917234,
+      "learning_rate": 5.443938587050186e-06,
+      "loss": 0.5172,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9404151119402985,
+      "grad_norm": 0.42296595646423346,
+      "learning_rate": 5.435404531627176e-06,
+      "loss": 0.4853,
+      "step": 8065
+    },
+    {
+      "epoch": 0.9409981343283582,
+      "grad_norm": 0.4573887758625825,
+      "learning_rate": 5.426952500055348e-06,
+      "loss": 0.4614,
+      "step": 8070
+    },
+    {
+      "epoch": 0.941581156716418,
+      "grad_norm": 0.4745039842823071,
+      "learning_rate": 5.41858252375464e-06,
+      "loss": 0.5061,
+      "step": 8075
+    },
+    {
+      "epoch": 0.9421641791044776,
+      "grad_norm": 0.47805447519039157,
+      "learning_rate": 5.410294633839949e-06,
+      "loss": 0.4735,
+      "step": 8080
+    },
+    {
+      "epoch": 0.9427472014925373,
+      "grad_norm": 0.43867449428861033,
+      "learning_rate": 5.402088861121025e-06,
+      "loss": 0.4718,
+      "step": 8085
+    },
+    {
+      "epoch": 0.9433302238805971,
+      "grad_norm": 0.4467252451572471,
+      "learning_rate": 5.393965236102353e-06,
+      "loss": 0.4798,
+      "step": 8090
+    },
+    {
+      "epoch": 0.9439132462686567,
+      "grad_norm": 0.5133360727759508,
+      "learning_rate": 5.385923788983034e-06,
+      "loss": 0.4894,
+      "step": 8095
+    },
+    {
+      "epoch": 0.9444962686567164,
+      "grad_norm": 0.4747210863179483,
+      "learning_rate": 5.377964549656685e-06,
+      "loss": 0.5098,
+      "step": 8100
+    },
+    {
+      "epoch": 0.9450792910447762,
+      "grad_norm": 0.44347953398302503,
+      "learning_rate": 5.370087547711307e-06,
+      "loss": 0.5105,
+      "step": 8105
+    },
+    {
+      "epoch": 0.9456623134328358,
+      "grad_norm": 0.4387472897639345,
+      "learning_rate": 5.362292812429207e-06,
+      "loss": 0.4815,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9462453358208955,
+      "grad_norm": 0.4722026626305582,
+      "learning_rate": 5.354580372786854e-06,
+      "loss": 0.4776,
+      "step": 8115
+    },
+    {
+      "epoch": 0.9468283582089553,
+      "grad_norm": 0.4566856777345421,
+      "learning_rate": 5.346950257454792e-06,
+      "loss": 0.5002,
+      "step": 8120
+    },
+    {
+      "epoch": 0.9474113805970149,
+      "grad_norm": 0.4601768931007766,
+      "learning_rate": 5.339402494797539e-06,
+      "loss": 0.4725,
+      "step": 8125
+    },
+    {
+      "epoch": 0.9479944029850746,
+      "grad_norm": 0.4422563608916944,
+      "learning_rate": 5.331937112873462e-06,
+      "loss": 0.4411,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9485774253731343,
+      "grad_norm": 0.44855925098469346,
+      "learning_rate": 5.324554139434679e-06,
+      "loss": 0.4941,
+      "step": 8135
+    },
+    {
+      "epoch": 0.949160447761194,
+      "grad_norm": 0.4437725309777671,
+      "learning_rate": 5.317253601926967e-06,
+      "loss": 0.4836,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9497434701492538,
+      "grad_norm": 0.4488409563199269,
+      "learning_rate": 5.310035527489651e-06,
+      "loss": 0.4958,
+      "step": 8145
+    },
+    {
+      "epoch": 0.9503264925373134,
+      "grad_norm": 0.46309779837786297,
+      "learning_rate": 5.3028999429555045e-06,
+      "loss": 0.493,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9509095149253731,
+      "grad_norm": 0.45673001540720787,
+      "learning_rate": 5.295846874850646e-06,
+      "loss": 0.4945,
+      "step": 8155
+    },
+    {
+      "epoch": 0.9514925373134329,
+      "grad_norm": 0.4481344375463158,
+      "learning_rate": 5.288876349394448e-06,
+      "loss": 0.4957,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9520755597014925,
+      "grad_norm": 0.45392704978409076,
+      "learning_rate": 5.281988392499431e-06,
+      "loss": 0.4721,
+      "step": 8165
+    },
+    {
+      "epoch": 0.9526585820895522,
+      "grad_norm": 0.45349195761765704,
+      "learning_rate": 5.275183029771177e-06,
+      "loss": 0.4741,
+      "step": 8170
+    },
+    {
+      "epoch": 0.953241604477612,
+      "grad_norm": 0.4411901617730967,
+      "learning_rate": 5.2684602865082255e-06,
+      "loss": 0.4966,
+      "step": 8175
+    },
+    {
+      "epoch": 0.9538246268656716,
+      "grad_norm": 0.4585915951637137,
+      "learning_rate": 5.261820187701984e-06,
+      "loss": 0.5453,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9544076492537313,
+      "grad_norm": 0.4593205732104069,
+      "learning_rate": 5.2552627580366334e-06,
+      "loss": 0.5013,
+      "step": 8185
+    },
+    {
+      "epoch": 0.9549906716417911,
+      "grad_norm": 0.42592095115888917,
+      "learning_rate": 5.248788021889036e-06,
+      "loss": 0.4797,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9555736940298507,
+      "grad_norm": 0.46962819902389563,
+      "learning_rate": 5.2423960033286505e-06,
+      "loss": 0.4763,
+      "step": 8195
+    },
+    {
+      "epoch": 0.9561567164179104,
+      "grad_norm": 0.40855238448922176,
+      "learning_rate": 5.236086726117433e-06,
+      "loss": 0.4743,
+      "step": 8200
+    },
+    {
+      "epoch": 0.9567397388059702,
+      "grad_norm": 0.43575271699567064,
+      "learning_rate": 5.229860213709753e-06,
+      "loss": 0.4773,
+      "step": 8205
+    },
+    {
+      "epoch": 0.9573227611940298,
+      "grad_norm": 0.4688749371758774,
+      "learning_rate": 5.223716489252311e-06,
+      "loss": 0.5166,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9579057835820896,
+      "grad_norm": 0.7670542201127664,
+      "learning_rate": 5.217655575584045e-06,
+      "loss": 0.493,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9584888059701493,
+      "grad_norm": 0.46421302891934385,
+      "learning_rate": 5.211677495236046e-06,
+      "loss": 0.5096,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9590718283582089,
+      "grad_norm": 0.4771862479599736,
+      "learning_rate": 5.205782270431484e-06,
+      "loss": 0.4693,
+      "step": 8225
+    },
+    {
+      "epoch": 0.9596548507462687,
+      "grad_norm": 0.44226860049129957,
+      "learning_rate": 5.199969923085515e-06,
+      "loss": 0.4459,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9602378731343284,
+      "grad_norm": 0.4294703525768145,
+      "learning_rate": 5.194240474805201e-06,
+      "loss": 0.481,
+      "step": 8235
+    },
+    {
+      "epoch": 0.960820895522388,
+      "grad_norm": 0.46086183427389743,
+      "learning_rate": 5.188593946889429e-06,
+      "loss": 0.4973,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9614039179104478,
+      "grad_norm": 0.4280169626175964,
+      "learning_rate": 5.183030360328846e-06,
+      "loss": 0.4698,
+      "step": 8245
+    },
+    {
+      "epoch": 0.9619869402985075,
+      "grad_norm": 0.4753252130786709,
+      "learning_rate": 5.177549735805758e-06,
+      "loss": 0.5015,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9625699626865671,
+      "grad_norm": 0.4620963248632002,
+      "learning_rate": 5.172152093694067e-06,
+      "loss": 0.4802,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9631529850746269,
+      "grad_norm": 0.5332881434749811,
+      "learning_rate": 5.166837454059193e-06,
+      "loss": 0.5106,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9637360074626866,
+      "grad_norm": 0.5025697757795641,
+      "learning_rate": 5.161605836658004e-06,
+      "loss": 0.4986,
+      "step": 8265
+    },
+    {
+      "epoch": 0.9643190298507462,
+      "grad_norm": 0.46525870817249937,
+      "learning_rate": 5.156457260938732e-06,
+      "loss": 0.4925,
+      "step": 8270
+    },
+    {
+      "epoch": 0.964902052238806,
+      "grad_norm": 0.4466358224458004,
+      "learning_rate": 5.151391746040905e-06,
+      "loss": 0.4896,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9654850746268657,
+      "grad_norm": 0.4722232462608345,
+      "learning_rate": 5.146409310795282e-06,
+      "loss": 0.5116,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9660680970149254,
+      "grad_norm": 0.4839383338646517,
+      "learning_rate": 5.14150997372378e-06,
+      "loss": 0.4704,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9666511194029851,
+      "grad_norm": 0.4598082477182381,
+      "learning_rate": 5.1366937530393955e-06,
+      "loss": 0.4675,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9672341417910447,
+      "grad_norm": 0.4571533268091363,
+      "learning_rate": 5.131960666646149e-06,
+      "loss": 0.4886,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9678171641791045,
+      "grad_norm": 0.4826350517333677,
+      "learning_rate": 5.127310732139018e-06,
+      "loss": 0.5239,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9684001865671642,
+      "grad_norm": 0.501035586719562,
+      "learning_rate": 5.122743966803858e-06,
+      "loss": 0.4845,
+      "step": 8305
+    },
+    {
+      "epoch": 0.9689832089552238,
+      "grad_norm": 0.45689428681207833,
+      "learning_rate": 5.118260387617359e-06,
+      "loss": 0.4802,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9695662313432836,
+      "grad_norm": 0.41839841626319924,
+      "learning_rate": 5.113860011246964e-06,
+      "loss": 0.4759,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9701492537313433,
+      "grad_norm": 0.4623224498637685,
+      "learning_rate": 5.109542854050814e-06,
+      "loss": 0.5191,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9707322761194029,
+      "grad_norm": 0.4400774444581771,
+      "learning_rate": 5.105308932077693e-06,
+      "loss": 0.4751,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9713152985074627,
+      "grad_norm": 0.5458461815433695,
+      "learning_rate": 5.101158261066959e-06,
+      "loss": 0.4934,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9718983208955224,
+      "grad_norm": 0.45228504839890776,
+      "learning_rate": 5.097090856448492e-06,
+      "loss": 0.5007,
+      "step": 8335
+    },
+    {
+      "epoch": 0.972481343283582,
+      "grad_norm": 0.467798541042544,
+      "learning_rate": 5.0931067333426275e-06,
+      "loss": 0.5082,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9730643656716418,
+      "grad_norm": 0.4266615826249801,
+      "learning_rate": 5.0892059065601145e-06,
+      "loss": 0.4974,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9736473880597015,
+      "grad_norm": 0.4621554435159516,
+      "learning_rate": 5.0853883906020525e-06,
+      "loss": 0.4965,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9742304104477612,
+      "grad_norm": 0.46073201489628657,
+      "learning_rate": 5.081654199659831e-06,
+      "loss": 0.5071,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9748134328358209,
+      "grad_norm": 0.5462060706132782,
+      "learning_rate": 5.07800334761509e-06,
+      "loss": 0.4938,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9753964552238806,
+      "grad_norm": 0.48250851908785264,
+      "learning_rate": 5.074435848039658e-06,
+      "loss": 0.5091,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9759794776119403,
+      "grad_norm": 0.44847740844622386,
+      "learning_rate": 5.070951714195508e-06,
+      "loss": 0.4957,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9765625,
+      "grad_norm": 0.4467178102848771,
+      "learning_rate": 5.067550959034707e-06,
+      "loss": 0.49,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9771455223880597,
+      "grad_norm": 0.46636700776310513,
+      "learning_rate": 5.064233595199362e-06,
+      "loss": 0.4884,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9777285447761194,
+      "grad_norm": 0.4615425374049883,
+      "learning_rate": 5.060999635021583e-06,
+      "loss": 0.4771,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9783115671641791,
+      "grad_norm": 0.45286825679857706,
+      "learning_rate": 5.057849090523426e-06,
+      "loss": 0.4959,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9788945895522388,
+      "grad_norm": 0.4569055516084358,
+      "learning_rate": 5.054781973416858e-06,
+      "loss": 0.4955,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9794776119402985,
+      "grad_norm": 0.4455395860822749,
+      "learning_rate": 5.051798295103711e-06,
+      "loss": 0.4752,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9800606343283582,
+      "grad_norm": 0.4183019666197293,
+      "learning_rate": 5.048898066675631e-06,
+      "loss": 0.4552,
+      "step": 8405
+    },
+    {
+      "epoch": 0.980643656716418,
+      "grad_norm": 0.43323200908430703,
+      "learning_rate": 5.046081298914053e-06,
+      "loss": 0.5092,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9812266791044776,
+      "grad_norm": 0.46606214824527303,
+      "learning_rate": 5.043348002290145e-06,
+      "loss": 0.4639,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9818097014925373,
+      "grad_norm": 0.4620950107319596,
+      "learning_rate": 5.0406981869647805e-06,
+      "loss": 0.5072,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9823927238805971,
+      "grad_norm": 0.44313034547356717,
+      "learning_rate": 5.038131862788491e-06,
+      "loss": 0.4765,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9829757462686567,
+      "grad_norm": 0.44702546793927656,
+      "learning_rate": 5.035649039301438e-06,
+      "loss": 0.4612,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9835587686567164,
+      "grad_norm": 0.45947101038805505,
+      "learning_rate": 5.033249725733377e-06,
+      "loss": 0.4967,
+      "step": 8435
+    },
+    {
+      "epoch": 0.9841417910447762,
+      "grad_norm": 0.42282849418109447,
+      "learning_rate": 5.0309339310036125e-06,
+      "loss": 0.507,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9847248134328358,
+      "grad_norm": 0.44692770292599415,
+      "learning_rate": 5.02870166372098e-06,
+      "loss": 0.4808,
+      "step": 8445
+    },
+    {
+      "epoch": 0.9853078358208955,
+      "grad_norm": 0.5582818670824954,
+      "learning_rate": 5.0265529321838004e-06,
+      "loss": 0.5405,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9858908582089553,
+      "grad_norm": 0.4665732954055579,
+      "learning_rate": 5.02448774437986e-06,
+      "loss": 0.5157,
+      "step": 8455
+    },
+    {
+      "epoch": 0.9864738805970149,
+      "grad_norm": 0.5202669052061701,
+      "learning_rate": 5.022506107986374e-06,
+      "loss": 0.5146,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9870569029850746,
+      "grad_norm": 0.4563878919577009,
+      "learning_rate": 5.020608030369962e-06,
+      "loss": 0.4614,
+      "step": 8465
+    },
+    {
+      "epoch": 0.9876399253731343,
+      "grad_norm": 0.45761657701588687,
+      "learning_rate": 5.018793518586616e-06,
+      "loss": 0.5007,
+      "step": 8470
+    },
+    {
+      "epoch": 0.988222947761194,
+      "grad_norm": 0.46136936074859125,
+      "learning_rate": 5.017062579381676e-06,
+      "loss": 0.5068,
+      "step": 8475
+    },
+    {
+      "epoch": 0.9888059701492538,
+      "grad_norm": 0.48777910519871126,
+      "learning_rate": 5.015415219189812e-06,
+      "loss": 0.4889,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9893889925373134,
+      "grad_norm": 0.47084471147391416,
+      "learning_rate": 5.013851444134987e-06,
+      "loss": 0.5022,
+      "step": 8485
+    },
+    {
+      "epoch": 0.9899720149253731,
+      "grad_norm": 0.432844624719787,
+      "learning_rate": 5.012371260030445e-06,
+      "loss": 0.491,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9905550373134329,
+      "grad_norm": 0.4649632477223049,
+      "learning_rate": 5.010974672378682e-06,
+      "loss": 0.4741,
+      "step": 8495
+    },
+    {
+      "epoch": 0.9911380597014925,
+      "grad_norm": 0.4764372199483869,
+      "learning_rate": 5.009661686371434e-06,
+      "loss": 0.4772,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9917210820895522,
+      "grad_norm": 0.4868588188840104,
+      "learning_rate": 5.008432306889652e-06,
+      "loss": 0.5214,
+      "step": 8505
+    },
+    {
+      "epoch": 0.992304104477612,
+      "grad_norm": 0.438360158275501,
+      "learning_rate": 5.0072865385034785e-06,
+      "loss": 0.4905,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9928871268656716,
+      "grad_norm": 0.48340344564902316,
+      "learning_rate": 5.006224385472242e-06,
+      "loss": 0.4927,
+      "step": 8515
+    },
+    {
+      "epoch": 0.9934701492537313,
+      "grad_norm": 0.45531955111325645,
+      "learning_rate": 5.0052458517444364e-06,
+      "loss": 0.4888,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9940531716417911,
+      "grad_norm": 0.4362698593831642,
+      "learning_rate": 5.004350940957703e-06,
+      "loss": 0.4818,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9946361940298507,
+      "grad_norm": 0.46810128935885154,
+      "learning_rate": 5.0035396564388184e-06,
+      "loss": 0.5101,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9952192164179104,
+      "grad_norm": 0.8386528644620809,
+      "learning_rate": 5.00281200120369e-06,
+      "loss": 0.4791,
+      "step": 8535
+    },
+    {
+      "epoch": 0.9958022388059702,
+      "grad_norm": 0.4834359055420526,
+      "learning_rate": 5.00216797795733e-06,
+      "loss": 0.4913,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9963852611940298,
+      "grad_norm": 0.4925810326491755,
+      "learning_rate": 5.001607589093861e-06,
+      "loss": 0.5076,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9969682835820896,
+      "grad_norm": 0.44173971825287517,
+      "learning_rate": 5.001130836696491e-06,
+      "loss": 0.4809,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9975513059701493,
+      "grad_norm": 0.48077397748529965,
+      "learning_rate": 5.000737722537526e-06,
+      "loss": 0.4974,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9981343283582089,
+      "grad_norm": 0.6138381620178532,
+      "learning_rate": 5.00042824807834e-06,
+      "loss": 0.4925,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9987173507462687,
+      "grad_norm": 0.43738249534157836,
+      "learning_rate": 5.000202414469386e-06,
+      "loss": 0.4573,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9993003731343284,
+      "grad_norm": 0.4562639975448534,
+      "learning_rate": 5.0000602225501925e-06,
+      "loss": 0.4941,
+      "step": 8570
+    },
+    {
+      "epoch": 0.999883395522388,
+      "grad_norm": 0.4592908023435152,
+      "learning_rate": 5.0000016728493425e-06,
+      "loss": 0.4962,
+      "step": 8575
+    },
+    {
+      "epoch": 1.0,
+      "step": 8576,
+      "total_flos": 488578517827584.0,
+      "train_loss": 0.5423156990836472,
+      "train_runtime": 27783.4456,
+      "train_samples_per_second": 1.235,
+      "train_steps_per_second": 0.309
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 8576,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 488578517827584.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}