diff --git "a/epoch4/trainer_state.json" "b/epoch4/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/epoch4/trainer_state.json"
@@ -0,0 +1,8203 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 11670,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001713796058269066,
+      "grad_norm": 0.43914204835891724,
+      "learning_rate": 4.996143958868895e-05,
+      "loss": 0.3312,
+      "step": 10
+    },
+    {
+      "epoch": 0.003427592116538132,
+      "grad_norm": 0.4367677569389343,
+      "learning_rate": 4.991859468723222e-05,
+      "loss": 0.3356,
+      "step": 20
+    },
+    {
+      "epoch": 0.005141388174807198,
+      "grad_norm": 0.40475136041641235,
+      "learning_rate": 4.987574978577549e-05,
+      "loss": 0.3271,
+      "step": 30
+    },
+    {
+      "epoch": 0.006855184233076264,
+      "grad_norm": 0.39824801683425903,
+      "learning_rate": 4.983290488431877e-05,
+      "loss": 0.3245,
+      "step": 40
+    },
+    {
+      "epoch": 0.00856898029134533,
+      "grad_norm": 0.42033132910728455,
+      "learning_rate": 4.9790059982862043e-05,
+      "loss": 0.3348,
+      "step": 50
+    },
+    {
+      "epoch": 0.010282776349614395,
+      "grad_norm": 0.3950645625591278,
+      "learning_rate": 4.974721508140531e-05,
+      "loss": 0.3194,
+      "step": 60
+    },
+    {
+      "epoch": 0.011996572407883462,
+      "grad_norm": 0.38537418842315674,
+      "learning_rate": 4.970437017994859e-05,
+      "loss": 0.3215,
+      "step": 70
+    },
+    {
+      "epoch": 0.013710368466152529,
+      "grad_norm": 0.4076625108718872,
+      "learning_rate": 4.9661525278491864e-05,
+      "loss": 0.3218,
+      "step": 80
+    },
+    {
+      "epoch": 0.015424164524421594,
+      "grad_norm": 0.4323882758617401,
+      "learning_rate": 4.961868037703514e-05,
+      "loss": 0.3307,
+      "step": 90
+    },
+    {
+      "epoch": 0.01713796058269066,
+      "grad_norm": 0.44885626435279846,
+      "learning_rate": 4.957583547557841e-05,
+      "loss": 0.3348,
+      "step": 100
+    },
+    {
+      "epoch": 0.018851756640959727,
+      "grad_norm": 0.4107305705547333,
+      "learning_rate": 4.953299057412168e-05,
+      "loss": 0.3182,
+      "step": 110
+    },
+    {
+      "epoch": 0.02056555269922879,
+      "grad_norm": 0.4244893789291382,
+      "learning_rate": 4.949014567266495e-05,
+      "loss": 0.3261,
+      "step": 120
+    },
+    {
+      "epoch": 0.022279348757497857,
+      "grad_norm": 0.4333065152168274,
+      "learning_rate": 4.944730077120823e-05,
+      "loss": 0.3334,
+      "step": 130
+    },
+    {
+      "epoch": 0.023993144815766924,
+      "grad_norm": 0.4291490912437439,
+      "learning_rate": 4.9404455869751504e-05,
+      "loss": 0.3455,
+      "step": 140
+    },
+    {
+      "epoch": 0.02570694087403599,
+      "grad_norm": 0.43335890769958496,
+      "learning_rate": 4.936161096829477e-05,
+      "loss": 0.3334,
+      "step": 150
+    },
+    {
+      "epoch": 0.027420736932305057,
+      "grad_norm": 0.4191276729106903,
+      "learning_rate": 4.931876606683805e-05,
+      "loss": 0.3466,
+      "step": 160
+    },
+    {
+      "epoch": 0.02913453299057412,
+      "grad_norm": 0.46823960542678833,
+      "learning_rate": 4.9275921165381324e-05,
+      "loss": 0.3379,
+      "step": 170
+    },
+    {
+      "epoch": 0.030848329048843187,
+      "grad_norm": 0.7344387769699097,
+      "learning_rate": 4.923307626392459e-05,
+      "loss": 0.3464,
+      "step": 180
+    },
+    {
+      "epoch": 0.032562125107112254,
+      "grad_norm": 0.5095270276069641,
+      "learning_rate": 4.919023136246787e-05,
+      "loss": 0.3462,
+      "step": 190
+    },
+    {
+      "epoch": 0.03427592116538132,
+      "grad_norm": 0.44180533289909363,
+      "learning_rate": 4.9147386461011144e-05,
+      "loss": 0.3275,
+      "step": 200
+    },
+    {
+      "epoch": 0.03598971722365039,
+      "grad_norm": 0.4259091913700104,
+      "learning_rate": 4.910454155955442e-05,
+      "loss": 0.3157,
+      "step": 210
+    },
+    {
+      "epoch": 0.037703513281919454,
+      "grad_norm": 0.5718210935592651,
+      "learning_rate": 4.906169665809769e-05,
+      "loss": 0.3423,
+      "step": 220
+    },
+    {
+      "epoch": 0.03941730934018852,
+      "grad_norm": 0.3788514733314514,
+      "learning_rate": 4.901885175664096e-05,
+      "loss": 0.3335,
+      "step": 230
+    },
+    {
+      "epoch": 0.04113110539845758,
+      "grad_norm": 0.42482584714889526,
+      "learning_rate": 4.897600685518424e-05,
+      "loss": 0.3447,
+      "step": 240
+    },
+    {
+      "epoch": 0.04284490145672665,
+      "grad_norm": 0.43104711174964905,
+      "learning_rate": 4.893316195372751e-05,
+      "loss": 0.3234,
+      "step": 250
+    },
+    {
+      "epoch": 0.044558697514995714,
+      "grad_norm": 0.3978526294231415,
+      "learning_rate": 4.8890317052270784e-05,
+      "loss": 0.342,
+      "step": 260
+    },
+    {
+      "epoch": 0.04627249357326478,
+      "grad_norm": 0.3912956118583679,
+      "learning_rate": 4.884747215081405e-05,
+      "loss": 0.3342,
+      "step": 270
+    },
+    {
+      "epoch": 0.04798628963153385,
+      "grad_norm": 0.5453451871871948,
+      "learning_rate": 4.880462724935733e-05,
+      "loss": 0.3272,
+      "step": 280
+    },
+    {
+      "epoch": 0.049700085689802914,
+      "grad_norm": 0.423217236995697,
+      "learning_rate": 4.8761782347900604e-05,
+      "loss": 0.3282,
+      "step": 290
+    },
+    {
+      "epoch": 0.05141388174807198,
+      "grad_norm": 0.38775932788848877,
+      "learning_rate": 4.871893744644387e-05,
+      "loss": 0.3279,
+      "step": 300
+    },
+    {
+      "epoch": 0.05312767780634105,
+      "grad_norm": 0.4705371856689453,
+      "learning_rate": 4.867609254498715e-05,
+      "loss": 0.3355,
+      "step": 310
+    },
+    {
+      "epoch": 0.054841473864610114,
+      "grad_norm": 0.4476145803928375,
+      "learning_rate": 4.8633247643530425e-05,
+      "loss": 0.3354,
+      "step": 320
+    },
+    {
+      "epoch": 0.056555269922879174,
+      "grad_norm": 0.4176316261291504,
+      "learning_rate": 4.859040274207369e-05,
+      "loss": 0.3292,
+      "step": 330
+    },
+    {
+      "epoch": 0.05826906598114824,
+      "grad_norm": 0.3761385381221771,
+      "learning_rate": 4.854755784061697e-05,
+      "loss": 0.3187,
+      "step": 340
+    },
+    {
+      "epoch": 0.05998286203941731,
+      "grad_norm": 0.4049933850765228,
+      "learning_rate": 4.850471293916024e-05,
+      "loss": 0.318,
+      "step": 350
+    },
+    {
+      "epoch": 0.061696658097686374,
+      "grad_norm": 0.42310062050819397,
+      "learning_rate": 4.846186803770352e-05,
+      "loss": 0.3437,
+      "step": 360
+    },
+    {
+      "epoch": 0.06341045415595545,
+      "grad_norm": 0.4141279458999634,
+      "learning_rate": 4.841902313624679e-05,
+      "loss": 0.3354,
+      "step": 370
+    },
+    {
+      "epoch": 0.06512425021422451,
+      "grad_norm": 0.393342524766922,
+      "learning_rate": 4.8376178234790065e-05,
+      "loss": 0.3173,
+      "step": 380
+    },
+    {
+      "epoch": 0.06683804627249357,
+      "grad_norm": 0.4489890933036804,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.338,
+      "step": 390
+    },
+    {
+      "epoch": 0.06855184233076264,
+      "grad_norm": 0.40590476989746094,
+      "learning_rate": 4.829048843187661e-05,
+      "loss": 0.3439,
+      "step": 400
+    },
+    {
+      "epoch": 0.0702656383890317,
+      "grad_norm": 0.4214794337749481,
+      "learning_rate": 4.8247643530419885e-05,
+      "loss": 0.3084,
+      "step": 410
+    },
+    {
+      "epoch": 0.07197943444730077,
+      "grad_norm": 0.4010763466358185,
+      "learning_rate": 4.8204798628963154e-05,
+      "loss": 0.3235,
+      "step": 420
+    },
+    {
+      "epoch": 0.07369323050556983,
+      "grad_norm": 0.3990427851676941,
+      "learning_rate": 4.816195372750643e-05,
+      "loss": 0.33,
+      "step": 430
+    },
+    {
+      "epoch": 0.07540702656383891,
+      "grad_norm": 0.3799406588077545,
+      "learning_rate": 4.8119108826049705e-05,
+      "loss": 0.3231,
+      "step": 440
+    },
+    {
+      "epoch": 0.07712082262210797,
+      "grad_norm": 0.42431640625,
+      "learning_rate": 4.8076263924592974e-05,
+      "loss": 0.3369,
+      "step": 450
+    },
+    {
+      "epoch": 0.07883461868037704,
+      "grad_norm": 0.4817958474159241,
+      "learning_rate": 4.803341902313625e-05,
+      "loss": 0.3261,
+      "step": 460
+    },
+    {
+      "epoch": 0.0805484147386461,
+      "grad_norm": 0.4291154146194458,
+      "learning_rate": 4.799057412167952e-05,
+      "loss": 0.3256,
+      "step": 470
+    },
+    {
+      "epoch": 0.08226221079691516,
+      "grad_norm": 0.41660982370376587,
+      "learning_rate": 4.79477292202228e-05,
+      "loss": 0.3312,
+      "step": 480
+    },
+    {
+      "epoch": 0.08397600685518423,
+      "grad_norm": 0.4013841152191162,
+      "learning_rate": 4.790488431876607e-05,
+      "loss": 0.3461,
+      "step": 490
+    },
+    {
+      "epoch": 0.0856898029134533,
+      "grad_norm": 0.4054085612297058,
+      "learning_rate": 4.786203941730934e-05,
+      "loss": 0.334,
+      "step": 500
+    },
+    {
+      "epoch": 0.08740359897172237,
+      "grad_norm": 0.410505086183548,
+      "learning_rate": 4.7819194515852614e-05,
+      "loss": 0.3105,
+      "step": 510
+    },
+    {
+      "epoch": 0.08911739502999143,
+      "grad_norm": 0.4129144847393036,
+      "learning_rate": 4.777634961439589e-05,
+      "loss": 0.3218,
+      "step": 520
+    },
+    {
+      "epoch": 0.0908311910882605,
+      "grad_norm": 0.4096630811691284,
+      "learning_rate": 4.7733504712939165e-05,
+      "loss": 0.3189,
+      "step": 530
+    },
+    {
+      "epoch": 0.09254498714652956,
+      "grad_norm": 0.38582107424736023,
+      "learning_rate": 4.7690659811482434e-05,
+      "loss": 0.3161,
+      "step": 540
+    },
+    {
+      "epoch": 0.09425878320479864,
+      "grad_norm": 0.38665908575057983,
+      "learning_rate": 4.764781491002571e-05,
+      "loss": 0.3433,
+      "step": 550
+    },
+    {
+      "epoch": 0.0959725792630677,
+      "grad_norm": 0.4843152165412903,
+      "learning_rate": 4.7604970008568985e-05,
+      "loss": 0.3278,
+      "step": 560
+    },
+    {
+      "epoch": 0.09768637532133675,
+      "grad_norm": 0.4557541310787201,
+      "learning_rate": 4.7562125107112254e-05,
+      "loss": 0.3298,
+      "step": 570
+    },
+    {
+      "epoch": 0.09940017137960583,
+      "grad_norm": 0.4062706232070923,
+      "learning_rate": 4.751928020565553e-05,
+      "loss": 0.336,
+      "step": 580
+    },
+    {
+      "epoch": 0.10111396743787489,
+      "grad_norm": 0.37997087836265564,
+      "learning_rate": 4.74764353041988e-05,
+      "loss": 0.3219,
+      "step": 590
+    },
+    {
+      "epoch": 0.10282776349614396,
+      "grad_norm": 0.3932053744792938,
+      "learning_rate": 4.743359040274208e-05,
+      "loss": 0.3182,
+      "step": 600
+    },
+    {
+      "epoch": 0.10454155955441302,
+      "grad_norm": 0.39029377698898315,
+      "learning_rate": 4.739074550128535e-05,
+      "loss": 0.3116,
+      "step": 610
+    },
+    {
+      "epoch": 0.1062553556126821,
+      "grad_norm": 0.4395703077316284,
+      "learning_rate": 4.734790059982862e-05,
+      "loss": 0.3231,
+      "step": 620
+    },
+    {
+      "epoch": 0.10796915167095116,
+      "grad_norm": 0.37828677892684937,
+      "learning_rate": 4.7305055698371895e-05,
+      "loss": 0.3419,
+      "step": 630
+    },
+    {
+      "epoch": 0.10968294772922023,
+      "grad_norm": 0.443380743265152,
+      "learning_rate": 4.726221079691517e-05,
+      "loss": 0.3304,
+      "step": 640
+    },
+    {
+      "epoch": 0.11139674378748929,
+      "grad_norm": 0.4121178686618805,
+      "learning_rate": 4.7219365895458446e-05,
+      "loss": 0.3408,
+      "step": 650
+    },
+    {
+      "epoch": 0.11311053984575835,
+      "grad_norm": 0.4037461280822754,
+      "learning_rate": 4.7176520994001715e-05,
+      "loss": 0.3264,
+      "step": 660
+    },
+    {
+      "epoch": 0.11482433590402742,
+      "grad_norm": 0.38691094517707825,
+      "learning_rate": 4.713367609254499e-05,
+      "loss": 0.342,
+      "step": 670
+    },
+    {
+      "epoch": 0.11653813196229648,
+      "grad_norm": 0.4495431184768677,
+      "learning_rate": 4.7090831191088266e-05,
+      "loss": 0.3323,
+      "step": 680
+    },
+    {
+      "epoch": 0.11825192802056556,
+      "grad_norm": 0.4214000403881073,
+      "learning_rate": 4.7047986289631535e-05,
+      "loss": 0.3161,
+      "step": 690
+    },
+    {
+      "epoch": 0.11996572407883462,
+      "grad_norm": 0.3833317458629608,
+      "learning_rate": 4.700514138817481e-05,
+      "loss": 0.3228,
+      "step": 700
+    },
+    {
+      "epoch": 0.12167952013710369,
+      "grad_norm": 0.4106675684452057,
+      "learning_rate": 4.696229648671808e-05,
+      "loss": 0.3214,
+      "step": 710
+    },
+    {
+      "epoch": 0.12339331619537275,
+      "grad_norm": 0.3777958154678345,
+      "learning_rate": 4.6919451585261355e-05,
+      "loss": 0.3227,
+      "step": 720
+    },
+    {
+      "epoch": 0.12510711225364182,
+      "grad_norm": 0.41527989506721497,
+      "learning_rate": 4.687660668380463e-05,
+      "loss": 0.3166,
+      "step": 730
+    },
+    {
+      "epoch": 0.1268209083119109,
+      "grad_norm": 0.4072493314743042,
+      "learning_rate": 4.68337617823479e-05,
+      "loss": 0.332,
+      "step": 740
+    },
+    {
+      "epoch": 0.12853470437017994,
+      "grad_norm": 0.4207071363925934,
+      "learning_rate": 4.6790916880891175e-05,
+      "loss": 0.3175,
+      "step": 750
+    },
+    {
+      "epoch": 0.13024850042844902,
+      "grad_norm": 0.3771101236343384,
+      "learning_rate": 4.674807197943445e-05,
+      "loss": 0.3389,
+      "step": 760
+    },
+    {
+      "epoch": 0.1319622964867181,
+      "grad_norm": 0.4102315306663513,
+      "learning_rate": 4.670522707797772e-05,
+      "loss": 0.3332,
+      "step": 770
+    },
+    {
+      "epoch": 0.13367609254498714,
+      "grad_norm": 0.4543974697589874,
+      "learning_rate": 4.6662382176520995e-05,
+      "loss": 0.3394,
+      "step": 780
+    },
+    {
+      "epoch": 0.1353898886032562,
+      "grad_norm": 0.4051818251609802,
+      "learning_rate": 4.661953727506427e-05,
+      "loss": 0.3212,
+      "step": 790
+    },
+    {
+      "epoch": 0.13710368466152528,
+      "grad_norm": 0.42799389362335205,
+      "learning_rate": 4.6576692373607546e-05,
+      "loss": 0.3362,
+      "step": 800
+    },
+    {
+      "epoch": 0.13881748071979436,
+      "grad_norm": 0.38577601313591003,
+      "learning_rate": 4.6533847472150815e-05,
+      "loss": 0.338,
+      "step": 810
+    },
+    {
+      "epoch": 0.1405312767780634,
+      "grad_norm": 0.4502124786376953,
+      "learning_rate": 4.649100257069409e-05,
+      "loss": 0.3282,
+      "step": 820
+    },
+    {
+      "epoch": 0.14224507283633248,
+      "grad_norm": 0.36490705609321594,
+      "learning_rate": 4.6448157669237366e-05,
+      "loss": 0.3139,
+      "step": 830
+    },
+    {
+      "epoch": 0.14395886889460155,
+      "grad_norm": 0.44979041814804077,
+      "learning_rate": 4.6405312767780635e-05,
+      "loss": 0.3157,
+      "step": 840
+    },
+    {
+      "epoch": 0.1456726649528706,
+      "grad_norm": 0.4342329502105713,
+      "learning_rate": 4.636246786632391e-05,
+      "loss": 0.3236,
+      "step": 850
+    },
+    {
+      "epoch": 0.14738646101113967,
+      "grad_norm": 0.38446012139320374,
+      "learning_rate": 4.631962296486718e-05,
+      "loss": 0.3401,
+      "step": 860
+    },
+    {
+      "epoch": 0.14910025706940874,
+      "grad_norm": 0.41796597838401794,
+      "learning_rate": 4.6276778063410455e-05,
+      "loss": 0.3321,
+      "step": 870
+    },
+    {
+      "epoch": 0.15081405312767782,
+      "grad_norm": 0.408687025308609,
+      "learning_rate": 4.623393316195373e-05,
+      "loss": 0.3164,
+      "step": 880
+    },
+    {
+      "epoch": 0.15252784918594686,
+      "grad_norm": 0.4135952889919281,
+      "learning_rate": 4.6191088260497e-05,
+      "loss": 0.3215,
+      "step": 890
+    },
+    {
+      "epoch": 0.15424164524421594,
+      "grad_norm": 0.40789395570755005,
+      "learning_rate": 4.6148243359040276e-05,
+      "loss": 0.3161,
+      "step": 900
+    },
+    {
+      "epoch": 0.155955441302485,
+      "grad_norm": 0.3803863823413849,
+      "learning_rate": 4.610539845758355e-05,
+      "loss": 0.3247,
+      "step": 910
+    },
+    {
+      "epoch": 0.15766923736075408,
+      "grad_norm": 0.3899743854999542,
+      "learning_rate": 4.606255355612683e-05,
+      "loss": 0.3261,
+      "step": 920
+    },
+    {
+      "epoch": 0.15938303341902313,
+      "grad_norm": 0.3962157070636749,
+      "learning_rate": 4.6019708654670096e-05,
+      "loss": 0.319,
+      "step": 930
+    },
+    {
+      "epoch": 0.1610968294772922,
+      "grad_norm": 0.3735091984272003,
+      "learning_rate": 4.5976863753213365e-05,
+      "loss": 0.3178,
+      "step": 940
+    },
+    {
+      "epoch": 0.16281062553556128,
+      "grad_norm": 0.37215062975883484,
+      "learning_rate": 4.593401885175665e-05,
+      "loss": 0.3235,
+      "step": 950
+    },
+    {
+      "epoch": 0.16452442159383032,
+      "grad_norm": 0.3859920799732208,
+      "learning_rate": 4.5891173950299916e-05,
+      "loss": 0.3185,
+      "step": 960
+    },
+    {
+      "epoch": 0.1662382176520994,
+      "grad_norm": 0.4166315495967865,
+      "learning_rate": 4.584832904884319e-05,
+      "loss": 0.3174,
+      "step": 970
+    },
+    {
+      "epoch": 0.16795201371036847,
+      "grad_norm": 0.4043422043323517,
+      "learning_rate": 4.580548414738646e-05,
+      "loss": 0.3243,
+      "step": 980
+    },
+    {
+      "epoch": 0.16966580976863754,
+      "grad_norm": 0.4373800754547119,
+      "learning_rate": 4.5762639245929736e-05,
+      "loss": 0.3176,
+      "step": 990
+    },
+    {
+      "epoch": 0.1713796058269066,
+      "grad_norm": 0.37483376264572144,
+      "learning_rate": 4.571979434447301e-05,
+      "loss": 0.3267,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17309340188517566,
+      "grad_norm": 0.4046979248523712,
+      "learning_rate": 4.567694944301628e-05,
+      "loss": 0.3224,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17480719794344474,
+      "grad_norm": 0.412712037563324,
+      "learning_rate": 4.5634104541559556e-05,
+      "loss": 0.3218,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17652099400171378,
+      "grad_norm": 0.38215428590774536,
+      "learning_rate": 4.559125964010283e-05,
+      "loss": 0.3204,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17823479005998286,
+      "grad_norm": 0.431820273399353,
+      "learning_rate": 4.554841473864611e-05,
+      "loss": 0.3244,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17994858611825193,
+      "grad_norm": 0.39680665731430054,
+      "learning_rate": 4.5505569837189376e-05,
+      "loss": 0.3292,
+      "step": 1050
+    },
+    {
+      "epoch": 0.181662382176521,
+      "grad_norm": 0.39560848474502563,
+      "learning_rate": 4.5462724935732645e-05,
+      "loss": 0.3161,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18337617823479005,
+      "grad_norm": 0.4183228015899658,
+      "learning_rate": 4.541988003427593e-05,
+      "loss": 0.3364,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18508997429305912,
+      "grad_norm": 0.43451449275016785,
+      "learning_rate": 4.5377035132819196e-05,
+      "loss": 0.3184,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1868037703513282,
+      "grad_norm": 0.4167250990867615,
+      "learning_rate": 4.533419023136247e-05,
+      "loss": 0.3337,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18851756640959727,
+      "grad_norm": 0.41648226976394653,
+      "learning_rate": 4.529134532990574e-05,
+      "loss": 0.3252,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19023136246786632,
+      "grad_norm": 0.46568888425827026,
+      "learning_rate": 4.5248500428449016e-05,
+      "loss": 0.3449,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1919451585261354,
+      "grad_norm": 0.3986925482749939,
+      "learning_rate": 4.520565552699229e-05,
+      "loss": 0.3173,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19365895458440446,
+      "grad_norm": 0.4106425642967224,
+      "learning_rate": 4.516281062553556e-05,
+      "loss": 0.3148,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1953727506426735,
+      "grad_norm": 0.4649380147457123,
+      "learning_rate": 4.5119965724078836e-05,
+      "loss": 0.321,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19708654670094258,
+      "grad_norm": 0.37514933943748474,
+      "learning_rate": 4.507712082262211e-05,
+      "loss": 0.3445,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19880034275921166,
+      "grad_norm": 0.40746593475341797,
+      "learning_rate": 4.503427592116538e-05,
+      "loss": 0.3238,
+      "step": 1160
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "grad_norm": 0.3904108703136444,
+      "learning_rate": 4.4991431019708657e-05,
+      "loss": 0.3224,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20222793487574978,
+      "grad_norm": 0.43391162157058716,
+      "learning_rate": 4.4948586118251925e-05,
+      "loss": 0.3171,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20394173093401885,
+      "grad_norm": 0.38622039556503296,
+      "learning_rate": 4.490574121679521e-05,
+      "loss": 0.3385,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20565552699228792,
+      "grad_norm": 0.3904707133769989,
+      "learning_rate": 4.486289631533848e-05,
+      "loss": 0.3166,
+      "step": 1200
+    },
+    {
+      "epoch": 0.207369323050557,
+      "grad_norm": 0.39762595295906067,
+      "learning_rate": 4.482005141388175e-05,
+      "loss": 0.3112,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20908311910882604,
+      "grad_norm": 0.36540812253952026,
+      "learning_rate": 4.477720651242502e-05,
+      "loss": 0.3277,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21079691516709512,
+      "grad_norm": 0.39610254764556885,
+      "learning_rate": 4.47343616109683e-05,
+      "loss": 0.3231,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2125107112253642,
+      "grad_norm": 0.3822088837623596,
+      "learning_rate": 4.469151670951157e-05,
+      "loss": 0.3314,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21422450728363324,
+      "grad_norm": 0.41252222657203674,
+      "learning_rate": 4.464867180805484e-05,
+      "loss": 0.3183,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2159383033419023,
+      "grad_norm": 0.3865353763103485,
+      "learning_rate": 4.460582690659812e-05,
+      "loss": 0.331,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21765209940017138,
+      "grad_norm": 0.42041000723838806,
+      "learning_rate": 4.456298200514139e-05,
+      "loss": 0.3204,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21936589545844046,
+      "grad_norm": 0.3931286633014679,
+      "learning_rate": 4.452013710368466e-05,
+      "loss": 0.316,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2210796915167095,
+      "grad_norm": 0.40770575404167175,
+      "learning_rate": 4.447729220222794e-05,
+      "loss": 0.3169,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22279348757497858,
+      "grad_norm": 0.4274751842021942,
+      "learning_rate": 4.4434447300771206e-05,
+      "loss": 0.3161,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22450728363324765,
+      "grad_norm": 0.4018862843513489,
+      "learning_rate": 4.439160239931449e-05,
+      "loss": 0.3165,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2262210796915167,
+      "grad_norm": 0.4179351329803467,
+      "learning_rate": 4.434875749785776e-05,
+      "loss": 0.3264,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22793487574978577,
+      "grad_norm": 0.3966600298881531,
+      "learning_rate": 4.4305912596401026e-05,
+      "loss": 0.3101,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22964867180805484,
+      "grad_norm": 0.41269004344940186,
+      "learning_rate": 4.42630676949443e-05,
+      "loss": 0.3304,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23136246786632392,
+      "grad_norm": 0.4152756929397583,
+      "learning_rate": 4.422022279348758e-05,
+      "loss": 0.3112,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23307626392459296,
+      "grad_norm": 0.3950376510620117,
+      "learning_rate": 4.417737789203085e-05,
+      "loss": 0.3231,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23479005998286204,
+      "grad_norm": 0.38274356722831726,
+      "learning_rate": 4.413453299057412e-05,
+      "loss": 0.3075,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2365038560411311,
+      "grad_norm": 0.4214038550853729,
+      "learning_rate": 4.40916880891174e-05,
+      "loss": 0.3117,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23821765209940018,
+      "grad_norm": 0.38462990522384644,
+      "learning_rate": 4.404884318766067e-05,
+      "loss": 0.3226,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23993144815766923,
+      "grad_norm": 0.39260828495025635,
+      "learning_rate": 4.400599828620394e-05,
+      "loss": 0.3161,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2416452442159383,
+      "grad_norm": 0.42813345789909363,
+      "learning_rate": 4.396315338474722e-05,
+      "loss": 0.3309,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24335904027420738,
+      "grad_norm": 0.408006489276886,
+      "learning_rate": 4.392030848329049e-05,
+      "loss": 0.3192,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24507283633247642,
+      "grad_norm": 0.3971213400363922,
+      "learning_rate": 4.387746358183377e-05,
+      "loss": 0.3224,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2467866323907455,
+      "grad_norm": 0.3895144462585449,
+      "learning_rate": 4.383461868037704e-05,
+      "loss": 0.3209,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24850042844901457,
+      "grad_norm": 0.5091983079910278,
+      "learning_rate": 4.3791773778920306e-05,
+      "loss": 0.3351,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25021422450728364,
+      "grad_norm": 0.43407779932022095,
+      "learning_rate": 4.374892887746358e-05,
+      "loss": 0.3141,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2519280205655527,
+      "grad_norm": 0.39077264070510864,
+      "learning_rate": 4.370608397600686e-05,
+      "loss": 0.3271,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2536418166238218,
+      "grad_norm": 0.4129246175289154,
+      "learning_rate": 4.366323907455013e-05,
+      "loss": 0.3161,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25535561268209084,
+      "grad_norm": 0.41544240713119507,
+      "learning_rate": 4.36203941730934e-05,
+      "loss": 0.3296,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2570694087403599,
+      "grad_norm": 0.3553474247455597,
+      "learning_rate": 4.357754927163668e-05,
+      "loss": 0.3247,
+      "step": 1500
+    },
+    {
+      "epoch": 0.258783204798629,
+      "grad_norm": 0.3913954496383667,
+      "learning_rate": 4.3534704370179953e-05,
+      "loss": 0.3223,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26049700085689803,
+      "grad_norm": 0.3749904930591583,
+      "learning_rate": 4.349185946872322e-05,
+      "loss": 0.329,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2622107969151671,
+      "grad_norm": 0.4127744436264038,
+      "learning_rate": 4.34490145672665e-05,
+      "loss": 0.313,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2639245929734362,
+      "grad_norm": 0.35376396775245667,
+      "learning_rate": 4.3406169665809774e-05,
+      "loss": 0.3182,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2656383890317052,
+      "grad_norm": 0.37427425384521484,
+      "learning_rate": 4.336332476435304e-05,
+      "loss": 0.3307,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26735218508997427,
+      "grad_norm": 0.453383207321167,
+      "learning_rate": 4.332047986289632e-05,
+      "loss": 0.3181,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26906598114824337,
+      "grad_norm": 0.3835713267326355,
+      "learning_rate": 4.327763496143959e-05,
+      "loss": 0.317,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2707797772065124,
+      "grad_norm": 0.3986583650112152,
+      "learning_rate": 4.323479005998286e-05,
+      "loss": 0.3042,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27249357326478146,
+      "grad_norm": 0.4538326859474182,
+      "learning_rate": 4.319194515852614e-05,
+      "loss": 0.3147,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27420736932305056,
+      "grad_norm": 0.39929988980293274,
+      "learning_rate": 4.314910025706941e-05,
+      "loss": 0.3303,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2759211653813196,
+      "grad_norm": 0.41685983538627625,
+      "learning_rate": 4.310625535561268e-05,
+      "loss": 0.3295,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2776349614395887,
+      "grad_norm": 0.40729179978370667,
+      "learning_rate": 4.306341045415596e-05,
+      "loss": 0.3156,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27934875749785776,
+      "grad_norm": 0.39685654640197754,
+      "learning_rate": 4.3020565552699234e-05,
+      "loss": 0.3219,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2810625535561268,
+      "grad_norm": 0.42483872175216675,
+      "learning_rate": 4.29777206512425e-05,
+      "loss": 0.3224,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2827763496143959,
+      "grad_norm": 0.40194588899612427,
+      "learning_rate": 4.293487574978578e-05,
+      "loss": 0.3214,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28449014567266495,
+      "grad_norm": 0.43271875381469727,
+      "learning_rate": 4.2892030848329054e-05,
+      "loss": 0.3258,
+      "step": 1660
+    },
+    {
+      "epoch": 0.286203941730934,
+      "grad_norm": 0.46178582310676575,
+      "learning_rate": 4.284918594687232e-05,
+      "loss": 0.3282,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2879177377892031,
+      "grad_norm": 0.37451499700546265,
+      "learning_rate": 4.28063410454156e-05,
+      "loss": 0.3298,
+      "step": 1680
+    },
+    {
+      "epoch": 0.28963153384747214,
+      "grad_norm": 0.4482177793979645,
+      "learning_rate": 4.276349614395887e-05,
+      "loss": 0.3128,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2913453299057412,
+      "grad_norm": 0.36960533261299133,
+      "learning_rate": 4.272065124250215e-05,
+      "loss": 0.3169,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2930591259640103,
+      "grad_norm": 0.4491027891635895,
+      "learning_rate": 4.267780634104542e-05,
+      "loss": 0.3317,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29477292202227934,
+      "grad_norm": 0.3663288652896881,
+      "learning_rate": 4.263496143958869e-05,
+      "loss": 0.3159,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29648671808054844,
+      "grad_norm": 0.43172726035118103,
+      "learning_rate": 4.259211653813196e-05,
+      "loss": 0.3172,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2982005141388175,
+      "grad_norm": 0.4410558342933655,
+      "learning_rate": 4.254927163667524e-05,
+      "loss": 0.3368,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29991431019708653,
+      "grad_norm": 0.40787994861602783,
+      "learning_rate": 4.2506426735218514e-05,
+      "loss": 0.3085,
+      "step": 1750
+    },
+    {
+      "epoch": 0.30162810625535563,
+      "grad_norm": 0.3924872875213623,
+      "learning_rate": 4.246358183376178e-05,
+      "loss": 0.3149,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3033419023136247,
+      "grad_norm": 0.40870195627212524,
+      "learning_rate": 4.242073693230505e-05,
+      "loss": 0.3136,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3050556983718937,
+      "grad_norm": 0.37982437014579773,
+      "learning_rate": 4.2377892030848335e-05,
+      "loss": 0.3222,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3067694944301628,
+      "grad_norm": 0.4050194025039673,
+      "learning_rate": 4.23350471293916e-05,
+      "loss": 0.334,
+      "step": 1790
+    },
+    {
+      "epoch": 0.30848329048843187,
+      "grad_norm": 0.41956526041030884,
+      "learning_rate": 4.229220222793488e-05,
+      "loss": 0.3157,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3101970865467009,
+      "grad_norm": 0.41833966970443726,
+      "learning_rate": 4.224935732647815e-05,
+      "loss": 0.3333,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31191088260497,
+      "grad_norm": 0.3904295265674591,
+      "learning_rate": 4.220651242502143e-05,
+      "loss": 0.3412,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31362467866323906,
+      "grad_norm": 0.41814124584198,
+      "learning_rate": 4.21636675235647e-05,
+      "loss": 0.3311,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31533847472150817,
+      "grad_norm": 0.38503211736679077,
+      "learning_rate": 4.212082262210797e-05,
+      "loss": 0.3185,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3170522707797772,
+      "grad_norm": 0.41230252385139465,
+      "learning_rate": 4.2077977720651244e-05,
+      "loss": 0.3138,
+      "step": 1850
+    },
+    {
+      "epoch": 0.31876606683804626,
+      "grad_norm": 0.4522543251514435,
+      "learning_rate": 4.203513281919452e-05,
+      "loss": 0.3363,
+      "step": 1860
+    },
+    {
+      "epoch": 0.32047986289631536,
+      "grad_norm": 0.41241583228111267,
+      "learning_rate": 4.1992287917737795e-05,
+      "loss": 0.3156,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3221936589545844,
+      "grad_norm": 0.3957996666431427,
+      "learning_rate": 4.1949443016281064e-05,
+      "loss": 0.3251,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32390745501285345,
+      "grad_norm": 0.43592071533203125,
+      "learning_rate": 4.190659811482433e-05,
+      "loss": 0.3188,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32562125107112255,
+      "grad_norm": 0.3835230767726898,
+      "learning_rate": 4.1863753213367615e-05,
+      "loss": 0.3128,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3273350471293916,
+      "grad_norm": 0.4197012484073639,
+      "learning_rate": 4.1820908311910884e-05,
+      "loss": 0.3117,
+      "step": 1910
+    },
+    {
+      "epoch": 0.32904884318766064,
+      "grad_norm": 0.4401513338088989,
+      "learning_rate": 4.177806341045416e-05,
+      "loss": 0.3118,
+      "step": 1920
+    },
+    {
+      "epoch": 0.33076263924592975,
+      "grad_norm": 0.4062804579734802,
+      "learning_rate": 4.173521850899743e-05,
+      "loss": 0.3088,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3324764353041988,
+      "grad_norm": 0.39562612771987915,
+      "learning_rate": 4.1692373607540704e-05,
+      "loss": 0.3291,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3341902313624679,
+      "grad_norm": 0.3845726251602173,
+      "learning_rate": 4.164952870608398e-05,
+      "loss": 0.319,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33590402742073694,
+      "grad_norm": 0.3665197789669037,
+      "learning_rate": 4.160668380462725e-05,
+      "loss": 0.3163,
+      "step": 1960
+    },
+    {
+      "epoch": 0.337617823479006,
+      "grad_norm": 0.39268404245376587,
+      "learning_rate": 4.1563838903170524e-05,
+      "loss": 0.3186,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3393316195372751,
+      "grad_norm": 0.4009098708629608,
+      "learning_rate": 4.15209940017138e-05,
+      "loss": 0.332,
+      "step": 1980
+    },
+    {
+      "epoch": 0.34104541559554413,
+      "grad_norm": 0.41637134552001953,
+      "learning_rate": 4.147814910025707e-05,
+      "loss": 0.3124,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3427592116538132,
+      "grad_norm": 0.3817277252674103,
+      "learning_rate": 4.1435304198800344e-05,
+      "loss": 0.3065,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3444730077120823,
+      "grad_norm": 0.40419119596481323,
+      "learning_rate": 4.139245929734362e-05,
+      "loss": 0.3058,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3461868037703513,
+      "grad_norm": 0.40321606397628784,
+      "learning_rate": 4.1349614395886895e-05,
+      "loss": 0.3215,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34790059982862037,
+      "grad_norm": 0.4105079472064972,
+      "learning_rate": 4.1306769494430164e-05,
+      "loss": 0.3222,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3496143958868895,
+      "grad_norm": 0.39861756563186646,
+      "learning_rate": 4.126392459297344e-05,
+      "loss": 0.3222,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3513281919451585,
+      "grad_norm": 0.3437749445438385,
+      "learning_rate": 4.122107969151671e-05,
+      "loss": 0.3248,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35304198800342756,
+      "grad_norm": 0.42124131321907043,
+      "learning_rate": 4.1178234790059984e-05,
+      "loss": 0.3222,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35475578406169667,
+      "grad_norm": 0.4235905110836029,
+      "learning_rate": 4.113538988860326e-05,
+      "loss": 0.3299,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3564695801199657,
+      "grad_norm": 0.3889576196670532,
+      "learning_rate": 4.109254498714653e-05,
+      "loss": 0.3168,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3581833761782348,
+      "grad_norm": 0.409067302942276,
+      "learning_rate": 4.1049700085689805e-05,
+      "loss": 0.3162,
+      "step": 2090
+    },
+    {
+      "epoch": 0.35989717223650386,
+      "grad_norm": 0.3821440041065216,
+      "learning_rate": 4.100685518423308e-05,
+      "loss": 0.3204,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3616109682947729,
+      "grad_norm": 0.4041682183742523,
+      "learning_rate": 4.096401028277635e-05,
+      "loss": 0.3336,
+      "step": 2110
+    },
+    {
+      "epoch": 0.363324764353042,
+      "grad_norm": 0.43424752354621887,
+      "learning_rate": 4.0921165381319625e-05,
+      "loss": 0.3199,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36503856041131105,
+      "grad_norm": 0.38477644324302673,
+      "learning_rate": 4.08783204798629e-05,
+      "loss": 0.3164,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3667523564695801,
+      "grad_norm": 0.4039604961872101,
+      "learning_rate": 4.0835475578406176e-05,
+      "loss": 0.302,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3684661525278492,
+      "grad_norm": 0.39774835109710693,
+      "learning_rate": 4.0792630676949445e-05,
+      "loss": 0.3357,
+      "step": 2150
+    },
+    {
+      "epoch": 0.37017994858611825,
+      "grad_norm": 0.44784727692604065,
+      "learning_rate": 4.0749785775492714e-05,
+      "loss": 0.3165,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3718937446443873,
+      "grad_norm": 0.40922826528549194,
+      "learning_rate": 4.070694087403599e-05,
+      "loss": 0.3117,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3736075407026564,
+      "grad_norm": 0.45614659786224365,
+      "learning_rate": 4.0664095972579265e-05,
+      "loss": 0.334,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37532133676092544,
+      "grad_norm": 0.4138086438179016,
+      "learning_rate": 4.062125107112254e-05,
+      "loss": 0.3089,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37703513281919454,
+      "grad_norm": 0.3747407793998718,
+      "learning_rate": 4.057840616966581e-05,
+      "loss": 0.3151,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3787489288774636,
+      "grad_norm": 0.40726715326309204,
+      "learning_rate": 4.0535561268209085e-05,
+      "loss": 0.3177,
+      "step": 2210
+    },
+    {
+      "epoch": 0.38046272493573263,
+      "grad_norm": 0.3894132375717163,
+      "learning_rate": 4.049271636675236e-05,
+      "loss": 0.2993,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38217652099400173,
+      "grad_norm": 0.4253694415092468,
+      "learning_rate": 4.044987146529563e-05,
+      "loss": 0.3178,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3838903170522708,
+      "grad_norm": 0.45020347833633423,
+      "learning_rate": 4.0407026563838905e-05,
+      "loss": 0.3177,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3856041131105398,
+      "grad_norm": 0.3648056983947754,
+      "learning_rate": 4.036418166238218e-05,
+      "loss": 0.3067,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3873179091688089,
+      "grad_norm": 0.3800349235534668,
+      "learning_rate": 4.0321336760925456e-05,
+      "loss": 0.311,
+      "step": 2260
+    },
+    {
+      "epoch": 0.389031705227078,
+      "grad_norm": 0.417293518781662,
+      "learning_rate": 4.0278491859468725e-05,
+      "loss": 0.3139,
+      "step": 2270
+    },
+    {
+      "epoch": 0.390745501285347,
+      "grad_norm": 0.4162093698978424,
+      "learning_rate": 4.0235646958011994e-05,
+      "loss": 0.3335,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3924592973436161,
+      "grad_norm": 0.4039321839809418,
+      "learning_rate": 4.0192802056555276e-05,
+      "loss": 0.3127,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39417309340188517,
+      "grad_norm": 0.40868979692459106,
+      "learning_rate": 4.0149957155098545e-05,
+      "loss": 0.299,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39588688946015427,
+      "grad_norm": 0.3893890082836151,
+      "learning_rate": 4.010711225364182e-05,
+      "loss": 0.3107,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3976006855184233,
+      "grad_norm": 0.4031045138835907,
+      "learning_rate": 4.006426735218509e-05,
+      "loss": 0.3054,
+      "step": 2320
+    },
+    {
+      "epoch": 0.39931448157669236,
+      "grad_norm": 0.38837090134620667,
+      "learning_rate": 4.0021422450728365e-05,
+      "loss": 0.3196,
+      "step": 2330
+    },
+    {
+      "epoch": 0.40102827763496146,
+      "grad_norm": 0.3399879038333893,
+      "learning_rate": 3.997857754927164e-05,
+      "loss": 0.3245,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4027420736932305,
+      "grad_norm": 0.3803098499774933,
+      "learning_rate": 3.993573264781491e-05,
+      "loss": 0.3083,
+      "step": 2350
+    },
+    {
+      "epoch": 0.40445586975149955,
+      "grad_norm": 0.4033030867576599,
+      "learning_rate": 3.9892887746358186e-05,
+      "loss": 0.318,
+      "step": 2360
+    },
+    {
+      "epoch": 0.40616966580976865,
+      "grad_norm": 0.40507230162620544,
+      "learning_rate": 3.985004284490146e-05,
+      "loss": 0.3329,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4078834618680377,
+      "grad_norm": 0.37352365255355835,
+      "learning_rate": 3.980719794344473e-05,
+      "loss": 0.3237,
+      "step": 2380
+    },
+    {
+      "epoch": 0.40959725792630675,
+      "grad_norm": 0.3635859787464142,
+      "learning_rate": 3.9764353041988006e-05,
+      "loss": 0.3178,
+      "step": 2390
+    },
+    {
+      "epoch": 0.41131105398457585,
+      "grad_norm": 0.4639202058315277,
+      "learning_rate": 3.9721508140531275e-05,
+      "loss": 0.314,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4130248500428449,
+      "grad_norm": 0.3707311451435089,
+      "learning_rate": 3.967866323907456e-05,
+      "loss": 0.3171,
+      "step": 2410
+    },
+    {
+      "epoch": 0.414738646101114,
+      "grad_norm": 0.37639179825782776,
+      "learning_rate": 3.9635818337617826e-05,
+      "loss": 0.2996,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41645244215938304,
+      "grad_norm": 0.40876850485801697,
+      "learning_rate": 3.9592973436161095e-05,
+      "loss": 0.3115,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4181662382176521,
+      "grad_norm": 0.4880918264389038,
+      "learning_rate": 3.955012853470437e-05,
+      "loss": 0.3193,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4198800342759212,
+      "grad_norm": 0.4018041789531708,
+      "learning_rate": 3.9507283633247646e-05,
+      "loss": 0.3067,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42159383033419023,
+      "grad_norm": 0.40657979249954224,
+      "learning_rate": 3.946443873179092e-05,
+      "loss": 0.3261,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4233076263924593,
+      "grad_norm": 0.3998405337333679,
+      "learning_rate": 3.942159383033419e-05,
+      "loss": 0.3027,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4250214224507284,
+      "grad_norm": 0.4458753764629364,
+      "learning_rate": 3.9378748928877466e-05,
+      "loss": 0.3165,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4267352185089974,
+      "grad_norm": 0.36847904324531555,
+      "learning_rate": 3.933590402742074e-05,
+      "loss": 0.3065,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4284490145672665,
+      "grad_norm": 0.41887134313583374,
+      "learning_rate": 3.929305912596401e-05,
+      "loss": 0.3202,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4301628106255356,
+      "grad_norm": 0.39404210448265076,
+      "learning_rate": 3.9250214224507286e-05,
+      "loss": 0.316,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4318766066838046,
+      "grad_norm": 0.35294997692108154,
+      "learning_rate": 3.9207369323050555e-05,
+      "loss": 0.3209,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43359040274207367,
+      "grad_norm": 0.47239357233047485,
+      "learning_rate": 3.916452442159384e-05,
+      "loss": 0.3151,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43530419880034277,
+      "grad_norm": 0.4055376648902893,
+      "learning_rate": 3.9121679520137106e-05,
+      "loss": 0.3158,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4370179948586118,
+      "grad_norm": 0.3814369738101959,
+      "learning_rate": 3.9078834618680375e-05,
+      "loss": 0.3142,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4387317909168809,
+      "grad_norm": 0.3641800880432129,
+      "learning_rate": 3.903598971722365e-05,
+      "loss": 0.3161,
+      "step": 2560
+    },
+    {
+      "epoch": 0.44044558697514996,
+      "grad_norm": 0.48547694087028503,
+      "learning_rate": 3.8993144815766926e-05,
+      "loss": 0.3197,
+      "step": 2570
+    },
+    {
+      "epoch": 0.442159383033419,
+      "grad_norm": 0.4065800905227661,
+      "learning_rate": 3.89502999143102e-05,
+      "loss": 0.3148,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4438731790916881,
+      "grad_norm": 0.41718825697898865,
+      "learning_rate": 3.890745501285347e-05,
+      "loss": 0.3114,
+      "step": 2590
+    },
+    {
+      "epoch": 0.44558697514995715,
+      "grad_norm": 0.40741679072380066,
+      "learning_rate": 3.8864610111396746e-05,
+      "loss": 0.3217,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4473007712082262,
+      "grad_norm": 0.3844093680381775,
+      "learning_rate": 3.882176520994002e-05,
+      "loss": 0.3209,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4490145672664953,
+      "grad_norm": 0.4430558681488037,
+      "learning_rate": 3.877892030848329e-05,
+      "loss": 0.3147,
+      "step": 2620
+    },
+    {
+      "epoch": 0.45072836332476435,
+      "grad_norm": 0.44319942593574524,
+      "learning_rate": 3.8736075407026567e-05,
+      "loss": 0.3227,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4524421593830334,
+      "grad_norm": 0.38565146923065186,
+      "learning_rate": 3.8693230505569835e-05,
+      "loss": 0.3133,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4541559554413025,
+      "grad_norm": 0.39172759652137756,
+      "learning_rate": 3.865038560411312e-05,
+      "loss": 0.3215,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45586975149957154,
+      "grad_norm": 0.42386338114738464,
+      "learning_rate": 3.860754070265639e-05,
+      "loss": 0.3288,
+      "step": 2660
+    },
+    {
+      "epoch": 0.45758354755784064,
+      "grad_norm": 0.430250883102417,
+      "learning_rate": 3.8564695801199656e-05,
+      "loss": 0.3208,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4592973436161097,
+      "grad_norm": 0.3613053262233734,
+      "learning_rate": 3.852185089974293e-05,
+      "loss": 0.3122,
+      "step": 2680
+    },
+    {
+      "epoch": 0.46101113967437873,
+      "grad_norm": 0.4381999671459198,
+      "learning_rate": 3.847900599828621e-05,
+      "loss": 0.3158,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46272493573264784,
+      "grad_norm": 0.382499098777771,
+      "learning_rate": 3.843616109682948e-05,
+      "loss": 0.3266,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4644387317909169,
+      "grad_norm": 0.41216859221458435,
+      "learning_rate": 3.839331619537275e-05,
+      "loss": 0.327,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4661525278491859,
+      "grad_norm": 0.3801523447036743,
+      "learning_rate": 3.835047129391603e-05,
+      "loss": 0.3067,
+      "step": 2720
+    },
+    {
+      "epoch": 0.46786632390745503,
+      "grad_norm": 0.3955176770687103,
+      "learning_rate": 3.83076263924593e-05,
+      "loss": 0.3323,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4695801199657241,
+      "grad_norm": 0.4152835011482239,
+      "learning_rate": 3.826478149100257e-05,
+      "loss": 0.3149,
+      "step": 2740
+    },
+    {
+      "epoch": 0.4712939160239931,
+      "grad_norm": 0.4680623710155487,
+      "learning_rate": 3.822193658954585e-05,
+      "loss": 0.3008,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4730077120822622,
+      "grad_norm": 0.4037633538246155,
+      "learning_rate": 3.8179091688089116e-05,
+      "loss": 0.31,
+      "step": 2760
+    },
+    {
+      "epoch": 0.47472150814053127,
+      "grad_norm": 0.3998965620994568,
+      "learning_rate": 3.813624678663239e-05,
+      "loss": 0.3255,
+      "step": 2770
+    },
+    {
+      "epoch": 0.47643530419880037,
+      "grad_norm": 0.4001263976097107,
+      "learning_rate": 3.809340188517567e-05,
+      "loss": 0.3119,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4781491002570694,
+      "grad_norm": 0.4194788932800293,
+      "learning_rate": 3.8050556983718936e-05,
+      "loss": 0.3014,
+      "step": 2790
+    },
+    {
+      "epoch": 0.47986289631533846,
+      "grad_norm": 0.4282771348953247,
+      "learning_rate": 3.800771208226221e-05,
+      "loss": 0.3166,
+      "step": 2800
+    },
+    {
+      "epoch": 0.48157669237360756,
+      "grad_norm": 0.36937791109085083,
+      "learning_rate": 3.796486718080549e-05,
+      "loss": 0.3177,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4832904884318766,
+      "grad_norm": 0.34598973393440247,
+      "learning_rate": 3.7922022279348756e-05,
+      "loss": 0.2978,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48500428449014565,
+      "grad_norm": 0.45177942514419556,
+      "learning_rate": 3.787917737789203e-05,
+      "loss": 0.3267,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48671808054841476,
+      "grad_norm": 0.37213265895843506,
+      "learning_rate": 3.783633247643531e-05,
+      "loss": 0.3021,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4884318766066838,
+      "grad_norm": 0.43076738715171814,
+      "learning_rate": 3.779348757497858e-05,
+      "loss": 0.3232,
+      "step": 2850
+    },
+    {
+      "epoch": 0.49014567266495285,
+      "grad_norm": 0.43916594982147217,
+      "learning_rate": 3.775064267352185e-05,
+      "loss": 0.3055,
+      "step": 2860
+    },
+    {
+      "epoch": 0.49185946872322195,
+      "grad_norm": 0.4522310197353363,
+      "learning_rate": 3.770779777206513e-05,
+      "loss": 0.3049,
+      "step": 2870
+    },
+    {
+      "epoch": 0.493573264781491,
+      "grad_norm": 0.42702722549438477,
+      "learning_rate": 3.76649528706084e-05,
+      "loss": 0.3282,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4952870608397601,
+      "grad_norm": 0.3864535689353943,
+      "learning_rate": 3.762210796915167e-05,
+      "loss": 0.3059,
+      "step": 2890
+    },
+    {
+      "epoch": 0.49700085689802914,
+      "grad_norm": 0.5631365180015564,
+      "learning_rate": 3.757926306769495e-05,
+      "loss": 0.316,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4987146529562982,
+      "grad_norm": 0.4108191728591919,
+      "learning_rate": 3.7536418166238216e-05,
+      "loss": 0.3193,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5004284490145673,
+      "grad_norm": 0.39087364077568054,
+      "learning_rate": 3.749357326478149e-05,
+      "loss": 0.3013,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5021422450728363,
+      "grad_norm": 0.3783542513847351,
+      "learning_rate": 3.745072836332477e-05,
+      "loss": 0.3202,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5038560411311054,
+      "grad_norm": 0.39396172761917114,
+      "learning_rate": 3.7407883461868037e-05,
+      "loss": 0.3061,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5055698371893744,
+      "grad_norm": 0.4316921532154083,
+      "learning_rate": 3.736503856041131e-05,
+      "loss": 0.2947,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5072836332476436,
+      "grad_norm": 0.40128785371780396,
+      "learning_rate": 3.732219365895459e-05,
+      "loss": 0.3082,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5089974293059126,
+      "grad_norm": 0.36867550015449524,
+      "learning_rate": 3.7279348757497864e-05,
+      "loss": 0.2975,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5107112253641817,
+      "grad_norm": 0.4272914230823517,
+      "learning_rate": 3.723650385604113e-05,
+      "loss": 0.308,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5124250214224507,
+      "grad_norm": 0.38395050168037415,
+      "learning_rate": 3.71936589545844e-05,
+      "loss": 0.3254,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5141388174807198,
+      "grad_norm": 0.3956032395362854,
+      "learning_rate": 3.7150814053127684e-05,
+      "loss": 0.3179,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5158526135389888,
+      "grad_norm": 0.4097507894039154,
+      "learning_rate": 3.710796915167095e-05,
+      "loss": 0.3021,
+      "step": 3010
+    },
+    {
+      "epoch": 0.517566409597258,
+      "grad_norm": 0.3923504650592804,
+      "learning_rate": 3.706512425021423e-05,
+      "loss": 0.2979,
+      "step": 3020
+    },
+    {
+      "epoch": 0.519280205655527,
+      "grad_norm": 0.409417062997818,
+      "learning_rate": 3.70222793487575e-05,
+      "loss": 0.3059,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5209940017137961,
+      "grad_norm": 0.44288355112075806,
+      "learning_rate": 3.697943444730077e-05,
+      "loss": 0.3203,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5227077977720651,
+      "grad_norm": 0.34440475702285767,
+      "learning_rate": 3.693658954584405e-05,
+      "loss": 0.3011,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5244215938303342,
+      "grad_norm": 0.3822530508041382,
+      "learning_rate": 3.689374464438732e-05,
+      "loss": 0.2997,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5261353898886033,
+      "grad_norm": 0.41341301798820496,
+      "learning_rate": 3.685089974293059e-05,
+      "loss": 0.301,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5278491859468724,
+      "grad_norm": 0.3852033317089081,
+      "learning_rate": 3.680805484147387e-05,
+      "loss": 0.305,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5295629820051414,
+      "grad_norm": 0.4288126826286316,
+      "learning_rate": 3.6765209940017144e-05,
+      "loss": 0.3129,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5312767780634104,
+      "grad_norm": 0.4876558184623718,
+      "learning_rate": 3.672236503856041e-05,
+      "loss": 0.3068,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5329905741216795,
+      "grad_norm": 0.37149110436439514,
+      "learning_rate": 3.667952013710368e-05,
+      "loss": 0.3106,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5347043701799485,
+      "grad_norm": 0.48074501752853394,
+      "learning_rate": 3.6636675235646964e-05,
+      "loss": 0.3193,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5364181662382177,
+      "grad_norm": 0.39209356904029846,
+      "learning_rate": 3.659383033419023e-05,
+      "loss": 0.3166,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5381319622964867,
+      "grad_norm": 0.39671874046325684,
+      "learning_rate": 3.655098543273351e-05,
+      "loss": 0.3196,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5398457583547558,
+      "grad_norm": 0.3712060749530792,
+      "learning_rate": 3.650814053127678e-05,
+      "loss": 0.3248,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5415595544130248,
+      "grad_norm": 0.41404080390930176,
+      "learning_rate": 3.646529562982005e-05,
+      "loss": 0.3061,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5432733504712939,
+      "grad_norm": 0.4329588711261749,
+      "learning_rate": 3.642245072836333e-05,
+      "loss": 0.317,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5449871465295629,
+      "grad_norm": 0.4148005247116089,
+      "learning_rate": 3.63796058269066e-05,
+      "loss": 0.3105,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5467009425878321,
+      "grad_norm": 0.5018778443336487,
+      "learning_rate": 3.633676092544987e-05,
+      "loss": 0.3139,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5484147386461011,
+      "grad_norm": 0.39833855628967285,
+      "learning_rate": 3.629391602399315e-05,
+      "loss": 0.2982,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5501285347043702,
+      "grad_norm": 0.43114879727363586,
+      "learning_rate": 3.625107112253642e-05,
+      "loss": 0.3158,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5518423307626392,
+      "grad_norm": 0.3683331310749054,
+      "learning_rate": 3.620822622107969e-05,
+      "loss": 0.3047,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5535561268209083,
+      "grad_norm": 0.40396639704704285,
+      "learning_rate": 3.616538131962296e-05,
+      "loss": 0.321,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5552699228791774,
+      "grad_norm": 0.4034627079963684,
+      "learning_rate": 3.6122536418166245e-05,
+      "loss": 0.3185,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5569837189374465,
+      "grad_norm": 0.3840987980365753,
+      "learning_rate": 3.6079691516709513e-05,
+      "loss": 0.3073,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5586975149957155,
+      "grad_norm": 0.4211840033531189,
+      "learning_rate": 3.603684661525278e-05,
+      "loss": 0.3198,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5604113110539846,
+      "grad_norm": 0.4210231900215149,
+      "learning_rate": 3.599400171379606e-05,
+      "loss": 0.28,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5621251071122536,
+      "grad_norm": 0.4034551680088043,
+      "learning_rate": 3.5951156812339334e-05,
+      "loss": 0.306,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5638389031705227,
+      "grad_norm": 0.41606375575065613,
+      "learning_rate": 3.590831191088261e-05,
+      "loss": 0.3238,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5655526992287918,
+      "grad_norm": 0.41806885600090027,
+      "learning_rate": 3.586546700942588e-05,
+      "loss": 0.3195,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5672664952870609,
+      "grad_norm": 0.3901151716709137,
+      "learning_rate": 3.5822622107969154e-05,
+      "loss": 0.3178,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5689802913453299,
+      "grad_norm": 0.4886183440685272,
+      "learning_rate": 3.577977720651243e-05,
+      "loss": 0.3058,
+      "step": 3320
+    },
+    {
+      "epoch": 0.570694087403599,
+      "grad_norm": 0.39053434133529663,
+      "learning_rate": 3.57369323050557e-05,
+      "loss": 0.2964,
+      "step": 3330
+    },
+    {
+      "epoch": 0.572407883461868,
+      "grad_norm": 0.42909854650497437,
+      "learning_rate": 3.5694087403598974e-05,
+      "loss": 0.3099,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5741216795201372,
+      "grad_norm": 0.3975629508495331,
+      "learning_rate": 3.565124250214224e-05,
+      "loss": 0.3014,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5758354755784062,
+      "grad_norm": 0.41938403248786926,
+      "learning_rate": 3.5608397600685525e-05,
+      "loss": 0.3081,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5775492716366752,
+      "grad_norm": 0.4035496413707733,
+      "learning_rate": 3.5565552699228794e-05,
+      "loss": 0.3009,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5792630676949443,
+      "grad_norm": 0.45365166664123535,
+      "learning_rate": 3.552270779777206e-05,
+      "loss": 0.32,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5809768637532133,
+      "grad_norm": 0.39258867502212524,
+      "learning_rate": 3.547986289631534e-05,
+      "loss": 0.3183,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5826906598114824,
+      "grad_norm": 0.5815192461013794,
+      "learning_rate": 3.5437017994858614e-05,
+      "loss": 0.3101,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5844044558697515,
+      "grad_norm": 0.42224058508872986,
+      "learning_rate": 3.539417309340189e-05,
+      "loss": 0.324,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5861182519280206,
+      "grad_norm": 0.4505104422569275,
+      "learning_rate": 3.535132819194516e-05,
+      "loss": 0.3245,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5878320479862896,
+      "grad_norm": 0.3851190507411957,
+      "learning_rate": 3.5308483290488434e-05,
+      "loss": 0.3244,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5895458440445587,
+      "grad_norm": 0.38795024156570435,
+      "learning_rate": 3.526563838903171e-05,
+      "loss": 0.3095,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5912596401028277,
+      "grad_norm": 0.4002437889575958,
+      "learning_rate": 3.522279348757498e-05,
+      "loss": 0.3081,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5929734361610969,
+      "grad_norm": 0.38535067439079285,
+      "learning_rate": 3.5179948586118254e-05,
+      "loss": 0.309,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5946872322193659,
+      "grad_norm": 0.4022165536880493,
+      "learning_rate": 3.513710368466153e-05,
+      "loss": 0.3204,
+      "step": 3470
+    },
+    {
+      "epoch": 0.596401028277635,
+      "grad_norm": 0.4585213363170624,
+      "learning_rate": 3.50942587832048e-05,
+      "loss": 0.3382,
+      "step": 3480
+    },
+    {
+      "epoch": 0.598114824335904,
+      "grad_norm": 0.3953900635242462,
+      "learning_rate": 3.5051413881748074e-05,
+      "loss": 0.3167,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5998286203941731,
+      "grad_norm": 0.42639920115470886,
+      "learning_rate": 3.500856898029134e-05,
+      "loss": 0.3028,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6015424164524421,
+      "grad_norm": 0.434978187084198,
+      "learning_rate": 3.496572407883462e-05,
+      "loss": 0.3076,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6032562125107113,
+      "grad_norm": 0.3918568789958954,
+      "learning_rate": 3.4922879177377894e-05,
+      "loss": 0.315,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6049700085689803,
+      "grad_norm": 0.4186408817768097,
+      "learning_rate": 3.488003427592117e-05,
+      "loss": 0.3139,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6066838046272494,
+      "grad_norm": 0.40361493825912476,
+      "learning_rate": 3.483718937446444e-05,
+      "loss": 0.3082,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6083976006855184,
+      "grad_norm": 0.4277793765068054,
+      "learning_rate": 3.4794344473007715e-05,
+      "loss": 0.3078,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6101113967437874,
+      "grad_norm": 0.44957688450813293,
+      "learning_rate": 3.475149957155099e-05,
+      "loss": 0.3101,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6118251928020566,
+      "grad_norm": 0.3873712718486786,
+      "learning_rate": 3.470865467009426e-05,
+      "loss": 0.3153,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6135389888603257,
+      "grad_norm": 0.4449580907821655,
+      "learning_rate": 3.4665809768637535e-05,
+      "loss": 0.3136,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6152527849185947,
+      "grad_norm": 0.3510473668575287,
+      "learning_rate": 3.462296486718081e-05,
+      "loss": 0.2965,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6169665809768637,
+      "grad_norm": 0.42599764466285706,
+      "learning_rate": 3.458011996572408e-05,
+      "loss": 0.3153,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6186803770351328,
+      "grad_norm": 0.36705848574638367,
+      "learning_rate": 3.4537275064267355e-05,
+      "loss": 0.3031,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6203941730934018,
+      "grad_norm": 0.3901403844356537,
+      "learning_rate": 3.4494430162810624e-05,
+      "loss": 0.3033,
+      "step": 3620
+    },
+    {
+      "epoch": 0.622107969151671,
+      "grad_norm": 0.4177183508872986,
+      "learning_rate": 3.44515852613539e-05,
+      "loss": 0.3119,
+      "step": 3630
+    },
+    {
+      "epoch": 0.62382176520994,
+      "grad_norm": 0.39418941736221313,
+      "learning_rate": 3.4408740359897175e-05,
+      "loss": 0.3118,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6255355612682091,
+      "grad_norm": 0.375946044921875,
+      "learning_rate": 3.4365895458440444e-05,
+      "loss": 0.3147,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6272493573264781,
+      "grad_norm": 0.4043296277523041,
+      "learning_rate": 3.432305055698372e-05,
+      "loss": 0.3142,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6289631533847472,
+      "grad_norm": 0.38900333642959595,
+      "learning_rate": 3.4280205655526995e-05,
+      "loss": 0.2958,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6306769494430163,
+      "grad_norm": 0.44235318899154663,
+      "learning_rate": 3.423736075407027e-05,
+      "loss": 0.3264,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6323907455012854,
+      "grad_norm": 0.3938853442668915,
+      "learning_rate": 3.419451585261354e-05,
+      "loss": 0.3102,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6341045415595544,
+      "grad_norm": 0.3863578140735626,
+      "learning_rate": 3.4151670951156815e-05,
+      "loss": 0.3091,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6358183376178235,
+      "grad_norm": 0.3701342046260834,
+      "learning_rate": 3.410882604970009e-05,
+      "loss": 0.3151,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6375321336760925,
+      "grad_norm": 0.3779924511909485,
+      "learning_rate": 3.406598114824336e-05,
+      "loss": 0.299,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6392459297343616,
+      "grad_norm": 0.39662858843803406,
+      "learning_rate": 3.4023136246786635e-05,
+      "loss": 0.3076,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6409597257926307,
+      "grad_norm": 0.35510873794555664,
+      "learning_rate": 3.3980291345329904e-05,
+      "loss": 0.316,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6426735218508998,
+      "grad_norm": 0.35568967461586,
+      "learning_rate": 3.3937446443873186e-05,
+      "loss": 0.308,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6443873179091688,
+      "grad_norm": 0.38227078318595886,
+      "learning_rate": 3.3894601542416455e-05,
+      "loss": 0.2913,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6461011139674379,
+      "grad_norm": 0.4040227234363556,
+      "learning_rate": 3.3851756640959724e-05,
+      "loss": 0.2976,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6478149100257069,
+      "grad_norm": 0.39672550559043884,
+      "learning_rate": 3.3808911739503e-05,
+      "loss": 0.3057,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6495287060839761,
+      "grad_norm": 0.4319610893726349,
+      "learning_rate": 3.3766066838046275e-05,
+      "loss": 0.313,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6512425021422451,
+      "grad_norm": 0.4290957748889923,
+      "learning_rate": 3.372322193658955e-05,
+      "loss": 0.3193,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6529562982005142,
+      "grad_norm": 0.40195539593696594,
+      "learning_rate": 3.368037703513282e-05,
+      "loss": 0.32,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6546700942587832,
+      "grad_norm": 0.3753972053527832,
+      "learning_rate": 3.363753213367609e-05,
+      "loss": 0.3149,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6563838903170522,
+      "grad_norm": 0.4367138147354126,
+      "learning_rate": 3.359468723221937e-05,
+      "loss": 0.3313,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6580976863753213,
+      "grad_norm": 0.4172726571559906,
+      "learning_rate": 3.355184233076264e-05,
+      "loss": 0.2958,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6598114824335904,
+      "grad_norm": 0.4136854410171509,
+      "learning_rate": 3.3508997429305916e-05,
+      "loss": 0.3271,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6615252784918595,
+      "grad_norm": 0.361028790473938,
+      "learning_rate": 3.3466152527849185e-05,
+      "loss": 0.3167,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6632390745501285,
+      "grad_norm": 0.37338533997535706,
+      "learning_rate": 3.342330762639246e-05,
+      "loss": 0.3101,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6649528706083976,
+      "grad_norm": 0.4406678080558777,
+      "learning_rate": 3.3380462724935736e-05,
+      "loss": 0.3008,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.39272168278694153,
+      "learning_rate": 3.3337617823479005e-05,
+      "loss": 0.2931,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6683804627249358,
+      "grad_norm": 0.40508198738098145,
+      "learning_rate": 3.329477292202228e-05,
+      "loss": 0.2997,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6700942587832048,
+      "grad_norm": 0.3926888406276703,
+      "learning_rate": 3.3251928020565556e-05,
+      "loss": 0.3158,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6718080548414739,
+      "grad_norm": 0.41264551877975464,
+      "learning_rate": 3.320908311910883e-05,
+      "loss": 0.3103,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6735218508997429,
+      "grad_norm": 0.4183017909526825,
+      "learning_rate": 3.31662382176521e-05,
+      "loss": 0.3157,
+      "step": 3930
+    },
+    {
+      "epoch": 0.675235646958012,
+      "grad_norm": 0.42990103363990784,
+      "learning_rate": 3.312339331619537e-05,
+      "loss": 0.3213,
+      "step": 3940
+    },
+    {
+      "epoch": 0.676949443016281,
+      "grad_norm": 0.3908301889896393,
+      "learning_rate": 3.308054841473865e-05,
+      "loss": 0.3097,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6786632390745502,
+      "grad_norm": 0.34192022681236267,
+      "learning_rate": 3.303770351328192e-05,
+      "loss": 0.2924,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6803770351328192,
+      "grad_norm": 0.3712366819381714,
+      "learning_rate": 3.2994858611825196e-05,
+      "loss": 0.3166,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6820908311910883,
+      "grad_norm": 0.38418450951576233,
+      "learning_rate": 3.2952013710368465e-05,
+      "loss": 0.3226,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6838046272493573,
+      "grad_norm": 0.35027676820755005,
+      "learning_rate": 3.290916880891174e-05,
+      "loss": 0.3169,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6855184233076264,
+      "grad_norm": 0.37734371423721313,
+      "learning_rate": 3.2866323907455016e-05,
+      "loss": 0.3208,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6872322193658955,
+      "grad_norm": 0.406653493642807,
+      "learning_rate": 3.2823479005998285e-05,
+      "loss": 0.2981,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6889460154241646,
+      "grad_norm": 0.3758227527141571,
+      "learning_rate": 3.278063410454156e-05,
+      "loss": 0.3106,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6906598114824336,
+      "grad_norm": 0.40474918484687805,
+      "learning_rate": 3.2737789203084836e-05,
+      "loss": 0.3088,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6923736075407027,
+      "grad_norm": 0.41509196162223816,
+      "learning_rate": 3.2694944301628105e-05,
+      "loss": 0.3076,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6940874035989717,
+      "grad_norm": 0.40685221552848816,
+      "learning_rate": 3.265209940017138e-05,
+      "loss": 0.3,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6958011996572407,
+      "grad_norm": 0.39293381571769714,
+      "learning_rate": 3.260925449871465e-05,
+      "loss": 0.297,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6975149957155099,
+      "grad_norm": 0.385626882314682,
+      "learning_rate": 3.256640959725793e-05,
+      "loss": 0.3112,
+      "step": 4070
+    },
+    {
+      "epoch": 0.699228791773779,
+      "grad_norm": 0.3609849512577057,
+      "learning_rate": 3.25235646958012e-05,
+      "loss": 0.3022,
+      "step": 4080
+    },
+    {
+      "epoch": 0.700942587832048,
+      "grad_norm": 0.45547643303871155,
+      "learning_rate": 3.248071979434447e-05,
+      "loss": 0.2966,
+      "step": 4090
+    },
+    {
+      "epoch": 0.702656383890317,
+      "grad_norm": 0.39011743664741516,
+      "learning_rate": 3.2437874892887745e-05,
+      "loss": 0.325,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7043701799485861,
+      "grad_norm": 0.4041966199874878,
+      "learning_rate": 3.239502999143102e-05,
+      "loss": 0.31,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7060839760068551,
+      "grad_norm": 0.4344090223312378,
+      "learning_rate": 3.23521850899743e-05,
+      "loss": 0.3157,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7077977720651243,
+      "grad_norm": 0.37727850675582886,
+      "learning_rate": 3.2309340188517566e-05,
+      "loss": 0.3144,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7095115681233933,
+      "grad_norm": 0.4806098937988281,
+      "learning_rate": 3.226649528706084e-05,
+      "loss": 0.308,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7112253641816624,
+      "grad_norm": 0.40128499269485474,
+      "learning_rate": 3.222365038560412e-05,
+      "loss": 0.2911,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7129391602399314,
+      "grad_norm": 0.42148104310035706,
+      "learning_rate": 3.2180805484147386e-05,
+      "loss": 0.3091,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7146529562982005,
+      "grad_norm": 0.41382914781570435,
+      "learning_rate": 3.213796058269066e-05,
+      "loss": 0.3102,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7163667523564696,
+      "grad_norm": 0.4522756040096283,
+      "learning_rate": 3.209511568123394e-05,
+      "loss": 0.3155,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7180805484147387,
+      "grad_norm": 0.40629592537879944,
+      "learning_rate": 3.205227077977721e-05,
+      "loss": 0.306,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7197943444730077,
+      "grad_norm": 0.3806251287460327,
+      "learning_rate": 3.200942587832048e-05,
+      "loss": 0.2924,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7215081405312768,
+      "grad_norm": 0.362578809261322,
+      "learning_rate": 3.196658097686375e-05,
+      "loss": 0.3073,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7232219365895458,
+      "grad_norm": 0.4149327576160431,
+      "learning_rate": 3.1923736075407026e-05,
+      "loss": 0.3078,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7249357326478149,
+      "grad_norm": 0.4010336995124817,
+      "learning_rate": 3.18808911739503e-05,
+      "loss": 0.2984,
+      "step": 4230
+    },
+    {
+      "epoch": 0.726649528706084,
+      "grad_norm": 0.3897489607334137,
+      "learning_rate": 3.183804627249358e-05,
+      "loss": 0.2962,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7283633247643531,
+      "grad_norm": 0.4140779376029968,
+      "learning_rate": 3.1795201371036846e-05,
+      "loss": 0.3113,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7300771208226221,
+      "grad_norm": 0.36506733298301697,
+      "learning_rate": 3.175235646958012e-05,
+      "loss": 0.3206,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7317909168808912,
+      "grad_norm": 0.377154678106308,
+      "learning_rate": 3.17095115681234e-05,
+      "loss": 0.2962,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7335047129391602,
+      "grad_norm": 0.4079926609992981,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.2976,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7352185089974294,
+      "grad_norm": 0.3581808805465698,
+      "learning_rate": 3.162382176520994e-05,
+      "loss": 0.3258,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7369323050556984,
+      "grad_norm": 0.41237592697143555,
+      "learning_rate": 3.158097686375322e-05,
+      "loss": 0.3078,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7386461011139674,
+      "grad_norm": 0.3892495036125183,
+      "learning_rate": 3.1538131962296486e-05,
+      "loss": 0.3222,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7403598971722365,
+      "grad_norm": 0.3892972767353058,
+      "learning_rate": 3.149528706083976e-05,
+      "loss": 0.3182,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7420736932305055,
+      "grad_norm": 0.35622286796569824,
+      "learning_rate": 3.145244215938303e-05,
+      "loss": 0.2935,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7437874892887746,
+      "grad_norm": 0.37672021985054016,
+      "learning_rate": 3.140959725792631e-05,
+      "loss": 0.3033,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7455012853470437,
+      "grad_norm": 0.39708781242370605,
+      "learning_rate": 3.136675235646958e-05,
+      "loss": 0.3037,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7472150814053128,
+      "grad_norm": 0.3636692464351654,
+      "learning_rate": 3.132390745501286e-05,
+      "loss": 0.3197,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7489288774635818,
+      "grad_norm": 0.412472665309906,
+      "learning_rate": 3.1281062553556126e-05,
+      "loss": 0.3147,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7506426735218509,
+      "grad_norm": 0.4187333881855011,
+      "learning_rate": 3.12382176520994e-05,
+      "loss": 0.321,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7523564695801199,
+      "grad_norm": 0.38632258772850037,
+      "learning_rate": 3.119537275064268e-05,
+      "loss": 0.3006,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7540702656383891,
+      "grad_norm": 0.4107763469219208,
+      "learning_rate": 3.1152527849185947e-05,
+      "loss": 0.3028,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7557840616966581,
+      "grad_norm": 0.3999440371990204,
+      "learning_rate": 3.110968294772922e-05,
+      "loss": 0.3056,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7574978577549272,
+      "grad_norm": 0.4522838294506073,
+      "learning_rate": 3.10668380462725e-05,
+      "loss": 0.3217,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7592116538131962,
+      "grad_norm": 0.3936906158924103,
+      "learning_rate": 3.102399314481577e-05,
+      "loss": 0.3263,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7609254498714653,
+      "grad_norm": 0.4072638750076294,
+      "learning_rate": 3.098114824335904e-05,
+      "loss": 0.3186,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7626392459297343,
+      "grad_norm": 0.4062633216381073,
+      "learning_rate": 3.093830334190231e-05,
+      "loss": 0.3312,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7643530419880035,
+      "grad_norm": 0.4110049605369568,
+      "learning_rate": 3.0895458440445594e-05,
+      "loss": 0.3163,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7660668380462725,
+      "grad_norm": 0.4055500328540802,
+      "learning_rate": 3.085261353898886e-05,
+      "loss": 0.2987,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7677806341045416,
+      "grad_norm": 0.4384361803531647,
+      "learning_rate": 3.080976863753213e-05,
+      "loss": 0.3097,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7694944301628106,
+      "grad_norm": 0.3911166191101074,
+      "learning_rate": 3.076692373607541e-05,
+      "loss": 0.3042,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7712082262210797,
+      "grad_norm": 0.381851464509964,
+      "learning_rate": 3.072407883461868e-05,
+      "loss": 0.3106,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7729220222793488,
+      "grad_norm": 0.4016900956630707,
+      "learning_rate": 3.068123393316196e-05,
+      "loss": 0.3145,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7746358183376179,
+      "grad_norm": 0.405513733625412,
+      "learning_rate": 3.063838903170523e-05,
+      "loss": 0.3007,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7763496143958869,
+      "grad_norm": 0.41820964217185974,
+      "learning_rate": 3.05955441302485e-05,
+      "loss": 0.3166,
+      "step": 4530
+    },
+    {
+      "epoch": 0.778063410454156,
+      "grad_norm": 0.4298486113548279,
+      "learning_rate": 3.055269922879178e-05,
+      "loss": 0.3014,
+      "step": 4540
+    },
+    {
+      "epoch": 0.779777206512425,
+      "grad_norm": 0.38170671463012695,
+      "learning_rate": 3.0509854327335047e-05,
+      "loss": 0.2908,
+      "step": 4550
+    },
+    {
+      "epoch": 0.781491002570694,
+      "grad_norm": 0.7947287559509277,
+      "learning_rate": 3.0467009425878323e-05,
+      "loss": 0.3115,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7832047986289632,
+      "grad_norm": 0.44346538186073303,
+      "learning_rate": 3.0424164524421595e-05,
+      "loss": 0.3043,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7849185946872322,
+      "grad_norm": 0.4573461413383484,
+      "learning_rate": 3.038131962296487e-05,
+      "loss": 0.3133,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7866323907455013,
+      "grad_norm": 0.38675206899642944,
+      "learning_rate": 3.0338474721508143e-05,
+      "loss": 0.3137,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7883461868037703,
+      "grad_norm": 0.40370744466781616,
+      "learning_rate": 3.0295629820051412e-05,
+      "loss": 0.3078,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7900599828620394,
+      "grad_norm": 0.3780510127544403,
+      "learning_rate": 3.025278491859469e-05,
+      "loss": 0.3051,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7917737789203085,
+      "grad_norm": 0.4316079914569855,
+      "learning_rate": 3.020994001713796e-05,
+      "loss": 0.3094,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7934875749785776,
+      "grad_norm": 0.38683196902275085,
+      "learning_rate": 3.016709511568124e-05,
+      "loss": 0.2837,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7952013710368466,
+      "grad_norm": 0.44551020860671997,
+      "learning_rate": 3.0124250214224508e-05,
+      "loss": 0.3027,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7969151670951157,
+      "grad_norm": 0.35879677534103394,
+      "learning_rate": 3.008140531276778e-05,
+      "loss": 0.3048,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7986289631533847,
+      "grad_norm": 0.3856838047504425,
+      "learning_rate": 3.0038560411311055e-05,
+      "loss": 0.3185,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8003427592116538,
+      "grad_norm": 0.40945693850517273,
+      "learning_rate": 2.9995715509854328e-05,
+      "loss": 0.3075,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8020565552699229,
+      "grad_norm": 0.3725194036960602,
+      "learning_rate": 2.9952870608397603e-05,
+      "loss": 0.3078,
+      "step": 4680
+    },
+    {
+      "epoch": 0.803770351328192,
+      "grad_norm": 0.3829960823059082,
+      "learning_rate": 2.9910025706940876e-05,
+      "loss": 0.3054,
+      "step": 4690
+    },
+    {
+      "epoch": 0.805484147386461,
+      "grad_norm": 0.4059649705886841,
+      "learning_rate": 2.9867180805484148e-05,
+      "loss": 0.3056,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8071979434447301,
+      "grad_norm": 0.4120868146419525,
+      "learning_rate": 2.9824335904027423e-05,
+      "loss": 0.3105,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8089117395029991,
+      "grad_norm": 0.3850419521331787,
+      "learning_rate": 2.9781491002570692e-05,
+      "loss": 0.2921,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8106255355612683,
+      "grad_norm": 0.408527672290802,
+      "learning_rate": 2.973864610111397e-05,
+      "loss": 0.3123,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8123393316195373,
+      "grad_norm": 0.4241873323917389,
+      "learning_rate": 2.969580119965724e-05,
+      "loss": 0.2894,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8140531276778064,
+      "grad_norm": 0.43134406208992004,
+      "learning_rate": 2.965295629820052e-05,
+      "loss": 0.315,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8157669237360754,
+      "grad_norm": 0.3915650546550751,
+      "learning_rate": 2.9610111396743788e-05,
+      "loss": 0.296,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8174807197943444,
+      "grad_norm": 0.3945915102958679,
+      "learning_rate": 2.956726649528706e-05,
+      "loss": 0.3131,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8191945158526135,
+      "grad_norm": 0.4121388792991638,
+      "learning_rate": 2.9524421593830336e-05,
+      "loss": 0.3026,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8209083119108826,
+      "grad_norm": 0.4124235510826111,
+      "learning_rate": 2.9481576692373608e-05,
+      "loss": 0.3089,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8226221079691517,
+      "grad_norm": 0.4459668695926666,
+      "learning_rate": 2.9438731790916884e-05,
+      "loss": 0.3144,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8243359040274207,
+      "grad_norm": 0.44301149249076843,
+      "learning_rate": 2.9395886889460156e-05,
+      "loss": 0.3083,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8260497000856898,
+      "grad_norm": 0.4339261054992676,
+      "learning_rate": 2.9353041988003428e-05,
+      "loss": 0.3142,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8277634961439588,
+      "grad_norm": 0.40257078409194946,
+      "learning_rate": 2.9310197086546704e-05,
+      "loss": 0.3035,
+      "step": 4830
+    },
+    {
+      "epoch": 0.829477292202228,
+      "grad_norm": 0.4084959328174591,
+      "learning_rate": 2.9267352185089976e-05,
+      "loss": 0.3049,
+      "step": 4840
+    },
+    {
+      "epoch": 0.831191088260497,
+      "grad_norm": 0.4485437273979187,
+      "learning_rate": 2.9224507283633252e-05,
+      "loss": 0.295,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8329048843187661,
+      "grad_norm": 0.42531272768974304,
+      "learning_rate": 2.918166238217652e-05,
+      "loss": 0.3008,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8346186803770351,
+      "grad_norm": 0.4186478853225708,
+      "learning_rate": 2.9138817480719793e-05,
+      "loss": 0.3045,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8363324764353042,
+      "grad_norm": 0.44759905338287354,
+      "learning_rate": 2.909597257926307e-05,
+      "loss": 0.3175,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8380462724935732,
+      "grad_norm": 0.39207273721694946,
+      "learning_rate": 2.905312767780634e-05,
+      "loss": 0.3006,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8397600685518424,
+      "grad_norm": 0.3415578305721283,
+      "learning_rate": 2.9010282776349616e-05,
+      "loss": 0.2952,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8414738646101114,
+      "grad_norm": 0.41721564531326294,
+      "learning_rate": 2.896743787489289e-05,
+      "loss": 0.3195,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8431876606683805,
+      "grad_norm": 0.40740966796875,
+      "learning_rate": 2.892459297343616e-05,
+      "loss": 0.3089,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8449014567266495,
+      "grad_norm": 0.48778238892555237,
+      "learning_rate": 2.8881748071979436e-05,
+      "loss": 0.311,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8466152527849186,
+      "grad_norm": 0.39738497138023376,
+      "learning_rate": 2.883890317052271e-05,
+      "loss": 0.3112,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8483290488431876,
+      "grad_norm": 0.3857996165752411,
+      "learning_rate": 2.8796058269065984e-05,
+      "loss": 0.3097,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8500428449014568,
+      "grad_norm": 0.37739014625549316,
+      "learning_rate": 2.8753213367609257e-05,
+      "loss": 0.3112,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8517566409597258,
+      "grad_norm": 0.3582662045955658,
+      "learning_rate": 2.8710368466152532e-05,
+      "loss": 0.3019,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8534704370179949,
+      "grad_norm": 0.4159857928752899,
+      "learning_rate": 2.8667523564695804e-05,
+      "loss": 0.3083,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8551842330762639,
+      "grad_norm": 0.41378700733184814,
+      "learning_rate": 2.8624678663239073e-05,
+      "loss": 0.3153,
+      "step": 4990
+    },
+    {
+      "epoch": 0.856898029134533,
+      "grad_norm": 0.4235714375972748,
+      "learning_rate": 2.8581833761782352e-05,
+      "loss": 0.2905,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8586118251928021,
+      "grad_norm": 0.45529866218566895,
+      "learning_rate": 2.853898886032562e-05,
+      "loss": 0.3119,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8603256212510711,
+      "grad_norm": 0.40263673663139343,
+      "learning_rate": 2.8496143958868897e-05,
+      "loss": 0.2947,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8620394173093402,
+      "grad_norm": 0.406290739774704,
+      "learning_rate": 2.845329905741217e-05,
+      "loss": 0.3095,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8637532133676092,
+      "grad_norm": 0.3786565065383911,
+      "learning_rate": 2.841045415595544e-05,
+      "loss": 0.3145,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8654670094258783,
+      "grad_norm": 0.37543046474456787,
+      "learning_rate": 2.8367609254498717e-05,
+      "loss": 0.3168,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8671808054841473,
+      "grad_norm": 0.4258265793323517,
+      "learning_rate": 2.832476435304199e-05,
+      "loss": 0.2966,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8688946015424165,
+      "grad_norm": 0.5114648938179016,
+      "learning_rate": 2.8281919451585265e-05,
+      "loss": 0.3057,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8706083976006855,
+      "grad_norm": 0.3751407563686371,
+      "learning_rate": 2.8239074550128537e-05,
+      "loss": 0.3065,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8723221936589546,
+      "grad_norm": 0.4122428297996521,
+      "learning_rate": 2.8196229648671806e-05,
+      "loss": 0.3121,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8740359897172236,
+      "grad_norm": 0.40011465549468994,
+      "learning_rate": 2.8153384747215085e-05,
+      "loss": 0.3062,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8757497857754927,
+      "grad_norm": 0.4549656808376312,
+      "learning_rate": 2.8110539845758354e-05,
+      "loss": 0.309,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8774635818337618,
+      "grad_norm": 0.40079987049102783,
+      "learning_rate": 2.8067694944301633e-05,
+      "loss": 0.3083,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8791773778920309,
+      "grad_norm": 0.4044526219367981,
+      "learning_rate": 2.80248500428449e-05,
+      "loss": 0.3078,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8808911739502999,
+      "grad_norm": 0.4119730293750763,
+      "learning_rate": 2.7982005141388174e-05,
+      "loss": 0.3126,
+      "step": 5140
+    },
+    {
+      "epoch": 0.882604970008569,
+      "grad_norm": 0.37520283460617065,
+      "learning_rate": 2.793916023993145e-05,
+      "loss": 0.3015,
+      "step": 5150
+    },
+    {
+      "epoch": 0.884318766066838,
+      "grad_norm": 0.4071120321750641,
+      "learning_rate": 2.7896315338474722e-05,
+      "loss": 0.3274,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8860325621251071,
+      "grad_norm": 0.38094213604927063,
+      "learning_rate": 2.7853470437017997e-05,
+      "loss": 0.32,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8877463581833762,
+      "grad_norm": 0.4089542627334595,
+      "learning_rate": 2.781062553556127e-05,
+      "loss": 0.2852,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8894601542416453,
+      "grad_norm": 0.42206066846847534,
+      "learning_rate": 2.7767780634104545e-05,
+      "loss": 0.3142,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8911739502999143,
+      "grad_norm": 0.400400847196579,
+      "learning_rate": 2.7724935732647817e-05,
+      "loss": 0.2995,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8928877463581834,
+      "grad_norm": 0.407252699136734,
+      "learning_rate": 2.7682090831191086e-05,
+      "loss": 0.3165,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8946015424164524,
+      "grad_norm": 0.42390021681785583,
+      "learning_rate": 2.7639245929734365e-05,
+      "loss": 0.3136,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8963153384747216,
+      "grad_norm": 0.37746793031692505,
+      "learning_rate": 2.7596401028277634e-05,
+      "loss": 0.2955,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8980291345329906,
+      "grad_norm": 0.40552595257759094,
+      "learning_rate": 2.7553556126820913e-05,
+      "loss": 0.3075,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8997429305912596,
+      "grad_norm": 0.38914138078689575,
+      "learning_rate": 2.7510711225364182e-05,
+      "loss": 0.3058,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9014567266495287,
+      "grad_norm": 0.39012840390205383,
+      "learning_rate": 2.7467866323907454e-05,
+      "loss": 0.3006,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9031705227077977,
+      "grad_norm": 0.40481963753700256,
+      "learning_rate": 2.742502142245073e-05,
+      "loss": 0.3051,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9048843187660668,
+      "grad_norm": 0.3763148784637451,
+      "learning_rate": 2.7382176520994002e-05,
+      "loss": 0.2963,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9065981148243359,
+      "grad_norm": 0.3657761216163635,
+      "learning_rate": 2.7339331619537278e-05,
+      "loss": 0.297,
+      "step": 5290
+    },
+    {
+      "epoch": 0.908311910882605,
+      "grad_norm": 0.41030603647232056,
+      "learning_rate": 2.729648671808055e-05,
+      "loss": 0.3033,
+      "step": 5300
+    },
+    {
+      "epoch": 0.910025706940874,
+      "grad_norm": 0.38039547204971313,
+      "learning_rate": 2.725364181662382e-05,
+      "loss": 0.3058,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9117395029991431,
+      "grad_norm": 0.3992447257041931,
+      "learning_rate": 2.7210796915167098e-05,
+      "loss": 0.3029,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9134532990574121,
+      "grad_norm": 0.40409329533576965,
+      "learning_rate": 2.7167952013710367e-05,
+      "loss": 0.3046,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9151670951156813,
+      "grad_norm": 0.383213609457016,
+      "learning_rate": 2.7125107112253646e-05,
+      "loss": 0.297,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9168808911739503,
+      "grad_norm": 0.42381173372268677,
+      "learning_rate": 2.7082262210796915e-05,
+      "loss": 0.3041,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9185946872322194,
+      "grad_norm": 0.4012395739555359,
+      "learning_rate": 2.7039417309340194e-05,
+      "loss": 0.3012,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9203084832904884,
+      "grad_norm": 0.4109445810317993,
+      "learning_rate": 2.6996572407883463e-05,
+      "loss": 0.3006,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9220222793487575,
+      "grad_norm": 0.4355178475379944,
+      "learning_rate": 2.6953727506426735e-05,
+      "loss": 0.3175,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9237360754070265,
+      "grad_norm": 0.40131524205207825,
+      "learning_rate": 2.691088260497001e-05,
+      "loss": 0.3167,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9254498714652957,
+      "grad_norm": 0.47020289301872253,
+      "learning_rate": 2.6868037703513283e-05,
+      "loss": 0.2983,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9271636675235647,
+      "grad_norm": 0.3975673317909241,
+      "learning_rate": 2.6825192802056558e-05,
+      "loss": 0.3276,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9288774635818338,
+      "grad_norm": 0.3912968039512634,
+      "learning_rate": 2.678234790059983e-05,
+      "loss": 0.3191,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9305912596401028,
+      "grad_norm": 0.4399542212486267,
+      "learning_rate": 2.6739502999143103e-05,
+      "loss": 0.3283,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9323050556983719,
+      "grad_norm": 0.38910576701164246,
+      "learning_rate": 2.669665809768638e-05,
+      "loss": 0.2955,
+      "step": 5440
+    },
+    {
+      "epoch": 0.934018851756641,
+      "grad_norm": 0.40711304545402527,
+      "learning_rate": 2.6653813196229647e-05,
+      "loss": 0.303,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9357326478149101,
+      "grad_norm": 0.38116922974586487,
+      "learning_rate": 2.6610968294772926e-05,
+      "loss": 0.3205,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9374464438731791,
+      "grad_norm": 0.366163432598114,
+      "learning_rate": 2.6568123393316195e-05,
+      "loss": 0.3088,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9391602399314481,
+      "grad_norm": 0.3540242910385132,
+      "learning_rate": 2.6525278491859467e-05,
+      "loss": 0.3033,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9408740359897172,
+      "grad_norm": 0.4145238995552063,
+      "learning_rate": 2.6482433590402743e-05,
+      "loss": 0.2987,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9425878320479862,
+      "grad_norm": 0.3904762864112854,
+      "learning_rate": 2.6439588688946015e-05,
+      "loss": 0.3064,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9443016281062554,
+      "grad_norm": 0.45036500692367554,
+      "learning_rate": 2.639674378748929e-05,
+      "loss": 0.3139,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9460154241645244,
+      "grad_norm": 0.3978581726551056,
+      "learning_rate": 2.6353898886032563e-05,
+      "loss": 0.3063,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9477292202227935,
+      "grad_norm": 0.41630059480667114,
+      "learning_rate": 2.6311053984575835e-05,
+      "loss": 0.3133,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9494430162810625,
+      "grad_norm": 0.38944846391677856,
+      "learning_rate": 2.626820908311911e-05,
+      "loss": 0.306,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9511568123393316,
+      "grad_norm": 0.3740023076534271,
+      "learning_rate": 2.6225364181662383e-05,
+      "loss": 0.3063,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9528706083976007,
+      "grad_norm": 0.4348326027393341,
+      "learning_rate": 2.618251928020566e-05,
+      "loss": 0.2943,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9545844044558698,
+      "grad_norm": 0.4485342502593994,
+      "learning_rate": 2.613967437874893e-05,
+      "loss": 0.3019,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9562982005141388,
+      "grad_norm": 0.43861231207847595,
+      "learning_rate": 2.6096829477292207e-05,
+      "loss": 0.312,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9580119965724079,
+      "grad_norm": 0.43278244137763977,
+      "learning_rate": 2.6053984575835476e-05,
+      "loss": 0.3068,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9597257926306769,
+      "grad_norm": 0.40838947892189026,
+      "learning_rate": 2.6011139674378748e-05,
+      "loss": 0.3145,
+      "step": 5600
+    },
+    {
+      "epoch": 0.961439588688946,
+      "grad_norm": 0.4251978099346161,
+      "learning_rate": 2.5968294772922023e-05,
+      "loss": 0.3015,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9631533847472151,
+      "grad_norm": 0.4422808885574341,
+      "learning_rate": 2.5925449871465296e-05,
+      "loss": 0.298,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9648671808054842,
+      "grad_norm": 0.4073251187801361,
+      "learning_rate": 2.588260497000857e-05,
+      "loss": 0.3065,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9665809768637532,
+      "grad_norm": 0.37306833267211914,
+      "learning_rate": 2.5839760068551844e-05,
+      "loss": 0.3211,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9682947729220223,
+      "grad_norm": 0.35942408442497253,
+      "learning_rate": 2.5796915167095116e-05,
+      "loss": 0.3135,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9700085689802913,
+      "grad_norm": 0.45881766080856323,
+      "learning_rate": 2.575407026563839e-05,
+      "loss": 0.3227,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9717223650385605,
+      "grad_norm": 0.39790812134742737,
+      "learning_rate": 2.5711225364181664e-05,
+      "loss": 0.2983,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9734361610968295,
+      "grad_norm": 0.44903764128685,
+      "learning_rate": 2.566838046272494e-05,
+      "loss": 0.305,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9751499571550986,
+      "grad_norm": 0.3822721540927887,
+      "learning_rate": 2.562553556126821e-05,
+      "loss": 0.3109,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9768637532133676,
+      "grad_norm": 0.42775872349739075,
+      "learning_rate": 2.558269065981148e-05,
+      "loss": 0.2989,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9785775492716366,
+      "grad_norm": 0.4138841927051544,
+      "learning_rate": 2.553984575835476e-05,
+      "loss": 0.3077,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9802913453299057,
+      "grad_norm": 0.4355012774467468,
+      "learning_rate": 2.5497000856898028e-05,
+      "loss": 0.3033,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9820051413881749,
+      "grad_norm": 0.3857831060886383,
+      "learning_rate": 2.5454155955441304e-05,
+      "loss": 0.3069,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9837189374464439,
+      "grad_norm": 0.4003482758998871,
+      "learning_rate": 2.5411311053984576e-05,
+      "loss": 0.2976,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9854327335047129,
+      "grad_norm": 0.41185998916625977,
+      "learning_rate": 2.536846615252785e-05,
+      "loss": 0.3097,
+      "step": 5750
+    },
+    {
+      "epoch": 0.987146529562982,
+      "grad_norm": 0.4086381196975708,
+      "learning_rate": 2.5325621251071124e-05,
+      "loss": 0.2938,
+      "step": 5760
+    },
+    {
+      "epoch": 0.988860325621251,
+      "grad_norm": 0.4063851535320282,
+      "learning_rate": 2.5282776349614396e-05,
+      "loss": 0.2911,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9905741216795202,
+      "grad_norm": 0.4704953134059906,
+      "learning_rate": 2.5239931448157672e-05,
+      "loss": 0.3084,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9922879177377892,
+      "grad_norm": 0.4330461025238037,
+      "learning_rate": 2.5197086546700944e-05,
+      "loss": 0.3028,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9940017137960583,
+      "grad_norm": 0.4061424434185028,
+      "learning_rate": 2.515424164524422e-05,
+      "loss": 0.318,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9957155098543273,
+      "grad_norm": 0.7861102819442749,
+      "learning_rate": 2.5111396743787492e-05,
+      "loss": 0.3197,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9974293059125964,
+      "grad_norm": 0.39462533593177795,
+      "learning_rate": 2.506855184233076e-05,
+      "loss": 0.3077,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9991431019708654,
+      "grad_norm": 0.4004395306110382,
+      "learning_rate": 2.502570694087404e-05,
+      "loss": 0.2802,
+      "step": 5830
+    },
+    {
+      "epoch": 1.0008568980291346,
+      "grad_norm": 0.4045383334159851,
+      "learning_rate": 2.498286203941731e-05,
+      "loss": 0.2908,
+      "step": 5840
+    },
+    {
+      "epoch": 1.0025706940874035,
+      "grad_norm": 0.4202393889427185,
+      "learning_rate": 2.4940017137960584e-05,
+      "loss": 0.2769,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0042844901456727,
+      "grad_norm": 0.45001935958862305,
+      "learning_rate": 2.4897172236503857e-05,
+      "loss": 0.2953,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0059982862039418,
+      "grad_norm": 0.44940000772476196,
+      "learning_rate": 2.4854327335047132e-05,
+      "loss": 0.2936,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0077120822622108,
+      "grad_norm": 0.42108064889907837,
+      "learning_rate": 2.4811482433590404e-05,
+      "loss": 0.2941,
+      "step": 5880
+    },
+    {
+      "epoch": 1.00942587832048,
+      "grad_norm": 0.3818753659725189,
+      "learning_rate": 2.4768637532133677e-05,
+      "loss": 0.2809,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0111396743787489,
+      "grad_norm": 0.40024736523628235,
+      "learning_rate": 2.472579263067695e-05,
+      "loss": 0.2936,
+      "step": 5900
+    },
+    {
+      "epoch": 1.012853470437018,
+      "grad_norm": 0.4132537245750427,
+      "learning_rate": 2.4682947729220225e-05,
+      "loss": 0.3074,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0145672664952872,
+      "grad_norm": 0.4422491788864136,
+      "learning_rate": 2.4640102827763497e-05,
+      "loss": 0.2946,
+      "step": 5920
+    },
+    {
+      "epoch": 1.016281062553556,
+      "grad_norm": 0.3496541976928711,
+      "learning_rate": 2.4597257926306772e-05,
+      "loss": 0.2833,
+      "step": 5930
+    },
+    {
+      "epoch": 1.0179948586118253,
+      "grad_norm": 0.4277006983757019,
+      "learning_rate": 2.4554413024850045e-05,
+      "loss": 0.289,
+      "step": 5940
+    },
+    {
+      "epoch": 1.0197086546700942,
+      "grad_norm": 0.4157383143901825,
+      "learning_rate": 2.4511568123393317e-05,
+      "loss": 0.2988,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0214224507283634,
+      "grad_norm": 0.4295196533203125,
+      "learning_rate": 2.446872322193659e-05,
+      "loss": 0.2871,
+      "step": 5960
+    },
+    {
+      "epoch": 1.0231362467866323,
+      "grad_norm": 0.40148210525512695,
+      "learning_rate": 2.4425878320479865e-05,
+      "loss": 0.2894,
+      "step": 5970
+    },
+    {
+      "epoch": 1.0248500428449014,
+      "grad_norm": 0.3696572780609131,
+      "learning_rate": 2.4383033419023137e-05,
+      "loss": 0.283,
+      "step": 5980
+    },
+    {
+      "epoch": 1.0265638389031706,
+      "grad_norm": 0.40655508637428284,
+      "learning_rate": 2.4340188517566413e-05,
+      "loss": 0.2727,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0282776349614395,
+      "grad_norm": 0.3886098563671112,
+      "learning_rate": 2.429734361610968e-05,
+      "loss": 0.3067,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0299914310197087,
+      "grad_norm": 0.3908361494541168,
+      "learning_rate": 2.4254498714652957e-05,
+      "loss": 0.2848,
+      "step": 6010
+    },
+    {
+      "epoch": 1.0317052270779776,
+      "grad_norm": 0.4567308723926544,
+      "learning_rate": 2.421165381319623e-05,
+      "loss": 0.2801,
+      "step": 6020
+    },
+    {
+      "epoch": 1.0334190231362468,
+      "grad_norm": 0.4428994357585907,
+      "learning_rate": 2.4168808911739505e-05,
+      "loss": 0.2974,
+      "step": 6030
+    },
+    {
+      "epoch": 1.035132819194516,
+      "grad_norm": 0.3713628053665161,
+      "learning_rate": 2.4125964010282777e-05,
+      "loss": 0.299,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0368466152527849,
+      "grad_norm": 0.39820927381515503,
+      "learning_rate": 2.4083119108826053e-05,
+      "loss": 0.2945,
+      "step": 6050
+    },
+    {
+      "epoch": 1.038560411311054,
+      "grad_norm": 0.4364137649536133,
+      "learning_rate": 2.4040274207369322e-05,
+      "loss": 0.2768,
+      "step": 6060
+    },
+    {
+      "epoch": 1.040274207369323,
+      "grad_norm": 0.3995022177696228,
+      "learning_rate": 2.3997429305912597e-05,
+      "loss": 0.2923,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0419880034275921,
+      "grad_norm": 0.3767637312412262,
+      "learning_rate": 2.395458440445587e-05,
+      "loss": 0.2952,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0437017994858613,
+      "grad_norm": 0.4112531840801239,
+      "learning_rate": 2.3911739502999145e-05,
+      "loss": 0.2864,
+      "step": 6090
+    },
+    {
+      "epoch": 1.0454155955441302,
+      "grad_norm": 0.4636956453323364,
+      "learning_rate": 2.3868894601542418e-05,
+      "loss": 0.2908,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0471293916023994,
+      "grad_norm": 0.45814594626426697,
+      "learning_rate": 2.382604970008569e-05,
+      "loss": 0.2925,
+      "step": 6110
+    },
+    {
+      "epoch": 1.0488431876606683,
+      "grad_norm": 0.3936929702758789,
+      "learning_rate": 2.3783204798628962e-05,
+      "loss": 0.2919,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0505569837189375,
+      "grad_norm": 0.4754982590675354,
+      "learning_rate": 2.3740359897172238e-05,
+      "loss": 0.2901,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0522707797772064,
+      "grad_norm": 0.42620086669921875,
+      "learning_rate": 2.369751499571551e-05,
+      "loss": 0.283,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0539845758354756,
+      "grad_norm": 0.4347045123577118,
+      "learning_rate": 2.3654670094258786e-05,
+      "loss": 0.284,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0556983718937447,
+      "grad_norm": 0.3772675395011902,
+      "learning_rate": 2.3611825192802058e-05,
+      "loss": 0.283,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0574121679520136,
+      "grad_norm": 0.4097364544868469,
+      "learning_rate": 2.356898029134533e-05,
+      "loss": 0.2858,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0591259640102828,
+      "grad_norm": 0.44972333312034607,
+      "learning_rate": 2.3526135389888602e-05,
+      "loss": 0.2865,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0608397600685517,
+      "grad_norm": 0.39057499170303345,
+      "learning_rate": 2.3483290488431878e-05,
+      "loss": 0.3002,
+      "step": 6190
+    },
+    {
+      "epoch": 1.062553556126821,
+      "grad_norm": 0.4321994483470917,
+      "learning_rate": 2.344044558697515e-05,
+      "loss": 0.2928,
+      "step": 6200
+    },
+    {
+      "epoch": 1.06426735218509,
+      "grad_norm": 0.4209161102771759,
+      "learning_rate": 2.3397600685518426e-05,
+      "loss": 0.3012,
+      "step": 6210
+    },
+    {
+      "epoch": 1.065981148243359,
+      "grad_norm": 0.40629979968070984,
+      "learning_rate": 2.3354755784061698e-05,
+      "loss": 0.294,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0676949443016281,
+      "grad_norm": 0.40791964530944824,
+      "learning_rate": 2.331191088260497e-05,
+      "loss": 0.2923,
+      "step": 6230
+    },
+    {
+      "epoch": 1.069408740359897,
+      "grad_norm": 0.4262239933013916,
+      "learning_rate": 2.3269065981148246e-05,
+      "loss": 0.2823,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0711225364181662,
+      "grad_norm": 0.41918402910232544,
+      "learning_rate": 2.3226221079691518e-05,
+      "loss": 0.2939,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0728363324764354,
+      "grad_norm": 0.3853673040866852,
+      "learning_rate": 2.318337617823479e-05,
+      "loss": 0.2833,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0745501285347043,
+      "grad_norm": 0.4460880160331726,
+      "learning_rate": 2.3140531276778066e-05,
+      "loss": 0.2898,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0762639245929735,
+      "grad_norm": 0.39993777871131897,
+      "learning_rate": 2.3097686375321338e-05,
+      "loss": 0.3101,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0779777206512424,
+      "grad_norm": 0.4199289381504059,
+      "learning_rate": 2.305484147386461e-05,
+      "loss": 0.3039,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0796915167095116,
+      "grad_norm": 0.42590129375457764,
+      "learning_rate": 2.3011996572407886e-05,
+      "loss": 0.2937,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0814053127677807,
+      "grad_norm": 0.4059952199459076,
+      "learning_rate": 2.296915167095116e-05,
+      "loss": 0.3043,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0831191088260497,
+      "grad_norm": 0.47935783863067627,
+      "learning_rate": 2.292630676949443e-05,
+      "loss": 0.2846,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0848329048843188,
+      "grad_norm": 0.45164498686790466,
+      "learning_rate": 2.2883461868037703e-05,
+      "loss": 0.3038,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0865467009425878,
+      "grad_norm": 0.4237392246723175,
+      "learning_rate": 2.284061696658098e-05,
+      "loss": 0.3054,
+      "step": 6340
+    },
+    {
+      "epoch": 1.088260497000857,
+      "grad_norm": 0.36155253648757935,
+      "learning_rate": 2.279777206512425e-05,
+      "loss": 0.2792,
+      "step": 6350
+    },
+    {
+      "epoch": 1.089974293059126,
+      "grad_norm": 0.3712191879749298,
+      "learning_rate": 2.2754927163667526e-05,
+      "loss": 0.2927,
+      "step": 6360
+    },
+    {
+      "epoch": 1.091688089117395,
+      "grad_norm": 0.422678142786026,
+      "learning_rate": 2.27120822622108e-05,
+      "loss": 0.283,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0934018851756642,
+      "grad_norm": 0.411361426115036,
+      "learning_rate": 2.2669237360754074e-05,
+      "loss": 0.2983,
+      "step": 6380
+    },
+    {
+      "epoch": 1.095115681233933,
+      "grad_norm": 0.4110596477985382,
+      "learning_rate": 2.2626392459297343e-05,
+      "loss": 0.2965,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0968294772922023,
+      "grad_norm": 0.4057029187679291,
+      "learning_rate": 2.258354755784062e-05,
+      "loss": 0.2961,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0985432733504712,
+      "grad_norm": 0.3869575262069702,
+      "learning_rate": 2.254070265638389e-05,
+      "loss": 0.2726,
+      "step": 6410
+    },
+    {
+      "epoch": 1.1002570694087404,
+      "grad_norm": 0.4176071584224701,
+      "learning_rate": 2.2497857754927167e-05,
+      "loss": 0.2973,
+      "step": 6420
+    },
+    {
+      "epoch": 1.1019708654670095,
+      "grad_norm": 0.4128270447254181,
+      "learning_rate": 2.245501285347044e-05,
+      "loss": 0.2887,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1036846615252784,
+      "grad_norm": 0.40935301780700684,
+      "learning_rate": 2.241216795201371e-05,
+      "loss": 0.2872,
+      "step": 6440
+    },
+    {
+      "epoch": 1.1053984575835476,
+      "grad_norm": 0.40727561712265015,
+      "learning_rate": 2.2369323050556983e-05,
+      "loss": 0.2942,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1071122536418165,
+      "grad_norm": 0.40293702483177185,
+      "learning_rate": 2.232647814910026e-05,
+      "loss": 0.3151,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1088260497000857,
+      "grad_norm": 0.4084971249103546,
+      "learning_rate": 2.228363324764353e-05,
+      "loss": 0.2875,
+      "step": 6470
+    },
+    {
+      "epoch": 1.1105398457583548,
+      "grad_norm": 0.3905606269836426,
+      "learning_rate": 2.2240788346186807e-05,
+      "loss": 0.288,
+      "step": 6480
+    },
+    {
+      "epoch": 1.1122536418166238,
+      "grad_norm": 0.4392050802707672,
+      "learning_rate": 2.219794344473008e-05,
+      "loss": 0.2912,
+      "step": 6490
+    },
+    {
+      "epoch": 1.113967437874893,
+      "grad_norm": 0.40402621030807495,
+      "learning_rate": 2.215509854327335e-05,
+      "loss": 0.2992,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1156812339331619,
+      "grad_norm": 0.42626067996025085,
+      "learning_rate": 2.2112253641816624e-05,
+      "loss": 0.2925,
+      "step": 6510
+    },
+    {
+      "epoch": 1.117395029991431,
+      "grad_norm": 0.4342977702617645,
+      "learning_rate": 2.20694087403599e-05,
+      "loss": 0.2966,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1191088260497002,
+      "grad_norm": 0.42427292466163635,
+      "learning_rate": 2.202656383890317e-05,
+      "loss": 0.2904,
+      "step": 6530
+    },
+    {
+      "epoch": 1.1208226221079691,
+      "grad_norm": 0.41280657052993774,
+      "learning_rate": 2.1983718937446447e-05,
+      "loss": 0.2999,
+      "step": 6540
+    },
+    {
+      "epoch": 1.1225364181662383,
+      "grad_norm": 0.41753578186035156,
+      "learning_rate": 2.1940874035989716e-05,
+      "loss": 0.2863,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1242502142245072,
+      "grad_norm": 0.392566442489624,
+      "learning_rate": 2.189802913453299e-05,
+      "loss": 0.3039,
+      "step": 6560
+    },
+    {
+      "epoch": 1.1259640102827764,
+      "grad_norm": 0.41189509630203247,
+      "learning_rate": 2.1855184233076264e-05,
+      "loss": 0.2907,
+      "step": 6570
+    },
+    {
+      "epoch": 1.1276778063410453,
+      "grad_norm": 0.420646607875824,
+      "learning_rate": 2.181233933161954e-05,
+      "loss": 0.2874,
+      "step": 6580
+    },
+    {
+      "epoch": 1.1293916023993145,
+      "grad_norm": 0.371812105178833,
+      "learning_rate": 2.176949443016281e-05,
+      "loss": 0.2961,
+      "step": 6590
+    },
+    {
+      "epoch": 1.1311053984575836,
+      "grad_norm": 0.3978787958621979,
+      "learning_rate": 2.1726649528706087e-05,
+      "loss": 0.2928,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1328191945158526,
+      "grad_norm": 0.4301498532295227,
+      "learning_rate": 2.1683804627249356e-05,
+      "loss": 0.284,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1345329905741217,
+      "grad_norm": 0.42761191725730896,
+      "learning_rate": 2.1640959725792632e-05,
+      "loss": 0.2977,
+      "step": 6620
+    },
+    {
+      "epoch": 1.1362467866323906,
+      "grad_norm": 0.45269954204559326,
+      "learning_rate": 2.1598114824335904e-05,
+      "loss": 0.2982,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1379605826906598,
+      "grad_norm": 0.4121401011943817,
+      "learning_rate": 2.155526992287918e-05,
+      "loss": 0.3003,
+      "step": 6640
+    },
+    {
+      "epoch": 1.139674378748929,
+      "grad_norm": 0.4257526695728302,
+      "learning_rate": 2.1512425021422452e-05,
+      "loss": 0.2893,
+      "step": 6650
+    },
+    {
+      "epoch": 1.141388174807198,
+      "grad_norm": 0.41644781827926636,
+      "learning_rate": 2.1469580119965727e-05,
+      "loss": 0.2896,
+      "step": 6660
+    },
+    {
+      "epoch": 1.143101970865467,
+      "grad_norm": 0.4130077362060547,
+      "learning_rate": 2.1426735218508996e-05,
+      "loss": 0.2926,
+      "step": 6670
+    },
+    {
+      "epoch": 1.144815766923736,
+      "grad_norm": 0.38322991132736206,
+      "learning_rate": 2.1383890317052272e-05,
+      "loss": 0.2985,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1465295629820051,
+      "grad_norm": 0.4365147054195404,
+      "learning_rate": 2.1341045415595544e-05,
+      "loss": 0.2944,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1482433590402743,
+      "grad_norm": 0.45200902223587036,
+      "learning_rate": 2.129820051413882e-05,
+      "loss": 0.289,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1499571550985432,
+      "grad_norm": 0.4541572034358978,
+      "learning_rate": 2.1255355612682092e-05,
+      "loss": 0.2965,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1516709511568124,
+      "grad_norm": 0.4185004234313965,
+      "learning_rate": 2.1212510711225364e-05,
+      "loss": 0.283,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1533847472150813,
+      "grad_norm": 0.4480586647987366,
+      "learning_rate": 2.1169665809768637e-05,
+      "loss": 0.2839,
+      "step": 6730
+    },
+    {
+      "epoch": 1.1550985432733505,
+      "grad_norm": 0.4363707900047302,
+      "learning_rate": 2.1126820908311912e-05,
+      "loss": 0.291,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1568123393316196,
+      "grad_norm": 0.43228283524513245,
+      "learning_rate": 2.1083976006855184e-05,
+      "loss": 0.2919,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1585261353898886,
+      "grad_norm": 0.38966530561447144,
+      "learning_rate": 2.104113110539846e-05,
+      "loss": 0.2941,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1602399314481577,
+      "grad_norm": 0.42686596512794495,
+      "learning_rate": 2.0998286203941732e-05,
+      "loss": 0.2952,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1619537275064267,
+      "grad_norm": 0.4043618440628052,
+      "learning_rate": 2.0955441302485005e-05,
+      "loss": 0.294,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1636675235646958,
+      "grad_norm": 0.44633689522743225,
+      "learning_rate": 2.0912596401028277e-05,
+      "loss": 0.2977,
+      "step": 6790
+    },
+    {
+      "epoch": 1.165381319622965,
+      "grad_norm": 0.42492836713790894,
+      "learning_rate": 2.0869751499571552e-05,
+      "loss": 0.3026,
+      "step": 6800
+    },
+    {
+      "epoch": 1.167095115681234,
+      "grad_norm": 0.40678128600120544,
+      "learning_rate": 2.0826906598114825e-05,
+      "loss": 0.2908,
+      "step": 6810
+    },
+    {
+      "epoch": 1.168808911739503,
+      "grad_norm": 0.4077664911746979,
+      "learning_rate": 2.07840616966581e-05,
+      "loss": 0.2808,
+      "step": 6820
+    },
+    {
+      "epoch": 1.170522707797772,
+      "grad_norm": 0.3967554569244385,
+      "learning_rate": 2.0741216795201373e-05,
+      "loss": 0.2958,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1722365038560412,
+      "grad_norm": 0.41393062472343445,
+      "learning_rate": 2.0698371893744645e-05,
+      "loss": 0.2944,
+      "step": 6840
+    },
+    {
+      "epoch": 1.17395029991431,
+      "grad_norm": 0.4049648344516754,
+      "learning_rate": 2.0655526992287917e-05,
+      "loss": 0.2881,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1756640959725793,
+      "grad_norm": 0.3909173011779785,
+      "learning_rate": 2.0612682090831193e-05,
+      "loss": 0.2863,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1773778920308484,
+      "grad_norm": 0.43796250224113464,
+      "learning_rate": 2.0569837189374465e-05,
+      "loss": 0.2825,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1790916880891174,
+      "grad_norm": 0.41272875666618347,
+      "learning_rate": 2.052699228791774e-05,
+      "loss": 0.2922,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1808054841473865,
+      "grad_norm": 0.37688764929771423,
+      "learning_rate": 2.0484147386461013e-05,
+      "loss": 0.2962,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1825192802056554,
+      "grad_norm": 0.40608006715774536,
+      "learning_rate": 2.0441302485004285e-05,
+      "loss": 0.2908,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1842330762639246,
+      "grad_norm": 0.43522876501083374,
+      "learning_rate": 2.0398457583547557e-05,
+      "loss": 0.2886,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1859468723221935,
+      "grad_norm": 0.3659178614616394,
+      "learning_rate": 2.0355612682090833e-05,
+      "loss": 0.2982,
+      "step": 6920
+    },
+    {
+      "epoch": 1.1876606683804627,
+      "grad_norm": 0.4055621027946472,
+      "learning_rate": 2.0312767780634105e-05,
+      "loss": 0.2956,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1893744644387318,
+      "grad_norm": 0.4199269115924835,
+      "learning_rate": 2.0269922879177377e-05,
+      "loss": 0.2956,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1910882604970008,
+      "grad_norm": 0.4305850863456726,
+      "learning_rate": 2.0227077977720653e-05,
+      "loss": 0.2982,
+      "step": 6950
+    },
+    {
+      "epoch": 1.19280205655527,
+      "grad_norm": 0.39645740389823914,
+      "learning_rate": 2.0184233076263925e-05,
+      "loss": 0.288,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1945158526135389,
+      "grad_norm": 0.40040698647499084,
+      "learning_rate": 2.01413881748072e-05,
+      "loss": 0.2898,
+      "step": 6970
+    },
+    {
+      "epoch": 1.196229648671808,
+      "grad_norm": 0.3971509039402008,
+      "learning_rate": 2.0098543273350473e-05,
+      "loss": 0.2985,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1979434447300772,
+      "grad_norm": 0.44253653287887573,
+      "learning_rate": 2.0055698371893745e-05,
+      "loss": 0.2835,
+      "step": 6990
+    },
+    {
+      "epoch": 1.1996572407883461,
+      "grad_norm": 0.45636850595474243,
+      "learning_rate": 2.0012853470437018e-05,
+      "loss": 0.2883,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2013710368466153,
+      "grad_norm": 0.4535350501537323,
+      "learning_rate": 1.9970008568980293e-05,
+      "loss": 0.293,
+      "step": 7010
+    },
+    {
+      "epoch": 1.2030848329048842,
+      "grad_norm": 0.4127025008201599,
+      "learning_rate": 1.9927163667523565e-05,
+      "loss": 0.2996,
+      "step": 7020
+    },
+    {
+      "epoch": 1.2047986289631534,
+      "grad_norm": 0.4382990598678589,
+      "learning_rate": 1.988431876606684e-05,
+      "loss": 0.297,
+      "step": 7030
+    },
+    {
+      "epoch": 1.2065124250214225,
+      "grad_norm": 0.39549946784973145,
+      "learning_rate": 1.9841473864610113e-05,
+      "loss": 0.2903,
+      "step": 7040
+    },
+    {
+      "epoch": 1.2082262210796915,
+      "grad_norm": 0.424176424741745,
+      "learning_rate": 1.9798628963153386e-05,
+      "loss": 0.2872,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2099400171379606,
+      "grad_norm": 0.4338185489177704,
+      "learning_rate": 1.9755784061696658e-05,
+      "loss": 0.3156,
+      "step": 7060
+    },
+    {
+      "epoch": 1.2116538131962296,
+      "grad_norm": 0.4188908636569977,
+      "learning_rate": 1.9712939160239933e-05,
+      "loss": 0.2807,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2133676092544987,
+      "grad_norm": 0.3906058073043823,
+      "learning_rate": 1.9670094258783206e-05,
+      "loss": 0.2879,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2150814053127679,
+      "grad_norm": 0.44863423705101013,
+      "learning_rate": 1.962724935732648e-05,
+      "loss": 0.2936,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2167952013710368,
+      "grad_norm": 0.39006105065345764,
+      "learning_rate": 1.9584404455869754e-05,
+      "loss": 0.2939,
+      "step": 7100
+    },
+    {
+      "epoch": 1.218508997429306,
+      "grad_norm": 0.40731120109558105,
+      "learning_rate": 1.9541559554413026e-05,
+      "loss": 0.2869,
+      "step": 7110
+    },
+    {
+      "epoch": 1.220222793487575,
+      "grad_norm": 0.4063948392868042,
+      "learning_rate": 1.9498714652956298e-05,
+      "loss": 0.2904,
+      "step": 7120
+    },
+    {
+      "epoch": 1.221936589545844,
+      "grad_norm": 0.41968026757240295,
+      "learning_rate": 1.9455869751499574e-05,
+      "loss": 0.2858,
+      "step": 7130
+    },
+    {
+      "epoch": 1.2236503856041132,
+      "grad_norm": 0.4371456801891327,
+      "learning_rate": 1.9413024850042846e-05,
+      "loss": 0.2976,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2253641816623821,
+      "grad_norm": 0.37145504355430603,
+      "learning_rate": 1.937017994858612e-05,
+      "loss": 0.278,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2270779777206513,
+      "grad_norm": 0.4246087074279785,
+      "learning_rate": 1.932733504712939e-05,
+      "loss": 0.2925,
+      "step": 7160
+    },
+    {
+      "epoch": 1.2287917737789202,
+      "grad_norm": 0.40554991364479065,
+      "learning_rate": 1.9284490145672666e-05,
+      "loss": 0.2963,
+      "step": 7170
+    },
+    {
+      "epoch": 1.2305055698371894,
+      "grad_norm": 0.45490774512290955,
+      "learning_rate": 1.9241645244215938e-05,
+      "loss": 0.2783,
+      "step": 7180
+    },
+    {
+      "epoch": 1.2322193658954586,
+      "grad_norm": 0.41756734251976013,
+      "learning_rate": 1.9198800342759214e-05,
+      "loss": 0.3114,
+      "step": 7190
+    },
+    {
+      "epoch": 1.2339331619537275,
+      "grad_norm": 0.45411577820777893,
+      "learning_rate": 1.9155955441302486e-05,
+      "loss": 0.2886,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2356469580119966,
+      "grad_norm": 0.38101446628570557,
+      "learning_rate": 1.9113110539845762e-05,
+      "loss": 0.2793,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2373607540702656,
+      "grad_norm": 0.42433956265449524,
+      "learning_rate": 1.907026563838903e-05,
+      "loss": 0.2864,
+      "step": 7220
+    },
+    {
+      "epoch": 1.2390745501285347,
+      "grad_norm": 0.4093148112297058,
+      "learning_rate": 1.9027420736932306e-05,
+      "loss": 0.2933,
+      "step": 7230
+    },
+    {
+      "epoch": 1.240788346186804,
+      "grad_norm": 0.40895602107048035,
+      "learning_rate": 1.898457583547558e-05,
+      "loss": 0.3055,
+      "step": 7240
+    },
+    {
+      "epoch": 1.2425021422450728,
+      "grad_norm": 0.47577911615371704,
+      "learning_rate": 1.8941730934018854e-05,
+      "loss": 0.2976,
+      "step": 7250
+    },
+    {
+      "epoch": 1.244215938303342,
+      "grad_norm": 0.3901151716709137,
+      "learning_rate": 1.8898886032562126e-05,
+      "loss": 0.2813,
+      "step": 7260
+    },
+    {
+      "epoch": 1.245929734361611,
+      "grad_norm": 0.4057775139808655,
+      "learning_rate": 1.88560411311054e-05,
+      "loss": 0.3085,
+      "step": 7270
+    },
+    {
+      "epoch": 1.24764353041988,
+      "grad_norm": 0.44583985209465027,
+      "learning_rate": 1.881319622964867e-05,
+      "loss": 0.3108,
+      "step": 7280
+    },
+    {
+      "epoch": 1.249357326478149,
+      "grad_norm": 0.43333137035369873,
+      "learning_rate": 1.8770351328191946e-05,
+      "loss": 0.2919,
+      "step": 7290
+    },
+    {
+      "epoch": 1.2510711225364182,
+      "grad_norm": 0.41508951783180237,
+      "learning_rate": 1.872750642673522e-05,
+      "loss": 0.3053,
+      "step": 7300
+    },
+    {
+      "epoch": 1.252784918594687,
+      "grad_norm": 0.4186965525150299,
+      "learning_rate": 1.8684661525278494e-05,
+      "loss": 0.2794,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2544987146529563,
+      "grad_norm": 0.4359688460826874,
+      "learning_rate": 1.8641816623821767e-05,
+      "loss": 0.2871,
+      "step": 7320
+    },
+    {
+      "epoch": 1.2562125107112254,
+      "grad_norm": 0.4597584009170532,
+      "learning_rate": 1.859897172236504e-05,
+      "loss": 0.2955,
+      "step": 7330
+    },
+    {
+      "epoch": 1.2579263067694944,
+      "grad_norm": 0.38643357157707214,
+      "learning_rate": 1.855612682090831e-05,
+      "loss": 0.2838,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2596401028277635,
+      "grad_norm": 0.3859046697616577,
+      "learning_rate": 1.8513281919451587e-05,
+      "loss": 0.2977,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2613538988860324,
+      "grad_norm": 0.5184519290924072,
+      "learning_rate": 1.847043701799486e-05,
+      "loss": 0.298,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2630676949443016,
+      "grad_norm": 0.41403108835220337,
+      "learning_rate": 1.8427592116538135e-05,
+      "loss": 0.2914,
+      "step": 7370
+    },
+    {
+      "epoch": 1.2647814910025708,
+      "grad_norm": 0.4143988788127899,
+      "learning_rate": 1.8384747215081403e-05,
+      "loss": 0.2847,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2664952870608397,
+      "grad_norm": 0.3805682361125946,
+      "learning_rate": 1.834190231362468e-05,
+      "loss": 0.2786,
+      "step": 7390
+    },
+    {
+      "epoch": 1.2682090831191088,
+      "grad_norm": 0.38515955209732056,
+      "learning_rate": 1.829905741216795e-05,
+      "loss": 0.2954,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2699228791773778,
+      "grad_norm": 0.39131253957748413,
+      "learning_rate": 1.8256212510711227e-05,
+      "loss": 0.2982,
+      "step": 7410
+    },
+    {
+      "epoch": 1.271636675235647,
+      "grad_norm": 0.4461085796356201,
+      "learning_rate": 1.82133676092545e-05,
+      "loss": 0.29,
+      "step": 7420
+    },
+    {
+      "epoch": 1.273350471293916,
+      "grad_norm": 0.4100518822669983,
+      "learning_rate": 1.8170522707797775e-05,
+      "loss": 0.2955,
+      "step": 7430
+    },
+    {
+      "epoch": 1.275064267352185,
+      "grad_norm": 0.44043442606925964,
+      "learning_rate": 1.8127677806341044e-05,
+      "loss": 0.2882,
+      "step": 7440
+    },
+    {
+      "epoch": 1.2767780634104542,
+      "grad_norm": 0.44252437353134155,
+      "learning_rate": 1.808483290488432e-05,
+      "loss": 0.2888,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2784918594687231,
+      "grad_norm": 0.4522548019886017,
+      "learning_rate": 1.804198800342759e-05,
+      "loss": 0.287,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2802056555269923,
+      "grad_norm": 0.4005347490310669,
+      "learning_rate": 1.7999143101970867e-05,
+      "loss": 0.3034,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2819194515852614,
+      "grad_norm": 0.3880985975265503,
+      "learning_rate": 1.795629820051414e-05,
+      "loss": 0.2667,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2836332476435304,
+      "grad_norm": 0.39406266808509827,
+      "learning_rate": 1.791345329905741e-05,
+      "loss": 0.2724,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2853470437017995,
+      "grad_norm": 0.37708085775375366,
+      "learning_rate": 1.7870608397600684e-05,
+      "loss": 0.3016,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2870608397600685,
+      "grad_norm": 0.4000243842601776,
+      "learning_rate": 1.782776349614396e-05,
+      "loss": 0.2895,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2887746358183376,
+      "grad_norm": 0.4599775969982147,
+      "learning_rate": 1.7784918594687232e-05,
+      "loss": 0.2927,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2904884318766068,
+      "grad_norm": 0.5392141938209534,
+      "learning_rate": 1.7742073693230507e-05,
+      "loss": 0.2964,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2922022279348757,
+      "grad_norm": 0.4110645651817322,
+      "learning_rate": 1.769922879177378e-05,
+      "loss": 0.3021,
+      "step": 7540
+    },
+    {
+      "epoch": 1.2939160239931449,
+      "grad_norm": 0.42544734477996826,
+      "learning_rate": 1.7656383890317052e-05,
+      "loss": 0.2927,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2956298200514138,
+      "grad_norm": 0.3835833966732025,
+      "learning_rate": 1.7613538988860324e-05,
+      "loss": 0.2801,
+      "step": 7560
+    },
+    {
+      "epoch": 1.297343616109683,
+      "grad_norm": 0.394071102142334,
+      "learning_rate": 1.75706940874036e-05,
+      "loss": 0.284,
+      "step": 7570
+    },
+    {
+      "epoch": 1.2990574121679521,
+      "grad_norm": 0.3885582387447357,
+      "learning_rate": 1.7527849185946872e-05,
+      "loss": 0.2992,
+      "step": 7580
+    },
+    {
+      "epoch": 1.300771208226221,
+      "grad_norm": 0.4305812120437622,
+      "learning_rate": 1.7485004284490148e-05,
+      "loss": 0.2939,
+      "step": 7590
+    },
+    {
+      "epoch": 1.3024850042844902,
+      "grad_norm": 0.40897855162620544,
+      "learning_rate": 1.744215938303342e-05,
+      "loss": 0.2985,
+      "step": 7600
+    },
+    {
+      "epoch": 1.3041988003427591,
+      "grad_norm": 0.4165668189525604,
+      "learning_rate": 1.7399314481576692e-05,
+      "loss": 0.2995,
+      "step": 7610
+    },
+    {
+      "epoch": 1.3059125964010283,
+      "grad_norm": 0.3743734657764435,
+      "learning_rate": 1.7356469580119968e-05,
+      "loss": 0.2783,
+      "step": 7620
+    },
+    {
+      "epoch": 1.3076263924592975,
+      "grad_norm": 0.44639477133750916,
+      "learning_rate": 1.731362467866324e-05,
+      "loss": 0.2971,
+      "step": 7630
+    },
+    {
+      "epoch": 1.3093401885175664,
+      "grad_norm": 0.4615757167339325,
+      "learning_rate": 1.7270779777206512e-05,
+      "loss": 0.2971,
+      "step": 7640
+    },
+    {
+      "epoch": 1.3110539845758356,
+      "grad_norm": 0.38079190254211426,
+      "learning_rate": 1.7227934875749788e-05,
+      "loss": 0.2895,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3127677806341045,
+      "grad_norm": 0.41293004155158997,
+      "learning_rate": 1.718508997429306e-05,
+      "loss": 0.2876,
+      "step": 7660
+    },
+    {
+      "epoch": 1.3144815766923736,
+      "grad_norm": 0.45726191997528076,
+      "learning_rate": 1.7142245072836332e-05,
+      "loss": 0.2905,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3161953727506428,
+      "grad_norm": 0.39152881503105164,
+      "learning_rate": 1.7099400171379608e-05,
+      "loss": 0.2862,
+      "step": 7680
+    },
+    {
+      "epoch": 1.3179091688089117,
+      "grad_norm": 0.45516616106033325,
+      "learning_rate": 1.705655526992288e-05,
+      "loss": 0.2869,
+      "step": 7690
+    },
+    {
+      "epoch": 1.3196229648671807,
+      "grad_norm": 0.3824329674243927,
+      "learning_rate": 1.7013710368466156e-05,
+      "loss": 0.2934,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3213367609254498,
+      "grad_norm": 0.41734811663627625,
+      "learning_rate": 1.6970865467009428e-05,
+      "loss": 0.2932,
+      "step": 7710
+    },
+    {
+      "epoch": 1.323050556983719,
+      "grad_norm": 0.3699459731578827,
+      "learning_rate": 1.69280205655527e-05,
+      "loss": 0.2939,
+      "step": 7720
+    },
+    {
+      "epoch": 1.3247643530419881,
+      "grad_norm": 0.44719019532203674,
+      "learning_rate": 1.6885175664095973e-05,
+      "loss": 0.3068,
+      "step": 7730
+    },
+    {
+      "epoch": 1.326478149100257,
+      "grad_norm": 0.39685600996017456,
+      "learning_rate": 1.6842330762639248e-05,
+      "loss": 0.2964,
+      "step": 7740
+    },
+    {
+      "epoch": 1.328191945158526,
+      "grad_norm": 0.4524918794631958,
+      "learning_rate": 1.679948586118252e-05,
+      "loss": 0.2975,
+      "step": 7750
+    },
+    {
+      "epoch": 1.3299057412167952,
+      "grad_norm": 0.45134490728378296,
+      "learning_rate": 1.6756640959725796e-05,
+      "loss": 0.2817,
+      "step": 7760
+    },
+    {
+      "epoch": 1.3316195372750643,
+      "grad_norm": 0.43576714396476746,
+      "learning_rate": 1.6713796058269065e-05,
+      "loss": 0.2906,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.3999752998352051,
+      "learning_rate": 1.667095115681234e-05,
+      "loss": 0.2892,
+      "step": 7780
+    },
+    {
+      "epoch": 1.3350471293916024,
+      "grad_norm": 0.4068009555339813,
+      "learning_rate": 1.6628106255355613e-05,
+      "loss": 0.3021,
+      "step": 7790
+    },
+    {
+      "epoch": 1.3367609254498714,
+      "grad_norm": 0.40887516736984253,
+      "learning_rate": 1.658526135389889e-05,
+      "loss": 0.2846,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3384747215081405,
+      "grad_norm": 0.413370817899704,
+      "learning_rate": 1.654241645244216e-05,
+      "loss": 0.2955,
+      "step": 7810
+    },
+    {
+      "epoch": 1.3401885175664097,
+      "grad_norm": 0.3994751572608948,
+      "learning_rate": 1.6499571550985436e-05,
+      "loss": 0.2958,
+      "step": 7820
+    },
+    {
+      "epoch": 1.3419023136246786,
+      "grad_norm": 0.44501566886901855,
+      "learning_rate": 1.6456726649528705e-05,
+      "loss": 0.3074,
+      "step": 7830
+    },
+    {
+      "epoch": 1.3436161096829478,
+      "grad_norm": 0.4057297110557556,
+      "learning_rate": 1.641388174807198e-05,
+      "loss": 0.2939,
+      "step": 7840
+    },
+    {
+      "epoch": 1.3453299057412167,
+      "grad_norm": 0.3912492096424103,
+      "learning_rate": 1.6371036846615253e-05,
+      "loss": 0.2979,
+      "step": 7850
+    },
+    {
+      "epoch": 1.3470437017994858,
+      "grad_norm": 0.4454675614833832,
+      "learning_rate": 1.632819194515853e-05,
+      "loss": 0.2921,
+      "step": 7860
+    },
+    {
+      "epoch": 1.348757497857755,
+      "grad_norm": 0.41817528009414673,
+      "learning_rate": 1.62853470437018e-05,
+      "loss": 0.2945,
+      "step": 7870
+    },
+    {
+      "epoch": 1.350471293916024,
+      "grad_norm": 0.4252879321575165,
+      "learning_rate": 1.6242502142245073e-05,
+      "loss": 0.3062,
+      "step": 7880
+    },
+    {
+      "epoch": 1.352185089974293,
+      "grad_norm": 0.3885163366794586,
+      "learning_rate": 1.6199657240788345e-05,
+      "loss": 0.2788,
+      "step": 7890
+    },
+    {
+      "epoch": 1.353898886032562,
+      "grad_norm": 0.45208239555358887,
+      "learning_rate": 1.615681233933162e-05,
+      "loss": 0.2935,
+      "step": 7900
+    },
+    {
+      "epoch": 1.3556126820908312,
+      "grad_norm": 0.4251621663570404,
+      "learning_rate": 1.6113967437874893e-05,
+      "loss": 0.2818,
+      "step": 7910
+    },
+    {
+      "epoch": 1.3573264781491003,
+      "grad_norm": 0.41676682233810425,
+      "learning_rate": 1.607112253641817e-05,
+      "loss": 0.2904,
+      "step": 7920
+    },
+    {
+      "epoch": 1.3590402742073693,
+      "grad_norm": 0.40959298610687256,
+      "learning_rate": 1.602827763496144e-05,
+      "loss": 0.2952,
+      "step": 7930
+    },
+    {
+      "epoch": 1.3607540702656384,
+      "grad_norm": 0.39400455355644226,
+      "learning_rate": 1.5985432733504713e-05,
+      "loss": 0.295,
+      "step": 7940
+    },
+    {
+      "epoch": 1.3624678663239074,
+      "grad_norm": 0.3885449469089508,
+      "learning_rate": 1.5942587832047986e-05,
+      "loss": 0.2763,
+      "step": 7950
+    },
+    {
+      "epoch": 1.3641816623821765,
+      "grad_norm": 0.41149842739105225,
+      "learning_rate": 1.589974293059126e-05,
+      "loss": 0.2985,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3658954584404457,
+      "grad_norm": 0.40511512756347656,
+      "learning_rate": 1.5856898029134534e-05,
+      "loss": 0.2836,
+      "step": 7970
+    },
+    {
+      "epoch": 1.3676092544987146,
+      "grad_norm": 0.4108186960220337,
+      "learning_rate": 1.581405312767781e-05,
+      "loss": 0.2871,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3693230505569838,
+      "grad_norm": 0.4291677474975586,
+      "learning_rate": 1.5771208226221078e-05,
+      "loss": 0.2933,
+      "step": 7990
+    },
+    {
+      "epoch": 1.3710368466152527,
+      "grad_norm": 0.3896220028400421,
+      "learning_rate": 1.5728363324764354e-05,
+      "loss": 0.3051,
+      "step": 8000
+    },
+    {
+      "epoch": 1.3727506426735219,
+      "grad_norm": 0.41442325711250305,
+      "learning_rate": 1.5685518423307626e-05,
+      "loss": 0.2942,
+      "step": 8010
+    },
+    {
+      "epoch": 1.374464438731791,
+      "grad_norm": 0.3902716040611267,
+      "learning_rate": 1.56426735218509e-05,
+      "loss": 0.2933,
+      "step": 8020
+    },
+    {
+      "epoch": 1.37617823479006,
+      "grad_norm": 0.4575580656528473,
+      "learning_rate": 1.5599828620394174e-05,
+      "loss": 0.2908,
+      "step": 8030
+    },
+    {
+      "epoch": 1.3778920308483291,
+      "grad_norm": 0.41972899436950684,
+      "learning_rate": 1.555698371893745e-05,
+      "loss": 0.2901,
+      "step": 8040
+    },
+    {
+      "epoch": 1.379605826906598,
+      "grad_norm": 0.49797460436820984,
+      "learning_rate": 1.5514138817480718e-05,
+      "loss": 0.2935,
+      "step": 8050
+    },
+    {
+      "epoch": 1.3813196229648672,
+      "grad_norm": 0.42879509925842285,
+      "learning_rate": 1.5471293916023994e-05,
+      "loss": 0.2981,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3830334190231364,
+      "grad_norm": 0.3812084496021271,
+      "learning_rate": 1.5428449014567266e-05,
+      "loss": 0.277,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3847472150814053,
+      "grad_norm": 0.41363996267318726,
+      "learning_rate": 1.5385604113110542e-05,
+      "loss": 0.2922,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3864610111396745,
+      "grad_norm": 0.4093998372554779,
+      "learning_rate": 1.5342759211653814e-05,
+      "loss": 0.2832,
+      "step": 8090
+    },
+    {
+      "epoch": 1.3881748071979434,
+      "grad_norm": 0.39057186245918274,
+      "learning_rate": 1.5299914310197086e-05,
+      "loss": 0.304,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3898886032562126,
+      "grad_norm": 0.4318443834781647,
+      "learning_rate": 1.525706940874036e-05,
+      "loss": 0.283,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3916023993144817,
+      "grad_norm": 0.3923870325088501,
+      "learning_rate": 1.5214224507283634e-05,
+      "loss": 0.2977,
+      "step": 8120
+    },
+    {
+      "epoch": 1.3933161953727506,
+      "grad_norm": 0.47978222370147705,
+      "learning_rate": 1.5171379605826908e-05,
+      "loss": 0.2842,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3950299914310196,
+      "grad_norm": 0.4198880195617676,
+      "learning_rate": 1.5128534704370182e-05,
+      "loss": 0.297,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3967437874892887,
+      "grad_norm": 0.41597798466682434,
+      "learning_rate": 1.5085689802913456e-05,
+      "loss": 0.2884,
+      "step": 8150
+    },
+    {
+      "epoch": 1.398457583547558,
+      "grad_norm": 0.43078944087028503,
+      "learning_rate": 1.5042844901456726e-05,
+      "loss": 0.2979,
+      "step": 8160
+    },
+    {
+      "epoch": 1.4001713796058268,
+      "grad_norm": 0.40881800651550293,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2897,
+      "step": 8170
+    },
+    {
+      "epoch": 1.401885175664096,
+      "grad_norm": 0.41258737444877625,
+      "learning_rate": 1.4957155098543274e-05,
+      "loss": 0.2893,
+      "step": 8180
+    },
+    {
+      "epoch": 1.403598971722365,
+      "grad_norm": 0.40783384442329407,
+      "learning_rate": 1.4914310197086548e-05,
+      "loss": 0.3117,
+      "step": 8190
+    },
+    {
+      "epoch": 1.405312767780634,
+      "grad_norm": 0.41735509037971497,
+      "learning_rate": 1.4871465295629822e-05,
+      "loss": 0.2975,
+      "step": 8200
+    },
+    {
+      "epoch": 1.4070265638389032,
+      "grad_norm": 0.3986874520778656,
+      "learning_rate": 1.4828620394173093e-05,
+      "loss": 0.2926,
+      "step": 8210
+    },
+    {
+      "epoch": 1.4087403598971722,
+      "grad_norm": 0.43188929557800293,
+      "learning_rate": 1.4785775492716367e-05,
+      "loss": 0.3003,
+      "step": 8220
+    },
+    {
+      "epoch": 1.4104541559554413,
+      "grad_norm": 0.4129735827445984,
+      "learning_rate": 1.474293059125964e-05,
+      "loss": 0.3015,
+      "step": 8230
+    },
+    {
+      "epoch": 1.4121679520137103,
+      "grad_norm": 0.4374343156814575,
+      "learning_rate": 1.4700085689802915e-05,
+      "loss": 0.2892,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4138817480719794,
+      "grad_norm": 0.4039885699748993,
+      "learning_rate": 1.4657240788346188e-05,
+      "loss": 0.2874,
+      "step": 8250
+    },
+    {
+      "epoch": 1.4155955441302486,
+      "grad_norm": 0.44015762209892273,
+      "learning_rate": 1.4614395886889462e-05,
+      "loss": 0.2904,
+      "step": 8260
+    },
+    {
+      "epoch": 1.4173093401885175,
+      "grad_norm": 0.4429149031639099,
+      "learning_rate": 1.4571550985432733e-05,
+      "loss": 0.2874,
+      "step": 8270
+    },
+    {
+      "epoch": 1.4190231362467867,
+      "grad_norm": 0.3884618878364563,
+      "learning_rate": 1.4528706083976007e-05,
+      "loss": 0.2996,
+      "step": 8280
+    },
+    {
+      "epoch": 1.4207369323050556,
+      "grad_norm": 0.42390188574790955,
+      "learning_rate": 1.448586118251928e-05,
+      "loss": 0.2884,
+      "step": 8290
+    },
+    {
+      "epoch": 1.4224507283633248,
+      "grad_norm": 0.43513190746307373,
+      "learning_rate": 1.4443016281062555e-05,
+      "loss": 0.2864,
+      "step": 8300
+    },
+    {
+      "epoch": 1.424164524421594,
+      "grad_norm": 0.4340656101703644,
+      "learning_rate": 1.4400171379605829e-05,
+      "loss": 0.2849,
+      "step": 8310
+    },
+    {
+      "epoch": 1.4258783204798628,
+      "grad_norm": 0.4230162799358368,
+      "learning_rate": 1.43573264781491e-05,
+      "loss": 0.2935,
+      "step": 8320
+    },
+    {
+      "epoch": 1.427592116538132,
+      "grad_norm": 0.4005115330219269,
+      "learning_rate": 1.4314481576692373e-05,
+      "loss": 0.2846,
+      "step": 8330
+    },
+    {
+      "epoch": 1.429305912596401,
+      "grad_norm": 0.45279762148857117,
+      "learning_rate": 1.4271636675235647e-05,
+      "loss": 0.3048,
+      "step": 8340
+    },
+    {
+      "epoch": 1.43101970865467,
+      "grad_norm": 0.3904445469379425,
+      "learning_rate": 1.4228791773778921e-05,
+      "loss": 0.2904,
+      "step": 8350
+    },
+    {
+      "epoch": 1.4327335047129393,
+      "grad_norm": 0.43565717339515686,
+      "learning_rate": 1.4185946872322195e-05,
+      "loss": 0.2925,
+      "step": 8360
+    },
+    {
+      "epoch": 1.4344473007712082,
+      "grad_norm": 0.45331141352653503,
+      "learning_rate": 1.4143101970865469e-05,
+      "loss": 0.2993,
+      "step": 8370
+    },
+    {
+      "epoch": 1.4361610968294773,
+      "grad_norm": 0.4167874753475189,
+      "learning_rate": 1.410025706940874e-05,
+      "loss": 0.2794,
+      "step": 8380
+    },
+    {
+      "epoch": 1.4378748928877463,
+      "grad_norm": 0.44904330372810364,
+      "learning_rate": 1.4057412167952013e-05,
+      "loss": 0.2831,
+      "step": 8390
+    },
+    {
+      "epoch": 1.4395886889460154,
+      "grad_norm": 0.5003662705421448,
+      "learning_rate": 1.4014567266495287e-05,
+      "loss": 0.3023,
+      "step": 8400
+    },
+    {
+      "epoch": 1.4413024850042846,
+      "grad_norm": 0.4118245244026184,
+      "learning_rate": 1.3971722365038561e-05,
+      "loss": 0.2908,
+      "step": 8410
+    },
+    {
+      "epoch": 1.4430162810625535,
+      "grad_norm": 0.41458413004875183,
+      "learning_rate": 1.3928877463581835e-05,
+      "loss": 0.2942,
+      "step": 8420
+    },
+    {
+      "epoch": 1.4447300771208227,
+      "grad_norm": 0.40198981761932373,
+      "learning_rate": 1.3886032562125106e-05,
+      "loss": 0.2946,
+      "step": 8430
+    },
+    {
+      "epoch": 1.4464438731790916,
+      "grad_norm": 0.40011557936668396,
+      "learning_rate": 1.384318766066838e-05,
+      "loss": 0.287,
+      "step": 8440
+    },
+    {
+      "epoch": 1.4481576692373608,
+      "grad_norm": 0.4027191698551178,
+      "learning_rate": 1.3800342759211654e-05,
+      "loss": 0.2829,
+      "step": 8450
+    },
+    {
+      "epoch": 1.44987146529563,
+      "grad_norm": 0.3991924822330475,
+      "learning_rate": 1.3757497857754928e-05,
+      "loss": 0.2812,
+      "step": 8460
+    },
+    {
+      "epoch": 1.4515852613538989,
+      "grad_norm": 0.4096791446208954,
+      "learning_rate": 1.3714652956298202e-05,
+      "loss": 0.2738,
+      "step": 8470
+    },
+    {
+      "epoch": 1.453299057412168,
+      "grad_norm": 0.3994709849357605,
+      "learning_rate": 1.3671808054841475e-05,
+      "loss": 0.2883,
+      "step": 8480
+    },
+    {
+      "epoch": 1.455012853470437,
+      "grad_norm": 0.40569868683815,
+      "learning_rate": 1.3628963153384746e-05,
+      "loss": 0.2874,
+      "step": 8490
+    },
+    {
+      "epoch": 1.4567266495287061,
+      "grad_norm": 0.4111444354057312,
+      "learning_rate": 1.358611825192802e-05,
+      "loss": 0.2789,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4584404455869753,
+      "grad_norm": 0.454795777797699,
+      "learning_rate": 1.3543273350471294e-05,
+      "loss": 0.2905,
+      "step": 8510
+    },
+    {
+      "epoch": 1.4601542416452442,
+      "grad_norm": 0.4353298544883728,
+      "learning_rate": 1.3500428449014568e-05,
+      "loss": 0.3012,
+      "step": 8520
+    },
+    {
+      "epoch": 1.4618680377035131,
+      "grad_norm": 0.399127334356308,
+      "learning_rate": 1.3457583547557842e-05,
+      "loss": 0.2985,
+      "step": 8530
+    },
+    {
+      "epoch": 1.4635818337617823,
+      "grad_norm": 0.37544289231300354,
+      "learning_rate": 1.3414738646101116e-05,
+      "loss": 0.301,
+      "step": 8540
+    },
+    {
+      "epoch": 1.4652956298200515,
+      "grad_norm": 0.37045615911483765,
+      "learning_rate": 1.3371893744644386e-05,
+      "loss": 0.2729,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4670094258783206,
+      "grad_norm": 0.43156683444976807,
+      "learning_rate": 1.332904884318766e-05,
+      "loss": 0.2963,
+      "step": 8560
+    },
+    {
+      "epoch": 1.4687232219365896,
+      "grad_norm": 0.403238981962204,
+      "learning_rate": 1.3286203941730934e-05,
+      "loss": 0.2936,
+      "step": 8570
+    },
+    {
+      "epoch": 1.4704370179948585,
+      "grad_norm": 0.4391413927078247,
+      "learning_rate": 1.3243359040274208e-05,
+      "loss": 0.2823,
+      "step": 8580
+    },
+    {
+      "epoch": 1.4721508140531276,
+      "grad_norm": 0.4374690055847168,
+      "learning_rate": 1.3200514138817482e-05,
+      "loss": 0.2966,
+      "step": 8590
+    },
+    {
+      "epoch": 1.4738646101113968,
+      "grad_norm": 0.40266698598861694,
+      "learning_rate": 1.3157669237360754e-05,
+      "loss": 0.2982,
+      "step": 8600
+    },
+    {
+      "epoch": 1.4755784061696657,
+      "grad_norm": 0.41828453540802,
+      "learning_rate": 1.3114824335904028e-05,
+      "loss": 0.3003,
+      "step": 8610
+    },
+    {
+      "epoch": 1.477292202227935,
+      "grad_norm": 0.4586716294288635,
+      "learning_rate": 1.30719794344473e-05,
+      "loss": 0.2913,
+      "step": 8620
+    },
+    {
+      "epoch": 1.4790059982862038,
+      "grad_norm": 0.3982478678226471,
+      "learning_rate": 1.3029134532990574e-05,
+      "loss": 0.2883,
+      "step": 8630
+    },
+    {
+      "epoch": 1.480719794344473,
+      "grad_norm": 0.3927355706691742,
+      "learning_rate": 1.2986289631533848e-05,
+      "loss": 0.2904,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4824335904027421,
+      "grad_norm": 0.43751057982444763,
+      "learning_rate": 1.2943444730077122e-05,
+      "loss": 0.2859,
+      "step": 8650
+    },
+    {
+      "epoch": 1.484147386461011,
+      "grad_norm": 0.4266740381717682,
+      "learning_rate": 1.2900599828620394e-05,
+      "loss": 0.2938,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4858611825192802,
+      "grad_norm": 0.4187754988670349,
+      "learning_rate": 1.2857754927163668e-05,
+      "loss": 0.2912,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4875749785775492,
+      "grad_norm": 0.40541937947273254,
+      "learning_rate": 1.2814910025706942e-05,
+      "loss": 0.3029,
+      "step": 8680
+    },
+    {
+      "epoch": 1.4892887746358183,
+      "grad_norm": 0.4201086759567261,
+      "learning_rate": 1.2772065124250215e-05,
+      "loss": 0.2961,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4910025706940875,
+      "grad_norm": 0.40029746294021606,
+      "learning_rate": 1.2729220222793489e-05,
+      "loss": 0.2832,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4927163667523564,
+      "grad_norm": 0.40658628940582275,
+      "learning_rate": 1.268637532133676e-05,
+      "loss": 0.2894,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4944301628106256,
+      "grad_norm": 0.36714300513267517,
+      "learning_rate": 1.2643530419880035e-05,
+      "loss": 0.2918,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4961439588688945,
+      "grad_norm": 0.4191112518310547,
+      "learning_rate": 1.2600685518423309e-05,
+      "loss": 0.2903,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4978577549271637,
+      "grad_norm": 0.41759270429611206,
+      "learning_rate": 1.2557840616966583e-05,
+      "loss": 0.293,
+      "step": 8740
+    },
+    {
+      "epoch": 1.4995715509854328,
+      "grad_norm": 0.41426923871040344,
+      "learning_rate": 1.2514995715509857e-05,
+      "loss": 0.281,
+      "step": 8750
+    },
+    {
+      "epoch": 1.5012853470437018,
+      "grad_norm": 0.4298325479030609,
+      "learning_rate": 1.2472150814053129e-05,
+      "loss": 0.293,
+      "step": 8760
+    },
+    {
+      "epoch": 1.502999143101971,
+      "grad_norm": 0.4791582524776459,
+      "learning_rate": 1.2429305912596403e-05,
+      "loss": 0.2857,
+      "step": 8770
+    },
+    {
+      "epoch": 1.5047129391602398,
+      "grad_norm": 0.4435204267501831,
+      "learning_rate": 1.2386461011139675e-05,
+      "loss": 0.293,
+      "step": 8780
+    },
+    {
+      "epoch": 1.506426735218509,
+      "grad_norm": 0.3941735029220581,
+      "learning_rate": 1.2343616109682949e-05,
+      "loss": 0.3003,
+      "step": 8790
+    },
+    {
+      "epoch": 1.5081405312767782,
+      "grad_norm": 0.44211193919181824,
+      "learning_rate": 1.2300771208226223e-05,
+      "loss": 0.2922,
+      "step": 8800
+    },
+    {
+      "epoch": 1.509854327335047,
+      "grad_norm": 0.3936205208301544,
+      "learning_rate": 1.2257926306769495e-05,
+      "loss": 0.2896,
+      "step": 8810
+    },
+    {
+      "epoch": 1.511568123393316,
+      "grad_norm": 0.43783196806907654,
+      "learning_rate": 1.2215081405312769e-05,
+      "loss": 0.2873,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5132819194515852,
+      "grad_norm": 0.4433899223804474,
+      "learning_rate": 1.2172236503856043e-05,
+      "loss": 0.2881,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5149957155098543,
+      "grad_norm": 0.4456241726875305,
+      "learning_rate": 1.2129391602399315e-05,
+      "loss": 0.3056,
+      "step": 8840
+    },
+    {
+      "epoch": 1.5167095115681235,
+      "grad_norm": 0.48001471161842346,
+      "learning_rate": 1.2086546700942589e-05,
+      "loss": 0.298,
+      "step": 8850
+    },
+    {
+      "epoch": 1.5184233076263924,
+      "grad_norm": 0.39325088262557983,
+      "learning_rate": 1.2043701799485861e-05,
+      "loss": 0.2981,
+      "step": 8860
+    },
+    {
+      "epoch": 1.5201371036846614,
+      "grad_norm": 0.4006384611129761,
+      "learning_rate": 1.2000856898029135e-05,
+      "loss": 0.2887,
+      "step": 8870
+    },
+    {
+      "epoch": 1.5218508997429305,
+      "grad_norm": 0.4165492653846741,
+      "learning_rate": 1.195801199657241e-05,
+      "loss": 0.2868,
+      "step": 8880
+    },
+    {
+      "epoch": 1.5235646958011997,
+      "grad_norm": 0.4088376760482788,
+      "learning_rate": 1.1915167095115681e-05,
+      "loss": 0.3074,
+      "step": 8890
+    },
+    {
+      "epoch": 1.5252784918594688,
+      "grad_norm": 0.4085462987422943,
+      "learning_rate": 1.1872322193658955e-05,
+      "loss": 0.2944,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5269922879177378,
+      "grad_norm": 0.40741875767707825,
+      "learning_rate": 1.182947729220223e-05,
+      "loss": 0.2794,
+      "step": 8910
+    },
+    {
+      "epoch": 1.5287060839760067,
+      "grad_norm": 0.41725581884384155,
+      "learning_rate": 1.1786632390745502e-05,
+      "loss": 0.2965,
+      "step": 8920
+    },
+    {
+      "epoch": 1.5304198800342759,
+      "grad_norm": 0.4100841283798218,
+      "learning_rate": 1.1743787489288775e-05,
+      "loss": 0.292,
+      "step": 8930
+    },
+    {
+      "epoch": 1.532133676092545,
+      "grad_norm": 0.39785757660865784,
+      "learning_rate": 1.170094258783205e-05,
+      "loss": 0.2791,
+      "step": 8940
+    },
+    {
+      "epoch": 1.5338474721508142,
+      "grad_norm": 0.4190772771835327,
+      "learning_rate": 1.1658097686375322e-05,
+      "loss": 0.2909,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5355612682090831,
+      "grad_norm": 0.4241431951522827,
+      "learning_rate": 1.1615252784918596e-05,
+      "loss": 0.3034,
+      "step": 8960
+    },
+    {
+      "epoch": 1.537275064267352,
+      "grad_norm": 0.4615418016910553,
+      "learning_rate": 1.1572407883461868e-05,
+      "loss": 0.2872,
+      "step": 8970
+    },
+    {
+      "epoch": 1.5389888603256212,
+      "grad_norm": 0.41917067766189575,
+      "learning_rate": 1.1529562982005142e-05,
+      "loss": 0.3044,
+      "step": 8980
+    },
+    {
+      "epoch": 1.5407026563838904,
+      "grad_norm": 0.4099867045879364,
+      "learning_rate": 1.1486718080548416e-05,
+      "loss": 0.2969,
+      "step": 8990
+    },
+    {
+      "epoch": 1.5424164524421595,
+      "grad_norm": 0.44426125288009644,
+      "learning_rate": 1.1443873179091688e-05,
+      "loss": 0.2853,
+      "step": 9000
+    },
+    {
+      "epoch": 1.5441302485004285,
+      "grad_norm": 0.4058697521686554,
+      "learning_rate": 1.1401028277634962e-05,
+      "loss": 0.2842,
+      "step": 9010
+    },
+    {
+      "epoch": 1.5458440445586974,
+      "grad_norm": 0.44107526540756226,
+      "learning_rate": 1.1358183376178236e-05,
+      "loss": 0.2787,
+      "step": 9020
+    },
+    {
+      "epoch": 1.5475578406169666,
+      "grad_norm": 0.41578471660614014,
+      "learning_rate": 1.1315338474721508e-05,
+      "loss": 0.2724,
+      "step": 9030
+    },
+    {
+      "epoch": 1.5492716366752357,
+      "grad_norm": 0.42603832483291626,
+      "learning_rate": 1.1272493573264782e-05,
+      "loss": 0.2806,
+      "step": 9040
+    },
+    {
+      "epoch": 1.5509854327335049,
+      "grad_norm": 0.4300294816493988,
+      "learning_rate": 1.1229648671808056e-05,
+      "loss": 0.2912,
+      "step": 9050
+    },
+    {
+      "epoch": 1.5526992287917738,
+      "grad_norm": 0.44864004850387573,
+      "learning_rate": 1.1186803770351328e-05,
+      "loss": 0.3061,
+      "step": 9060
+    },
+    {
+      "epoch": 1.5544130248500427,
+      "grad_norm": 0.42053478956222534,
+      "learning_rate": 1.1143958868894602e-05,
+      "loss": 0.287,
+      "step": 9070
+    },
+    {
+      "epoch": 1.556126820908312,
+      "grad_norm": 0.407592236995697,
+      "learning_rate": 1.1101113967437874e-05,
+      "loss": 0.2873,
+      "step": 9080
+    },
+    {
+      "epoch": 1.557840616966581,
+      "grad_norm": 0.4154520332813263,
+      "learning_rate": 1.1058269065981148e-05,
+      "loss": 0.2972,
+      "step": 9090
+    },
+    {
+      "epoch": 1.5595544130248502,
+      "grad_norm": 0.474953830242157,
+      "learning_rate": 1.1015424164524422e-05,
+      "loss": 0.2971,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5612682090831191,
+      "grad_norm": 0.4030286967754364,
+      "learning_rate": 1.0972579263067694e-05,
+      "loss": 0.2851,
+      "step": 9110
+    },
+    {
+      "epoch": 1.562982005141388,
+      "grad_norm": 0.4235098958015442,
+      "learning_rate": 1.0929734361610968e-05,
+      "loss": 0.2863,
+      "step": 9120
+    },
+    {
+      "epoch": 1.5646958011996572,
+      "grad_norm": 0.4046483337879181,
+      "learning_rate": 1.0886889460154242e-05,
+      "loss": 0.2872,
+      "step": 9130
+    },
+    {
+      "epoch": 1.5664095972579264,
+      "grad_norm": 0.42266711592674255,
+      "learning_rate": 1.0844044558697515e-05,
+      "loss": 0.2962,
+      "step": 9140
+    },
+    {
+      "epoch": 1.5681233933161953,
+      "grad_norm": 0.3989273011684418,
+      "learning_rate": 1.0801199657240789e-05,
+      "loss": 0.2975,
+      "step": 9150
+    },
+    {
+      "epoch": 1.5698371893744645,
+      "grad_norm": 0.3651719391345978,
+      "learning_rate": 1.0758354755784062e-05,
+      "loss": 0.2774,
+      "step": 9160
+    },
+    {
+      "epoch": 1.5715509854327334,
+      "grad_norm": 0.429797500371933,
+      "learning_rate": 1.0715509854327335e-05,
+      "loss": 0.2906,
+      "step": 9170
+    },
+    {
+      "epoch": 1.5732647814910026,
+      "grad_norm": 0.41648441553115845,
+      "learning_rate": 1.0672664952870609e-05,
+      "loss": 0.2879,
+      "step": 9180
+    },
+    {
+      "epoch": 1.5749785775492717,
+      "grad_norm": 0.41190940141677856,
+      "learning_rate": 1.0629820051413883e-05,
+      "loss": 0.2781,
+      "step": 9190
+    },
+    {
+      "epoch": 1.5766923736075407,
+      "grad_norm": 0.4211117625236511,
+      "learning_rate": 1.0586975149957155e-05,
+      "loss": 0.286,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5784061696658098,
+      "grad_norm": 0.4373011291027069,
+      "learning_rate": 1.0544130248500429e-05,
+      "loss": 0.2806,
+      "step": 9210
+    },
+    {
+      "epoch": 1.5801199657240788,
+      "grad_norm": 0.45145246386528015,
+      "learning_rate": 1.0501285347043701e-05,
+      "loss": 0.2868,
+      "step": 9220
+    },
+    {
+      "epoch": 1.581833761782348,
+      "grad_norm": 0.43550366163253784,
+      "learning_rate": 1.0458440445586975e-05,
+      "loss": 0.2798,
+      "step": 9230
+    },
+    {
+      "epoch": 1.583547557840617,
+      "grad_norm": 0.4000817835330963,
+      "learning_rate": 1.0415595544130249e-05,
+      "loss": 0.2816,
+      "step": 9240
+    },
+    {
+      "epoch": 1.585261353898886,
+      "grad_norm": 0.39046937227249146,
+      "learning_rate": 1.0372750642673521e-05,
+      "loss": 0.2869,
+      "step": 9250
+    },
+    {
+      "epoch": 1.586975149957155,
+      "grad_norm": 0.42524218559265137,
+      "learning_rate": 1.0329905741216795e-05,
+      "loss": 0.2806,
+      "step": 9260
+    },
+    {
+      "epoch": 1.588688946015424,
+      "grad_norm": 0.42042532563209534,
+      "learning_rate": 1.0287060839760069e-05,
+      "loss": 0.2779,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5904027420736933,
+      "grad_norm": 0.42058244347572327,
+      "learning_rate": 1.0244215938303341e-05,
+      "loss": 0.2964,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5921165381319624,
+      "grad_norm": 0.40325188636779785,
+      "learning_rate": 1.0201371036846615e-05,
+      "loss": 0.2958,
+      "step": 9290
+    },
+    {
+      "epoch": 1.5938303341902313,
+      "grad_norm": 0.38157641887664795,
+      "learning_rate": 1.0158526135389889e-05,
+      "loss": 0.3041,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5955441302485003,
+      "grad_norm": 0.4275236129760742,
+      "learning_rate": 1.0115681233933161e-05,
+      "loss": 0.3231,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5972579263067694,
+      "grad_norm": 0.4074757397174835,
+      "learning_rate": 1.0072836332476435e-05,
+      "loss": 0.2736,
+      "step": 9320
+    },
+    {
+      "epoch": 1.5989717223650386,
+      "grad_norm": 0.43237921595573425,
+      "learning_rate": 1.002999143101971e-05,
+      "loss": 0.2781,
+      "step": 9330
+    },
+    {
+      "epoch": 1.6006855184233078,
+      "grad_norm": 0.4433465301990509,
+      "learning_rate": 9.987146529562983e-06,
+      "loss": 0.2914,
+      "step": 9340
+    },
+    {
+      "epoch": 1.6023993144815767,
+      "grad_norm": 0.5035930275917053,
+      "learning_rate": 9.944301628106255e-06,
+      "loss": 0.2931,
+      "step": 9350
+    },
+    {
+      "epoch": 1.6041131105398456,
+      "grad_norm": 0.410807728767395,
+      "learning_rate": 9.90145672664953e-06,
+      "loss": 0.2979,
+      "step": 9360
+    },
+    {
+      "epoch": 1.6058269065981148,
+      "grad_norm": 0.4296032190322876,
+      "learning_rate": 9.858611825192803e-06,
+      "loss": 0.289,
+      "step": 9370
+    },
+    {
+      "epoch": 1.607540702656384,
+      "grad_norm": 0.41497138142585754,
+      "learning_rate": 9.815766923736076e-06,
+      "loss": 0.2807,
+      "step": 9380
+    },
+    {
+      "epoch": 1.609254498714653,
+      "grad_norm": 0.42269009351730347,
+      "learning_rate": 9.77292202227935e-06,
+      "loss": 0.2872,
+      "step": 9390
+    },
+    {
+      "epoch": 1.610968294772922,
+      "grad_norm": 0.4518378674983978,
+      "learning_rate": 9.730077120822623e-06,
+      "loss": 0.2843,
+      "step": 9400
+    },
+    {
+      "epoch": 1.612682090831191,
+      "grad_norm": 0.42029905319213867,
+      "learning_rate": 9.687232219365897e-06,
+      "loss": 0.2884,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6143958868894601,
+      "grad_norm": 0.3936188220977783,
+      "learning_rate": 9.64438731790917e-06,
+      "loss": 0.2878,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6161096829477293,
+      "grad_norm": 0.4592173397541046,
+      "learning_rate": 9.601542416452444e-06,
+      "loss": 0.2881,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6178234790059984,
+      "grad_norm": 0.4086572825908661,
+      "learning_rate": 9.558697514995716e-06,
+      "loss": 0.291,
+      "step": 9440
+    },
+    {
+      "epoch": 1.6195372750642674,
+      "grad_norm": 0.3873860538005829,
+      "learning_rate": 9.51585261353899e-06,
+      "loss": 0.2873,
+      "step": 9450
+    },
+    {
+      "epoch": 1.6212510711225363,
+      "grad_norm": 0.4031946063041687,
+      "learning_rate": 9.473007712082264e-06,
+      "loss": 0.2987,
+      "step": 9460
+    },
+    {
+      "epoch": 1.6229648671808055,
+      "grad_norm": 0.4394548833370209,
+      "learning_rate": 9.430162810625536e-06,
+      "loss": 0.2836,
+      "step": 9470
+    },
+    {
+      "epoch": 1.6246786632390746,
+      "grad_norm": 0.42085668444633484,
+      "learning_rate": 9.38731790916881e-06,
+      "loss": 0.2945,
+      "step": 9480
+    },
+    {
+      "epoch": 1.6263924592973438,
+      "grad_norm": 0.4487452208995819,
+      "learning_rate": 9.344473007712084e-06,
+      "loss": 0.2874,
+      "step": 9490
+    },
+    {
+      "epoch": 1.6281062553556127,
+      "grad_norm": 0.412774920463562,
+      "learning_rate": 9.301628106255356e-06,
+      "loss": 0.2968,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6298200514138816,
+      "grad_norm": 0.4071948230266571,
+      "learning_rate": 9.25878320479863e-06,
+      "loss": 0.2852,
+      "step": 9510
+    },
+    {
+      "epoch": 1.6315338474721508,
+      "grad_norm": 0.4378093183040619,
+      "learning_rate": 9.215938303341904e-06,
+      "loss": 0.2973,
+      "step": 9520
+    },
+    {
+      "epoch": 1.63324764353042,
+      "grad_norm": 0.4330248534679413,
+      "learning_rate": 9.173093401885176e-06,
+      "loss": 0.3026,
+      "step": 9530
+    },
+    {
+      "epoch": 1.634961439588689,
+      "grad_norm": 0.4182138741016388,
+      "learning_rate": 9.13024850042845e-06,
+      "loss": 0.2827,
+      "step": 9540
+    },
+    {
+      "epoch": 1.636675235646958,
+      "grad_norm": 0.4200056791305542,
+      "learning_rate": 9.087403598971722e-06,
+      "loss": 0.2914,
+      "step": 9550
+    },
+    {
+      "epoch": 1.638389031705227,
+      "grad_norm": 0.4373012185096741,
+      "learning_rate": 9.044558697514996e-06,
+      "loss": 0.2894,
+      "step": 9560
+    },
+    {
+      "epoch": 1.6401028277634961,
+      "grad_norm": 0.39351126551628113,
+      "learning_rate": 9.00171379605827e-06,
+      "loss": 0.3022,
+      "step": 9570
+    },
+    {
+      "epoch": 1.6418166238217653,
+      "grad_norm": 0.43978285789489746,
+      "learning_rate": 8.958868894601542e-06,
+      "loss": 0.3008,
+      "step": 9580
+    },
+    {
+      "epoch": 1.6435304198800342,
+      "grad_norm": 0.4187815189361572,
+      "learning_rate": 8.916023993144816e-06,
+      "loss": 0.2947,
+      "step": 9590
+    },
+    {
+      "epoch": 1.6452442159383034,
+      "grad_norm": 0.43810713291168213,
+      "learning_rate": 8.87317909168809e-06,
+      "loss": 0.3014,
+      "step": 9600
+    },
+    {
+      "epoch": 1.6469580119965723,
+      "grad_norm": 0.44445452094078064,
+      "learning_rate": 8.830334190231363e-06,
+      "loss": 0.2946,
+      "step": 9610
+    },
+    {
+      "epoch": 1.6486718080548415,
+      "grad_norm": 0.41234463453292847,
+      "learning_rate": 8.787489288774636e-06,
+      "loss": 0.2751,
+      "step": 9620
+    },
+    {
+      "epoch": 1.6503856041131106,
+      "grad_norm": 0.40182796120643616,
+      "learning_rate": 8.74464438731791e-06,
+      "loss": 0.3161,
+      "step": 9630
+    },
+    {
+      "epoch": 1.6520994001713796,
+      "grad_norm": 0.38951143622398376,
+      "learning_rate": 8.701799485861183e-06,
+      "loss": 0.2872,
+      "step": 9640
+    },
+    {
+      "epoch": 1.6538131962296485,
+      "grad_norm": 0.4077799320220947,
+      "learning_rate": 8.658954584404457e-06,
+      "loss": 0.3104,
+      "step": 9650
+    },
+    {
+      "epoch": 1.6555269922879177,
+      "grad_norm": 0.40131694078445435,
+      "learning_rate": 8.61610968294773e-06,
+      "loss": 0.2771,
+      "step": 9660
+    },
+    {
+      "epoch": 1.6572407883461868,
+      "grad_norm": 0.4214296042919159,
+      "learning_rate": 8.573264781491003e-06,
+      "loss": 0.296,
+      "step": 9670
+    },
+    {
+      "epoch": 1.658954584404456,
+      "grad_norm": 0.4252249598503113,
+      "learning_rate": 8.530419880034277e-06,
+      "loss": 0.3164,
+      "step": 9680
+    },
+    {
+      "epoch": 1.660668380462725,
+      "grad_norm": 0.44432786107063293,
+      "learning_rate": 8.487574978577549e-06,
+      "loss": 0.2807,
+      "step": 9690
+    },
+    {
+      "epoch": 1.6623821765209938,
+      "grad_norm": 0.4420909881591797,
+      "learning_rate": 8.444730077120823e-06,
+      "loss": 0.2884,
+      "step": 9700
+    },
+    {
+      "epoch": 1.664095972579263,
+      "grad_norm": 0.395749568939209,
+      "learning_rate": 8.401885175664097e-06,
+      "loss": 0.2882,
+      "step": 9710
+    },
+    {
+      "epoch": 1.6658097686375322,
+      "grad_norm": 0.4272911548614502,
+      "learning_rate": 8.359040274207369e-06,
+      "loss": 0.2973,
+      "step": 9720
+    },
+    {
+      "epoch": 1.6675235646958013,
+      "grad_norm": 0.40362003445625305,
+      "learning_rate": 8.316195372750643e-06,
+      "loss": 0.3032,
+      "step": 9730
+    },
+    {
+      "epoch": 1.6692373607540703,
+      "grad_norm": 0.4511494040489197,
+      "learning_rate": 8.273350471293917e-06,
+      "loss": 0.3025,
+      "step": 9740
+    },
+    {
+      "epoch": 1.6709511568123392,
+      "grad_norm": 0.45694980025291443,
+      "learning_rate": 8.230505569837189e-06,
+      "loss": 0.2877,
+      "step": 9750
+    },
+    {
+      "epoch": 1.6726649528706083,
+      "grad_norm": 0.4160903990268707,
+      "learning_rate": 8.187660668380463e-06,
+      "loss": 0.2975,
+      "step": 9760
+    },
+    {
+      "epoch": 1.6743787489288775,
+      "grad_norm": 0.3811100125312805,
+      "learning_rate": 8.144815766923737e-06,
+      "loss": 0.2824,
+      "step": 9770
+    },
+    {
+      "epoch": 1.6760925449871467,
+      "grad_norm": 0.4177766442298889,
+      "learning_rate": 8.10197086546701e-06,
+      "loss": 0.3008,
+      "step": 9780
+    },
+    {
+      "epoch": 1.6778063410454156,
+      "grad_norm": 0.3982120454311371,
+      "learning_rate": 8.059125964010283e-06,
+      "loss": 0.3066,
+      "step": 9790
+    },
+    {
+      "epoch": 1.6795201371036845,
+      "grad_norm": 0.42550134658813477,
+      "learning_rate": 8.016281062553555e-06,
+      "loss": 0.3025,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6812339331619537,
+      "grad_norm": 0.4049496054649353,
+      "learning_rate": 7.97343616109683e-06,
+      "loss": 0.2859,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6829477292202228,
+      "grad_norm": 0.4004078507423401,
+      "learning_rate": 7.930591259640103e-06,
+      "loss": 0.2779,
+      "step": 9820
+    },
+    {
+      "epoch": 1.684661525278492,
+      "grad_norm": 0.4874005913734436,
+      "learning_rate": 7.887746358183376e-06,
+      "loss": 0.2926,
+      "step": 9830
+    },
+    {
+      "epoch": 1.686375321336761,
+      "grad_norm": 0.4065563678741455,
+      "learning_rate": 7.84490145672665e-06,
+      "loss": 0.2764,
+      "step": 9840
+    },
+    {
+      "epoch": 1.6880891173950299,
+      "grad_norm": 0.44135749340057373,
+      "learning_rate": 7.802056555269923e-06,
+      "loss": 0.2833,
+      "step": 9850
+    },
+    {
+      "epoch": 1.689802913453299,
+      "grad_norm": 0.4260938763618469,
+      "learning_rate": 7.759211653813196e-06,
+      "loss": 0.2928,
+      "step": 9860
+    },
+    {
+      "epoch": 1.6915167095115682,
+      "grad_norm": 0.45824310183525085,
+      "learning_rate": 7.71636675235647e-06,
+      "loss": 0.2794,
+      "step": 9870
+    },
+    {
+      "epoch": 1.6932305055698373,
+      "grad_norm": 0.4385504126548767,
+      "learning_rate": 7.673521850899744e-06,
+      "loss": 0.2808,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6949443016281063,
+      "grad_norm": 0.4629712402820587,
+      "learning_rate": 7.630676949443016e-06,
+      "loss": 0.2886,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6966580976863752,
+      "grad_norm": 0.3871168792247772,
+      "learning_rate": 7.5878320479862906e-06,
+      "loss": 0.2873,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6983718937446444,
+      "grad_norm": 0.42369934916496277,
+      "learning_rate": 7.544987146529563e-06,
+      "loss": 0.2897,
+      "step": 9910
+    },
+    {
+      "epoch": 1.7000856898029135,
+      "grad_norm": 0.41781407594680786,
+      "learning_rate": 7.502142245072837e-06,
+      "loss": 0.3079,
+      "step": 9920
+    },
+    {
+      "epoch": 1.7017994858611827,
+      "grad_norm": 0.44442740082740784,
+      "learning_rate": 7.459297343616111e-06,
+      "loss": 0.2994,
+      "step": 9930
+    },
+    {
+      "epoch": 1.7035132819194516,
+      "grad_norm": 0.44649460911750793,
+      "learning_rate": 7.416452442159383e-06,
+      "loss": 0.2929,
+      "step": 9940
+    },
+    {
+      "epoch": 1.7052270779777206,
+      "grad_norm": 0.3989729881286621,
+      "learning_rate": 7.373607540702657e-06,
+      "loss": 0.2828,
+      "step": 9950
+    },
+    {
+      "epoch": 1.7069408740359897,
+      "grad_norm": 0.4344664514064789,
+      "learning_rate": 7.330762639245931e-06,
+      "loss": 0.2978,
+      "step": 9960
+    },
+    {
+      "epoch": 1.7086546700942589,
+      "grad_norm": 0.4076600968837738,
+      "learning_rate": 7.287917737789203e-06,
+      "loss": 0.2977,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7103684661525278,
+      "grad_norm": 0.41427338123321533,
+      "learning_rate": 7.245072836332477e-06,
+      "loss": 0.2841,
+      "step": 9980
+    },
+    {
+      "epoch": 1.712082262210797,
+      "grad_norm": 0.3814539313316345,
+      "learning_rate": 7.202227934875751e-06,
+      "loss": 0.2858,
+      "step": 9990
+    },
+    {
+      "epoch": 1.713796058269066,
+      "grad_norm": 0.42108654975891113,
+      "learning_rate": 7.159383033419023e-06,
+      "loss": 0.2894,
+      "step": 10000
+    },
+    {
+      "epoch": 1.715509854327335,
+      "grad_norm": 0.3633783757686615,
+      "learning_rate": 7.116538131962297e-06,
+      "loss": 0.2925,
+      "step": 10010
+    },
+    {
+      "epoch": 1.7172236503856042,
+      "grad_norm": 0.3946066200733185,
+      "learning_rate": 7.073693230505571e-06,
+      "loss": 0.2994,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7189374464438731,
+      "grad_norm": 0.4146474301815033,
+      "learning_rate": 7.030848329048843e-06,
+      "loss": 0.2953,
+      "step": 10030
+    },
+    {
+      "epoch": 1.7206512425021423,
+      "grad_norm": 0.3993196189403534,
+      "learning_rate": 6.988003427592117e-06,
+      "loss": 0.2911,
+      "step": 10040
+    },
+    {
+      "epoch": 1.7223650385604112,
+      "grad_norm": 0.40329092741012573,
+      "learning_rate": 6.9451585261353894e-06,
+      "loss": 0.2881,
+      "step": 10050
+    },
+    {
+      "epoch": 1.7240788346186804,
+      "grad_norm": 0.5011234879493713,
+      "learning_rate": 6.902313624678663e-06,
+      "loss": 0.2932,
+      "step": 10060
+    },
+    {
+      "epoch": 1.7257926306769495,
+      "grad_norm": 0.38307321071624756,
+      "learning_rate": 6.859468723221937e-06,
+      "loss": 0.3053,
+      "step": 10070
+    },
+    {
+      "epoch": 1.7275064267352185,
+      "grad_norm": 0.4366379380226135,
+      "learning_rate": 6.8166238217652096e-06,
+      "loss": 0.2808,
+      "step": 10080
+    },
+    {
+      "epoch": 1.7292202227934874,
+      "grad_norm": 0.4010637402534485,
+      "learning_rate": 6.7737789203084835e-06,
+      "loss": 0.2751,
+      "step": 10090
+    },
+    {
+      "epoch": 1.7309340188517566,
+      "grad_norm": 0.40287262201309204,
+      "learning_rate": 6.7309340188517574e-06,
+      "loss": 0.3002,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7326478149100257,
+      "grad_norm": 0.41935181617736816,
+      "learning_rate": 6.68808911739503e-06,
+      "loss": 0.2877,
+      "step": 10110
+    },
+    {
+      "epoch": 1.7343616109682949,
+      "grad_norm": 0.4558008015155792,
+      "learning_rate": 6.645244215938304e-06,
+      "loss": 0.2899,
+      "step": 10120
+    },
+    {
+      "epoch": 1.7360754070265638,
+      "grad_norm": 0.4290597140789032,
+      "learning_rate": 6.6023993144815776e-06,
+      "loss": 0.2883,
+      "step": 10130
+    },
+    {
+      "epoch": 1.7377892030848328,
+      "grad_norm": 0.4628410041332245,
+      "learning_rate": 6.55955441302485e-06,
+      "loss": 0.2836,
+      "step": 10140
+    },
+    {
+      "epoch": 1.739502999143102,
+      "grad_norm": 0.4033065438270569,
+      "learning_rate": 6.516709511568124e-06,
+      "loss": 0.2887,
+      "step": 10150
+    },
+    {
+      "epoch": 1.741216795201371,
+      "grad_norm": 0.45962148904800415,
+      "learning_rate": 6.473864610111397e-06,
+      "loss": 0.3033,
+      "step": 10160
+    },
+    {
+      "epoch": 1.7429305912596402,
+      "grad_norm": 0.4227752685546875,
+      "learning_rate": 6.43101970865467e-06,
+      "loss": 0.2865,
+      "step": 10170
+    },
+    {
+      "epoch": 1.7446443873179092,
+      "grad_norm": 0.4542335867881775,
+      "learning_rate": 6.388174807197944e-06,
+      "loss": 0.299,
+      "step": 10180
+    },
+    {
+      "epoch": 1.746358183376178,
+      "grad_norm": 0.4056103527545929,
+      "learning_rate": 6.345329905741217e-06,
+      "loss": 0.2821,
+      "step": 10190
+    },
+    {
+      "epoch": 1.7480719794344473,
+      "grad_norm": 0.40837472677230835,
+      "learning_rate": 6.302485004284491e-06,
+      "loss": 0.292,
+      "step": 10200
+    },
+    {
+      "epoch": 1.7497857754927164,
+      "grad_norm": 0.4009834825992584,
+      "learning_rate": 6.259640102827764e-06,
+      "loss": 0.2795,
+      "step": 10210
+    },
+    {
+      "epoch": 1.7514995715509856,
+      "grad_norm": 0.4178604781627655,
+      "learning_rate": 6.216795201371037e-06,
+      "loss": 0.2816,
+      "step": 10220
+    },
+    {
+      "epoch": 1.7532133676092545,
+      "grad_norm": 0.4082658588886261,
+      "learning_rate": 6.173950299914311e-06,
+      "loss": 0.2924,
+      "step": 10230
+    },
+    {
+      "epoch": 1.7549271636675234,
+      "grad_norm": 0.42404434084892273,
+      "learning_rate": 6.131105398457584e-06,
+      "loss": 0.2906,
+      "step": 10240
+    },
+    {
+      "epoch": 1.7566409597257926,
+      "grad_norm": 0.4004674255847931,
+      "learning_rate": 6.088260497000857e-06,
+      "loss": 0.2941,
+      "step": 10250
+    },
+    {
+      "epoch": 1.7583547557840618,
+      "grad_norm": 0.42199867963790894,
+      "learning_rate": 6.045415595544131e-06,
+      "loss": 0.2821,
+      "step": 10260
+    },
+    {
+      "epoch": 1.760068551842331,
+      "grad_norm": 0.3495849072933197,
+      "learning_rate": 6.002570694087404e-06,
+      "loss": 0.3087,
+      "step": 10270
+    },
+    {
+      "epoch": 1.7617823479005998,
+      "grad_norm": 0.43507277965545654,
+      "learning_rate": 5.959725792630677e-06,
+      "loss": 0.2986,
+      "step": 10280
+    },
+    {
+      "epoch": 1.7634961439588688,
+      "grad_norm": 0.40910395979881287,
+      "learning_rate": 5.91688089117395e-06,
+      "loss": 0.2951,
+      "step": 10290
+    },
+    {
+      "epoch": 1.765209940017138,
+      "grad_norm": 0.4194374978542328,
+      "learning_rate": 5.874035989717224e-06,
+      "loss": 0.2787,
+      "step": 10300
+    },
+    {
+      "epoch": 1.766923736075407,
+      "grad_norm": 0.38423359394073486,
+      "learning_rate": 5.831191088260497e-06,
+      "loss": 0.2735,
+      "step": 10310
+    },
+    {
+      "epoch": 1.7686375321336762,
+      "grad_norm": 0.41085579991340637,
+      "learning_rate": 5.7883461868037705e-06,
+      "loss": 0.2938,
+      "step": 10320
+    },
+    {
+      "epoch": 1.7703513281919452,
+      "grad_norm": 0.46077996492385864,
+      "learning_rate": 5.745501285347044e-06,
+      "loss": 0.2779,
+      "step": 10330
+    },
+    {
+      "epoch": 1.7720651242502141,
+      "grad_norm": 0.36527299880981445,
+      "learning_rate": 5.7026563838903175e-06,
+      "loss": 0.2914,
+      "step": 10340
+    },
+    {
+      "epoch": 1.7737789203084833,
+      "grad_norm": 0.393930047750473,
+      "learning_rate": 5.664095972579264e-06,
+      "loss": 0.3094,
+      "step": 10350
+    },
+    {
+      "epoch": 1.7754927163667524,
+      "grad_norm": 0.4599752426147461,
+      "learning_rate": 5.621251071122537e-06,
+      "loss": 0.2842,
+      "step": 10360
+    },
+    {
+      "epoch": 1.7772065124250214,
+      "grad_norm": 0.40328943729400635,
+      "learning_rate": 5.57840616966581e-06,
+      "loss": 0.2835,
+      "step": 10370
+    },
+    {
+      "epoch": 1.7789203084832905,
+      "grad_norm": 0.43173113465309143,
+      "learning_rate": 5.535561268209084e-06,
+      "loss": 0.2846,
+      "step": 10380
+    },
+    {
+      "epoch": 1.7806341045415595,
+      "grad_norm": 0.42086324095726013,
+      "learning_rate": 5.492716366752357e-06,
+      "loss": 0.2895,
+      "step": 10390
+    },
+    {
+      "epoch": 1.7823479005998286,
+      "grad_norm": 0.4806514382362366,
+      "learning_rate": 5.44987146529563e-06,
+      "loss": 0.3024,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7840616966580978,
+      "grad_norm": 0.4030762314796448,
+      "learning_rate": 5.407026563838903e-06,
+      "loss": 0.2791,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7857754927163667,
+      "grad_norm": 0.39070066809654236,
+      "learning_rate": 5.364181662382177e-06,
+      "loss": 0.2938,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7874892887746359,
+      "grad_norm": 0.4071260988712311,
+      "learning_rate": 5.32133676092545e-06,
+      "loss": 0.3055,
+      "step": 10430
+    },
+    {
+      "epoch": 1.7892030848329048,
+      "grad_norm": 0.40182167291641235,
+      "learning_rate": 5.278491859468723e-06,
+      "loss": 0.2862,
+      "step": 10440
+    },
+    {
+      "epoch": 1.790916880891174,
+      "grad_norm": 0.38437891006469727,
+      "learning_rate": 5.235646958011997e-06,
+      "loss": 0.2971,
+      "step": 10450
+    },
+    {
+      "epoch": 1.7926306769494431,
+      "grad_norm": 0.3624264597892761,
+      "learning_rate": 5.19280205655527e-06,
+      "loss": 0.2856,
+      "step": 10460
+    },
+    {
+      "epoch": 1.794344473007712,
+      "grad_norm": 0.42766231298446655,
+      "learning_rate": 5.149957155098543e-06,
+      "loss": 0.3043,
+      "step": 10470
+    },
+    {
+      "epoch": 1.796058269065981,
+      "grad_norm": 0.40614619851112366,
+      "learning_rate": 5.107112253641816e-06,
+      "loss": 0.2839,
+      "step": 10480
+    },
+    {
+      "epoch": 1.7977720651242501,
+      "grad_norm": 0.395358145236969,
+      "learning_rate": 5.06426735218509e-06,
+      "loss": 0.2803,
+      "step": 10490
+    },
+    {
+      "epoch": 1.7994858611825193,
+      "grad_norm": 0.41832295060157776,
+      "learning_rate": 5.0214224507283634e-06,
+      "loss": 0.298,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8011996572407885,
+      "grad_norm": 0.40189501643180847,
+      "learning_rate": 4.9785775492716365e-06,
+      "loss": 0.2687,
+      "step": 10510
+    },
+    {
+      "epoch": 1.8029134532990574,
+      "grad_norm": 0.38964855670928955,
+      "learning_rate": 4.9357326478149105e-06,
+      "loss": 0.2902,
+      "step": 10520
+    },
+    {
+      "epoch": 1.8046272493573263,
+      "grad_norm": 0.44363847374916077,
+      "learning_rate": 4.8928877463581836e-06,
+      "loss": 0.3003,
+      "step": 10530
+    },
+    {
+      "epoch": 1.8063410454155955,
+      "grad_norm": 0.37253981828689575,
+      "learning_rate": 4.850042844901457e-06,
+      "loss": 0.2941,
+      "step": 10540
+    },
+    {
+      "epoch": 1.8080548414738646,
+      "grad_norm": 0.4161553978919983,
+      "learning_rate": 4.80719794344473e-06,
+      "loss": 0.3002,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8097686375321338,
+      "grad_norm": 0.4097243547439575,
+      "learning_rate": 4.764353041988004e-06,
+      "loss": 0.2836,
+      "step": 10560
+    },
+    {
+      "epoch": 1.8114824335904027,
+      "grad_norm": 0.4094318449497223,
+      "learning_rate": 4.721508140531277e-06,
+      "loss": 0.2954,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8131962296486717,
+      "grad_norm": 0.4131515324115753,
+      "learning_rate": 4.67866323907455e-06,
+      "loss": 0.2883,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8149100257069408,
+      "grad_norm": 0.4214491546154022,
+      "learning_rate": 4.635818337617824e-06,
+      "loss": 0.3,
+      "step": 10590
+    },
+    {
+      "epoch": 1.81662382176521,
+      "grad_norm": 0.443494588136673,
+      "learning_rate": 4.592973436161097e-06,
+      "loss": 0.3067,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8183376178234791,
+      "grad_norm": 0.3683464825153351,
+      "learning_rate": 4.550128534704371e-06,
+      "loss": 0.2904,
+      "step": 10610
+    },
+    {
+      "epoch": 1.820051413881748,
+      "grad_norm": 0.41147130727767944,
+      "learning_rate": 4.507283633247644e-06,
+      "loss": 0.2877,
+      "step": 10620
+    },
+    {
+      "epoch": 1.821765209940017,
+      "grad_norm": 0.3929702639579773,
+      "learning_rate": 4.464438731790917e-06,
+      "loss": 0.285,
+      "step": 10630
+    },
+    {
+      "epoch": 1.8234790059982862,
+      "grad_norm": 0.5559737086296082,
+      "learning_rate": 4.421593830334191e-06,
+      "loss": 0.3104,
+      "step": 10640
+    },
+    {
+      "epoch": 1.8251928020565553,
+      "grad_norm": 0.41679906845092773,
+      "learning_rate": 4.378748928877464e-06,
+      "loss": 0.2858,
+      "step": 10650
+    },
+    {
+      "epoch": 1.8269065981148245,
+      "grad_norm": 0.4101051688194275,
+      "learning_rate": 4.335904027420737e-06,
+      "loss": 0.297,
+      "step": 10660
+    },
+    {
+      "epoch": 1.8286203941730934,
+      "grad_norm": 0.5653268098831177,
+      "learning_rate": 4.293059125964011e-06,
+      "loss": 0.2988,
+      "step": 10670
+    },
+    {
+      "epoch": 1.8303341902313623,
+      "grad_norm": 0.4039577543735504,
+      "learning_rate": 4.250214224507284e-06,
+      "loss": 0.2993,
+      "step": 10680
+    },
+    {
+      "epoch": 1.8320479862896315,
+      "grad_norm": 0.4027394652366638,
+      "learning_rate": 4.207369323050557e-06,
+      "loss": 0.3027,
+      "step": 10690
+    },
+    {
+      "epoch": 1.8337617823479007,
+      "grad_norm": 0.4310206174850464,
+      "learning_rate": 4.16452442159383e-06,
+      "loss": 0.2907,
+      "step": 10700
+    },
+    {
+      "epoch": 1.8354755784061698,
+      "grad_norm": 0.43067917227745056,
+      "learning_rate": 4.121679520137104e-06,
+      "loss": 0.2913,
+      "step": 10710
+    },
+    {
+      "epoch": 1.8371893744644388,
+      "grad_norm": 0.36390581727027893,
+      "learning_rate": 4.078834618680377e-06,
+      "loss": 0.284,
+      "step": 10720
+    },
+    {
+      "epoch": 1.8389031705227077,
+      "grad_norm": 0.39374905824661255,
+      "learning_rate": 4.0359897172236504e-06,
+      "loss": 0.2892,
+      "step": 10730
+    },
+    {
+      "epoch": 1.8406169665809768,
+      "grad_norm": 0.4211897850036621,
+      "learning_rate": 3.993144815766924e-06,
+      "loss": 0.2994,
+      "step": 10740
+    },
+    {
+      "epoch": 1.842330762639246,
+      "grad_norm": 0.446594774723053,
+      "learning_rate": 3.9502999143101974e-06,
+      "loss": 0.3033,
+      "step": 10750
+    },
+    {
+      "epoch": 1.8440445586975152,
+      "grad_norm": 0.4168258309364319,
+      "learning_rate": 3.9074550128534705e-06,
+      "loss": 0.2864,
+      "step": 10760
+    },
+    {
+      "epoch": 1.845758354755784,
+      "grad_norm": 0.44400787353515625,
+      "learning_rate": 3.864610111396744e-06,
+      "loss": 0.2848,
+      "step": 10770
+    },
+    {
+      "epoch": 1.847472150814053,
+      "grad_norm": 0.4246983826160431,
+      "learning_rate": 3.8217652099400176e-06,
+      "loss": 0.2975,
+      "step": 10780
+    },
+    {
+      "epoch": 1.8491859468723222,
+      "grad_norm": 0.45496413111686707,
+      "learning_rate": 3.7789203084832907e-06,
+      "loss": 0.3042,
+      "step": 10790
+    },
+    {
+      "epoch": 1.8508997429305913,
+      "grad_norm": 0.4220747649669647,
+      "learning_rate": 3.7360754070265637e-06,
+      "loss": 0.2943,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8526135389888603,
+      "grad_norm": 0.4112527668476105,
+      "learning_rate": 3.6932305055698377e-06,
+      "loss": 0.291,
+      "step": 10810
+    },
+    {
+      "epoch": 1.8543273350471294,
+      "grad_norm": 0.38883769512176514,
+      "learning_rate": 3.6503856041131108e-06,
+      "loss": 0.2972,
+      "step": 10820
+    },
+    {
+      "epoch": 1.8560411311053984,
+      "grad_norm": 0.5144733786582947,
+      "learning_rate": 3.607540702656384e-06,
+      "loss": 0.2884,
+      "step": 10830
+    },
+    {
+      "epoch": 1.8577549271636675,
+      "grad_norm": 0.3674341142177582,
+      "learning_rate": 3.564695801199657e-06,
+      "loss": 0.3013,
+      "step": 10840
+    },
+    {
+      "epoch": 1.8594687232219367,
+      "grad_norm": 0.43376797437667847,
+      "learning_rate": 3.521850899742931e-06,
+      "loss": 0.3009,
+      "step": 10850
+    },
+    {
+      "epoch": 1.8611825192802056,
+      "grad_norm": 0.4120640456676483,
+      "learning_rate": 3.479005998286204e-06,
+      "loss": 0.2944,
+      "step": 10860
+    },
+    {
+      "epoch": 1.8628963153384748,
+      "grad_norm": 0.3969920575618744,
+      "learning_rate": 3.4361610968294775e-06,
+      "loss": 0.2871,
+      "step": 10870
+    },
+    {
+      "epoch": 1.8646101113967437,
+      "grad_norm": 0.40410271286964417,
+      "learning_rate": 3.3933161953727506e-06,
+      "loss": 0.2918,
+      "step": 10880
+    },
+    {
+      "epoch": 1.8663239074550129,
+      "grad_norm": 0.37988319993019104,
+      "learning_rate": 3.3504712939160245e-06,
+      "loss": 0.289,
+      "step": 10890
+    },
+    {
+      "epoch": 1.868037703513282,
+      "grad_norm": 0.4151453971862793,
+      "learning_rate": 3.3076263924592976e-06,
+      "loss": 0.2874,
+      "step": 10900
+    },
+    {
+      "epoch": 1.869751499571551,
+      "grad_norm": 0.46920526027679443,
+      "learning_rate": 3.2647814910025707e-06,
+      "loss": 0.2887,
+      "step": 10910
+    },
+    {
+      "epoch": 1.87146529562982,
+      "grad_norm": 0.376041054725647,
+      "learning_rate": 3.2219365895458446e-06,
+      "loss": 0.2785,
+      "step": 10920
+    },
+    {
+      "epoch": 1.873179091688089,
+      "grad_norm": 0.4450885057449341,
+      "learning_rate": 3.1790916880891177e-06,
+      "loss": 0.281,
+      "step": 10930
+    },
+    {
+      "epoch": 1.8748928877463582,
+      "grad_norm": 0.4607630968093872,
+      "learning_rate": 3.136246786632391e-06,
+      "loss": 0.2846,
+      "step": 10940
+    },
+    {
+      "epoch": 1.8766066838046274,
+      "grad_norm": 0.43728557229042053,
+      "learning_rate": 3.0934018851756643e-06,
+      "loss": 0.289,
+      "step": 10950
+    },
+    {
+      "epoch": 1.8783204798628963,
+      "grad_norm": 0.3537221848964691,
+      "learning_rate": 3.0505569837189374e-06,
+      "loss": 0.2838,
+      "step": 10960
+    },
+    {
+      "epoch": 1.8800342759211652,
+      "grad_norm": 0.3849254846572876,
+      "learning_rate": 3.007712082262211e-06,
+      "loss": 0.2727,
+      "step": 10970
+    },
+    {
+      "epoch": 1.8817480719794344,
+      "grad_norm": 0.38857021927833557,
+      "learning_rate": 2.964867180805484e-06,
+      "loss": 0.2798,
+      "step": 10980
+    },
+    {
+      "epoch": 1.8834618680377035,
+      "grad_norm": 0.4325757920742035,
+      "learning_rate": 2.9220222793487575e-06,
+      "loss": 0.2907,
+      "step": 10990
+    },
+    {
+      "epoch": 1.8851756640959727,
+      "grad_norm": 0.4491504430770874,
+      "learning_rate": 2.879177377892031e-06,
+      "loss": 0.299,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8868894601542416,
+      "grad_norm": 0.3967271149158478,
+      "learning_rate": 2.836332476435304e-06,
+      "loss": 0.287,
+      "step": 11010
+    },
+    {
+      "epoch": 1.8886032562125106,
+      "grad_norm": 0.4465595781803131,
+      "learning_rate": 2.7934875749785776e-06,
+      "loss": 0.2811,
+      "step": 11020
+    },
+    {
+      "epoch": 1.8903170522707797,
+      "grad_norm": 0.418375700712204,
+      "learning_rate": 2.750642673521851e-06,
+      "loss": 0.2951,
+      "step": 11030
+    },
+    {
+      "epoch": 1.8920308483290489,
+      "grad_norm": 0.4344941973686218,
+      "learning_rate": 2.7077977720651242e-06,
+      "loss": 0.2914,
+      "step": 11040
+    },
+    {
+      "epoch": 1.893744644387318,
+      "grad_norm": 0.4044998586177826,
+      "learning_rate": 2.6649528706083978e-06,
+      "loss": 0.2829,
+      "step": 11050
+    },
+    {
+      "epoch": 1.895458440445587,
+      "grad_norm": 0.4352487623691559,
+      "learning_rate": 2.6221079691516713e-06,
+      "loss": 0.2969,
+      "step": 11060
+    },
+    {
+      "epoch": 1.897172236503856,
+      "grad_norm": 0.4365026354789734,
+      "learning_rate": 2.5792630676949444e-06,
+      "loss": 0.2862,
+      "step": 11070
+    },
+    {
+      "epoch": 1.898886032562125,
+      "grad_norm": 0.4220198094844818,
+      "learning_rate": 2.536418166238218e-06,
+      "loss": 0.3022,
+      "step": 11080
+    },
+    {
+      "epoch": 1.9005998286203942,
+      "grad_norm": 0.4542985260486603,
+      "learning_rate": 2.4935732647814914e-06,
+      "loss": 0.2917,
+      "step": 11090
+    },
+    {
+      "epoch": 1.9023136246786634,
+      "grad_norm": 0.4488210678100586,
+      "learning_rate": 2.4507283633247645e-06,
+      "loss": 0.2845,
+      "step": 11100
+    },
+    {
+      "epoch": 1.9040274207369323,
+      "grad_norm": 0.3916579782962799,
+      "learning_rate": 2.407883461868038e-06,
+      "loss": 0.2853,
+      "step": 11110
+    },
+    {
+      "epoch": 1.9057412167952013,
+      "grad_norm": 0.4579949378967285,
+      "learning_rate": 2.365038560411311e-06,
+      "loss": 0.2843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.9074550128534704,
+      "grad_norm": 0.4021014869213104,
+      "learning_rate": 2.3221936589545846e-06,
+      "loss": 0.2972,
+      "step": 11130
+    },
+    {
+      "epoch": 1.9091688089117396,
+      "grad_norm": 0.39919403195381165,
+      "learning_rate": 2.2793487574978577e-06,
+      "loss": 0.2857,
+      "step": 11140
+    },
+    {
+      "epoch": 1.9108826049700087,
+      "grad_norm": 0.42908406257629395,
+      "learning_rate": 2.236503856041131e-06,
+      "loss": 0.2962,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9125964010282777,
+      "grad_norm": 0.3778807818889618,
+      "learning_rate": 2.1936589545844043e-06,
+      "loss": 0.2755,
+      "step": 11160
+    },
+    {
+      "epoch": 1.9143101970865466,
+      "grad_norm": 0.3933291733264923,
+      "learning_rate": 2.1508140531276778e-06,
+      "loss": 0.2912,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9160239931448158,
+      "grad_norm": 0.4091528356075287,
+      "learning_rate": 2.1079691516709513e-06,
+      "loss": 0.2863,
+      "step": 11180
+    },
+    {
+      "epoch": 1.917737789203085,
+      "grad_norm": 0.41364723443984985,
+      "learning_rate": 2.0651242502142244e-06,
+      "loss": 0.3019,
+      "step": 11190
+    },
+    {
+      "epoch": 1.919451585261354,
+      "grad_norm": 0.39539095759391785,
+      "learning_rate": 2.022279348757498e-06,
+      "loss": 0.2941,
+      "step": 11200
+    },
+    {
+      "epoch": 1.921165381319623,
+      "grad_norm": 0.408882737159729,
+      "learning_rate": 1.9794344473007714e-06,
+      "loss": 0.3004,
+      "step": 11210
+    },
+    {
+      "epoch": 1.922879177377892,
+      "grad_norm": 0.43997281789779663,
+      "learning_rate": 1.936589545844045e-06,
+      "loss": 0.3063,
+      "step": 11220
+    },
+    {
+      "epoch": 1.924592973436161,
+      "grad_norm": 0.37508559226989746,
+      "learning_rate": 1.8937446443873178e-06,
+      "loss": 0.2851,
+      "step": 11230
+    },
+    {
+      "epoch": 1.9263067694944302,
+      "grad_norm": 0.45042285323143005,
+      "learning_rate": 1.8508997429305913e-06,
+      "loss": 0.2945,
+      "step": 11240
+    },
+    {
+      "epoch": 1.9280205655526992,
+      "grad_norm": 0.39290183782577515,
+      "learning_rate": 1.8080548414738646e-06,
+      "loss": 0.2896,
+      "step": 11250
+    },
+    {
+      "epoch": 1.9297343616109683,
+      "grad_norm": 0.43095430731773376,
+      "learning_rate": 1.7652099400171381e-06,
+      "loss": 0.2835,
+      "step": 11260
+    },
+    {
+      "epoch": 1.9314481576692373,
+      "grad_norm": 0.402061402797699,
+      "learning_rate": 1.7223650385604116e-06,
+      "loss": 0.273,
+      "step": 11270
+    },
+    {
+      "epoch": 1.9331619537275064,
+      "grad_norm": 0.41067150235176086,
+      "learning_rate": 1.6795201371036847e-06,
+      "loss": 0.2931,
+      "step": 11280
+    },
+    {
+      "epoch": 1.9348757497857756,
+      "grad_norm": 0.40099552273750305,
+      "learning_rate": 1.6366752356469582e-06,
+      "loss": 0.293,
+      "step": 11290
+    },
+    {
+      "epoch": 1.9365895458440445,
+      "grad_norm": 0.433057963848114,
+      "learning_rate": 1.5938303341902313e-06,
+      "loss": 0.2978,
+      "step": 11300
+    },
+    {
+      "epoch": 1.9383033419023135,
+      "grad_norm": 0.3876418471336365,
+      "learning_rate": 1.5509854327335048e-06,
+      "loss": 0.2892,
+      "step": 11310
+    },
+    {
+      "epoch": 1.9400171379605826,
+      "grad_norm": 0.4423709809780121,
+      "learning_rate": 1.5081405312767782e-06,
+      "loss": 0.2781,
+      "step": 11320
+    },
+    {
+      "epoch": 1.9417309340188518,
+      "grad_norm": 0.37492167949676514,
+      "learning_rate": 1.4652956298200515e-06,
+      "loss": 0.2882,
+      "step": 11330
+    },
+    {
+      "epoch": 1.943444730077121,
+      "grad_norm": 0.4402335584163666,
+      "learning_rate": 1.422450728363325e-06,
+      "loss": 0.3049,
+      "step": 11340
+    },
+    {
+      "epoch": 1.9451585261353899,
+      "grad_norm": 0.4293029308319092,
+      "learning_rate": 1.3796058269065983e-06,
+      "loss": 0.2836,
+      "step": 11350
+    },
+    {
+      "epoch": 1.9468723221936588,
+      "grad_norm": 0.4107189178466797,
+      "learning_rate": 1.3367609254498716e-06,
+      "loss": 0.2897,
+      "step": 11360
+    },
+    {
+      "epoch": 1.948586118251928,
+      "grad_norm": 0.44577571749687195,
+      "learning_rate": 1.2939160239931449e-06,
+      "loss": 0.2894,
+      "step": 11370
+    },
+    {
+      "epoch": 1.9502999143101971,
+      "grad_norm": 0.3917602002620697,
+      "learning_rate": 1.2510711225364182e-06,
+      "loss": 0.3023,
+      "step": 11380
+    },
+    {
+      "epoch": 1.9520137103684663,
+      "grad_norm": 0.3750052750110626,
+      "learning_rate": 1.2082262210796915e-06,
+      "loss": 0.3,
+      "step": 11390
+    },
+    {
+      "epoch": 1.9537275064267352,
+      "grad_norm": 0.40812334418296814,
+      "learning_rate": 1.1653813196229648e-06,
+      "loss": 0.297,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9554413024850041,
+      "grad_norm": 0.3837462067604065,
+      "learning_rate": 1.1225364181662383e-06,
+      "loss": 0.2914,
+      "step": 11410
+    },
+    {
+      "epoch": 1.9571550985432733,
+      "grad_norm": 0.4080122709274292,
+      "learning_rate": 1.0796915167095116e-06,
+      "loss": 0.3031,
+      "step": 11420
+    },
+    {
+      "epoch": 1.9588688946015425,
+      "grad_norm": 0.4329179525375366,
+      "learning_rate": 1.036846615252785e-06,
+      "loss": 0.2788,
+      "step": 11430
+    },
+    {
+      "epoch": 1.9605826906598116,
+      "grad_norm": 0.41013336181640625,
+      "learning_rate": 9.940017137960584e-07,
+      "loss": 0.2871,
+      "step": 11440
+    },
+    {
+      "epoch": 1.9622964867180805,
+      "grad_norm": 0.3924817442893982,
+      "learning_rate": 9.511568123393317e-07,
+      "loss": 0.2957,
+      "step": 11450
+    },
+    {
+      "epoch": 1.9640102827763495,
+      "grad_norm": 0.3885457217693329,
+      "learning_rate": 9.08311910882605e-07,
+      "loss": 0.2963,
+      "step": 11460
+    },
+    {
+      "epoch": 1.9657240788346186,
+      "grad_norm": 0.4009818136692047,
+      "learning_rate": 8.654670094258783e-07,
+      "loss": 0.2899,
+      "step": 11470
+    },
+    {
+      "epoch": 1.9674378748928878,
+      "grad_norm": 0.40517812967300415,
+      "learning_rate": 8.226221079691516e-07,
+      "loss": 0.2842,
+      "step": 11480
+    },
+    {
+      "epoch": 1.969151670951157,
+      "grad_norm": 0.39424851536750793,
+      "learning_rate": 7.797772065124251e-07,
+      "loss": 0.2844,
+      "step": 11490
+    },
+    {
+      "epoch": 1.9708654670094259,
+      "grad_norm": 0.4126230776309967,
+      "learning_rate": 7.369323050556984e-07,
+      "loss": 0.2882,
+      "step": 11500
+    },
+    {
+      "epoch": 1.9725792630676948,
+      "grad_norm": 0.4105775058269501,
+      "learning_rate": 6.940874035989717e-07,
+      "loss": 0.2879,
+      "step": 11510
+    },
+    {
+      "epoch": 1.974293059125964,
+      "grad_norm": 0.39232248067855835,
+      "learning_rate": 6.512425021422451e-07,
+      "loss": 0.2942,
+      "step": 11520
+    },
+    {
+      "epoch": 1.9760068551842331,
+      "grad_norm": 0.45345765352249146,
+      "learning_rate": 6.083976006855184e-07,
+      "loss": 0.3061,
+      "step": 11530
+    },
+    {
+      "epoch": 1.9777206512425023,
+      "grad_norm": 0.40826553106307983,
+      "learning_rate": 5.655526992287918e-07,
+      "loss": 0.3054,
+      "step": 11540
+    },
+    {
+      "epoch": 1.9794344473007712,
+      "grad_norm": 0.45879650115966797,
+      "learning_rate": 5.227077977720651e-07,
+      "loss": 0.298,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9811482433590402,
+      "grad_norm": 0.41387298703193665,
+      "learning_rate": 4.798628963153384e-07,
+      "loss": 0.2964,
+      "step": 11560
+    },
+    {
+      "epoch": 1.9828620394173093,
+      "grad_norm": 0.42677220702171326,
+      "learning_rate": 4.370179948586119e-07,
+      "loss": 0.2959,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9845758354755785,
+      "grad_norm": 0.3857152462005615,
+      "learning_rate": 3.941730934018852e-07,
+      "loss": 0.3018,
+      "step": 11580
+    },
+    {
+      "epoch": 1.9862896315338476,
+      "grad_norm": 0.4266795217990875,
+      "learning_rate": 3.5132819194515855e-07,
+      "loss": 0.2923,
+      "step": 11590
+    },
+    {
+      "epoch": 1.9880034275921166,
+      "grad_norm": 0.4096374213695526,
+      "learning_rate": 3.084832904884319e-07,
+      "loss": 0.2899,
+      "step": 11600
+    },
+    {
+      "epoch": 1.9897172236503855,
+      "grad_norm": 0.40352872014045715,
+      "learning_rate": 2.6563838903170526e-07,
+      "loss": 0.2965,
+      "step": 11610
+    },
+    {
+      "epoch": 1.9914310197086547,
+      "grad_norm": 0.4364762008190155,
+      "learning_rate": 2.2279348757497856e-07,
+      "loss": 0.3046,
+      "step": 11620
+    },
+    {
+      "epoch": 1.9931448157669238,
+      "grad_norm": 0.4287553131580353,
+      "learning_rate": 1.7994858611825194e-07,
+      "loss": 0.2866,
+      "step": 11630
+    },
+    {
+      "epoch": 1.9948586118251928,
+      "grad_norm": 0.4138564467430115,
+      "learning_rate": 1.3710368466152527e-07,
+      "loss": 0.2979,
+      "step": 11640
+    },
+    {
+      "epoch": 1.996572407883462,
+      "grad_norm": 0.39418089389801025,
+      "learning_rate": 9.425878320479863e-08,
+      "loss": 0.2869,
+      "step": 11650
+    },
+    {
+      "epoch": 1.9982862039417308,
+      "grad_norm": 0.43074512481689453,
+      "learning_rate": 5.1413881748071975e-08,
+      "loss": 0.2838,
+      "step": 11660
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8341490030288696,
+      "learning_rate": 8.56898029134533e-09,
+      "loss": 0.2892,
+      "step": 11670
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 11670,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.7565399580672e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}