diff --git "a/adapters/checkpoint-11240/trainer_state.json" "b/adapters/checkpoint-11240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/adapters/checkpoint-11240/trainer_state.json"
@@ -0,0 +1,8078 @@
+{
+  "best_global_step": 11000,
+  "best_metric": 0.6082284450531006,
+  "best_model_checkpoint": "/workspace/rails-finetune/adapters-qwen3-8b/checkpoint-11000",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 11240,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0017793594306049821,
+      "grad_norm": 1.1857829093933105,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.7382530212402343,
+      "step": 10
+    },
+    {
+      "epoch": 0.0035587188612099642,
+      "grad_norm": 1.1971900463104248,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.7001752853393555,
+      "step": 20
+    },
+    {
+      "epoch": 0.005338078291814947,
+      "grad_norm": 1.2064718008041382,
+      "learning_rate": 2.9e-06,
+      "loss": 1.7585294723510743,
+      "step": 30
+    },
+    {
+      "epoch": 0.0071174377224199285,
+      "grad_norm": 1.1466728448867798,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.6992141723632812,
+      "step": 40
+    },
+    {
+      "epoch": 0.008896797153024912,
+      "grad_norm": 1.0449944734573364,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.7329919815063477,
+      "step": 50
+    },
+    {
+      "epoch": 0.010676156583629894,
+      "grad_norm": 1.25763738155365,
+      "learning_rate": 5.9e-06,
+      "loss": 1.5657649040222168,
+      "step": 60
+    },
+    {
+      "epoch": 0.012455516014234875,
+      "grad_norm": 0.9314267039299011,
+      "learning_rate": 6.9e-06,
+      "loss": 1.5626873970031738,
+      "step": 70
+    },
+    {
+      "epoch": 0.014234875444839857,
+      "grad_norm": 0.8502155542373657,
+      "learning_rate": 7.9e-06,
+      "loss": 1.3948446273803712,
+      "step": 80
+    },
+    {
+      "epoch": 0.01601423487544484,
+      "grad_norm": 0.505824625492096,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.318073558807373,
+      "step": 90
+    },
+    {
+      "epoch": 0.017793594306049824,
+      "grad_norm": 0.4828319251537323,
+      "learning_rate": 9.9e-06,
+      "loss": 1.2034348487854003,
+      "step": 100
+    },
+    {
+      "epoch": 0.019572953736654804,
+      "grad_norm": 0.4351405203342438,
+      "learning_rate": 1.0900000000000002e-05,
+      "loss": 1.2164586067199707,
+      "step": 110
+    },
+    {
+      "epoch": 0.021352313167259787,
+      "grad_norm": 0.3060351014137268,
+      "learning_rate": 1.1900000000000001e-05,
+      "loss": 1.1403413772583009,
+      "step": 120
+    },
+    {
+      "epoch": 0.023131672597864767,
+      "grad_norm": 0.30274662375450134,
+      "learning_rate": 1.2900000000000002e-05,
+      "loss": 1.0329069137573241,
+      "step": 130
+    },
+    {
+      "epoch": 0.02491103202846975,
+      "grad_norm": 0.27452540397644043,
+      "learning_rate": 1.39e-05,
+      "loss": 1.0137288093566894,
+      "step": 140
+    },
+    {
+      "epoch": 0.026690391459074734,
+      "grad_norm": 0.22874200344085693,
+      "learning_rate": 1.4900000000000001e-05,
+      "loss": 0.9552822113037109,
+      "step": 150
+    },
+    {
+      "epoch": 0.028469750889679714,
+      "grad_norm": 0.20503273606300354,
+      "learning_rate": 1.5900000000000004e-05,
+      "loss": 0.9580234527587891,
+      "step": 160
+    },
+    {
+      "epoch": 0.030249110320284697,
+      "grad_norm": 0.20734967291355133,
+      "learning_rate": 1.69e-05,
+      "loss": 0.9269493103027344,
+      "step": 170
+    },
+    {
+      "epoch": 0.03202846975088968,
+      "grad_norm": 0.2046293467283249,
+      "learning_rate": 1.79e-05,
+      "loss": 0.898008155822754,
+      "step": 180
+    },
+    {
+      "epoch": 0.033807829181494664,
+      "grad_norm": 0.20902672410011292,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 0.8822259902954102,
+      "step": 190
+    },
+    {
+      "epoch": 0.03558718861209965,
+      "grad_norm": 0.265747606754303,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 0.9367627143859864,
+      "step": 200
+    },
+    {
+      "epoch": 0.037366548042704624,
+      "grad_norm": 0.21941998600959778,
+      "learning_rate": 1.9999967204339314e-05,
+      "loss": 0.8784577369689941,
+      "step": 210
+    },
+    {
+      "epoch": 0.03914590747330961,
+      "grad_norm": 0.23806284368038177,
+      "learning_rate": 1.999985383689953e-05,
+      "loss": 0.8498974800109863,
+      "step": 220
+    },
+    {
+      "epoch": 0.04092526690391459,
+      "grad_norm": 0.2539118826389313,
+      "learning_rate": 1.9999659493713742e-05,
+      "loss": 0.8517349243164063,
+      "step": 230
+    },
+    {
+      "epoch": 0.042704626334519574,
+      "grad_norm": 0.279153436422348,
+      "learning_rate": 1.9999384176355685e-05,
+      "loss": 0.8303974151611329,
+      "step": 240
+    },
+    {
+      "epoch": 0.04448398576512456,
+      "grad_norm": 0.2661009132862091,
+      "learning_rate": 1.9999027887054793e-05,
+      "loss": 0.8362269401550293,
+      "step": 250
+    },
+    {
+      "epoch": 0.046263345195729534,
+      "grad_norm": 0.2643781006336212,
+      "learning_rate": 1.9998590628696186e-05,
+      "loss": 0.7995978832244873,
+      "step": 260
+    },
+    {
+      "epoch": 0.04804270462633452,
+      "grad_norm": 0.22854603826999664,
+      "learning_rate": 1.9998072404820648e-05,
+      "loss": 0.8313385009765625,
+      "step": 270
+    },
+    {
+      "epoch": 0.0498220640569395,
+      "grad_norm": 0.25225695967674255,
+      "learning_rate": 1.99974732196246e-05,
+      "loss": 0.8484455108642578,
+      "step": 280
+    },
+    {
+      "epoch": 0.051601423487544484,
+      "grad_norm": 0.2538894712924957,
+      "learning_rate": 1.999679307796006e-05,
+      "loss": 0.8153322219848633,
+      "step": 290
+    },
+    {
+      "epoch": 0.05338078291814947,
+      "grad_norm": 0.257697194814682,
+      "learning_rate": 1.9996031985334604e-05,
+      "loss": 0.8145216941833496,
+      "step": 300
+    },
+    {
+      "epoch": 0.05516014234875445,
+      "grad_norm": 0.2465839385986328,
+      "learning_rate": 1.9995189947911323e-05,
+      "loss": 0.8290293693542481,
+      "step": 310
+    },
+    {
+      "epoch": 0.05693950177935943,
+      "grad_norm": 0.2330036759376526,
+      "learning_rate": 1.9994266972508786e-05,
+      "loss": 0.8094453811645508,
+      "step": 320
+    },
+    {
+      "epoch": 0.05871886120996441,
+      "grad_norm": 0.24966399371623993,
+      "learning_rate": 1.9993263066600956e-05,
+      "loss": 0.7964043140411377,
+      "step": 330
+    },
+    {
+      "epoch": 0.060498220640569395,
+      "grad_norm": 0.2663300335407257,
+      "learning_rate": 1.9992178238317156e-05,
+      "loss": 0.7674037456512451,
+      "step": 340
+    },
+    {
+      "epoch": 0.06227758007117438,
+      "grad_norm": 0.29889997839927673,
+      "learning_rate": 1.9991012496442e-05,
+      "loss": 0.7757882118225098,
+      "step": 350
+    },
+    {
+      "epoch": 0.06405693950177936,
+      "grad_norm": 0.2283693104982376,
+      "learning_rate": 1.9989765850415303e-05,
+      "loss": 0.7695906162261963,
+      "step": 360
+    },
+    {
+      "epoch": 0.06583629893238434,
+      "grad_norm": 0.2109132707118988,
+      "learning_rate": 1.9988438310332015e-05,
+      "loss": 0.7833895683288574,
+      "step": 370
+    },
+    {
+      "epoch": 0.06761565836298933,
+      "grad_norm": 0.2546916604042053,
+      "learning_rate": 1.998702988694216e-05,
+      "loss": 0.8128045082092286,
+      "step": 380
+    },
+    {
+      "epoch": 0.0693950177935943,
+      "grad_norm": 0.2531549632549286,
+      "learning_rate": 1.998554059165071e-05,
+      "loss": 0.7902643203735351,
+      "step": 390
+    },
+    {
+      "epoch": 0.0711743772241993,
+      "grad_norm": 0.21756362915039062,
+      "learning_rate": 1.9983970436517523e-05,
+      "loss": 0.7454084873199462,
+      "step": 400
+    },
+    {
+      "epoch": 0.07295373665480427,
+      "grad_norm": 0.2294779121875763,
+      "learning_rate": 1.9982319434257236e-05,
+      "loss": 0.7487451553344726,
+      "step": 410
+    },
+    {
+      "epoch": 0.07473309608540925,
+      "grad_norm": 0.26356714963912964,
+      "learning_rate": 1.9980587598239155e-05,
+      "loss": 0.748570442199707,
+      "step": 420
+    },
+    {
+      "epoch": 0.07651245551601424,
+      "grad_norm": 0.22382904589176178,
+      "learning_rate": 1.9978774942487155e-05,
+      "loss": 0.7261887073516846,
+      "step": 430
+    },
+    {
+      "epoch": 0.07829181494661921,
+      "grad_norm": 0.22544927895069122,
+      "learning_rate": 1.997688148167957e-05,
+      "loss": 0.7761906623840332,
+      "step": 440
+    },
+    {
+      "epoch": 0.0800711743772242,
+      "grad_norm": 0.2679840922355652,
+      "learning_rate": 1.997490723114906e-05,
+      "loss": 0.7872249603271484,
+      "step": 450
+    },
+    {
+      "epoch": 0.08185053380782918,
+      "grad_norm": 0.20914621651172638,
+      "learning_rate": 1.9972852206882504e-05,
+      "loss": 0.7876029014587402,
+      "step": 460
+    },
+    {
+      "epoch": 0.08362989323843416,
+      "grad_norm": 0.36782026290893555,
+      "learning_rate": 1.9970716425520854e-05,
+      "loss": 0.7812703132629395,
+      "step": 470
+    },
+    {
+      "epoch": 0.08540925266903915,
+      "grad_norm": 0.2317405343055725,
+      "learning_rate": 1.9968499904359017e-05,
+      "loss": 0.7694793701171875,
+      "step": 480
+    },
+    {
+      "epoch": 0.08718861209964412,
+      "grad_norm": 0.23587286472320557,
+      "learning_rate": 1.9966202661345694e-05,
+      "loss": 0.7593471050262451,
+      "step": 490
+    },
+    {
+      "epoch": 0.08896797153024912,
+      "grad_norm": 0.3179067075252533,
+      "learning_rate": 1.9963824715083255e-05,
+      "loss": 0.804302978515625,
+      "step": 500
+    },
+    {
+      "epoch": 0.08896797153024912,
+      "eval_loss": 0.7566477656364441,
+      "eval_runtime": 406.3608,
+      "eval_samples_per_second": 12.376,
+      "eval_steps_per_second": 6.189,
+      "step": 500
+    },
+    {
+      "epoch": 0.09074733096085409,
+      "grad_norm": 0.2583659291267395,
+      "learning_rate": 1.9961366084827584e-05,
+      "loss": 0.7231699466705322,
+      "step": 510
+    },
+    {
+      "epoch": 0.09252669039145907,
+      "grad_norm": 0.3117372393608093,
+      "learning_rate": 1.9958826790487905e-05,
+      "loss": 0.7830834865570069,
+      "step": 520
+    },
+    {
+      "epoch": 0.09430604982206406,
+      "grad_norm": 0.26633507013320923,
+      "learning_rate": 1.995620685262665e-05,
+      "loss": 0.7565378189086914,
+      "step": 530
+    },
+    {
+      "epoch": 0.09608540925266904,
+      "grad_norm": 0.2491220384836197,
+      "learning_rate": 1.9953506292459275e-05,
+      "loss": 0.770139503479004,
+      "step": 540
+    },
+    {
+      "epoch": 0.09786476868327403,
+      "grad_norm": 0.33897528052330017,
+      "learning_rate": 1.9950725131854082e-05,
+      "loss": 0.8139609336853028,
+      "step": 550
+    },
+    {
+      "epoch": 0.099644128113879,
+      "grad_norm": 0.27655312418937683,
+      "learning_rate": 1.9947863393332053e-05,
+      "loss": 0.7632327556610108,
+      "step": 560
+    },
+    {
+      "epoch": 0.10142348754448399,
+      "grad_norm": 0.2421133816242218,
+      "learning_rate": 1.994492110006667e-05,
+      "loss": 0.7782410621643067,
+      "step": 570
+    },
+    {
+      "epoch": 0.10320284697508897,
+      "grad_norm": 0.26968705654144287,
+      "learning_rate": 1.994189827588372e-05,
+      "loss": 0.7438684940338135,
+      "step": 580
+    },
+    {
+      "epoch": 0.10498220640569395,
+      "grad_norm": 0.28697794675827026,
+      "learning_rate": 1.993879494526111e-05,
+      "loss": 0.7320491313934326,
+      "step": 590
+    },
+    {
+      "epoch": 0.10676156583629894,
+      "grad_norm": 0.327286034822464,
+      "learning_rate": 1.9935611133328657e-05,
+      "loss": 0.7248147010803223,
+      "step": 600
+    },
+    {
+      "epoch": 0.10854092526690391,
+      "grad_norm": 0.3679325580596924,
+      "learning_rate": 1.9932346865867885e-05,
+      "loss": 0.7461765766143799,
+      "step": 610
+    },
+    {
+      "epoch": 0.1103202846975089,
+      "grad_norm": 0.3118029832839966,
+      "learning_rate": 1.992900216931184e-05,
+      "loss": 0.764728593826294,
+      "step": 620
+    },
+    {
+      "epoch": 0.11209964412811388,
+      "grad_norm": 0.2683190107345581,
+      "learning_rate": 1.992557707074484e-05,
+      "loss": 0.7256179809570312,
+      "step": 630
+    },
+    {
+      "epoch": 0.11387900355871886,
+      "grad_norm": 0.32056325674057007,
+      "learning_rate": 1.9922071597902286e-05,
+      "loss": 0.7553198337554932,
+      "step": 640
+    },
+    {
+      "epoch": 0.11565836298932385,
+      "grad_norm": 0.4004197418689728,
+      "learning_rate": 1.9918485779170417e-05,
+      "loss": 0.7440505027770996,
+      "step": 650
+    },
+    {
+      "epoch": 0.11743772241992882,
+      "grad_norm": 0.3497128486633301,
+      "learning_rate": 1.9914819643586096e-05,
+      "loss": 0.7672629833221436,
+      "step": 660
+    },
+    {
+      "epoch": 0.11921708185053381,
+      "grad_norm": 0.3187069594860077,
+      "learning_rate": 1.9911073220836562e-05,
+      "loss": 0.7420164585113526,
+      "step": 670
+    },
+    {
+      "epoch": 0.12099644128113879,
+      "grad_norm": 0.29900017380714417,
+      "learning_rate": 1.9907246541259194e-05,
+      "loss": 0.7458691120147705,
+      "step": 680
+    },
+    {
+      "epoch": 0.12277580071174377,
+      "grad_norm": 0.298501193523407,
+      "learning_rate": 1.9903339635841274e-05,
+      "loss": 0.7346842765808106,
+      "step": 690
+    },
+    {
+      "epoch": 0.12455516014234876,
+      "grad_norm": 0.38077300786972046,
+      "learning_rate": 1.9899352536219713e-05,
+      "loss": 0.7980701446533203,
+      "step": 700
+    },
+    {
+      "epoch": 0.12633451957295375,
+      "grad_norm": 0.2950842082500458,
+      "learning_rate": 1.9895285274680826e-05,
+      "loss": 0.7282920837402344,
+      "step": 710
+    },
+    {
+      "epoch": 0.12811387900355872,
+      "grad_norm": 0.2746928632259369,
+      "learning_rate": 1.989113788416005e-05,
+      "loss": 0.7119527816772461,
+      "step": 720
+    },
+    {
+      "epoch": 0.1298932384341637,
+      "grad_norm": 0.30003389716148376,
+      "learning_rate": 1.9886910398241673e-05,
+      "loss": 0.7120148181915283,
+      "step": 730
+    },
+    {
+      "epoch": 0.13167259786476868,
+      "grad_norm": 0.3238595128059387,
+      "learning_rate": 1.9882602851158584e-05,
+      "loss": 0.7219894886016845,
+      "step": 740
+    },
+    {
+      "epoch": 0.13345195729537365,
+      "grad_norm": 0.4421483278274536,
+      "learning_rate": 1.9878215277791977e-05,
+      "loss": 0.6878085613250733,
+      "step": 750
+    },
+    {
+      "epoch": 0.13523131672597866,
+      "grad_norm": 0.3009251058101654,
+      "learning_rate": 1.9873747713671073e-05,
+      "loss": 0.777537488937378,
+      "step": 760
+    },
+    {
+      "epoch": 0.13701067615658363,
+      "grad_norm": 0.3649790287017822,
+      "learning_rate": 1.9869200194972828e-05,
+      "loss": 0.7608931541442872,
+      "step": 770
+    },
+    {
+      "epoch": 0.1387900355871886,
+      "grad_norm": 0.38479843735694885,
+      "learning_rate": 1.986457275852166e-05,
+      "loss": 0.7644641876220704,
+      "step": 780
+    },
+    {
+      "epoch": 0.14056939501779359,
+      "grad_norm": 0.3267346918582916,
+      "learning_rate": 1.9859865441789126e-05,
+      "loss": 0.7437977313995361,
+      "step": 790
+    },
+    {
+      "epoch": 0.1423487544483986,
+      "grad_norm": 0.3238297402858734,
+      "learning_rate": 1.985507828289363e-05,
+      "loss": 0.7596290111541748,
+      "step": 800
+    },
+    {
+      "epoch": 0.14412811387900357,
+      "grad_norm": 0.4067242443561554,
+      "learning_rate": 1.985021132060012e-05,
+      "loss": 0.723486328125,
+      "step": 810
+    },
+    {
+      "epoch": 0.14590747330960854,
+      "grad_norm": 0.3294743299484253,
+      "learning_rate": 1.9845264594319755e-05,
+      "loss": 0.8035991668701172,
+      "step": 820
+    },
+    {
+      "epoch": 0.14768683274021352,
+      "grad_norm": 0.2861204445362091,
+      "learning_rate": 1.9840238144109613e-05,
+      "loss": 0.7118996620178223,
+      "step": 830
+    },
+    {
+      "epoch": 0.1494661921708185,
+      "grad_norm": 0.3281143605709076,
+      "learning_rate": 1.9835132010672334e-05,
+      "loss": 0.7610855102539062,
+      "step": 840
+    },
+    {
+      "epoch": 0.1512455516014235,
+      "grad_norm": 0.3176390826702118,
+      "learning_rate": 1.982994623535583e-05,
+      "loss": 0.7489484310150146,
+      "step": 850
+    },
+    {
+      "epoch": 0.15302491103202848,
+      "grad_norm": 0.34338897466659546,
+      "learning_rate": 1.9824680860152914e-05,
+      "loss": 0.7180755615234375,
+      "step": 860
+    },
+    {
+      "epoch": 0.15480427046263345,
+      "grad_norm": 0.41952571272850037,
+      "learning_rate": 1.9819335927700975e-05,
+      "loss": 0.6884951591491699,
+      "step": 870
+    },
+    {
+      "epoch": 0.15658362989323843,
+      "grad_norm": 0.29155367612838745,
+      "learning_rate": 1.9813911481281637e-05,
+      "loss": 0.683270263671875,
+      "step": 880
+    },
+    {
+      "epoch": 0.1583629893238434,
+      "grad_norm": 0.3815101981163025,
+      "learning_rate": 1.98084075648204e-05,
+      "loss": 0.714734411239624,
+      "step": 890
+    },
+    {
+      "epoch": 0.1601423487544484,
+      "grad_norm": 0.364044189453125,
+      "learning_rate": 1.980282422288629e-05,
+      "loss": 0.7451518535614013,
+      "step": 900
+    },
+    {
+      "epoch": 0.1619217081850534,
+      "grad_norm": 0.4181061387062073,
+      "learning_rate": 1.9797161500691496e-05,
+      "loss": 0.7484359741210938,
+      "step": 910
+    },
+    {
+      "epoch": 0.16370106761565836,
+      "grad_norm": 0.3803650140762329,
+      "learning_rate": 1.9791419444091006e-05,
+      "loss": 0.7413453578948974,
+      "step": 920
+    },
+    {
+      "epoch": 0.16548042704626334,
+      "grad_norm": 0.3450170159339905,
+      "learning_rate": 1.9785598099582225e-05,
+      "loss": 0.732274341583252,
+      "step": 930
+    },
+    {
+      "epoch": 0.16725978647686832,
+      "grad_norm": 0.31141409277915955,
+      "learning_rate": 1.9779697514304624e-05,
+      "loss": 0.768674898147583,
+      "step": 940
+    },
+    {
+      "epoch": 0.16903914590747332,
+      "grad_norm": 0.33961221575737,
+      "learning_rate": 1.977371773603932e-05,
+      "loss": 0.7539153099060059,
+      "step": 950
+    },
+    {
+      "epoch": 0.1708185053380783,
+      "grad_norm": 0.3468526303768158,
+      "learning_rate": 1.9767658813208725e-05,
+      "loss": 0.6995216369628906,
+      "step": 960
+    },
+    {
+      "epoch": 0.17259786476868327,
+      "grad_norm": 0.3776821196079254,
+      "learning_rate": 1.976152079487614e-05,
+      "loss": 0.7344133853912354,
+      "step": 970
+    },
+    {
+      "epoch": 0.17437722419928825,
+      "grad_norm": 0.43669673800468445,
+      "learning_rate": 1.9755303730745344e-05,
+      "loss": 0.7037209510803223,
+      "step": 980
+    },
+    {
+      "epoch": 0.17615658362989323,
+      "grad_norm": 0.43197065591812134,
+      "learning_rate": 1.9749007671160223e-05,
+      "loss": 0.7632620334625244,
+      "step": 990
+    },
+    {
+      "epoch": 0.17793594306049823,
+      "grad_norm": 0.36352500319480896,
+      "learning_rate": 1.9742632667104332e-05,
+      "loss": 0.7493629455566406,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17793594306049823,
+      "eval_loss": 0.7127183079719543,
+      "eval_runtime": 407.5587,
+      "eval_samples_per_second": 12.339,
+      "eval_steps_per_second": 6.171,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1797153024911032,
+      "grad_norm": 0.4177298843860626,
+      "learning_rate": 1.9736178770200492e-05,
+      "loss": 0.7822850227355957,
+      "step": 1010
+    },
+    {
+      "epoch": 0.18149466192170818,
+      "grad_norm": 0.4624420404434204,
+      "learning_rate": 1.972964603271038e-05,
+      "loss": 0.7450732707977294,
+      "step": 1020
+    },
+    {
+      "epoch": 0.18327402135231316,
+      "grad_norm": 0.35002920031547546,
+      "learning_rate": 1.97230345075341e-05,
+      "loss": 0.7509373188018799,
+      "step": 1030
+    },
+    {
+      "epoch": 0.18505338078291814,
+      "grad_norm": 0.4136241674423218,
+      "learning_rate": 1.9716344248209754e-05,
+      "loss": 0.6727419853210449,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18683274021352314,
+      "grad_norm": 0.41807156801223755,
+      "learning_rate": 1.9709575308913004e-05,
+      "loss": 0.7258425712585449,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18861209964412812,
+      "grad_norm": 0.4383244514465332,
+      "learning_rate": 1.9702727744456645e-05,
+      "loss": 0.7470430374145508,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1903914590747331,
+      "grad_norm": 0.44305190443992615,
+      "learning_rate": 1.969580161029015e-05,
+      "loss": 0.6925637722015381,
+      "step": 1070
+    },
+    {
+      "epoch": 0.19217081850533807,
+      "grad_norm": 0.45914319157600403,
+      "learning_rate": 1.9688796962499228e-05,
+      "loss": 0.7320804595947266,
+      "step": 1080
+    },
+    {
+      "epoch": 0.19395017793594305,
+      "grad_norm": 0.37220245599746704,
+      "learning_rate": 1.9681713857805367e-05,
+      "loss": 0.6863605499267578,
+      "step": 1090
+    },
+    {
+      "epoch": 0.19572953736654805,
+      "grad_norm": 0.473320871591568,
+      "learning_rate": 1.9674552353565374e-05,
+      "loss": 0.7069521427154541,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19750889679715303,
+      "grad_norm": 0.34435564279556274,
+      "learning_rate": 1.9667312507770905e-05,
+      "loss": 0.7488323211669922,
+      "step": 1110
+    },
+    {
+      "epoch": 0.199288256227758,
+      "grad_norm": 0.4495692551136017,
+      "learning_rate": 1.9659994379048015e-05,
+      "loss": 0.7306941032409668,
+      "step": 1120
+    },
+    {
+      "epoch": 0.20106761565836298,
+      "grad_norm": 0.37990689277648926,
+      "learning_rate": 1.9652598026656666e-05,
+      "loss": 0.6781065464019775,
+      "step": 1130
+    },
+    {
+      "epoch": 0.20284697508896798,
+      "grad_norm": 0.374970942735672,
+      "learning_rate": 1.9645123510490242e-05,
+      "loss": 0.732900333404541,
+      "step": 1140
+    },
+    {
+      "epoch": 0.20462633451957296,
+      "grad_norm": 0.42331087589263916,
+      "learning_rate": 1.963757089107508e-05,
+      "loss": 0.7150296211242676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.20640569395017794,
+      "grad_norm": 0.39357990026474,
+      "learning_rate": 1.962994022956998e-05,
+      "loss": 0.70610032081604,
+      "step": 1160
+    },
+    {
+      "epoch": 0.20818505338078291,
+      "grad_norm": 0.3616408407688141,
+      "learning_rate": 1.9622231587765688e-05,
+      "loss": 0.6808771610260009,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2099644128113879,
+      "grad_norm": 0.4579455554485321,
+      "learning_rate": 1.9614445028084424e-05,
+      "loss": 0.7024923801422119,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2117437722419929,
+      "grad_norm": 0.47354868054389954,
+      "learning_rate": 1.9606580613579352e-05,
+      "loss": 0.7523046493530273,
+      "step": 1190
+    },
+    {
+      "epoch": 0.21352313167259787,
+      "grad_norm": 0.3599710166454315,
+      "learning_rate": 1.9598638407934096e-05,
+      "loss": 0.7396236419677734,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21530249110320285,
+      "grad_norm": 0.37311187386512756,
+      "learning_rate": 1.959061847546219e-05,
+      "loss": 0.6676182746887207,
+      "step": 1210
+    },
+    {
+      "epoch": 0.21708185053380782,
+      "grad_norm": 0.4571494460105896,
+      "learning_rate": 1.9582520881106585e-05,
+      "loss": 0.7176971435546875,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2188612099644128,
+      "grad_norm": 0.39556118845939636,
+      "learning_rate": 1.9574345690439113e-05,
+      "loss": 0.6899125576019287,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2206405693950178,
+      "grad_norm": 0.4152670204639435,
+      "learning_rate": 1.9566092969659964e-05,
+      "loss": 0.7056239128112793,
+      "step": 1240
+    },
+    {
+      "epoch": 0.22241992882562278,
+      "grad_norm": 0.5533052086830139,
+      "learning_rate": 1.9557762785597133e-05,
+      "loss": 0.7325549602508545,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22419928825622776,
+      "grad_norm": 0.457738995552063,
+      "learning_rate": 1.9549355205705895e-05,
+      "loss": 0.7097938060760498,
+      "step": 1260
+    },
+    {
+      "epoch": 0.22597864768683273,
+      "grad_norm": 0.4812074303627014,
+      "learning_rate": 1.9540870298068247e-05,
+      "loss": 0.6999053478240966,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2277580071174377,
+      "grad_norm": 0.42034661769866943,
+      "learning_rate": 1.9532308131392365e-05,
+      "loss": 0.7124747276306153,
+      "step": 1280
+    },
+    {
+      "epoch": 0.22953736654804271,
+      "grad_norm": 0.44799792766571045,
+      "learning_rate": 1.9523668775012053e-05,
+      "loss": 0.7096034049987793,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2313167259786477,
+      "grad_norm": 0.4658997654914856,
+      "learning_rate": 1.9514952298886157e-05,
+      "loss": 0.7080921649932861,
+      "step": 1300
+    },
+    {
+      "epoch": 0.23309608540925267,
+      "grad_norm": 0.5269479751586914,
+      "learning_rate": 1.9506158773598035e-05,
+      "loss": 0.6962251663208008,
+      "step": 1310
+    },
+    {
+      "epoch": 0.23487544483985764,
+      "grad_norm": 0.45521607995033264,
+      "learning_rate": 1.9497288270354944e-05,
+      "loss": 0.7323726177215576,
+      "step": 1320
+    },
+    {
+      "epoch": 0.23665480427046262,
+      "grad_norm": 0.4322509169578552,
+      "learning_rate": 1.9488340860987504e-05,
+      "loss": 0.7227589607238769,
+      "step": 1330
+    },
+    {
+      "epoch": 0.23843416370106763,
+      "grad_norm": 0.47730758786201477,
+      "learning_rate": 1.9479316617949084e-05,
+      "loss": 0.702051305770874,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2402135231316726,
+      "grad_norm": 0.42763814330101013,
+      "learning_rate": 1.9470215614315232e-05,
+      "loss": 0.7380130767822266,
+      "step": 1350
+    },
+    {
+      "epoch": 0.24199288256227758,
+      "grad_norm": 0.4630064070224762,
+      "learning_rate": 1.9461037923783087e-05,
+      "loss": 0.7470481395721436,
+      "step": 1360
+    },
+    {
+      "epoch": 0.24377224199288255,
+      "grad_norm": 0.367767870426178,
+      "learning_rate": 1.9451783620670767e-05,
+      "loss": 0.7009376049041748,
+      "step": 1370
+    },
+    {
+      "epoch": 0.24555160142348753,
+      "grad_norm": 0.38925161957740784,
+      "learning_rate": 1.9442452779916775e-05,
+      "loss": 0.7043869972229004,
+      "step": 1380
+    },
+    {
+      "epoch": 0.24733096085409254,
+      "grad_norm": 0.37959718704223633,
+      "learning_rate": 1.943304547707939e-05,
+      "loss": 0.7024062633514404,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2491103202846975,
+      "grad_norm": 0.3639119267463684,
+      "learning_rate": 1.9423561788336073e-05,
+      "loss": 0.7025011539459228,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2508896797153025,
+      "grad_norm": 0.5123258233070374,
+      "learning_rate": 1.9414001790482815e-05,
+      "loss": 0.7173181533813476,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2526690391459075,
+      "grad_norm": 0.5145444273948669,
+      "learning_rate": 1.940436556093355e-05,
+      "loss": 0.6975203514099121,
+      "step": 1420
+    },
+    {
+      "epoch": 0.25444839857651247,
+      "grad_norm": 0.41073641180992126,
+      "learning_rate": 1.93946531777195e-05,
+      "loss": 0.7116940021514893,
+      "step": 1430
+    },
+    {
+      "epoch": 0.25622775800711745,
+      "grad_norm": 0.5241482853889465,
+      "learning_rate": 1.9384864719488562e-05,
+      "loss": 0.7322525024414063,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2580071174377224,
+      "grad_norm": 0.41522547602653503,
+      "learning_rate": 1.9375000265504673e-05,
+      "loss": 0.7099958419799804,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2597864768683274,
+      "grad_norm": 0.4306512773036957,
+      "learning_rate": 1.9365059895647146e-05,
+      "loss": 0.6963861465454102,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2615658362989324,
+      "grad_norm": 0.4515567421913147,
+      "learning_rate": 1.935504369041004e-05,
+      "loss": 0.7317886829376221,
+      "step": 1470
+    },
+    {
+      "epoch": 0.26334519572953735,
+      "grad_norm": 0.46133843064308167,
+      "learning_rate": 1.9344951730901523e-05,
+      "loss": 0.6936720371246338,
+      "step": 1480
+    },
+    {
+      "epoch": 0.26512455516014233,
+      "grad_norm": 0.4332071840763092,
+      "learning_rate": 1.933478409884317e-05,
+      "loss": 0.7092292308807373,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2669039145907473,
+      "grad_norm": 0.44672346115112305,
+      "learning_rate": 1.9324540876569356e-05,
+      "loss": 0.7138745784759521,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2669039145907473,
+      "eval_loss": 0.6928849816322327,
+      "eval_runtime": 409.2608,
+      "eval_samples_per_second": 12.288,
+      "eval_steps_per_second": 6.145,
+      "step": 1500
+    },
+    {
+      "epoch": 0.26868327402135234,
+      "grad_norm": 0.4207611680030823,
+      "learning_rate": 1.9314222147026538e-05,
+      "loss": 0.682267141342163,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2704626334519573,
+      "grad_norm": 0.4715620279312134,
+      "learning_rate": 1.9303827993772627e-05,
+      "loss": 0.6928095340728759,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2722419928825623,
+      "grad_norm": 0.5061793327331543,
+      "learning_rate": 1.9293358500976284e-05,
+      "loss": 0.7393241405487061,
+      "step": 1530
+    },
+    {
+      "epoch": 0.27402135231316727,
+      "grad_norm": 0.46995067596435547,
+      "learning_rate": 1.9282813753416247e-05,
+      "loss": 0.7240311622619628,
+      "step": 1540
+    },
+    {
+      "epoch": 0.27580071174377224,
+      "grad_norm": 0.35846948623657227,
+      "learning_rate": 1.927219383648064e-05,
+      "loss": 0.6961266994476318,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2775800711743772,
+      "grad_norm": 0.4667574465274811,
+      "learning_rate": 1.9261498836166297e-05,
+      "loss": 0.7295796871185303,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2793594306049822,
+      "grad_norm": 0.5029377341270447,
+      "learning_rate": 1.9250728839078043e-05,
+      "loss": 0.7046424865722656,
+      "step": 1570
+    },
+    {
+      "epoch": 0.28113879003558717,
+      "grad_norm": 0.46402570605278015,
+      "learning_rate": 1.9239883932428002e-05,
+      "loss": 0.7341272830963135,
+      "step": 1580
+    },
+    {
+      "epoch": 0.28291814946619215,
+      "grad_norm": 0.45273569226264954,
+      "learning_rate": 1.9228964204034906e-05,
+      "loss": 0.7231084823608398,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2846975088967972,
+      "grad_norm": 0.44853758811950684,
+      "learning_rate": 1.9217969742323358e-05,
+      "loss": 0.7228631019592285,
+      "step": 1600
+    },
+    {
+      "epoch": 0.28647686832740216,
+      "grad_norm": 0.4073372483253479,
+      "learning_rate": 1.9206900636323138e-05,
+      "loss": 0.7120136737823486,
+      "step": 1610
+    },
+    {
+      "epoch": 0.28825622775800713,
+      "grad_norm": 0.4837400019168854,
+      "learning_rate": 1.9195756975668463e-05,
+      "loss": 0.7246061325073242,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2900355871886121,
+      "grad_norm": 0.48374027013778687,
+      "learning_rate": 1.918453885059728e-05,
+      "loss": 0.6975275039672851,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2918149466192171,
+      "grad_norm": 0.40771543979644775,
+      "learning_rate": 1.9173246351950515e-05,
+      "loss": 0.7376579284667969,
+      "step": 1640
+    },
+    {
+      "epoch": 0.29359430604982206,
+      "grad_norm": 0.47605007886886597,
+      "learning_rate": 1.916187957117136e-05,
+      "loss": 0.7509649753570556,
+      "step": 1650
+    },
+    {
+      "epoch": 0.29537366548042704,
+      "grad_norm": 0.4246625006198883,
+      "learning_rate": 1.9150438600304514e-05,
+      "loss": 0.7086214065551758,
+      "step": 1660
+    },
+    {
+      "epoch": 0.297153024911032,
+      "grad_norm": 0.4986574053764343,
+      "learning_rate": 1.9138923531995448e-05,
+      "loss": 0.6780657768249512,
+      "step": 1670
+    },
+    {
+      "epoch": 0.298932384341637,
+      "grad_norm": 0.5241477489471436,
+      "learning_rate": 1.912733445948965e-05,
+      "loss": 0.7178135395050049,
+      "step": 1680
+    },
+    {
+      "epoch": 0.30071174377224197,
+      "grad_norm": 0.48822927474975586,
+      "learning_rate": 1.9115671476631865e-05,
+      "loss": 0.6914261817932129,
+      "step": 1690
+    },
+    {
+      "epoch": 0.302491103202847,
+      "grad_norm": 0.4603191018104553,
+      "learning_rate": 1.910393467786535e-05,
+      "loss": 0.7189798355102539,
+      "step": 1700
+    },
+    {
+      "epoch": 0.304270462633452,
+      "grad_norm": 0.5957316160202026,
+      "learning_rate": 1.90921241582311e-05,
+      "loss": 0.6929316520690918,
+      "step": 1710
+    },
+    {
+      "epoch": 0.30604982206405695,
+      "grad_norm": 0.4563639163970947,
+      "learning_rate": 1.9080240013367075e-05,
+      "loss": 0.6874090194702148,
+      "step": 1720
+    },
+    {
+      "epoch": 0.30782918149466193,
+      "grad_norm": 0.4933296740055084,
+      "learning_rate": 1.9068282339507433e-05,
+      "loss": 0.7047487258911133,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3096085409252669,
+      "grad_norm": 0.46199893951416016,
+      "learning_rate": 1.9056251233481747e-05,
+      "loss": 0.6585260391235351,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3113879003558719,
+      "grad_norm": 0.4224775433540344,
+      "learning_rate": 1.904414679271421e-05,
+      "loss": 0.7062453269958496,
+      "step": 1750
+    },
+    {
+      "epoch": 0.31316725978647686,
+      "grad_norm": 0.5592194199562073,
+      "learning_rate": 1.9031969115222876e-05,
+      "loss": 0.7372538089752197,
+      "step": 1760
+    },
+    {
+      "epoch": 0.31494661921708184,
+      "grad_norm": 0.447518527507782,
+      "learning_rate": 1.9019718299618836e-05,
+      "loss": 0.6815076351165772,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3167259786476868,
+      "grad_norm": 0.4872969686985016,
+      "learning_rate": 1.9007394445105433e-05,
+      "loss": 0.6925329685211181,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3185053380782918,
+      "grad_norm": 0.5796515941619873,
+      "learning_rate": 1.8994997651477457e-05,
+      "loss": 0.6911862850189209,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3202846975088968,
+      "grad_norm": 0.45499077439308167,
+      "learning_rate": 1.8982528019120335e-05,
+      "loss": 0.689606761932373,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3220640569395018,
+      "grad_norm": 0.5108110308647156,
+      "learning_rate": 1.8969985649009325e-05,
+      "loss": 0.698862886428833,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3238434163701068,
+      "grad_norm": 0.5635261535644531,
+      "learning_rate": 1.8957370642708682e-05,
+      "loss": 0.7093265533447266,
+      "step": 1820
+    },
+    {
+      "epoch": 0.32562277580071175,
+      "grad_norm": 0.5035433769226074,
+      "learning_rate": 1.8944683102370862e-05,
+      "loss": 0.6829179763793946,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3274021352313167,
+      "grad_norm": 0.5030480623245239,
+      "learning_rate": 1.8931923130735667e-05,
+      "loss": 0.6931506633758545,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3291814946619217,
+      "grad_norm": 0.42200684547424316,
+      "learning_rate": 1.891909083112943e-05,
+      "loss": 0.7182873725891114,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3309608540925267,
+      "grad_norm": 0.5390200614929199,
+      "learning_rate": 1.8906186307464168e-05,
+      "loss": 0.7040542602539063,
+      "step": 1860
+    },
+    {
+      "epoch": 0.33274021352313166,
+      "grad_norm": 0.4840771555900574,
+      "learning_rate": 1.889320966423676e-05,
+      "loss": 0.71949143409729,
+      "step": 1870
+    },
+    {
+      "epoch": 0.33451957295373663,
+      "grad_norm": 0.46363523602485657,
+      "learning_rate": 1.8880161006528075e-05,
+      "loss": 0.7095215797424317,
+      "step": 1880
+    },
+    {
+      "epoch": 0.33629893238434166,
+      "grad_norm": 0.5143831372261047,
+      "learning_rate": 1.8867040440002137e-05,
+      "loss": 0.6844244003295898,
+      "step": 1890
+    },
+    {
+      "epoch": 0.33807829181494664,
+      "grad_norm": 0.42344361543655396,
+      "learning_rate": 1.8853848070905264e-05,
+      "loss": 0.7266733169555664,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3398576512455516,
+      "grad_norm": 0.5283271074295044,
+      "learning_rate": 1.884058400606521e-05,
+      "loss": 0.6759650707244873,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3416370106761566,
+      "grad_norm": 0.6041154861450195,
+      "learning_rate": 1.88272483528903e-05,
+      "loss": 0.7138604164123535,
+      "step": 1920
+    },
+    {
+      "epoch": 0.34341637010676157,
+      "grad_norm": 0.4607132077217102,
+      "learning_rate": 1.8813841219368562e-05,
+      "loss": 0.7108243465423584,
+      "step": 1930
+    },
+    {
+      "epoch": 0.34519572953736655,
+      "grad_norm": 0.4770022928714752,
+      "learning_rate": 1.880036271406684e-05,
+      "loss": 0.700477123260498,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3469750889679715,
+      "grad_norm": 0.5347304940223694,
+      "learning_rate": 1.8786812946129934e-05,
+      "loss": 0.7176999092102051,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3487544483985765,
+      "grad_norm": 0.4638780355453491,
+      "learning_rate": 1.8773192025279712e-05,
+      "loss": 0.6859623908996582,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3505338078291815,
+      "grad_norm": 0.4621883034706116,
+      "learning_rate": 1.87595000618142e-05,
+      "loss": 0.6809545516967773,
+      "step": 1970
+    },
+    {
+      "epoch": 0.35231316725978645,
+      "grad_norm": 0.4492979943752289,
+      "learning_rate": 1.8745737166606716e-05,
+      "loss": 0.7484791278839111,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3540925266903915,
+      "grad_norm": 0.5058820843696594,
+      "learning_rate": 1.873190345110496e-05,
+      "loss": 0.6860589027404785,
+      "step": 1990
+    },
+    {
+      "epoch": 0.35587188612099646,
+      "grad_norm": 0.467847615480423,
+      "learning_rate": 1.8717999027330114e-05,
+      "loss": 0.6946381092071533,
+      "step": 2000
+    },
+    {
+      "epoch": 0.35587188612099646,
+      "eval_loss": 0.6786053776741028,
+      "eval_runtime": 408.2053,
+      "eval_samples_per_second": 12.32,
+      "eval_steps_per_second": 6.161,
+      "step": 2000
+    },
+    {
+      "epoch": 0.35765124555160144,
+      "grad_norm": 0.4924924373626709,
+      "learning_rate": 1.870402400787593e-05,
+      "loss": 0.7328177452087402,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3594306049822064,
+      "grad_norm": 0.529029369354248,
+      "learning_rate": 1.8689978505907828e-05,
+      "loss": 0.6788232803344727,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3612099644128114,
+      "grad_norm": 0.5123940110206604,
+      "learning_rate": 1.8675862635161968e-05,
+      "loss": 0.6493151664733887,
+      "step": 2030
+    },
+    {
+      "epoch": 0.36298932384341637,
+      "grad_norm": 0.4411087930202484,
+      "learning_rate": 1.866167650994434e-05,
+      "loss": 0.6652609825134277,
+      "step": 2040
+    },
+    {
+      "epoch": 0.36476868327402134,
+      "grad_norm": 0.5237643122673035,
+      "learning_rate": 1.8647420245129822e-05,
+      "loss": 0.6707428932189942,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3665480427046263,
+      "grad_norm": 0.5926516056060791,
+      "learning_rate": 1.863309395616128e-05,
+      "loss": 0.7144505023956299,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3683274021352313,
+      "grad_norm": 0.5331065058708191,
+      "learning_rate": 1.86186977590486e-05,
+      "loss": 0.6472876071929932,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3701067615658363,
+      "grad_norm": 0.6008846163749695,
+      "learning_rate": 1.860423177036776e-05,
+      "loss": 0.6628296852111817,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3718861209964413,
+      "grad_norm": 0.49397945404052734,
+      "learning_rate": 1.85896961072599e-05,
+      "loss": 0.7032846927642822,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3736654804270463,
+      "grad_norm": 0.5322824716567993,
+      "learning_rate": 1.8575090887430354e-05,
+      "loss": 0.6903145790100098,
+      "step": 2100
+    },
+    {
+      "epoch": 0.37544483985765126,
+      "grad_norm": 0.45118311047554016,
+      "learning_rate": 1.8560416229147718e-05,
+      "loss": 0.7186954975128174,
+      "step": 2110
+    },
+    {
+      "epoch": 0.37722419928825623,
+      "grad_norm": 0.47973230481147766,
+      "learning_rate": 1.8545672251242855e-05,
+      "loss": 0.677478837966919,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3790035587188612,
+      "grad_norm": 0.5959491729736328,
+      "learning_rate": 1.8530859073107973e-05,
+      "loss": 0.7285795211791992,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3807829181494662,
+      "grad_norm": 0.5627938508987427,
+      "learning_rate": 1.851597681469565e-05,
+      "loss": 0.6971286773681641,
+      "step": 2140
+    },
+    {
+      "epoch": 0.38256227758007116,
+      "grad_norm": 0.5175175666809082,
+      "learning_rate": 1.850102559651784e-05,
+      "loss": 0.7243900775909424,
+      "step": 2150
+    },
+    {
+      "epoch": 0.38434163701067614,
+      "grad_norm": 0.5537509918212891,
+      "learning_rate": 1.848600553964491e-05,
+      "loss": 0.6653794765472412,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3861209964412811,
+      "grad_norm": 0.6098501086235046,
+      "learning_rate": 1.847091676570468e-05,
+      "loss": 0.6803691864013672,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3879003558718861,
+      "grad_norm": 0.5026776790618896,
+      "learning_rate": 1.8455759396881402e-05,
+      "loss": 0.7138524055480957,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3896797153024911,
+      "grad_norm": 0.5244454741477966,
+      "learning_rate": 1.8440533555914795e-05,
+      "loss": 0.7204444885253907,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3914590747330961,
+      "grad_norm": 0.5315225124359131,
+      "learning_rate": 1.842523936609905e-05,
+      "loss": 0.7181321144104004,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3932384341637011,
+      "grad_norm": 0.5039063096046448,
+      "learning_rate": 1.8409876951281814e-05,
+      "loss": 0.7282841205596924,
+      "step": 2210
+    },
+    {
+      "epoch": 0.39501779359430605,
+      "grad_norm": 0.47619393467903137,
+      "learning_rate": 1.8394446435863206e-05,
+      "loss": 0.7016836643218994,
+      "step": 2220
+    },
+    {
+      "epoch": 0.39679715302491103,
+      "grad_norm": 0.5733043551445007,
+      "learning_rate": 1.8378947944794806e-05,
+      "loss": 0.6978562355041504,
+      "step": 2230
+    },
+    {
+      "epoch": 0.398576512455516,
+      "grad_norm": 0.5004534125328064,
+      "learning_rate": 1.8363381603578628e-05,
+      "loss": 0.6713067531585694,
+      "step": 2240
+    },
+    {
+      "epoch": 0.400355871886121,
+      "grad_norm": 0.6110662817955017,
+      "learning_rate": 1.8347747538266133e-05,
+      "loss": 0.6862231254577636,
+      "step": 2250
+    },
+    {
+      "epoch": 0.40213523131672596,
+      "grad_norm": 0.5308374166488647,
+      "learning_rate": 1.8332045875457174e-05,
+      "loss": 0.6602601528167724,
+      "step": 2260
+    },
+    {
+      "epoch": 0.40391459074733094,
+      "grad_norm": 0.6257572174072266,
+      "learning_rate": 1.8316276742299e-05,
+      "loss": 0.6580804347991943,
+      "step": 2270
+    },
+    {
+      "epoch": 0.40569395017793597,
+      "grad_norm": 0.6403529047966003,
+      "learning_rate": 1.830044026648521e-05,
+      "loss": 0.6985883712768555,
+      "step": 2280
+    },
+    {
+      "epoch": 0.40747330960854095,
+      "grad_norm": 0.5609093308448792,
+      "learning_rate": 1.828453657625472e-05,
+      "loss": 0.7306273937225342,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4092526690391459,
+      "grad_norm": 0.5089350342750549,
+      "learning_rate": 1.8268565800390733e-05,
+      "loss": 0.6552363395690918,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4110320284697509,
+      "grad_norm": 0.5988994240760803,
+      "learning_rate": 1.8252528068219683e-05,
+      "loss": 0.6836632251739502,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4128113879003559,
+      "grad_norm": 0.558757483959198,
+      "learning_rate": 1.8236423509610207e-05,
+      "loss": 0.6945361137390137,
+      "step": 2320
+    },
+    {
+      "epoch": 0.41459074733096085,
+      "grad_norm": 0.5903414487838745,
+      "learning_rate": 1.8220252254972077e-05,
+      "loss": 0.6987195014953613,
+      "step": 2330
+    },
+    {
+      "epoch": 0.41637010676156583,
+      "grad_norm": 0.5187668800354004,
+      "learning_rate": 1.8204014435255136e-05,
+      "loss": 0.7381620407104492,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4181494661921708,
+      "grad_norm": 0.5883176922798157,
+      "learning_rate": 1.8187710181948274e-05,
+      "loss": 0.6803225994110107,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4199288256227758,
+      "grad_norm": 0.5465121269226074,
+      "learning_rate": 1.817133962707833e-05,
+      "loss": 0.6955878257751464,
+      "step": 2360
+    },
+    {
+      "epoch": 0.42170818505338076,
+      "grad_norm": 0.5035154223442078,
+      "learning_rate": 1.815490290320902e-05,
+      "loss": 0.6446046829223633,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4234875444839858,
+      "grad_norm": 0.561638593673706,
+      "learning_rate": 1.8138400143439892e-05,
+      "loss": 0.6994437217712403,
+      "step": 2380
+    },
+    {
+      "epoch": 0.42526690391459077,
+      "grad_norm": 0.547978937625885,
+      "learning_rate": 1.812183148140523e-05,
+      "loss": 0.6748724937438965,
+      "step": 2390
+    },
+    {
+      "epoch": 0.42704626334519574,
+      "grad_norm": 0.7230744957923889,
+      "learning_rate": 1.8105197051272974e-05,
+      "loss": 0.6960064888000488,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4288256227758007,
+      "grad_norm": 0.6158634424209595,
+      "learning_rate": 1.8088496987743623e-05,
+      "loss": 0.6599089622497558,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4306049822064057,
+      "grad_norm": 0.5255537629127502,
+      "learning_rate": 1.807173142604917e-05,
+      "loss": 0.7354787349700928,
+      "step": 2420
+    },
+    {
+      "epoch": 0.43238434163701067,
+      "grad_norm": 0.5891067385673523,
+      "learning_rate": 1.8054900501951988e-05,
+      "loss": 0.6725128650665283,
+      "step": 2430
+    },
+    {
+      "epoch": 0.43416370106761565,
+      "grad_norm": 0.6183168888092041,
+      "learning_rate": 1.8038004351743726e-05,
+      "loss": 0.6672306537628174,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4359430604982206,
+      "grad_norm": 0.5241208672523499,
+      "learning_rate": 1.8021043112244222e-05,
+      "loss": 0.6435032367706299,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4377224199288256,
+      "grad_norm": 0.47406890988349915,
+      "learning_rate": 1.8004016920800392e-05,
+      "loss": 0.7117865085601807,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4395017793594306,
+      "grad_norm": 0.45818889141082764,
+      "learning_rate": 1.7986925915285098e-05,
+      "loss": 0.7240960597991943,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4412811387900356,
+      "grad_norm": 0.5167953372001648,
+      "learning_rate": 1.796977023409606e-05,
+      "loss": 0.6828316211700439,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4430604982206406,
+      "grad_norm": 0.48911231756210327,
+      "learning_rate": 1.795255001615472e-05,
+      "loss": 0.6907057285308837,
+      "step": 2490
+    },
+    {
+      "epoch": 0.44483985765124556,
+      "grad_norm": 0.4384164810180664,
+      "learning_rate": 1.7935265400905107e-05,
+      "loss": 0.6647283554077148,
+      "step": 2500
+    },
+    {
+      "epoch": 0.44483985765124556,
+      "eval_loss": 0.667682409286499,
+      "eval_runtime": 408.4893,
+      "eval_samples_per_second": 12.311,
+      "eval_steps_per_second": 6.157,
+      "step": 2500
+    },
+    {
+      "epoch": 0.44661921708185054,
+      "grad_norm": 0.5545853972434998,
+      "learning_rate": 1.7917916528312735e-05,
+      "loss": 0.6821264743804931,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4483985765124555,
+      "grad_norm": 0.5834780335426331,
+      "learning_rate": 1.7900503538863446e-05,
+      "loss": 0.6818623542785645,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4501779359430605,
+      "grad_norm": 0.6038497090339661,
+      "learning_rate": 1.7883026573562278e-05,
+      "loss": 0.6842663764953614,
+      "step": 2530
+    },
+    {
+      "epoch": 0.45195729537366547,
+      "grad_norm": 0.5673312544822693,
+      "learning_rate": 1.7865485773932336e-05,
+      "loss": 0.6788026809692382,
+      "step": 2540
+    },
+    {
+      "epoch": 0.45373665480427045,
+      "grad_norm": 0.6115579605102539,
+      "learning_rate": 1.7847881282013623e-05,
+      "loss": 0.724776029586792,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4555160142348754,
+      "grad_norm": 0.6251904964447021,
+      "learning_rate": 1.7830213240361916e-05,
+      "loss": 0.6907171249389649,
+      "step": 2560
+    },
+    {
+      "epoch": 0.45729537366548045,
+      "grad_norm": 0.5565618872642517,
+      "learning_rate": 1.7812481792047587e-05,
+      "loss": 0.6535940647125245,
+      "step": 2570
+    },
+    {
+      "epoch": 0.45907473309608543,
+      "grad_norm": 0.5508270263671875,
+      "learning_rate": 1.7794687080654462e-05,
+      "loss": 0.6779204845428467,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4608540925266904,
+      "grad_norm": 0.6121755242347717,
+      "learning_rate": 1.777682925027865e-05,
+      "loss": 0.6833740234375,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4626334519572954,
+      "grad_norm": 0.5710318684577942,
+      "learning_rate": 1.7758908445527376e-05,
+      "loss": 0.6820470333099365,
+      "step": 2600
+    },
+    {
+      "epoch": 0.46441281138790036,
+      "grad_norm": 0.624721884727478,
+      "learning_rate": 1.774092481151782e-05,
+      "loss": 0.7131676197052002,
+      "step": 2610
+    },
+    {
+      "epoch": 0.46619217081850534,
+      "grad_norm": 0.5539907813072205,
+      "learning_rate": 1.7722878493875922e-05,
+      "loss": 0.6834908962249756,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4679715302491103,
+      "grad_norm": 0.7775362730026245,
+      "learning_rate": 1.7704769638735225e-05,
+      "loss": 0.7060842990875245,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4697508896797153,
+      "grad_norm": 0.5198112726211548,
+      "learning_rate": 1.7686598392735678e-05,
+      "loss": 0.6912449836730957,
+      "step": 2640
+    },
+    {
+      "epoch": 0.47153024911032027,
+      "grad_norm": 0.5231760144233704,
+      "learning_rate": 1.766836490302245e-05,
+      "loss": 0.7053235054016114,
+      "step": 2650
+    },
+    {
+      "epoch": 0.47330960854092524,
+      "grad_norm": 0.7802064418792725,
+      "learning_rate": 1.7650069317244743e-05,
+      "loss": 0.6698042392730713,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4750889679715303,
+      "grad_norm": 0.5244280099868774,
+      "learning_rate": 1.763171178355459e-05,
+      "loss": 0.6572854042053222,
+      "step": 2670
+    },
+    {
+      "epoch": 0.47686832740213525,
+      "grad_norm": 0.6323163509368896,
+      "learning_rate": 1.761329245060567e-05,
+      "loss": 0.6909306526184082,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4786476868327402,
+      "grad_norm": 0.636985182762146,
+      "learning_rate": 1.7594811467552076e-05,
+      "loss": 0.7432591915130615,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4804270462633452,
+      "grad_norm": 0.5935755372047424,
+      "learning_rate": 1.7576268984047146e-05,
+      "loss": 0.6589434146881104,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4822064056939502,
+      "grad_norm": 0.5581539869308472,
+      "learning_rate": 1.7557665150242205e-05,
+      "loss": 0.6235795974731445,
+      "step": 2710
+    },
+    {
+      "epoch": 0.48398576512455516,
+      "grad_norm": 0.5456061363220215,
+      "learning_rate": 1.7539000116785402e-05,
+      "loss": 0.6965986251831054,
+      "step": 2720
+    },
+    {
+      "epoch": 0.48576512455516013,
+      "grad_norm": 0.6723548769950867,
+      "learning_rate": 1.752027403482043e-05,
+      "loss": 0.6792643070220947,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4875444839857651,
+      "grad_norm": 0.5704858899116516,
+      "learning_rate": 1.7501487055985364e-05,
+      "loss": 0.6697192192077637,
+      "step": 2740
+    },
+    {
+      "epoch": 0.4893238434163701,
+      "grad_norm": 0.559579074382782,
+      "learning_rate": 1.7482639332411386e-05,
+      "loss": 0.6839393615722656,
+      "step": 2750
+    },
+    {
+      "epoch": 0.49110320284697506,
+      "grad_norm": 0.6328978538513184,
+      "learning_rate": 1.7463731016721574e-05,
+      "loss": 0.6757444858551025,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4928825622775801,
+      "grad_norm": 0.5843029022216797,
+      "learning_rate": 1.744476226202966e-05,
+      "loss": 0.6616747379302979,
+      "step": 2770
+    },
+    {
+      "epoch": 0.49466192170818507,
+      "grad_norm": 0.593262791633606,
+      "learning_rate": 1.7425733221938802e-05,
+      "loss": 0.6773699283599853,
+      "step": 2780
+    },
+    {
+      "epoch": 0.49644128113879005,
+      "grad_norm": 0.4597082734107971,
+      "learning_rate": 1.740664405054032e-05,
+      "loss": 0.6922731876373291,
+      "step": 2790
+    },
+    {
+      "epoch": 0.498220640569395,
+      "grad_norm": 0.5115553140640259,
+      "learning_rate": 1.7387494902412462e-05,
+      "loss": 0.6650360107421875,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5256341695785522,
+      "learning_rate": 1.7368285932619152e-05,
+      "loss": 0.6744340896606446,
+      "step": 2810
+    },
+    {
+      "epoch": 0.501779359430605,
+      "grad_norm": 0.5476647019386292,
+      "learning_rate": 1.734901729670873e-05,
+      "loss": 0.6778338432312012,
+      "step": 2820
+    },
+    {
+      "epoch": 0.50355871886121,
+      "grad_norm": 0.5267303586006165,
+      "learning_rate": 1.7329689150712692e-05,
+      "loss": 0.6595091342926025,
+      "step": 2830
+    },
+    {
+      "epoch": 0.505338078291815,
+      "grad_norm": 0.49511632323265076,
+      "learning_rate": 1.7310301651144427e-05,
+      "loss": 0.6996967315673828,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5071174377224199,
+      "grad_norm": 0.5894142985343933,
+      "learning_rate": 1.729085495499796e-05,
+      "loss": 0.7025831699371338,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5088967971530249,
+      "grad_norm": 0.5248763561248779,
+      "learning_rate": 1.727134921974666e-05,
+      "loss": 0.6990334510803222,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5106761565836299,
+      "grad_norm": 0.6369486451148987,
+      "learning_rate": 1.7251784603341984e-05,
+      "loss": 0.6775143146514893,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5124555160142349,
+      "grad_norm": 0.7855329513549805,
+      "learning_rate": 1.723216126421219e-05,
+      "loss": 0.6654418468475342,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5142348754448398,
+      "grad_norm": 0.5383808016777039,
+      "learning_rate": 1.7212479361261047e-05,
+      "loss": 0.7224104404449463,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5160142348754448,
+      "grad_norm": 0.5875473618507385,
+      "learning_rate": 1.7192739053866568e-05,
+      "loss": 0.6993866920471191,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5177935943060499,
+      "grad_norm": 0.5973880290985107,
+      "learning_rate": 1.7172940501879702e-05,
+      "loss": 0.6883405685424805,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5195729537366548,
+      "grad_norm": 0.7815128564834595,
+      "learning_rate": 1.715308386562304e-05,
+      "loss": 0.6863636493682861,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5213523131672598,
+      "grad_norm": 0.5661780834197998,
+      "learning_rate": 1.7133169305889526e-05,
+      "loss": 0.6976628303527832,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5231316725978647,
+      "grad_norm": 0.6116606593132019,
+      "learning_rate": 1.7113196983941152e-05,
+      "loss": 0.7353427410125732,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5249110320284698,
+      "grad_norm": 0.5908172726631165,
+      "learning_rate": 1.709316706150765e-05,
+      "loss": 0.7118365287780761,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5266903914590747,
+      "grad_norm": 0.5671530365943909,
+      "learning_rate": 1.707307970078518e-05,
+      "loss": 0.6777832508087158,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5284697508896797,
+      "grad_norm": 0.635502815246582,
+      "learning_rate": 1.7052935064435023e-05,
+      "loss": 0.6588366985321045,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5302491103202847,
+      "grad_norm": 0.674394428730011,
+      "learning_rate": 1.7032733315582254e-05,
+      "loss": 0.6987817764282227,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5320284697508897,
+      "grad_norm": 0.6007933020591736,
+      "learning_rate": 1.7012474617814433e-05,
+      "loss": 0.6640087604522705,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5338078291814946,
+      "grad_norm": 0.5124489665031433,
+      "learning_rate": 1.6992159135180283e-05,
+      "loss": 0.642765474319458,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5338078291814946,
+      "eval_loss": 0.6583885550498962,
+      "eval_runtime": 409.5482,
+      "eval_samples_per_second": 12.279,
+      "eval_steps_per_second": 6.141,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5355871886120996,
+      "grad_norm": 0.6412521004676819,
+      "learning_rate": 1.6971787032188336e-05,
+      "loss": 0.6574789047241211,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5373665480427047,
+      "grad_norm": 0.5690603852272034,
+      "learning_rate": 1.6951358473805633e-05,
+      "loss": 0.7070611476898193,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5391459074733096,
+      "grad_norm": 0.5345459580421448,
+      "learning_rate": 1.6930873625456362e-05,
+      "loss": 0.6888082027435303,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5409252669039146,
+      "grad_norm": 0.67616868019104,
+      "learning_rate": 1.6910332653020536e-05,
+      "loss": 0.6946770668029785,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5427046263345195,
+      "grad_norm": 0.6721035838127136,
+      "learning_rate": 1.6889735722832643e-05,
+      "loss": 0.6662865161895752,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5444839857651246,
+      "grad_norm": 0.6241431832313538,
+      "learning_rate": 1.6869083001680304e-05,
+      "loss": 0.6311937808990479,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5462633451957295,
+      "grad_norm": 0.6454595923423767,
+      "learning_rate": 1.6848374656802912e-05,
+      "loss": 0.6600630283355713,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5480427046263345,
+      "grad_norm": 0.5625023245811462,
+      "learning_rate": 1.6827610855890278e-05,
+      "loss": 0.7191053867340088,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5498220640569395,
+      "grad_norm": 0.6803082823753357,
+      "learning_rate": 1.6806791767081296e-05,
+      "loss": 0.6553170680999756,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5516014234875445,
+      "grad_norm": 0.6465727686882019,
+      "learning_rate": 1.6785917558962552e-05,
+      "loss": 0.6733019351959229,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5533807829181495,
+      "grad_norm": 0.5595579743385315,
+      "learning_rate": 1.6764988400566973e-05,
+      "loss": 0.6531811714172363,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5551601423487544,
+      "grad_norm": 0.6368974447250366,
+      "learning_rate": 1.6744004461372455e-05,
+      "loss": 0.6765477657318115,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5569395017793595,
+      "grad_norm": 0.5315082669258118,
+      "learning_rate": 1.67229659113005e-05,
+      "loss": 0.6876490116119385,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5587188612099644,
+      "grad_norm": 0.6011325716972351,
+      "learning_rate": 1.6701872920714822e-05,
+      "loss": 0.671757698059082,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5604982206405694,
+      "grad_norm": 0.5254577994346619,
+      "learning_rate": 1.6680725660419987e-05,
+      "loss": 0.6709373950958252,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5622775800711743,
+      "grad_norm": 0.5257652401924133,
+      "learning_rate": 1.6659524301660014e-05,
+      "loss": 0.6314177513122559,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5640569395017794,
+      "grad_norm": 0.577984094619751,
+      "learning_rate": 1.6638269016116995e-05,
+      "loss": 0.6948809623718262,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5658362989323843,
+      "grad_norm": 0.5697906613349915,
+      "learning_rate": 1.6616959975909706e-05,
+      "loss": 0.6561762809753418,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5676156583629893,
+      "grad_norm": 0.5052287578582764,
+      "learning_rate": 1.6595597353592216e-05,
+      "loss": 0.6577863693237305,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5693950177935944,
+      "grad_norm": 0.5690486431121826,
+      "learning_rate": 1.6574181322152477e-05,
+      "loss": 0.6753673553466797,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5711743772241993,
+      "grad_norm": 0.6763755679130554,
+      "learning_rate": 1.6552712055010935e-05,
+      "loss": 0.6881022453308105,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5729537366548043,
+      "grad_norm": 0.7030187249183655,
+      "learning_rate": 1.6531189726019127e-05,
+      "loss": 0.6971624851226806,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5747330960854092,
+      "grad_norm": 0.5229400992393494,
+      "learning_rate": 1.6509614509458263e-05,
+      "loss": 0.6652966976165772,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5765124555160143,
+      "grad_norm": 0.5338436961174011,
+      "learning_rate": 1.6487986580037822e-05,
+      "loss": 0.6946625709533691,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5782918149466192,
+      "grad_norm": 0.5740528702735901,
+      "learning_rate": 1.646630611289414e-05,
+      "loss": 0.6570149421691894,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5800711743772242,
+      "grad_norm": 0.5782309174537659,
+      "learning_rate": 1.6444573283588977e-05,
+      "loss": 0.657336950302124,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5818505338078291,
+      "grad_norm": 0.5372493863105774,
+      "learning_rate": 1.6422788268108112e-05,
+      "loss": 0.7015529632568359,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5836298932384342,
+      "grad_norm": 0.734725832939148,
+      "learning_rate": 1.6400951242859915e-05,
+      "loss": 0.7008133888244629,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5854092526690391,
+      "grad_norm": 0.6092361211776733,
+      "learning_rate": 1.6379062384673914e-05,
+      "loss": 0.6785021305084229,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5871886120996441,
+      "grad_norm": 0.5179940462112427,
+      "learning_rate": 1.6357121870799357e-05,
+      "loss": 0.6846163749694825,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5889679715302492,
+      "grad_norm": 0.6822018027305603,
+      "learning_rate": 1.633512987890379e-05,
+      "loss": 0.6979034423828125,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5907473309608541,
+      "grad_norm": 0.6708216667175293,
+      "learning_rate": 1.631308658707161e-05,
+      "loss": 0.616365623474121,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5925266903914591,
+      "grad_norm": 0.5766979455947876,
+      "learning_rate": 1.6290992173802628e-05,
+      "loss": 0.6930478096008301,
+      "step": 3330
+    },
+    {
+      "epoch": 0.594306049822064,
+      "grad_norm": 0.5902812480926514,
+      "learning_rate": 1.6268846818010615e-05,
+      "loss": 0.6487626552581787,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5960854092526691,
+      "grad_norm": 0.45683762431144714,
+      "learning_rate": 1.6246650699021866e-05,
+      "loss": 0.670246696472168,
+      "step": 3350
+    },
+    {
+      "epoch": 0.597864768683274,
+      "grad_norm": 0.4878489077091217,
+      "learning_rate": 1.6224403996573743e-05,
+      "loss": 0.6784106254577636,
+      "step": 3360
+    },
+    {
+      "epoch": 0.599644128113879,
+      "grad_norm": 0.6697226166725159,
+      "learning_rate": 1.62021068908132e-05,
+      "loss": 0.7080551147460937,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6014234875444839,
+      "grad_norm": 0.6667253971099854,
+      "learning_rate": 1.6179759562295356e-05,
+      "loss": 0.6681561470031738,
+      "step": 3380
+    },
+    {
+      "epoch": 0.603202846975089,
+      "grad_norm": 0.560409426689148,
+      "learning_rate": 1.6157362191982025e-05,
+      "loss": 0.7013211727142334,
+      "step": 3390
+    },
+    {
+      "epoch": 0.604982206405694,
+      "grad_norm": 0.6660729050636292,
+      "learning_rate": 1.6134914961240224e-05,
+      "loss": 0.6561143398284912,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6067615658362989,
+      "grad_norm": 0.5322085618972778,
+      "learning_rate": 1.6112418051840745e-05,
+      "loss": 0.6231056213378906,
+      "step": 3410
+    },
+    {
+      "epoch": 0.608540925266904,
+      "grad_norm": 0.6110619902610779,
+      "learning_rate": 1.6089871645956644e-05,
+      "loss": 0.696910810470581,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6103202846975089,
+      "grad_norm": 0.7193135619163513,
+      "learning_rate": 1.6067275926161792e-05,
+      "loss": 0.6709835052490234,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6120996441281139,
+      "grad_norm": 0.5821454524993896,
+      "learning_rate": 1.6044631075429406e-05,
+      "loss": 0.6750143527984619,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6138790035587188,
+      "grad_norm": 0.6266823410987854,
+      "learning_rate": 1.6021937277130516e-05,
+      "loss": 0.6660190105438233,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6156583629893239,
+      "grad_norm": 0.6361811757087708,
+      "learning_rate": 1.5999194715032543e-05,
+      "loss": 0.6241551399230957,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6174377224199288,
+      "grad_norm": 0.5690405964851379,
+      "learning_rate": 1.5976403573297767e-05,
+      "loss": 0.6768513679504394,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6192170818505338,
+      "grad_norm": 0.6438109278678894,
+      "learning_rate": 1.595356403648186e-05,
+      "loss": 0.6852968215942383,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6209964412811388,
+      "grad_norm": 0.5949437022209167,
+      "learning_rate": 1.5930676289532373e-05,
+      "loss": 0.6737981319427491,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6227758007117438,
+      "grad_norm": 0.6436291933059692,
+      "learning_rate": 1.590774051778726e-05,
+      "loss": 0.6780194282531739,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6227758007117438,
+      "eval_loss": 0.6505803465843201,
+      "eval_runtime": 412.7337,
+      "eval_samples_per_second": 12.185,
+      "eval_steps_per_second": 6.094,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6245551601423488,
+      "grad_norm": 0.5366589426994324,
+      "learning_rate": 1.588475690697335e-05,
+      "loss": 0.6628384590148926,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6263345195729537,
+      "grad_norm": 0.5512357950210571,
+      "learning_rate": 1.5861725643204876e-05,
+      "loss": 0.6656132221221924,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6281138790035588,
+      "grad_norm": 0.5239390134811401,
+      "learning_rate": 1.5838646912981937e-05,
+      "loss": 0.6585372447967529,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6298932384341637,
+      "grad_norm": 0.611584484577179,
+      "learning_rate": 1.5815520903188998e-05,
+      "loss": 0.6700050354003906,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6316725978647687,
+      "grad_norm": 0.5967355966567993,
+      "learning_rate": 1.5792347801093393e-05,
+      "loss": 0.6763735294342041,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6334519572953736,
+      "grad_norm": 0.5486634373664856,
+      "learning_rate": 1.576912779434379e-05,
+      "loss": 0.6555093765258789,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6352313167259787,
+      "grad_norm": 0.6155397891998291,
+      "learning_rate": 1.5745861070968667e-05,
+      "loss": 0.660044813156128,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6370106761565836,
+      "grad_norm": 0.6666173338890076,
+      "learning_rate": 1.5722547819374807e-05,
+      "loss": 0.6919036388397217,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6387900355871886,
+      "grad_norm": 0.628685474395752,
+      "learning_rate": 1.5699188228345765e-05,
+      "loss": 0.6886429786682129,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6405693950177936,
+      "grad_norm": 0.5950825810432434,
+      "learning_rate": 1.5675782487040337e-05,
+      "loss": 0.6522699356079101,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6423487544483986,
+      "grad_norm": 0.6900691390037537,
+      "learning_rate": 1.565233078499103e-05,
+      "loss": 0.6545799255371094,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6441281138790036,
+      "grad_norm": 0.6600697040557861,
+      "learning_rate": 1.5628833312102526e-05,
+      "loss": 0.690045976638794,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6459074733096085,
+      "grad_norm": 0.7045279145240784,
+      "learning_rate": 1.5605290258650144e-05,
+      "loss": 0.6638887405395508,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6476868327402135,
+      "grad_norm": 0.6415700316429138,
+      "learning_rate": 1.5581701815278302e-05,
+      "loss": 0.6451498508453369,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6494661921708185,
+      "grad_norm": 0.6846213340759277,
+      "learning_rate": 1.5558068172998977e-05,
+      "loss": 0.6113666534423828,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6512455516014235,
+      "grad_norm": 0.584511935710907,
+      "learning_rate": 1.5534389523190142e-05,
+      "loss": 0.7053666591644288,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6530249110320284,
+      "grad_norm": 0.7102410793304443,
+      "learning_rate": 1.551066605759424e-05,
+      "loss": 0.6806889533996582,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6548042704626335,
+      "grad_norm": 0.5693709254264832,
+      "learning_rate": 1.5486897968316604e-05,
+      "loss": 0.6717594623565674,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6565836298932385,
+      "grad_norm": 0.6451858878135681,
+      "learning_rate": 1.546308544782392e-05,
+      "loss": 0.69324951171875,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6583629893238434,
+      "grad_norm": 0.5983235836029053,
+      "learning_rate": 1.543922868894268e-05,
+      "loss": 0.6728285312652588,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6601423487544484,
+      "grad_norm": 0.5725272297859192,
+      "learning_rate": 1.541532788485758e-05,
+      "loss": 0.6560508728027343,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6619217081850534,
+      "grad_norm": 0.6461522579193115,
+      "learning_rate": 1.5391383229110005e-05,
+      "loss": 0.6883309364318848,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6637010676156584,
+      "grad_norm": 0.6512270569801331,
+      "learning_rate": 1.5367394915596414e-05,
+      "loss": 0.6622300624847413,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6654804270462633,
+      "grad_norm": 0.6106224656105042,
+      "learning_rate": 1.534336313856681e-05,
+      "loss": 0.6878883361816406,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6672597864768683,
+      "grad_norm": 0.6902608275413513,
+      "learning_rate": 1.5319288092623142e-05,
+      "loss": 0.6434782981872559,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6690391459074733,
+      "grad_norm": 0.5784608721733093,
+      "learning_rate": 1.5295169972717743e-05,
+      "loss": 0.6367124557495117,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6708185053380783,
+      "grad_norm": 0.6913280487060547,
+      "learning_rate": 1.5271008974151744e-05,
+      "loss": 0.630396556854248,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6725978647686833,
+      "grad_norm": 0.7483834624290466,
+      "learning_rate": 1.5246805292573487e-05,
+      "loss": 0.6295557975769043,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6743772241992882,
+      "grad_norm": 0.7149993777275085,
+      "learning_rate": 1.5222559123976962e-05,
+      "loss": 0.6769547939300538,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6761565836298933,
+      "grad_norm": 0.6534271836280823,
+      "learning_rate": 1.5198270664700187e-05,
+      "loss": 0.7023432731628418,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6779359430604982,
+      "grad_norm": 0.6507487297058105,
+      "learning_rate": 1.5173940111423657e-05,
+      "loss": 0.6808289527893067,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6797153024911032,
+      "grad_norm": 0.5977747440338135,
+      "learning_rate": 1.5149567661168715e-05,
+      "loss": 0.6398194313049317,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6814946619217082,
+      "grad_norm": 0.6036016941070557,
+      "learning_rate": 1.5125153511295989e-05,
+      "loss": 0.6317630767822265,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6832740213523132,
+      "grad_norm": 0.8481062054634094,
+      "learning_rate": 1.5100697859503762e-05,
+      "loss": 0.6741732597351074,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6850533807829181,
+      "grad_norm": 0.5988038182258606,
+      "learning_rate": 1.5076200903826391e-05,
+      "loss": 0.6638431072235107,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6868327402135231,
+      "grad_norm": 0.6785585880279541,
+      "learning_rate": 1.5051662842632709e-05,
+      "loss": 0.6523032665252686,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6886120996441281,
+      "grad_norm": 0.6749284863471985,
+      "learning_rate": 1.5027083874624392e-05,
+      "loss": 0.6517146110534668,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6903914590747331,
+      "grad_norm": 0.6880838871002197,
+      "learning_rate": 1.5002464198834383e-05,
+      "loss": 0.6895311355590821,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6921708185053381,
+      "grad_norm": 0.6296641826629639,
+      "learning_rate": 1.4977804014625257e-05,
+      "loss": 0.691303300857544,
+      "step": 3890
+    },
+    {
+      "epoch": 0.693950177935943,
+      "grad_norm": 0.62867671251297,
+      "learning_rate": 1.4953103521687612e-05,
+      "loss": 0.6549241065979003,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6957295373665481,
+      "grad_norm": 0.6810404658317566,
+      "learning_rate": 1.4928362920038455e-05,
+      "loss": 0.6266399383544922,
+      "step": 3910
+    },
+    {
+      "epoch": 0.697508896797153,
+      "grad_norm": 0.7017929553985596,
+      "learning_rate": 1.4903582410019586e-05,
+      "loss": 0.6477387905120849,
+      "step": 3920
+    },
+    {
+      "epoch": 0.699288256227758,
+      "grad_norm": 0.5611311197280884,
+      "learning_rate": 1.487876219229596e-05,
+      "loss": 0.6562673568725585,
+      "step": 3930
+    },
+    {
+      "epoch": 0.701067615658363,
+      "grad_norm": 0.5652351975440979,
+      "learning_rate": 1.4853902467854084e-05,
+      "loss": 0.6768126487731934,
+      "step": 3940
+    },
+    {
+      "epoch": 0.702846975088968,
+      "grad_norm": 0.7303992509841919,
+      "learning_rate": 1.4829003438000374e-05,
+      "loss": 0.6467350959777832,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7046263345195729,
+      "grad_norm": 0.5587636232376099,
+      "learning_rate": 1.4804065304359525e-05,
+      "loss": 0.6700319766998291,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7064056939501779,
+      "grad_norm": 0.7100324630737305,
+      "learning_rate": 1.477908826887289e-05,
+      "loss": 0.6796034812927246,
+      "step": 3970
+    },
+    {
+      "epoch": 0.708185053380783,
+      "grad_norm": 0.6912830471992493,
+      "learning_rate": 1.4754072533796833e-05,
+      "loss": 0.6907623291015625,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7099644128113879,
+      "grad_norm": 0.744030773639679,
+      "learning_rate": 1.4729018301701093e-05,
+      "loss": 0.6747735023498536,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7117437722419929,
+      "grad_norm": 0.6915676593780518,
+      "learning_rate": 1.4703925775467149e-05,
+      "loss": 0.6172840595245361,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7117437722419929,
+      "eval_loss": 0.6436223387718201,
+      "eval_runtime": 412.5315,
+      "eval_samples_per_second": 12.191,
+      "eval_steps_per_second": 6.097,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7135231316725978,
+      "grad_norm": 0.5881261229515076,
+      "learning_rate": 1.4678795158286579e-05,
+      "loss": 0.6607324123382569,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7153024911032029,
+      "grad_norm": 0.7214511036872864,
+      "learning_rate": 1.4653626653659399e-05,
+      "loss": 0.6573444366455078,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7170818505338078,
+      "grad_norm": 0.6598398089408875,
+      "learning_rate": 1.4628420465392432e-05,
+      "loss": 0.6589923858642578,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7188612099644128,
+      "grad_norm": 0.6202049255371094,
+      "learning_rate": 1.4603176797597654e-05,
+      "loss": 0.6494176387786865,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7206405693950177,
+      "grad_norm": 0.6085755825042725,
+      "learning_rate": 1.4577895854690536e-05,
+      "loss": 0.6625244140625,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7224199288256228,
+      "grad_norm": 0.799647331237793,
+      "learning_rate": 1.4552577841388388e-05,
+      "loss": 0.6426272392272949,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7241992882562278,
+      "grad_norm": 0.5650002956390381,
+      "learning_rate": 1.4527222962708714e-05,
+      "loss": 0.6512536525726318,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7259786476868327,
+      "grad_norm": 0.7105417251586914,
+      "learning_rate": 1.4501831423967531e-05,
+      "loss": 0.639466667175293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7277580071174378,
+      "grad_norm": 0.6997768878936768,
+      "learning_rate": 1.4476403430777729e-05,
+      "loss": 0.6274962425231934,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7295373665480427,
+      "grad_norm": 0.6622691750526428,
+      "learning_rate": 1.4450939189047379e-05,
+      "loss": 0.6450656890869141,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7313167259786477,
+      "grad_norm": 0.669937789440155,
+      "learning_rate": 1.4425438904978103e-05,
+      "loss": 0.6467487335205078,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7330960854092526,
+      "grad_norm": 0.6928410530090332,
+      "learning_rate": 1.4399902785063366e-05,
+      "loss": 0.6678302764892579,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7348754448398577,
+      "grad_norm": 0.7608679533004761,
+      "learning_rate": 1.4374331036086831e-05,
+      "loss": 0.6973666191101074,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7366548042704626,
+      "grad_norm": 0.7009332180023193,
+      "learning_rate": 1.4348723865120663e-05,
+      "loss": 0.6456516742706299,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7384341637010676,
+      "grad_norm": 0.7158817648887634,
+      "learning_rate": 1.4323081479523878e-05,
+      "loss": 0.6287565708160401,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7402135231316725,
+      "grad_norm": 0.5314562916755676,
+      "learning_rate": 1.4297404086940635e-05,
+      "loss": 0.6649733543395996,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7419928825622776,
+      "grad_norm": 0.6606214642524719,
+      "learning_rate": 1.4271691895298573e-05,
+      "loss": 0.6175446033477783,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7437722419928826,
+      "grad_norm": 0.618217945098877,
+      "learning_rate": 1.4245945112807133e-05,
+      "loss": 0.6604631423950196,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7455516014234875,
+      "grad_norm": 0.5399601459503174,
+      "learning_rate": 1.422016394795585e-05,
+      "loss": 0.6682997226715088,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7473309608540926,
+      "grad_norm": 0.5808553695678711,
+      "learning_rate": 1.419434860951268e-05,
+      "loss": 0.6625009059906006,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7491103202846975,
+      "grad_norm": 0.670625627040863,
+      "learning_rate": 1.416849930652231e-05,
+      "loss": 0.6778800010681152,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7508896797153025,
+      "grad_norm": 0.6508112549781799,
+      "learning_rate": 1.4142616248304459e-05,
+      "loss": 0.6265085220336915,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7526690391459074,
+      "grad_norm": 0.5993587970733643,
+      "learning_rate": 1.4116699644452182e-05,
+      "loss": 0.656840181350708,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7544483985765125,
+      "grad_norm": 0.6819363236427307,
+      "learning_rate": 1.4090749704830184e-05,
+      "loss": 0.6575029373168946,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7562277580071174,
+      "grad_norm": 0.6625942587852478,
+      "learning_rate": 1.4064766639573104e-05,
+      "loss": 0.6340457916259765,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7580071174377224,
+      "grad_norm": 0.8185866475105286,
+      "learning_rate": 1.4038750659083831e-05,
+      "loss": 0.6835246086120605,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7597864768683275,
+      "grad_norm": 0.8232684135437012,
+      "learning_rate": 1.4012701974031782e-05,
+      "loss": 0.6450761795043946,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7615658362989324,
+      "grad_norm": 0.6913644671440125,
+      "learning_rate": 1.3986620795351214e-05,
+      "loss": 0.6313485145568848,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7633451957295374,
+      "grad_norm": 0.641167163848877,
+      "learning_rate": 1.3960507334239501e-05,
+      "loss": 0.6450904846191406,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7651245551601423,
+      "grad_norm": 0.5932332873344421,
+      "learning_rate": 1.3934361802155436e-05,
+      "loss": 0.6665386199951172,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7669039145907474,
+      "grad_norm": 0.5669364929199219,
+      "learning_rate": 1.3908184410817511e-05,
+      "loss": 0.6564301013946533,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7686832740213523,
+      "grad_norm": 0.6267620325088501,
+      "learning_rate": 1.3881975372202201e-05,
+      "loss": 0.6448751449584961,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7704626334519573,
+      "grad_norm": 0.8893764615058899,
+      "learning_rate": 1.3855734898542252e-05,
+      "loss": 0.6395359516143799,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7722419928825622,
+      "grad_norm": 0.6805179119110107,
+      "learning_rate": 1.3829463202324967e-05,
+      "loss": 0.6458981990814209,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7740213523131673,
+      "grad_norm": 0.6416231393814087,
+      "learning_rate": 1.3803160496290472e-05,
+      "loss": 0.6462121963500976,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7758007117437722,
+      "grad_norm": 0.7718709707260132,
+      "learning_rate": 1.3776826993430006e-05,
+      "loss": 0.6271074295043946,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7775800711743772,
+      "grad_norm": 0.6111568212509155,
+      "learning_rate": 1.375046290698419e-05,
+      "loss": 0.6282791137695313,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7793594306049823,
+      "grad_norm": 0.7178627252578735,
+      "learning_rate": 1.3724068450441303e-05,
+      "loss": 0.6567965507507324,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7811387900355872,
+      "grad_norm": 0.6303468346595764,
+      "learning_rate": 1.3697643837535546e-05,
+      "loss": 0.6168845653533935,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7829181494661922,
+      "grad_norm": 0.6654033660888672,
+      "learning_rate": 1.3671189282245326e-05,
+      "loss": 0.6331443309783935,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7846975088967971,
+      "grad_norm": 0.7051156163215637,
+      "learning_rate": 1.3644704998791501e-05,
+      "loss": 0.6948952198028564,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7864768683274022,
+      "grad_norm": 0.5967740416526794,
+      "learning_rate": 1.361819120163567e-05,
+      "loss": 0.6140963077545166,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7882562277580071,
+      "grad_norm": 0.688831090927124,
+      "learning_rate": 1.3591648105478423e-05,
+      "loss": 0.6627942085266113,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7900355871886121,
+      "grad_norm": 0.5357785820960999,
+      "learning_rate": 1.3565075925257605e-05,
+      "loss": 0.6558830261230468,
+      "step": 4440
+    },
+    {
+      "epoch": 0.791814946619217,
+      "grad_norm": 0.6449471712112427,
+      "learning_rate": 1.3538474876146567e-05,
+      "loss": 0.6528484344482421,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7935943060498221,
+      "grad_norm": 0.6996321678161621,
+      "learning_rate": 1.3511845173552446e-05,
+      "loss": 0.6519684314727783,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7953736654804271,
+      "grad_norm": 0.6798763871192932,
+      "learning_rate": 1.348518703311439e-05,
+      "loss": 0.6224774360656739,
+      "step": 4470
+    },
+    {
+      "epoch": 0.797153024911032,
+      "grad_norm": 0.698722243309021,
+      "learning_rate": 1.3458500670701833e-05,
+      "loss": 0.6481215953826904,
+      "step": 4480
+    },
+    {
+      "epoch": 0.798932384341637,
+      "grad_norm": 0.6421968340873718,
+      "learning_rate": 1.3431786302412749e-05,
+      "loss": 0.6016243934631348,
+      "step": 4490
+    },
+    {
+      "epoch": 0.800711743772242,
+      "grad_norm": 0.6529412269592285,
+      "learning_rate": 1.3405044144571888e-05,
+      "loss": 0.6958633899688721,
+      "step": 4500
+    },
+    {
+      "epoch": 0.800711743772242,
+      "eval_loss": 0.6375713348388672,
+      "eval_runtime": 411.392,
+      "eval_samples_per_second": 12.224,
+      "eval_steps_per_second": 6.113,
+      "step": 4500
+    },
+    {
+      "epoch": 0.802491103202847,
+      "grad_norm": 0.7270268797874451,
+      "learning_rate": 1.3378274413729036e-05,
+      "loss": 0.6834945201873779,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8042704626334519,
+      "grad_norm": 0.6442169547080994,
+      "learning_rate": 1.335147732665725e-05,
+      "loss": 0.6535075187683106,
+      "step": 4520
+    },
+    {
+      "epoch": 0.806049822064057,
+      "grad_norm": 0.7817319631576538,
+      "learning_rate": 1.3324653100351117e-05,
+      "loss": 0.6588070869445801,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8078291814946619,
+      "grad_norm": 0.6025936603546143,
+      "learning_rate": 1.3297801952024983e-05,
+      "loss": 0.6654253482818604,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8096085409252669,
+      "grad_norm": 0.6981809735298157,
+      "learning_rate": 1.3270924099111204e-05,
+      "loss": 0.6565302848815918,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8113879003558719,
+      "grad_norm": 0.7100082635879517,
+      "learning_rate": 1.3244019759258378e-05,
+      "loss": 0.6451606273651123,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8131672597864769,
+      "grad_norm": 0.7197165489196777,
+      "learning_rate": 1.3217089150329589e-05,
+      "loss": 0.6707216739654541,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8149466192170819,
+      "grad_norm": 0.6494991183280945,
+      "learning_rate": 1.3190132490400642e-05,
+      "loss": 0.6538206577301026,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8167259786476868,
+      "grad_norm": 0.610467255115509,
+      "learning_rate": 1.316314999775829e-05,
+      "loss": 0.6370253562927246,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8185053380782918,
+      "grad_norm": 0.765133261680603,
+      "learning_rate": 1.3136141890898473e-05,
+      "loss": 0.6969471454620362,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8202846975088968,
+      "grad_norm": 0.6949911117553711,
+      "learning_rate": 1.3109108388524551e-05,
+      "loss": 0.6275270462036133,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8220640569395018,
+      "grad_norm": 0.6720005869865417,
+      "learning_rate": 1.3082049709545524e-05,
+      "loss": 0.6380300521850586,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8238434163701067,
+      "grad_norm": 0.7357544898986816,
+      "learning_rate": 1.3054966073074264e-05,
+      "loss": 0.6391136169433593,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8256227758007118,
+      "grad_norm": 0.6352887749671936,
+      "learning_rate": 1.3027857698425748e-05,
+      "loss": 0.6527080059051513,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8274021352313167,
+      "grad_norm": 0.6615159511566162,
+      "learning_rate": 1.3000724805115265e-05,
+      "loss": 0.6786240100860595,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8291814946619217,
+      "grad_norm": 0.7004484534263611,
+      "learning_rate": 1.2973567612856659e-05,
+      "loss": 0.6390516281127929,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8309608540925267,
+      "grad_norm": 0.7463746070861816,
+      "learning_rate": 1.294638634156053e-05,
+      "loss": 0.6174628734588623,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8327402135231317,
+      "grad_norm": 0.7666842341423035,
+      "learning_rate": 1.2919181211332474e-05,
+      "loss": 0.6169525623321533,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8345195729537367,
+      "grad_norm": 0.6636411547660828,
+      "learning_rate": 1.2891952442471274e-05,
+      "loss": 0.6446310043334961,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8362989323843416,
+      "grad_norm": 0.9001480937004089,
+      "learning_rate": 1.2864700255467148e-05,
+      "loss": 0.6689016342163085,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8380782918149466,
+      "grad_norm": 0.551397442817688,
+      "learning_rate": 1.2837424870999933e-05,
+      "loss": 0.6616122245788574,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8398576512455516,
+      "grad_norm": 0.7142363786697388,
+      "learning_rate": 1.281012650993732e-05,
+      "loss": 0.690334415435791,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8416370106761566,
+      "grad_norm": 0.701836347579956,
+      "learning_rate": 1.2782805393333054e-05,
+      "loss": 0.6755175113677978,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8434163701067615,
+      "grad_norm": 0.6752656102180481,
+      "learning_rate": 1.2755461742425147e-05,
+      "loss": 0.6132485866546631,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8451957295373665,
+      "grad_norm": 0.6739411354064941,
+      "learning_rate": 1.2728095778634094e-05,
+      "loss": 0.6823519706726074,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8469750889679716,
+      "grad_norm": 0.6728504300117493,
+      "learning_rate": 1.2700707723561064e-05,
+      "loss": 0.6264513492584228,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8487544483985765,
+      "grad_norm": 0.6379041075706482,
+      "learning_rate": 1.2673297798986118e-05,
+      "loss": 0.6372091770172119,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8505338078291815,
+      "grad_norm": 0.6896198987960815,
+      "learning_rate": 1.2645866226866405e-05,
+      "loss": 0.6795585632324219,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8523131672597865,
+      "grad_norm": 0.6732060313224792,
+      "learning_rate": 1.261841322933438e-05,
+      "loss": 0.6782153129577637,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8540925266903915,
+      "grad_norm": 0.7170758247375488,
+      "learning_rate": 1.2590939028695987e-05,
+      "loss": 0.6489102840423584,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8558718861209964,
+      "grad_norm": 0.7446131110191345,
+      "learning_rate": 1.2563443847428862e-05,
+      "loss": 0.6556113719940185,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8576512455516014,
+      "grad_norm": 0.5690045952796936,
+      "learning_rate": 1.2535927908180547e-05,
+      "loss": 0.6207235813140869,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8594306049822064,
+      "grad_norm": 0.7088342308998108,
+      "learning_rate": 1.2508391433766667e-05,
+      "loss": 0.6439788341522217,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8612099644128114,
+      "grad_norm": 0.573747992515564,
+      "learning_rate": 1.2480834647169134e-05,
+      "loss": 0.6493591785430908,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8629893238434164,
+      "grad_norm": 0.6367696523666382,
+      "learning_rate": 1.2453257771534348e-05,
+      "loss": 0.6573034286499023,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8647686832740213,
+      "grad_norm": 0.6127105355262756,
+      "learning_rate": 1.2425661030171382e-05,
+      "loss": 0.6419090747833252,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8665480427046264,
+      "grad_norm": 0.6644669771194458,
+      "learning_rate": 1.2398044646550167e-05,
+      "loss": 0.6521550178527832,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8683274021352313,
+      "grad_norm": 0.6144851446151733,
+      "learning_rate": 1.2370408844299705e-05,
+      "loss": 0.6388635158538818,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8701067615658363,
+      "grad_norm": 0.748590886592865,
+      "learning_rate": 1.2342753847206236e-05,
+      "loss": 0.6553171157836915,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8718861209964412,
+      "grad_norm": 0.7603781819343567,
+      "learning_rate": 1.2315079879211435e-05,
+      "loss": 0.639424467086792,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8736654804270463,
+      "grad_norm": 0.60029137134552,
+      "learning_rate": 1.2287387164410597e-05,
+      "loss": 0.6580222606658935,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8754448398576512,
+      "grad_norm": 0.5586308836936951,
+      "learning_rate": 1.2259675927050829e-05,
+      "loss": 0.6166585445404053,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8772241992882562,
+      "grad_norm": 0.6883022785186768,
+      "learning_rate": 1.2231946391529213e-05,
+      "loss": 0.651360559463501,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8790035587188612,
+      "grad_norm": 0.7347468733787537,
+      "learning_rate": 1.2204198782391018e-05,
+      "loss": 0.6587865352630615,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8807829181494662,
+      "grad_norm": 0.8330339193344116,
+      "learning_rate": 1.2176433324327868e-05,
+      "loss": 0.6858234405517578,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8825622775800712,
+      "grad_norm": 0.6569383144378662,
+      "learning_rate": 1.2148650242175908e-05,
+      "loss": 0.7256248950958252,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8843416370106761,
+      "grad_norm": 0.7147039771080017,
+      "learning_rate": 1.2120849760914013e-05,
+      "loss": 0.6601557254791259,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8861209964412812,
+      "grad_norm": 0.7715662717819214,
+      "learning_rate": 1.2093032105661944e-05,
+      "loss": 0.6112511634826661,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8879003558718861,
+      "grad_norm": 0.7267486453056335,
+      "learning_rate": 1.2065197501678529e-05,
+      "loss": 0.6230842590332031,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8896797153024911,
+      "grad_norm": 0.7366806268692017,
+      "learning_rate": 1.203734617435985e-05,
+      "loss": 0.6849907875061035,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8896797153024911,
+      "eval_loss": 0.6321616172790527,
+      "eval_runtime": 411.8441,
+      "eval_samples_per_second": 12.211,
+      "eval_steps_per_second": 6.107,
+      "step": 5000
+    },
+    {
+      "epoch": 0.891459074733096,
+      "grad_norm": 0.7812427282333374,
+      "learning_rate": 1.2009478349237397e-05,
+      "loss": 0.6636211395263671,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8932384341637011,
+      "grad_norm": 0.6625553965568542,
+      "learning_rate": 1.1981594251976265e-05,
+      "loss": 0.6543920040130615,
+      "step": 5020
+    },
+    {
+      "epoch": 0.895017793594306,
+      "grad_norm": 0.6391006112098694,
+      "learning_rate": 1.1953694108373313e-05,
+      "loss": 0.653505516052246,
+      "step": 5030
+    },
+    {
+      "epoch": 0.896797153024911,
+      "grad_norm": 0.763852059841156,
+      "learning_rate": 1.1925778144355338e-05,
+      "loss": 0.6871216773986817,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8985765124555161,
+      "grad_norm": 0.7798473834991455,
+      "learning_rate": 1.189784658597724e-05,
+      "loss": 0.6243946552276611,
+      "step": 5050
+    },
+    {
+      "epoch": 0.900355871886121,
+      "grad_norm": 0.6348104476928711,
+      "learning_rate": 1.1869899659420208e-05,
+      "loss": 0.6567151069641113,
+      "step": 5060
+    },
+    {
+      "epoch": 0.902135231316726,
+      "grad_norm": 0.6433530449867249,
+      "learning_rate": 1.1841937590989873e-05,
+      "loss": 0.677039909362793,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9039145907473309,
+      "grad_norm": 0.8366072177886963,
+      "learning_rate": 1.1813960607114476e-05,
+      "loss": 0.6381283760070801,
+      "step": 5080
+    },
+    {
+      "epoch": 0.905693950177936,
+      "grad_norm": 0.7176342606544495,
+      "learning_rate": 1.1785968934343045e-05,
+      "loss": 0.6601722717285157,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9074733096085409,
+      "grad_norm": 0.6531491279602051,
+      "learning_rate": 1.1757962799343548e-05,
+      "loss": 0.6662145137786866,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9092526690391459,
+      "grad_norm": 0.6918512582778931,
+      "learning_rate": 1.1729942428901068e-05,
+      "loss": 0.6182730674743653,
+      "step": 5110
+    },
+    {
+      "epoch": 0.9110320284697508,
+      "grad_norm": 0.5597010850906372,
+      "learning_rate": 1.1701908049915964e-05,
+      "loss": 0.6443261623382568,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9128113879003559,
+      "grad_norm": 0.6807442307472229,
+      "learning_rate": 1.1673859889402028e-05,
+      "loss": 0.6334492206573487,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9145907473309609,
+      "grad_norm": 0.6249061226844788,
+      "learning_rate": 1.1645798174484653e-05,
+      "loss": 0.614483642578125,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9163701067615658,
+      "grad_norm": 0.5527177453041077,
+      "learning_rate": 1.161772313239899e-05,
+      "loss": 0.650807237625122,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9181494661921709,
+      "grad_norm": 0.7590207457542419,
+      "learning_rate": 1.1589634990488107e-05,
+      "loss": 0.6382132053375245,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9199288256227758,
+      "grad_norm": 0.7238802909851074,
+      "learning_rate": 1.1561533976201157e-05,
+      "loss": 0.6336176872253418,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9217081850533808,
+      "grad_norm": 0.8075314164161682,
+      "learning_rate": 1.1533420317091519e-05,
+      "loss": 0.6544758796691894,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9234875444839857,
+      "grad_norm": 0.7770307064056396,
+      "learning_rate": 1.1505294240814979e-05,
+      "loss": 0.6220970153808594,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9252669039145908,
+      "grad_norm": 0.5794429779052734,
+      "learning_rate": 1.1477155975127866e-05,
+      "loss": 0.6260251045227051,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9270462633451957,
+      "grad_norm": 0.6863322257995605,
+      "learning_rate": 1.1449005747885212e-05,
+      "loss": 0.6177238464355469,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9288256227758007,
+      "grad_norm": 0.6806472539901733,
+      "learning_rate": 1.142084378703892e-05,
+      "loss": 0.6731130599975585,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9306049822064056,
+      "grad_norm": 0.6430651545524597,
+      "learning_rate": 1.1392670320635894e-05,
+      "loss": 0.6535952091217041,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9323843416370107,
+      "grad_norm": 0.6920055150985718,
+      "learning_rate": 1.1364485576816225e-05,
+      "loss": 0.6301285266876221,
+      "step": 5240
+    },
+    {
+      "epoch": 0.9341637010676157,
+      "grad_norm": 0.7043541669845581,
+      "learning_rate": 1.1336289783811303e-05,
+      "loss": 0.6545778751373291,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9359430604982206,
+      "grad_norm": 0.7749147415161133,
+      "learning_rate": 1.130808316994201e-05,
+      "loss": 0.6474496841430664,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9377224199288257,
+      "grad_norm": 0.6907658576965332,
+      "learning_rate": 1.127986596361684e-05,
+      "loss": 0.6155929565429688,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9395017793594306,
+      "grad_norm": 0.6255580186843872,
+      "learning_rate": 1.125163839333007e-05,
+      "loss": 0.6644204139709473,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9412811387900356,
+      "grad_norm": 0.5674806833267212,
+      "learning_rate": 1.1223400687659898e-05,
+      "loss": 0.6196834564208984,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9430604982206405,
+      "grad_norm": 0.7296446561813354,
+      "learning_rate": 1.1195153075266591e-05,
+      "loss": 0.6685366153717041,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9448398576512456,
+      "grad_norm": 0.5960925817489624,
+      "learning_rate": 1.1166895784890644e-05,
+      "loss": 0.6431320190429688,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9466192170818505,
+      "grad_norm": 0.6691383719444275,
+      "learning_rate": 1.1138629045350911e-05,
+      "loss": 0.6147652626037597,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9483985765124555,
+      "grad_norm": 0.7819799184799194,
+      "learning_rate": 1.1110353085542778e-05,
+      "loss": 0.6021285057067871,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9501779359430605,
+      "grad_norm": 0.690168023109436,
+      "learning_rate": 1.1082068134436281e-05,
+      "loss": 0.6405185699462891,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9519572953736655,
+      "grad_norm": 0.6121358871459961,
+      "learning_rate": 1.1053774421074272e-05,
+      "loss": 0.6638952255249023,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9537366548042705,
+      "grad_norm": 0.6227392554283142,
+      "learning_rate": 1.1025472174570554e-05,
+      "loss": 0.6735183715820312,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9555160142348754,
+      "grad_norm": 0.7254964113235474,
+      "learning_rate": 1.099716162410803e-05,
+      "loss": 0.6874561309814453,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9572953736654805,
+      "grad_norm": 0.916182816028595,
+      "learning_rate": 1.0968842998936843e-05,
+      "loss": 0.6294379234313965,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9590747330960854,
+      "grad_norm": 0.7955939769744873,
+      "learning_rate": 1.0940516528372527e-05,
+      "loss": 0.648458194732666,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9608540925266904,
+      "grad_norm": 0.5914123058319092,
+      "learning_rate": 1.0912182441794141e-05,
+      "loss": 0.641319465637207,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9626334519572953,
+      "grad_norm": 0.8009176850318909,
+      "learning_rate": 1.0883840968642416e-05,
+      "loss": 0.6294963836669922,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9644128113879004,
+      "grad_norm": 0.8268579244613647,
+      "learning_rate": 1.0855492338417905e-05,
+      "loss": 0.6227912425994873,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9661921708185054,
+      "grad_norm": 0.7227009534835815,
+      "learning_rate": 1.0827136780679109e-05,
+      "loss": 0.6420284271240234,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9679715302491103,
+      "grad_norm": 0.7250238060951233,
+      "learning_rate": 1.0798774525040628e-05,
+      "loss": 0.6643415927886963,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9697508896797153,
+      "grad_norm": 0.6804770827293396,
+      "learning_rate": 1.07704058011713e-05,
+      "loss": 0.6299617290496826,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9715302491103203,
+      "grad_norm": 0.8592752814292908,
+      "learning_rate": 1.0742030838792343e-05,
+      "loss": 0.6287535667419434,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9733096085409253,
+      "grad_norm": 0.6754792928695679,
+      "learning_rate": 1.0713649867675483e-05,
+      "loss": 0.6633755207061768,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9750889679715302,
+      "grad_norm": 0.8593308925628662,
+      "learning_rate": 1.0685263117641117e-05,
+      "loss": 0.6187657833099365,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9768683274021353,
+      "grad_norm": 0.6328932642936707,
+      "learning_rate": 1.065687081855643e-05,
+      "loss": 0.618121862411499,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9786476868327402,
+      "grad_norm": 0.7381909489631653,
+      "learning_rate": 1.062847320033354e-05,
+      "loss": 0.6680717945098877,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9786476868327402,
+      "eval_loss": 0.6279275417327881,
+      "eval_runtime": 413.1424,
+      "eval_samples_per_second": 12.173,
+      "eval_steps_per_second": 6.087,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9804270462633452,
+      "grad_norm": 0.8082507848739624,
+      "learning_rate": 1.0600070492927644e-05,
+      "loss": 0.619424295425415,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9822064056939501,
+      "grad_norm": 0.782515823841095,
+      "learning_rate": 1.0571662926335138e-05,
+      "loss": 0.6681380271911621,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9839857651245552,
+      "grad_norm": 0.756851851940155,
+      "learning_rate": 1.0543250730591778e-05,
+      "loss": 0.648841381072998,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9857651245551602,
+      "grad_norm": 0.6018344759941101,
+      "learning_rate": 1.0514834135770805e-05,
+      "loss": 0.6303011417388916,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9875444839857651,
+      "grad_norm": 0.6782673597335815,
+      "learning_rate": 1.0486413371981077e-05,
+      "loss": 0.6756137371063232,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9893238434163701,
+      "grad_norm": 0.680172860622406,
+      "learning_rate": 1.045798866936521e-05,
+      "loss": 0.62680082321167,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9911032028469751,
+      "grad_norm": 0.7947545647621155,
+      "learning_rate": 1.042956025809772e-05,
+      "loss": 0.6698834419250488,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9928825622775801,
+      "grad_norm": 0.6484361290931702,
+      "learning_rate": 1.0401128368383151e-05,
+      "loss": 0.6242072582244873,
+      "step": 5580
+    },
+    {
+      "epoch": 0.994661921708185,
+      "grad_norm": 0.8095009326934814,
+      "learning_rate": 1.037269323045422e-05,
+      "loss": 0.6368332862854004,
+      "step": 5590
+    },
+    {
+      "epoch": 0.99644128113879,
+      "grad_norm": 0.7750351428985596,
+      "learning_rate": 1.034425507456994e-05,
+      "loss": 0.6159787654876709,
+      "step": 5600
+    },
+    {
+      "epoch": 0.998220640569395,
+      "grad_norm": 0.6969480514526367,
+      "learning_rate": 1.0315814131013768e-05,
+      "loss": 0.6518408298492432,
+      "step": 5610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7439318299293518,
+      "learning_rate": 1.0287370630091731e-05,
+      "loss": 0.6812664031982422,
+      "step": 5620
+    },
+    {
+      "epoch": 1.001779359430605,
+      "grad_norm": 0.7261248826980591,
+      "learning_rate": 1.0258924802130565e-05,
+      "loss": 0.6570511817932129,
+      "step": 5630
+    },
+    {
+      "epoch": 1.00355871886121,
+      "grad_norm": 0.6743627190589905,
+      "learning_rate": 1.0230476877475854e-05,
+      "loss": 0.6013021469116211,
+      "step": 5640
+    },
+    {
+      "epoch": 1.0053380782918149,
+      "grad_norm": 0.698407769203186,
+      "learning_rate": 1.0202027086490154e-05,
+      "loss": 0.6084145069122314,
+      "step": 5650
+    },
+    {
+      "epoch": 1.00711743772242,
+      "grad_norm": 0.7206895351409912,
+      "learning_rate": 1.0173575659551137e-05,
+      "loss": 0.642551326751709,
+      "step": 5660
+    },
+    {
+      "epoch": 1.008896797153025,
+      "grad_norm": 0.8103647232055664,
+      "learning_rate": 1.0145122827049725e-05,
+      "loss": 0.5881889820098877,
+      "step": 5670
+    },
+    {
+      "epoch": 1.01067615658363,
+      "grad_norm": 0.7496983408927917,
+      "learning_rate": 1.0116668819388218e-05,
+      "loss": 0.6021871089935302,
+      "step": 5680
+    },
+    {
+      "epoch": 1.0124555160142348,
+      "grad_norm": 0.6843361258506775,
+      "learning_rate": 1.0088213866978435e-05,
+      "loss": 0.6468405723571777,
+      "step": 5690
+    },
+    {
+      "epoch": 1.0142348754448398,
+      "grad_norm": 0.6858223676681519,
+      "learning_rate": 1.0059758200239842e-05,
+      "loss": 0.6358915328979492,
+      "step": 5700
+    },
+    {
+      "epoch": 1.0160142348754448,
+      "grad_norm": 0.9082915186882019,
+      "learning_rate": 1.0031302049597691e-05,
+      "loss": 0.6340798377990723,
+      "step": 5710
+    },
+    {
+      "epoch": 1.0177935943060499,
+      "grad_norm": 0.6842049360275269,
+      "learning_rate": 1.0002845645481152e-05,
+      "loss": 0.6286153793334961,
+      "step": 5720
+    },
+    {
+      "epoch": 1.019572953736655,
+      "grad_norm": 0.7649215459823608,
+      "learning_rate": 9.974389218321453e-06,
+      "loss": 0.6225019931793213,
+      "step": 5730
+    },
+    {
+      "epoch": 1.0213523131672597,
+      "grad_norm": 0.6515229344367981,
+      "learning_rate": 9.945932998549996e-06,
+      "loss": 0.6120016574859619,
+      "step": 5740
+    },
+    {
+      "epoch": 1.0231316725978647,
+      "grad_norm": 0.5917989015579224,
+      "learning_rate": 9.917477216596521e-06,
+      "loss": 0.6404437065124512,
+      "step": 5750
+    },
+    {
+      "epoch": 1.0249110320284698,
+      "grad_norm": 0.7759416699409485,
+      "learning_rate": 9.889022102887205e-06,
+      "loss": 0.6269901275634766,
+      "step": 5760
+    },
+    {
+      "epoch": 1.0266903914590748,
+      "grad_norm": 0.7149049043655396,
+      "learning_rate": 9.860567887842827e-06,
+      "loss": 0.6032210350036621,
+      "step": 5770
+    },
+    {
+      "epoch": 1.0284697508896796,
+      "grad_norm": 0.6775475144386292,
+      "learning_rate": 9.832114801876877e-06,
+      "loss": 0.6163185119628907,
+      "step": 5780
+    },
+    {
+      "epoch": 1.0302491103202847,
+      "grad_norm": 0.7221093773841858,
+      "learning_rate": 9.80366307539372e-06,
+      "loss": 0.5833635330200195,
+      "step": 5790
+    },
+    {
+      "epoch": 1.0320284697508897,
+      "grad_norm": 0.7127135396003723,
+      "learning_rate": 9.775212938786689e-06,
+      "loss": 0.6052863597869873,
+      "step": 5800
+    },
+    {
+      "epoch": 1.0338078291814947,
+      "grad_norm": 0.7285045981407166,
+      "learning_rate": 9.746764622436265e-06,
+      "loss": 0.6380964756011963,
+      "step": 5810
+    },
+    {
+      "epoch": 1.0355871886120998,
+      "grad_norm": 0.7396540641784668,
+      "learning_rate": 9.718318356708172e-06,
+      "loss": 0.6616352081298829,
+      "step": 5820
+    },
+    {
+      "epoch": 1.0373665480427046,
+      "grad_norm": 0.6225027441978455,
+      "learning_rate": 9.689874371951541e-06,
+      "loss": 0.6689383029937744,
+      "step": 5830
+    },
+    {
+      "epoch": 1.0391459074733096,
+      "grad_norm": 0.8353135585784912,
+      "learning_rate": 9.661432898497024e-06,
+      "loss": 0.6042938232421875,
+      "step": 5840
+    },
+    {
+      "epoch": 1.0409252669039146,
+      "grad_norm": 0.6610475778579712,
+      "learning_rate": 9.632994166654941e-06,
+      "loss": 0.6156399726867676,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0427046263345197,
+      "grad_norm": 0.7322497963905334,
+      "learning_rate": 9.604558406713413e-06,
+      "loss": 0.6352502822875976,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0444839857651245,
+      "grad_norm": 0.7838056087493896,
+      "learning_rate": 9.576125848936484e-06,
+      "loss": 0.6298691272735596,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0462633451957295,
+      "grad_norm": 0.81569504737854,
+      "learning_rate": 9.547696723562289e-06,
+      "loss": 0.6307301998138428,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0480427046263345,
+      "grad_norm": 0.8897931575775146,
+      "learning_rate": 9.519271260801146e-06,
+      "loss": 0.6494285106658936,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0498220640569396,
+      "grad_norm": 0.7175611257553101,
+      "learning_rate": 9.490849690833732e-06,
+      "loss": 0.5930293083190918,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0516014234875444,
+      "grad_norm": 0.6708109378814697,
+      "learning_rate": 9.462432243809191e-06,
+      "loss": 0.595799732208252,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0533807829181494,
+      "grad_norm": 0.6808570623397827,
+      "learning_rate": 9.434019149843285e-06,
+      "loss": 0.6564432621002197,
+      "step": 5920
+    },
+    {
+      "epoch": 1.0551601423487544,
+      "grad_norm": 0.8115124106407166,
+      "learning_rate": 9.405610639016522e-06,
+      "loss": 0.6025516033172608,
+      "step": 5930
+    },
+    {
+      "epoch": 1.0569395017793595,
+      "grad_norm": 0.7574615478515625,
+      "learning_rate": 9.377206941372306e-06,
+      "loss": 0.6670190811157226,
+      "step": 5940
+    },
+    {
+      "epoch": 1.0587188612099645,
+      "grad_norm": 0.8795053362846375,
+      "learning_rate": 9.34880828691505e-06,
+      "loss": 0.6606448650360107,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0604982206405693,
+      "grad_norm": 0.7287918925285339,
+      "learning_rate": 9.320414905608348e-06,
+      "loss": 0.6295819759368897,
+      "step": 5960
+    },
+    {
+      "epoch": 1.0622775800711743,
+      "grad_norm": 0.7272006869316101,
+      "learning_rate": 9.292027027373075e-06,
+      "loss": 0.6425396919250488,
+      "step": 5970
+    },
+    {
+      "epoch": 1.0640569395017794,
+      "grad_norm": 0.6797536015510559,
+      "learning_rate": 9.263644882085564e-06,
+      "loss": 0.5847621917724609,
+      "step": 5980
+    },
+    {
+      "epoch": 1.0658362989323844,
+      "grad_norm": 0.6698694229125977,
+      "learning_rate": 9.235268699575704e-06,
+      "loss": 0.6509596824645996,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0676156583629894,
+      "grad_norm": 0.7406882643699646,
+      "learning_rate": 9.206898709625109e-06,
+      "loss": 0.6330312252044678,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0676156583629894,
+      "eval_loss": 0.6239920854568481,
+      "eval_runtime": 411.471,
+      "eval_samples_per_second": 12.222,
+      "eval_steps_per_second": 6.112,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0693950177935942,
+      "grad_norm": 0.695398211479187,
+      "learning_rate": 9.17853514196525e-06,
+      "loss": 0.6093502998352051,
+      "step": 6010
+    },
+    {
+      "epoch": 1.0711743772241993,
+      "grad_norm": 0.839644730091095,
+      "learning_rate": 9.150178226275584e-06,
+      "loss": 0.6351033687591553,
+      "step": 6020
+    },
+    {
+      "epoch": 1.0729537366548043,
+      "grad_norm": 0.6793309450149536,
+      "learning_rate": 9.121828192181716e-06,
+      "loss": 0.6527360916137696,
+      "step": 6030
+    },
+    {
+      "epoch": 1.0747330960854093,
+      "grad_norm": 0.8601770401000977,
+      "learning_rate": 9.093485269253508e-06,
+      "loss": 0.6375271320343018,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0765124555160142,
+      "grad_norm": 0.8268119692802429,
+      "learning_rate": 9.065149687003256e-06,
+      "loss": 0.6043989181518554,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0782918149466192,
+      "grad_norm": 0.6872331500053406,
+      "learning_rate": 9.0368216748838e-06,
+      "loss": 0.5880636215209961,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0800711743772242,
+      "grad_norm": 0.7637034058570862,
+      "learning_rate": 9.008501462286682e-06,
+      "loss": 0.6080229759216309,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0818505338078293,
+      "grad_norm": 0.7357282042503357,
+      "learning_rate": 8.980189278540294e-06,
+      "loss": 0.6133236408233642,
+      "step": 6080
+    },
+    {
+      "epoch": 1.083629893238434,
+      "grad_norm": 0.7250421643257141,
+      "learning_rate": 8.951885352908006e-06,
+      "loss": 0.6319203853607178,
+      "step": 6090
+    },
+    {
+      "epoch": 1.085409252669039,
+      "grad_norm": 0.6590055227279663,
+      "learning_rate": 8.92358991458631e-06,
+      "loss": 0.649744987487793,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0871886120996441,
+      "grad_norm": 0.7239564061164856,
+      "learning_rate": 8.895303192702988e-06,
+      "loss": 0.6277432918548584,
+      "step": 6110
+    },
+    {
+      "epoch": 1.0889679715302492,
+      "grad_norm": 0.8234860301017761,
+      "learning_rate": 8.867025416315221e-06,
+      "loss": 0.6343496799468994,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0907473309608542,
+      "grad_norm": 0.8464725613594055,
+      "learning_rate": 8.838756814407766e-06,
+      "loss": 0.6384105205535888,
+      "step": 6130
+    },
+    {
+      "epoch": 1.092526690391459,
+      "grad_norm": 0.8043434619903564,
+      "learning_rate": 8.810497615891078e-06,
+      "loss": 0.6212152957916259,
+      "step": 6140
+    },
+    {
+      "epoch": 1.094306049822064,
+      "grad_norm": 0.7087031602859497,
+      "learning_rate": 8.782248049599468e-06,
+      "loss": 0.5984238624572754,
+      "step": 6150
+    },
+    {
+      "epoch": 1.096085409252669,
+      "grad_norm": 0.7631204724311829,
+      "learning_rate": 8.754008344289253e-06,
+      "loss": 0.6375543594360351,
+      "step": 6160
+    },
+    {
+      "epoch": 1.097864768683274,
+      "grad_norm": 0.6403396725654602,
+      "learning_rate": 8.725778728636893e-06,
+      "loss": 0.6312360763549805,
+      "step": 6170
+    },
+    {
+      "epoch": 1.099644128113879,
+      "grad_norm": 0.8257510662078857,
+      "learning_rate": 8.697559431237153e-06,
+      "loss": 0.6133918762207031,
+      "step": 6180
+    },
+    {
+      "epoch": 1.101423487544484,
+      "grad_norm": 0.6782218813896179,
+      "learning_rate": 8.66935068060123e-06,
+      "loss": 0.6308522701263428,
+      "step": 6190
+    },
+    {
+      "epoch": 1.103202846975089,
+      "grad_norm": 0.7291736602783203,
+      "learning_rate": 8.641152705154935e-06,
+      "loss": 0.6390894889831543,
+      "step": 6200
+    },
+    {
+      "epoch": 1.104982206405694,
+      "grad_norm": 0.7014942169189453,
+      "learning_rate": 8.612965733236811e-06,
+      "loss": 0.6089354038238526,
+      "step": 6210
+    },
+    {
+      "epoch": 1.106761565836299,
+      "grad_norm": 0.6617055535316467,
+      "learning_rate": 8.584789993096303e-06,
+      "loss": 0.6288234710693359,
+      "step": 6220
+    },
+    {
+      "epoch": 1.1085409252669038,
+      "grad_norm": 0.7303082942962646,
+      "learning_rate": 8.5566257128919e-06,
+      "loss": 0.6383206367492675,
+      "step": 6230
+    },
+    {
+      "epoch": 1.1103202846975089,
+      "grad_norm": 0.7110517024993896,
+      "learning_rate": 8.528473120689302e-06,
+      "loss": 0.6590532779693603,
+      "step": 6240
+    },
+    {
+      "epoch": 1.112099644128114,
+      "grad_norm": 0.779214084148407,
+      "learning_rate": 8.50033244445955e-06,
+      "loss": 0.6261368274688721,
+      "step": 6250
+    },
+    {
+      "epoch": 1.113879003558719,
+      "grad_norm": 0.819240152835846,
+      "learning_rate": 8.472203912077205e-06,
+      "loss": 0.6361266136169433,
+      "step": 6260
+    },
+    {
+      "epoch": 1.1156583629893237,
+      "grad_norm": 0.7131257653236389,
+      "learning_rate": 8.444087751318484e-06,
+      "loss": 0.6407057285308838,
+      "step": 6270
+    },
+    {
+      "epoch": 1.1174377224199288,
+      "grad_norm": 0.9261246919631958,
+      "learning_rate": 8.415984189859418e-06,
+      "loss": 0.5832745552062988,
+      "step": 6280
+    },
+    {
+      "epoch": 1.1192170818505338,
+      "grad_norm": 0.8466585278511047,
+      "learning_rate": 8.38789345527403e-06,
+      "loss": 0.6238195419311523,
+      "step": 6290
+    },
+    {
+      "epoch": 1.1209964412811388,
+      "grad_norm": 0.7208281755447388,
+      "learning_rate": 8.359815775032457e-06,
+      "loss": 0.6571403980255127,
+      "step": 6300
+    },
+    {
+      "epoch": 1.1227758007117439,
+      "grad_norm": 0.6916828155517578,
+      "learning_rate": 8.331751376499131e-06,
+      "loss": 0.6353542327880859,
+      "step": 6310
+    },
+    {
+      "epoch": 1.1245551601423487,
+      "grad_norm": 0.804423987865448,
+      "learning_rate": 8.303700486930935e-06,
+      "loss": 0.6343297481536865,
+      "step": 6320
+    },
+    {
+      "epoch": 1.1263345195729537,
+      "grad_norm": 0.8437130451202393,
+      "learning_rate": 8.275663333475365e-06,
+      "loss": 0.6212164878845214,
+      "step": 6330
+    },
+    {
+      "epoch": 1.1281138790035588,
+      "grad_norm": 0.8472525477409363,
+      "learning_rate": 8.247640143168675e-06,
+      "loss": 0.6239662647247315,
+      "step": 6340
+    },
+    {
+      "epoch": 1.1298932384341638,
+      "grad_norm": 0.696539044380188,
+      "learning_rate": 8.219631142934062e-06,
+      "loss": 0.6344510078430176,
+      "step": 6350
+    },
+    {
+      "epoch": 1.1316725978647686,
+      "grad_norm": 0.7422542572021484,
+      "learning_rate": 8.191636559579802e-06,
+      "loss": 0.6504097938537597,
+      "step": 6360
+    },
+    {
+      "epoch": 1.1334519572953736,
+      "grad_norm": 0.801693856716156,
+      "learning_rate": 8.163656619797444e-06,
+      "loss": 0.6314446449279785,
+      "step": 6370
+    },
+    {
+      "epoch": 1.1352313167259787,
+      "grad_norm": 0.786159098148346,
+      "learning_rate": 8.135691550159943e-06,
+      "loss": 0.6034214973449707,
+      "step": 6380
+    },
+    {
+      "epoch": 1.1370106761565837,
+      "grad_norm": 0.7948549389839172,
+      "learning_rate": 8.107741577119853e-06,
+      "loss": 0.653587532043457,
+      "step": 6390
+    },
+    {
+      "epoch": 1.1387900355871885,
+      "grad_norm": 0.6635374426841736,
+      "learning_rate": 8.079806927007469e-06,
+      "loss": 0.6300637722015381,
+      "step": 6400
+    },
+    {
+      "epoch": 1.1405693950177935,
+      "grad_norm": 0.7938548922538757,
+      "learning_rate": 8.051887826029006e-06,
+      "loss": 0.6417815685272217,
+      "step": 6410
+    },
+    {
+      "epoch": 1.1423487544483986,
+      "grad_norm": 0.7719388008117676,
+      "learning_rate": 8.023984500264782e-06,
+      "loss": 0.624653434753418,
+      "step": 6420
+    },
+    {
+      "epoch": 1.1441281138790036,
+      "grad_norm": 0.8489981293678284,
+      "learning_rate": 7.996097175667352e-06,
+      "loss": 0.6066908836364746,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1459074733096086,
+      "grad_norm": 0.7810853123664856,
+      "learning_rate": 7.968226078059716e-06,
+      "loss": 0.5920337677001953,
+      "step": 6440
+    },
+    {
+      "epoch": 1.1476868327402134,
+      "grad_norm": 0.7887650728225708,
+      "learning_rate": 7.940371433133459e-06,
+      "loss": 0.6641497135162353,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1494661921708185,
+      "grad_norm": 0.7652443051338196,
+      "learning_rate": 7.912533466446947e-06,
+      "loss": 0.6473179340362549,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1512455516014235,
+      "grad_norm": 0.7206646800041199,
+      "learning_rate": 7.88471240342348e-06,
+      "loss": 0.5841953754425049,
+      "step": 6470
+    },
+    {
+      "epoch": 1.1530249110320285,
+      "grad_norm": 0.7867498993873596,
+      "learning_rate": 7.856908469349495e-06,
+      "loss": 0.6486004829406739,
+      "step": 6480
+    },
+    {
+      "epoch": 1.1548042704626336,
+      "grad_norm": 0.8296671509742737,
+      "learning_rate": 7.829121889372702e-06,
+      "loss": 0.6224873065948486,
+      "step": 6490
+    },
+    {
+      "epoch": 1.1565836298932384,
+      "grad_norm": 0.8746348023414612,
+      "learning_rate": 7.801352888500304e-06,
+      "loss": 0.640509843826294,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1565836298932384,
+      "eval_loss": 0.620580792427063,
+      "eval_runtime": 411.917,
+      "eval_samples_per_second": 12.209,
+      "eval_steps_per_second": 6.106,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1583629893238434,
+      "grad_norm": 0.7378506660461426,
+      "learning_rate": 7.773601691597135e-06,
+      "loss": 0.6328612327575683,
+      "step": 6510
+    },
+    {
+      "epoch": 1.1601423487544484,
+      "grad_norm": 0.7263290882110596,
+      "learning_rate": 7.745868523383876e-06,
+      "loss": 0.6071421146392822,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1619217081850535,
+      "grad_norm": 0.8653813004493713,
+      "learning_rate": 7.7181536084352e-06,
+      "loss": 0.6067211627960205,
+      "step": 6530
+    },
+    {
+      "epoch": 1.1637010676156583,
+      "grad_norm": 0.9452764987945557,
+      "learning_rate": 7.690457171177984e-06,
+      "loss": 0.6291831970214844,
+      "step": 6540
+    },
+    {
+      "epoch": 1.1654804270462633,
+      "grad_norm": 0.65375155210495,
+      "learning_rate": 7.66277943588947e-06,
+      "loss": 0.6606705665588379,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1672597864768683,
+      "grad_norm": 0.978212833404541,
+      "learning_rate": 7.635120626695456e-06,
+      "loss": 0.5956210613250732,
+      "step": 6560
+    },
+    {
+      "epoch": 1.1690391459074734,
+      "grad_norm": 0.7400839328765869,
+      "learning_rate": 7.607480967568497e-06,
+      "loss": 0.6242622852325439,
+      "step": 6570
+    },
+    {
+      "epoch": 1.1708185053380782,
+      "grad_norm": 1.0911694765090942,
+      "learning_rate": 7.579860682326055e-06,
+      "loss": 0.6239947319030762,
+      "step": 6580
+    },
+    {
+      "epoch": 1.1725978647686832,
+      "grad_norm": 0.7858565449714661,
+      "learning_rate": 7.552259994628728e-06,
+      "loss": 0.6366580963134766,
+      "step": 6590
+    },
+    {
+      "epoch": 1.1743772241992882,
+      "grad_norm": 0.8578958511352539,
+      "learning_rate": 7.5246791279784056e-06,
+      "loss": 0.6067252635955811,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1761565836298933,
+      "grad_norm": 0.7615208029747009,
+      "learning_rate": 7.4971183057164785e-06,
+      "loss": 0.6032675266265869,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1779359430604983,
+      "grad_norm": 0.8318284749984741,
+      "learning_rate": 7.469577751022024e-06,
+      "loss": 0.5974441528320312,
+      "step": 6620
+    },
+    {
+      "epoch": 1.1797153024911031,
+      "grad_norm": 0.882125198841095,
+      "learning_rate": 7.442057686909998e-06,
+      "loss": 0.6573843002319336,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1814946619217082,
+      "grad_norm": 0.7975313663482666,
+      "learning_rate": 7.414558336229426e-06,
+      "loss": 0.6807239055633545,
+      "step": 6640
+    },
+    {
+      "epoch": 1.1832740213523132,
+      "grad_norm": 0.7822281122207642,
+      "learning_rate": 7.387079921661614e-06,
+      "loss": 0.6663013458251953,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1850533807829182,
+      "grad_norm": 0.8564392328262329,
+      "learning_rate": 7.359622665718317e-06,
+      "loss": 0.6218353271484375,
+      "step": 6660
+    },
+    {
+      "epoch": 1.1868327402135233,
+      "grad_norm": 0.7377424836158752,
+      "learning_rate": 7.332186790739973e-06,
+      "loss": 0.5968504905700683,
+      "step": 6670
+    },
+    {
+      "epoch": 1.188612099644128,
+      "grad_norm": 0.730351984500885,
+      "learning_rate": 7.304772518893866e-06,
+      "loss": 0.6158394813537598,
+      "step": 6680
+    },
+    {
+      "epoch": 1.190391459074733,
+      "grad_norm": 0.7883412837982178,
+      "learning_rate": 7.277380072172354e-06,
+      "loss": 0.6515108585357666,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1921708185053381,
+      "grad_norm": 0.9434972405433655,
+      "learning_rate": 7.250009672391063e-06,
+      "loss": 0.6276324272155762,
+      "step": 6700
+    },
+    {
+      "epoch": 1.193950177935943,
+      "grad_norm": 0.8042780756950378,
+      "learning_rate": 7.2226615411870796e-06,
+      "loss": 0.6224194526672363,
+      "step": 6710
+    },
+    {
+      "epoch": 1.195729537366548,
+      "grad_norm": 0.8094464540481567,
+      "learning_rate": 7.195335900017181e-06,
+      "loss": 0.6250015735626221,
+      "step": 6720
+    },
+    {
+      "epoch": 1.197508896797153,
+      "grad_norm": 0.8359760046005249,
+      "learning_rate": 7.168032970156011e-06,
+      "loss": 0.6454795837402344,
+      "step": 6730
+    },
+    {
+      "epoch": 1.199288256227758,
+      "grad_norm": 0.728487491607666,
+      "learning_rate": 7.140752972694325e-06,
+      "loss": 0.6341996192932129,
+      "step": 6740
+    },
+    {
+      "epoch": 1.201067615658363,
+      "grad_norm": 0.8493318557739258,
+      "learning_rate": 7.113496128537154e-06,
+      "loss": 0.6078832626342774,
+      "step": 6750
+    },
+    {
+      "epoch": 1.2028469750889679,
+      "grad_norm": 0.7177495360374451,
+      "learning_rate": 7.086262658402059e-06,
+      "loss": 0.6244680881500244,
+      "step": 6760
+    },
+    {
+      "epoch": 1.204626334519573,
+      "grad_norm": 0.7511991858482361,
+      "learning_rate": 7.059052782817317e-06,
+      "loss": 0.6233652591705322,
+      "step": 6770
+    },
+    {
+      "epoch": 1.206405693950178,
+      "grad_norm": 1.0159603357315063,
+      "learning_rate": 7.0318667221201505e-06,
+      "loss": 0.6203540802001953,
+      "step": 6780
+    },
+    {
+      "epoch": 1.208185053380783,
+      "grad_norm": 0.9800230264663696,
+      "learning_rate": 7.004704696454924e-06,
+      "loss": 0.621929931640625,
+      "step": 6790
+    },
+    {
+      "epoch": 1.209964412811388,
+      "grad_norm": 0.740306556224823,
+      "learning_rate": 6.97756692577139e-06,
+      "loss": 0.5935471534729004,
+      "step": 6800
+    },
+    {
+      "epoch": 1.2117437722419928,
+      "grad_norm": 0.7026015520095825,
+      "learning_rate": 6.950453629822874e-06,
+      "loss": 0.5996420383453369,
+      "step": 6810
+    },
+    {
+      "epoch": 1.2135231316725978,
+      "grad_norm": 0.7063615918159485,
+      "learning_rate": 6.923365028164532e-06,
+      "loss": 0.6140129089355468,
+      "step": 6820
+    },
+    {
+      "epoch": 1.2153024911032029,
+      "grad_norm": 0.6547632217407227,
+      "learning_rate": 6.8963013401515365e-06,
+      "loss": 0.6896812915802002,
+      "step": 6830
+    },
+    {
+      "epoch": 1.217081850533808,
+      "grad_norm": 0.7407922744750977,
+      "learning_rate": 6.8692627849373226e-06,
+      "loss": 0.6203756332397461,
+      "step": 6840
+    },
+    {
+      "epoch": 1.2188612099644127,
+      "grad_norm": 0.7740945219993591,
+      "learning_rate": 6.842249581471814e-06,
+      "loss": 0.6261481761932373,
+      "step": 6850
+    },
+    {
+      "epoch": 1.2206405693950177,
+      "grad_norm": 0.8180769085884094,
+      "learning_rate": 6.815261948499628e-06,
+      "loss": 0.6112605094909668,
+      "step": 6860
+    },
+    {
+      "epoch": 1.2224199288256228,
+      "grad_norm": 0.7909367084503174,
+      "learning_rate": 6.788300104558341e-06,
+      "loss": 0.6290336132049561,
+      "step": 6870
+    },
+    {
+      "epoch": 1.2241992882562278,
+      "grad_norm": 0.7958800792694092,
+      "learning_rate": 6.76136426797668e-06,
+      "loss": 0.6461628913879395,
+      "step": 6880
+    },
+    {
+      "epoch": 1.2259786476868326,
+      "grad_norm": 0.83876633644104,
+      "learning_rate": 6.734454656872784e-06,
+      "loss": 0.6532453060150146,
+      "step": 6890
+    },
+    {
+      "epoch": 1.2277580071174377,
+      "grad_norm": 0.8222100138664246,
+      "learning_rate": 6.707571489152418e-06,
+      "loss": 0.6066458225250244,
+      "step": 6900
+    },
+    {
+      "epoch": 1.2295373665480427,
+      "grad_norm": 0.7338365316390991,
+      "learning_rate": 6.680714982507219e-06,
+      "loss": 0.6187913417816162,
+      "step": 6910
+    },
+    {
+      "epoch": 1.2313167259786477,
+      "grad_norm": 1.0070669651031494,
+      "learning_rate": 6.653885354412935e-06,
+      "loss": 0.638453197479248,
+      "step": 6920
+    },
+    {
+      "epoch": 1.2330960854092528,
+      "grad_norm": 0.8759623169898987,
+      "learning_rate": 6.627082822127657e-06,
+      "loss": 0.6360817909240722,
+      "step": 6930
+    },
+    {
+      "epoch": 1.2348754448398576,
+      "grad_norm": 0.7884645462036133,
+      "learning_rate": 6.600307602690057e-06,
+      "loss": 0.6332939147949219,
+      "step": 6940
+    },
+    {
+      "epoch": 1.2366548042704626,
+      "grad_norm": 0.7928617596626282,
+      "learning_rate": 6.573559912917648e-06,
+      "loss": 0.6071663856506347,
+      "step": 6950
+    },
+    {
+      "epoch": 1.2384341637010676,
+      "grad_norm": 0.8969078063964844,
+      "learning_rate": 6.546839969405003e-06,
+      "loss": 0.6088360786437989,
+      "step": 6960
+    },
+    {
+      "epoch": 1.2402135231316727,
+      "grad_norm": 0.6923132538795471,
+      "learning_rate": 6.520147988522021e-06,
+      "loss": 0.602042293548584,
+      "step": 6970
+    },
+    {
+      "epoch": 1.2419928825622777,
+      "grad_norm": 0.8759524822235107,
+      "learning_rate": 6.493484186412172e-06,
+      "loss": 0.6699591636657715,
+      "step": 6980
+    },
+    {
+      "epoch": 1.2437722419928825,
+      "grad_norm": 0.9501714706420898,
+      "learning_rate": 6.466848778990734e-06,
+      "loss": 0.631440258026123,
+      "step": 6990
+    },
+    {
+      "epoch": 1.2455516014234875,
+      "grad_norm": 0.8564425706863403,
+      "learning_rate": 6.440241981943063e-06,
+      "loss": 0.6175348281860351,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2455516014234875,
+      "eval_loss": 0.6177652478218079,
+      "eval_runtime": 410.1105,
+      "eval_samples_per_second": 12.263,
+      "eval_steps_per_second": 6.132,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2473309608540926,
+      "grad_norm": 0.8425670266151428,
+      "learning_rate": 6.413664010722825e-06,
+      "loss": 0.6334220886230468,
+      "step": 7010
+    },
+    {
+      "epoch": 1.2491103202846976,
+      "grad_norm": 0.6265377998352051,
+      "learning_rate": 6.3871150805502765e-06,
+      "loss": 0.6206889629364014,
+      "step": 7020
+    },
+    {
+      "epoch": 1.2508896797153026,
+      "grad_norm": 0.7120670676231384,
+      "learning_rate": 6.360595406410497e-06,
+      "loss": 0.609261417388916,
+      "step": 7030
+    },
+    {
+      "epoch": 1.2526690391459074,
+      "grad_norm": 0.7951711416244507,
+      "learning_rate": 6.334105203051673e-06,
+      "loss": 0.6361230850219727,
+      "step": 7040
+    },
+    {
+      "epoch": 1.2544483985765125,
+      "grad_norm": 0.9487454891204834,
+      "learning_rate": 6.307644684983329e-06,
+      "loss": 0.5716474533081055,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2562277580071175,
+      "grad_norm": 0.7447927594184875,
+      "learning_rate": 6.281214066474623e-06,
+      "loss": 0.6536383152008056,
+      "step": 7060
+    },
+    {
+      "epoch": 1.2580071174377223,
+      "grad_norm": 0.8137961626052856,
+      "learning_rate": 6.254813561552585e-06,
+      "loss": 0.6361209869384765,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2597864768683273,
+      "grad_norm": 0.7065114378929138,
+      "learning_rate": 6.2284433840004e-06,
+      "loss": 0.6489431858062744,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2615658362989324,
+      "grad_norm": 0.7409806847572327,
+      "learning_rate": 6.202103747355667e-06,
+      "loss": 0.6485908031463623,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2633451957295374,
+      "grad_norm": 0.662041962146759,
+      "learning_rate": 6.175794864908672e-06,
+      "loss": 0.6204521179199218,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2651245551601424,
+      "grad_norm": 0.7687442898750305,
+      "learning_rate": 6.1495169497006755e-06,
+      "loss": 0.6361071109771729,
+      "step": 7110
+    },
+    {
+      "epoch": 1.2669039145907472,
+      "grad_norm": 0.9222472906112671,
+      "learning_rate": 6.123270214522159e-06,
+      "loss": 0.6162978172302246,
+      "step": 7120
+    },
+    {
+      "epoch": 1.2686832740213523,
+      "grad_norm": 0.819275975227356,
+      "learning_rate": 6.097054871911132e-06,
+      "loss": 0.6023138523101806,
+      "step": 7130
+    },
+    {
+      "epoch": 1.2704626334519573,
+      "grad_norm": 0.8842992782592773,
+      "learning_rate": 6.070871134151386e-06,
+      "loss": 0.658526086807251,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2722419928825623,
+      "grad_norm": 0.7101578116416931,
+      "learning_rate": 6.044719213270791e-06,
+      "loss": 0.6004268169403076,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2740213523131674,
+      "grad_norm": 0.7537800669670105,
+      "learning_rate": 6.018599321039569e-06,
+      "loss": 0.6223325729370117,
+      "step": 7160
+    },
+    {
+      "epoch": 1.2758007117437722,
+      "grad_norm": 0.6810758709907532,
+      "learning_rate": 5.9925116689685925e-06,
+      "loss": 0.6302705764770508,
+      "step": 7170
+    },
+    {
+      "epoch": 1.2775800711743772,
+      "grad_norm": 0.8179675936698914,
+      "learning_rate": 5.966456468307653e-06,
+      "loss": 0.6177227020263671,
+      "step": 7180
+    },
+    {
+      "epoch": 1.2793594306049823,
+      "grad_norm": 0.7833004593849182,
+      "learning_rate": 5.940433930043772e-06,
+      "loss": 0.6414045333862305,
+      "step": 7190
+    },
+    {
+      "epoch": 1.281138790035587,
+      "grad_norm": 0.9323675036430359,
+      "learning_rate": 5.914444264899466e-06,
+      "loss": 0.603922176361084,
+      "step": 7200
+    },
+    {
+      "epoch": 1.282918149466192,
+      "grad_norm": 0.844142735004425,
+      "learning_rate": 5.888487683331072e-06,
+      "loss": 0.6308831691741943,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2846975088967971,
+      "grad_norm": 0.8182924389839172,
+      "learning_rate": 5.862564395527013e-06,
+      "loss": 0.6259811401367188,
+      "step": 7220
+    },
+    {
+      "epoch": 1.2864768683274022,
+      "grad_norm": 0.8091204166412354,
+      "learning_rate": 5.836674611406117e-06,
+      "loss": 0.6158743858337402,
+      "step": 7230
+    },
+    {
+      "epoch": 1.2882562277580072,
+      "grad_norm": 0.9385861754417419,
+      "learning_rate": 5.810818540615903e-06,
+      "loss": 0.6304144382476806,
+      "step": 7240
+    },
+    {
+      "epoch": 1.290035587188612,
+      "grad_norm": 1.0828267335891724,
+      "learning_rate": 5.784996392530892e-06,
+      "loss": 0.6548801898956299,
+      "step": 7250
+    },
+    {
+      "epoch": 1.291814946619217,
+      "grad_norm": 0.7489603161811829,
+      "learning_rate": 5.759208376250916e-06,
+      "loss": 0.6102576732635498,
+      "step": 7260
+    },
+    {
+      "epoch": 1.293594306049822,
+      "grad_norm": 0.7216728329658508,
+      "learning_rate": 5.733454700599408e-06,
+      "loss": 0.6362377166748047,
+      "step": 7270
+    },
+    {
+      "epoch": 1.295373665480427,
+      "grad_norm": 0.8442836403846741,
+      "learning_rate": 5.707735574121732e-06,
+      "loss": 0.6299904346466064,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2971530249110321,
+      "grad_norm": 0.6905636787414551,
+      "learning_rate": 5.68205120508347e-06,
+      "loss": 0.6162694454193115,
+      "step": 7290
+    },
+    {
+      "epoch": 1.298932384341637,
+      "grad_norm": 0.8710734844207764,
+      "learning_rate": 5.656401801468764e-06,
+      "loss": 0.5905096054077148,
+      "step": 7300
+    },
+    {
+      "epoch": 1.300711743772242,
+      "grad_norm": 0.7497642040252686,
+      "learning_rate": 5.630787570978601e-06,
+      "loss": 0.6081714630126953,
+      "step": 7310
+    },
+    {
+      "epoch": 1.302491103202847,
+      "grad_norm": 0.7930024266242981,
+      "learning_rate": 5.605208721029162e-06,
+      "loss": 0.6393218517303467,
+      "step": 7320
+    },
+    {
+      "epoch": 1.304270462633452,
+      "grad_norm": 0.622921347618103,
+      "learning_rate": 5.579665458750119e-06,
+      "loss": 0.6169804096221924,
+      "step": 7330
+    },
+    {
+      "epoch": 1.306049822064057,
+      "grad_norm": 0.7385269403457642,
+      "learning_rate": 5.554157990982964e-06,
+      "loss": 0.6091341495513916,
+      "step": 7340
+    },
+    {
+      "epoch": 1.3078291814946619,
+      "grad_norm": 0.839229941368103,
+      "learning_rate": 5.528686524279336e-06,
+      "loss": 0.6216320037841797,
+      "step": 7350
+    },
+    {
+      "epoch": 1.309608540925267,
+      "grad_norm": 0.8395034670829773,
+      "learning_rate": 5.503251264899361e-06,
+      "loss": 0.6268163204193116,
+      "step": 7360
+    },
+    {
+      "epoch": 1.311387900355872,
+      "grad_norm": 0.7301527261734009,
+      "learning_rate": 5.477852418809955e-06,
+      "loss": 0.6226423740386963,
+      "step": 7370
+    },
+    {
+      "epoch": 1.3131672597864767,
+      "grad_norm": 0.9010869860649109,
+      "learning_rate": 5.452490191683171e-06,
+      "loss": 0.6155654907226562,
+      "step": 7380
+    },
+    {
+      "epoch": 1.3149466192170818,
+      "grad_norm": 0.7506121397018433,
+      "learning_rate": 5.427164788894543e-06,
+      "loss": 0.6015159130096436,
+      "step": 7390
+    },
+    {
+      "epoch": 1.3167259786476868,
+      "grad_norm": 0.8243475556373596,
+      "learning_rate": 5.401876415521402e-06,
+      "loss": 0.6120582103729248,
+      "step": 7400
+    },
+    {
+      "epoch": 1.3185053380782918,
+      "grad_norm": 0.877004086971283,
+      "learning_rate": 5.3766252763412375e-06,
+      "loss": 0.6368375301361084,
+      "step": 7410
+    },
+    {
+      "epoch": 1.3202846975088969,
+      "grad_norm": 0.6776864528656006,
+      "learning_rate": 5.351411575830015e-06,
+      "loss": 0.5798074722290039,
+      "step": 7420
+    },
+    {
+      "epoch": 1.3220640569395017,
+      "grad_norm": 0.7413604259490967,
+      "learning_rate": 5.326235518160543e-06,
+      "loss": 0.6417413711547851,
+      "step": 7430
+    },
+    {
+      "epoch": 1.3238434163701067,
+      "grad_norm": 0.680547297000885,
+      "learning_rate": 5.301097307200804e-06,
+      "loss": 0.6709810256958008,
+      "step": 7440
+    },
+    {
+      "epoch": 1.3256227758007118,
+      "grad_norm": 0.8982633352279663,
+      "learning_rate": 5.275997146512317e-06,
+      "loss": 0.6244979858398437,
+      "step": 7450
+    },
+    {
+      "epoch": 1.3274021352313168,
+      "grad_norm": 0.7878574728965759,
+      "learning_rate": 5.250935239348469e-06,
+      "loss": 0.6555228233337402,
+      "step": 7460
+    },
+    {
+      "epoch": 1.3291814946619218,
+      "grad_norm": 0.7958076000213623,
+      "learning_rate": 5.2259117886528974e-06,
+      "loss": 0.6336334228515625,
+      "step": 7470
+    },
+    {
+      "epoch": 1.3309608540925266,
+      "grad_norm": 0.8020284175872803,
+      "learning_rate": 5.200926997057818e-06,
+      "loss": 0.6250798225402832,
+      "step": 7480
+    },
+    {
+      "epoch": 1.3327402135231317,
+      "grad_norm": 0.9492243528366089,
+      "learning_rate": 5.175981066882403e-06,
+      "loss": 0.5951490879058838,
+      "step": 7490
+    },
+    {
+      "epoch": 1.3345195729537367,
+      "grad_norm": 0.8778184056282043,
+      "learning_rate": 5.15107420013113e-06,
+      "loss": 0.5949514865875244,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3345195729537367,
+      "eval_loss": 0.6148350238800049,
+      "eval_runtime": 411.3736,
+      "eval_samples_per_second": 12.225,
+      "eval_steps_per_second": 6.114,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3362989323843417,
+      "grad_norm": 0.888278603553772,
+      "learning_rate": 5.126206598492167e-06,
+      "loss": 0.6134849071502686,
+      "step": 7510
+    },
+    {
+      "epoch": 1.3380782918149468,
+      "grad_norm": 0.8123595714569092,
+      "learning_rate": 5.101378463335713e-06,
+      "loss": 0.6279965877532959,
+      "step": 7520
+    },
+    {
+      "epoch": 1.3398576512455516,
+      "grad_norm": 0.7082033753395081,
+      "learning_rate": 5.07658999571238e-06,
+      "loss": 0.6239530086517334,
+      "step": 7530
+    },
+    {
+      "epoch": 1.3416370106761566,
+      "grad_norm": 0.8400927782058716,
+      "learning_rate": 5.051841396351574e-06,
+      "loss": 0.5917630195617676,
+      "step": 7540
+    },
+    {
+      "epoch": 1.3434163701067616,
+      "grad_norm": 0.8962631225585938,
+      "learning_rate": 5.027132865659847e-06,
+      "loss": 0.6083420753479004,
+      "step": 7550
+    },
+    {
+      "epoch": 1.3451957295373664,
+      "grad_norm": 0.9693284630775452,
+      "learning_rate": 5.0024646037193025e-06,
+      "loss": 0.606035566329956,
+      "step": 7560
+    },
+    {
+      "epoch": 1.3469750889679715,
+      "grad_norm": 0.8629393577575684,
+      "learning_rate": 4.977836810285942e-06,
+      "loss": 0.6264309406280517,
+      "step": 7570
+    },
+    {
+      "epoch": 1.3487544483985765,
+      "grad_norm": 0.7961446642875671,
+      "learning_rate": 4.953249684788084e-06,
+      "loss": 0.6090545654296875,
+      "step": 7580
+    },
+    {
+      "epoch": 1.3505338078291815,
+      "grad_norm": 0.7653380036354065,
+      "learning_rate": 4.928703426324712e-06,
+      "loss": 0.6106627941131592,
+      "step": 7590
+    },
+    {
+      "epoch": 1.3523131672597866,
+      "grad_norm": 0.8642345070838928,
+      "learning_rate": 4.9041982336639004e-06,
+      "loss": 0.5813540458679199,
+      "step": 7600
+    },
+    {
+      "epoch": 1.3540925266903914,
+      "grad_norm": 0.7263400554656982,
+      "learning_rate": 4.8797343052411675e-06,
+      "loss": 0.685100507736206,
+      "step": 7610
+    },
+    {
+      "epoch": 1.3558718861209964,
+      "grad_norm": 0.8232746720314026,
+      "learning_rate": 4.855311839157899e-06,
+      "loss": 0.6377671718597412,
+      "step": 7620
+    },
+    {
+      "epoch": 1.3576512455516014,
+      "grad_norm": 0.73633873462677,
+      "learning_rate": 4.830931033179725e-06,
+      "loss": 0.6779125690460205,
+      "step": 7630
+    },
+    {
+      "epoch": 1.3594306049822065,
+      "grad_norm": 0.8550395965576172,
+      "learning_rate": 4.806592084734928e-06,
+      "loss": 0.6058492660522461,
+      "step": 7640
+    },
+    {
+      "epoch": 1.3612099644128115,
+      "grad_norm": 0.7828453779220581,
+      "learning_rate": 4.782295190912831e-06,
+      "loss": 0.6215229034423828,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3629893238434163,
+      "grad_norm": 0.8796820640563965,
+      "learning_rate": 4.758040548462233e-06,
+      "loss": 0.5914856433868408,
+      "step": 7660
+    },
+    {
+      "epoch": 1.3647686832740213,
+      "grad_norm": 0.7559328675270081,
+      "learning_rate": 4.733828353789772e-06,
+      "loss": 0.6324697017669678,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3665480427046264,
+      "grad_norm": 0.9165831804275513,
+      "learning_rate": 4.709658802958366e-06,
+      "loss": 0.6462049961090088,
+      "step": 7680
+    },
+    {
+      "epoch": 1.3683274021352312,
+      "grad_norm": 0.9339064359664917,
+      "learning_rate": 4.685532091685625e-06,
+      "loss": 0.6461282253265381,
+      "step": 7690
+    },
+    {
+      "epoch": 1.3701067615658362,
+      "grad_norm": 0.8191850781440735,
+      "learning_rate": 4.66144841534224e-06,
+      "loss": 0.6682465553283692,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3718861209964412,
+      "grad_norm": 0.9715204238891602,
+      "learning_rate": 4.637407968950434e-06,
+      "loss": 0.5913180351257324,
+      "step": 7710
+    },
+    {
+      "epoch": 1.3736654804270463,
+      "grad_norm": 0.7584412693977356,
+      "learning_rate": 4.613410947182354e-06,
+      "loss": 0.5695419311523438,
+      "step": 7720
+    },
+    {
+      "epoch": 1.3754448398576513,
+      "grad_norm": 0.6991924047470093,
+      "learning_rate": 4.589457544358521e-06,
+      "loss": 0.6356719493865967,
+      "step": 7730
+    },
+    {
+      "epoch": 1.3772241992882561,
+      "grad_norm": 0.8555135726928711,
+      "learning_rate": 4.56554795444623e-06,
+      "loss": 0.5793665409088135,
+      "step": 7740
+    },
+    {
+      "epoch": 1.3790035587188612,
+      "grad_norm": 0.9497565031051636,
+      "learning_rate": 4.5416823710580046e-06,
+      "loss": 0.6186023712158203,
+      "step": 7750
+    },
+    {
+      "epoch": 1.3807829181494662,
+      "grad_norm": 0.8282408118247986,
+      "learning_rate": 4.517860987450007e-06,
+      "loss": 0.6287422180175781,
+      "step": 7760
+    },
+    {
+      "epoch": 1.3825622775800712,
+      "grad_norm": 0.7512138485908508,
+      "learning_rate": 4.494083996520492e-06,
+      "loss": 0.6511450290679932,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3843416370106763,
+      "grad_norm": 0.8342724442481995,
+      "learning_rate": 4.470351590808228e-06,
+      "loss": 0.6355900764465332,
+      "step": 7780
+    },
+    {
+      "epoch": 1.386120996441281,
+      "grad_norm": 0.9951474666595459,
+      "learning_rate": 4.446663962490951e-06,
+      "loss": 0.6075921058654785,
+      "step": 7790
+    },
+    {
+      "epoch": 1.387900355871886,
+      "grad_norm": 0.7905099391937256,
+      "learning_rate": 4.423021303383799e-06,
+      "loss": 0.5925492286682129,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3896797153024911,
+      "grad_norm": 0.8992406129837036,
+      "learning_rate": 4.3994238049377715e-06,
+      "loss": 0.6204987049102784,
+      "step": 7810
+    },
+    {
+      "epoch": 1.3914590747330962,
+      "grad_norm": 1.124568223953247,
+      "learning_rate": 4.375871658238163e-06,
+      "loss": 0.6461961746215821,
+      "step": 7820
+    },
+    {
+      "epoch": 1.3932384341637012,
+      "grad_norm": 0.7916883230209351,
+      "learning_rate": 4.352365054003022e-06,
+      "loss": 0.6402833461761475,
+      "step": 7830
+    },
+    {
+      "epoch": 1.395017793594306,
+      "grad_norm": 1.0332368612289429,
+      "learning_rate": 4.328904182581619e-06,
+      "loss": 0.6191961288452148,
+      "step": 7840
+    },
+    {
+      "epoch": 1.396797153024911,
+      "grad_norm": 1.0284537076950073,
+      "learning_rate": 4.305489233952881e-06,
+      "loss": 0.634841012954712,
+      "step": 7850
+    },
+    {
+      "epoch": 1.398576512455516,
+      "grad_norm": 0.8202362656593323,
+      "learning_rate": 4.282120397723879e-06,
+      "loss": 0.5904129028320313,
+      "step": 7860
+    },
+    {
+      "epoch": 1.4003558718861209,
+      "grad_norm": 0.8149864077568054,
+      "learning_rate": 4.258797863128266e-06,
+      "loss": 0.6212067127227783,
+      "step": 7870
+    },
+    {
+      "epoch": 1.402135231316726,
+      "grad_norm": 0.750307023525238,
+      "learning_rate": 4.235521819024776e-06,
+      "loss": 0.6193465232849121,
+      "step": 7880
+    },
+    {
+      "epoch": 1.403914590747331,
+      "grad_norm": 0.719935417175293,
+      "learning_rate": 4.212292453895658e-06,
+      "loss": 0.6318532466888428,
+      "step": 7890
+    },
+    {
+      "epoch": 1.405693950177936,
+      "grad_norm": 0.8568369150161743,
+      "learning_rate": 4.189109955845186e-06,
+      "loss": 0.6267857074737548,
+      "step": 7900
+    },
+    {
+      "epoch": 1.407473309608541,
+      "grad_norm": 0.8335803151130676,
+      "learning_rate": 4.165974512598102e-06,
+      "loss": 0.6253261089324951,
+      "step": 7910
+    },
+    {
+      "epoch": 1.4092526690391458,
+      "grad_norm": 0.9222790598869324,
+      "learning_rate": 4.142886311498133e-06,
+      "loss": 0.6295660972595215,
+      "step": 7920
+    },
+    {
+      "epoch": 1.4110320284697508,
+      "grad_norm": 0.7769973874092102,
+      "learning_rate": 4.119845539506436e-06,
+      "loss": 0.6295949935913085,
+      "step": 7930
+    },
+    {
+      "epoch": 1.4128113879003559,
+      "grad_norm": 0.8693393468856812,
+      "learning_rate": 4.096852383200106e-06,
+      "loss": 0.5903539180755615,
+      "step": 7940
+    },
+    {
+      "epoch": 1.414590747330961,
+      "grad_norm": 0.8232077956199646,
+      "learning_rate": 4.073907028770671e-06,
+      "loss": 0.5986703395843506,
+      "step": 7950
+    },
+    {
+      "epoch": 1.416370106761566,
+      "grad_norm": 0.9507026672363281,
+      "learning_rate": 4.051009662022559e-06,
+      "loss": 0.6164099693298339,
+      "step": 7960
+    },
+    {
+      "epoch": 1.4181494661921707,
+      "grad_norm": 0.9032939672470093,
+      "learning_rate": 4.028160468371618e-06,
+      "loss": 0.5873546600341797,
+      "step": 7970
+    },
+    {
+      "epoch": 1.4199288256227758,
+      "grad_norm": 0.7442892789840698,
+      "learning_rate": 4.005359632843598e-06,
+      "loss": 0.6300751209259033,
+      "step": 7980
+    },
+    {
+      "epoch": 1.4217081850533808,
+      "grad_norm": 0.870611310005188,
+      "learning_rate": 3.982607340072673e-06,
+      "loss": 0.6464887142181397,
+      "step": 7990
+    },
+    {
+      "epoch": 1.4234875444839858,
+      "grad_norm": 0.9346739053726196,
+      "learning_rate": 3.959903774299914e-06,
+      "loss": 0.5978512763977051,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4234875444839858,
+      "eval_loss": 0.6126887202262878,
+      "eval_runtime": 408.0866,
+      "eval_samples_per_second": 12.323,
+      "eval_steps_per_second": 6.163,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4252669039145909,
+      "grad_norm": 0.7768172025680542,
+      "learning_rate": 3.937249119371837e-06,
+      "loss": 0.6360022544860839,
+      "step": 8010
+    },
+    {
+      "epoch": 1.4270462633451957,
+      "grad_norm": 0.9232711791992188,
+      "learning_rate": 3.914643558738871e-06,
+      "loss": 0.6255642414093018,
+      "step": 8020
+    },
+    {
+      "epoch": 1.4288256227758007,
+      "grad_norm": 0.6983018517494202,
+      "learning_rate": 3.892087275453913e-06,
+      "loss": 0.5994945526123047,
+      "step": 8030
+    },
+    {
+      "epoch": 1.4306049822064058,
+      "grad_norm": 1.073339819908142,
+      "learning_rate": 3.869580452170813e-06,
+      "loss": 0.6356189250946045,
+      "step": 8040
+    },
+    {
+      "epoch": 1.4323843416370106,
+      "grad_norm": 0.6503099799156189,
+      "learning_rate": 3.847123271142921e-06,
+      "loss": 0.6280940055847168,
+      "step": 8050
+    },
+    {
+      "epoch": 1.4341637010676156,
+      "grad_norm": 0.7652536630630493,
+      "learning_rate": 3.824715914221593e-06,
+      "loss": 0.633913803100586,
+      "step": 8060
+    },
+    {
+      "epoch": 1.4359430604982206,
+      "grad_norm": 0.7776927351951599,
+      "learning_rate": 3.802358562854719e-06,
+      "loss": 0.6014345169067383,
+      "step": 8070
+    },
+    {
+      "epoch": 1.4377224199288257,
+      "grad_norm": 0.9608566164970398,
+      "learning_rate": 3.780051398085274e-06,
+      "loss": 0.6504060745239257,
+      "step": 8080
+    },
+    {
+      "epoch": 1.4395017793594307,
+      "grad_norm": 1.0493940114974976,
+      "learning_rate": 3.7577946005498224e-06,
+      "loss": 0.6023014545440674,
+      "step": 8090
+    },
+    {
+      "epoch": 1.4412811387900355,
+      "grad_norm": 0.6732901930809021,
+      "learning_rate": 3.735588350477083e-06,
+      "loss": 0.618662166595459,
+      "step": 8100
+    },
+    {
+      "epoch": 1.4430604982206405,
+      "grad_norm": 0.7948004007339478,
+      "learning_rate": 3.7134328276864474e-06,
+      "loss": 0.6384446144104003,
+      "step": 8110
+    },
+    {
+      "epoch": 1.4448398576512456,
+      "grad_norm": 0.8854271769523621,
+      "learning_rate": 3.691328211586537e-06,
+      "loss": 0.6317539215087891,
+      "step": 8120
+    },
+    {
+      "epoch": 1.4466192170818506,
+      "grad_norm": 0.849312961101532,
+      "learning_rate": 3.669274681173741e-06,
+      "loss": 0.632611894607544,
+      "step": 8130
+    },
+    {
+      "epoch": 1.4483985765124556,
+      "grad_norm": 0.8737841844558716,
+      "learning_rate": 3.647272415030787e-06,
+      "loss": 0.6378211498260498,
+      "step": 8140
+    },
+    {
+      "epoch": 1.4501779359430604,
+      "grad_norm": 0.9041767716407776,
+      "learning_rate": 3.6253215913252614e-06,
+      "loss": 0.6225557804107666,
+      "step": 8150
+    },
+    {
+      "epoch": 1.4519572953736655,
+      "grad_norm": 0.9240919947624207,
+      "learning_rate": 3.603422387808203e-06,
+      "loss": 0.6011790752410888,
+      "step": 8160
+    },
+    {
+      "epoch": 1.4537366548042705,
+      "grad_norm": 0.7381271719932556,
+      "learning_rate": 3.5815749818126298e-06,
+      "loss": 0.6100322723388671,
+      "step": 8170
+    },
+    {
+      "epoch": 1.4555160142348753,
+      "grad_norm": 0.766522228717804,
+      "learning_rate": 3.559779550252135e-06,
+      "loss": 0.6223373413085938,
+      "step": 8180
+    },
+    {
+      "epoch": 1.4572953736654806,
+      "grad_norm": 0.722597599029541,
+      "learning_rate": 3.5380362696194246e-06,
+      "loss": 0.6096580982208252,
+      "step": 8190
+    },
+    {
+      "epoch": 1.4590747330960854,
+      "grad_norm": 0.848345935344696,
+      "learning_rate": 3.5163453159849158e-06,
+      "loss": 0.5932037830352783,
+      "step": 8200
+    },
+    {
+      "epoch": 1.4608540925266904,
+      "grad_norm": 0.8649978637695312,
+      "learning_rate": 3.4947068649952875e-06,
+      "loss": 0.616939926147461,
+      "step": 8210
+    },
+    {
+      "epoch": 1.4626334519572954,
+      "grad_norm": 1.0290734767913818,
+      "learning_rate": 3.473121091872068e-06,
+      "loss": 0.5814998149871826,
+      "step": 8220
+    },
+    {
+      "epoch": 1.4644128113879002,
+      "grad_norm": 1.0387070178985596,
+      "learning_rate": 3.4515881714102283e-06,
+      "loss": 0.602289867401123,
+      "step": 8230
+    },
+    {
+      "epoch": 1.4661921708185053,
+      "grad_norm": 0.7532253861427307,
+      "learning_rate": 3.4301082779767394e-06,
+      "loss": 0.6335249423980713,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4679715302491103,
+      "grad_norm": 0.8629032373428345,
+      "learning_rate": 3.4086815855091906e-06,
+      "loss": 0.6227278709411621,
+      "step": 8250
+    },
+    {
+      "epoch": 1.4697508896797153,
+      "grad_norm": 0.9435595870018005,
+      "learning_rate": 3.3873082675143533e-06,
+      "loss": 0.6437982559204102,
+      "step": 8260
+    },
+    {
+      "epoch": 1.4715302491103204,
+      "grad_norm": 0.9069509506225586,
+      "learning_rate": 3.3659884970667955e-06,
+      "loss": 0.5948871612548828,
+      "step": 8270
+    },
+    {
+      "epoch": 1.4733096085409252,
+      "grad_norm": 0.7571467757225037,
+      "learning_rate": 3.344722446807469e-06,
+      "loss": 0.6302636623382568,
+      "step": 8280
+    },
+    {
+      "epoch": 1.4750889679715302,
+      "grad_norm": 0.7709717154502869,
+      "learning_rate": 3.3235102889423263e-06,
+      "loss": 0.6287346839904785,
+      "step": 8290
+    },
+    {
+      "epoch": 1.4768683274021353,
+      "grad_norm": 0.9050244688987732,
+      "learning_rate": 3.302352195240901e-06,
+      "loss": 0.6431692123413086,
+      "step": 8300
+    },
+    {
+      "epoch": 1.4786476868327403,
+      "grad_norm": 0.8429788947105408,
+      "learning_rate": 3.281248337034947e-06,
+      "loss": 0.6204410076141358,
+      "step": 8310
+    },
+    {
+      "epoch": 1.4804270462633453,
+      "grad_norm": 0.7680916786193848,
+      "learning_rate": 3.2601988852170207e-06,
+      "loss": 0.6185726642608642,
+      "step": 8320
+    },
+    {
+      "epoch": 1.4822064056939501,
+      "grad_norm": 0.7400604486465454,
+      "learning_rate": 3.2392040102391278e-06,
+      "loss": 0.711566686630249,
+      "step": 8330
+    },
+    {
+      "epoch": 1.4839857651245552,
+      "grad_norm": 0.7004432082176208,
+      "learning_rate": 3.2182638821113156e-06,
+      "loss": 0.5973163604736328,
+      "step": 8340
+    },
+    {
+      "epoch": 1.4857651245551602,
+      "grad_norm": 0.8438174724578857,
+      "learning_rate": 3.1973786704003086e-06,
+      "loss": 0.6273365497589112,
+      "step": 8350
+    },
+    {
+      "epoch": 1.487544483985765,
+      "grad_norm": 0.9420303702354431,
+      "learning_rate": 3.1765485442281453e-06,
+      "loss": 0.6080061435699463,
+      "step": 8360
+    },
+    {
+      "epoch": 1.48932384341637,
+      "grad_norm": 0.8402919769287109,
+      "learning_rate": 3.1557736722707843e-06,
+      "loss": 0.6027824401855468,
+      "step": 8370
+    },
+    {
+      "epoch": 1.491103202846975,
+      "grad_norm": 1.0150177478790283,
+      "learning_rate": 3.1350542227567693e-06,
+      "loss": 0.629277515411377,
+      "step": 8380
+    },
+    {
+      "epoch": 1.49288256227758,
+      "grad_norm": 0.8253095149993896,
+      "learning_rate": 3.1143903634658314e-06,
+      "loss": 0.6055563926696778,
+      "step": 8390
+    },
+    {
+      "epoch": 1.4946619217081851,
+      "grad_norm": 0.7219249606132507,
+      "learning_rate": 3.093782261727567e-06,
+      "loss": 0.6305870532989502,
+      "step": 8400
+    },
+    {
+      "epoch": 1.49644128113879,
+      "grad_norm": 0.7794970870018005,
+      "learning_rate": 3.073230084420051e-06,
+      "loss": 0.6157556533813476,
+      "step": 8410
+    },
+    {
+      "epoch": 1.498220640569395,
+      "grad_norm": 0.8257132172584534,
+      "learning_rate": 3.0527339979685068e-06,
+      "loss": 0.628666877746582,
+      "step": 8420
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7853142619132996,
+      "learning_rate": 3.0322941683439455e-06,
+      "loss": 0.6387495040893555,
+      "step": 8430
+    },
+    {
+      "epoch": 1.501779359430605,
+      "grad_norm": 0.9834029674530029,
+      "learning_rate": 3.011910761061837e-06,
+      "loss": 0.6235998153686524,
+      "step": 8440
+    },
+    {
+      "epoch": 1.50355871886121,
+      "grad_norm": 0.7815040349960327,
+      "learning_rate": 2.99158394118075e-06,
+      "loss": 0.6229785919189453,
+      "step": 8450
+    },
+    {
+      "epoch": 1.5053380782918149,
+      "grad_norm": 0.6993287801742554,
+      "learning_rate": 2.9713138733010373e-06,
+      "loss": 0.6498080730438233,
+      "step": 8460
+    },
+    {
+      "epoch": 1.50711743772242,
+      "grad_norm": 0.8505419492721558,
+      "learning_rate": 2.951100721563479e-06,
+      "loss": 0.6163151264190674,
+      "step": 8470
+    },
+    {
+      "epoch": 1.508896797153025,
+      "grad_norm": 0.7740962505340576,
+      "learning_rate": 2.930944649647971e-06,
+      "loss": 0.6463754177093506,
+      "step": 8480
+    },
+    {
+      "epoch": 1.5106761565836297,
+      "grad_norm": 0.9081360101699829,
+      "learning_rate": 2.9108458207722013e-06,
+      "loss": 0.5963332653045654,
+      "step": 8490
+    },
+    {
+      "epoch": 1.512455516014235,
+      "grad_norm": 0.7486013770103455,
+      "learning_rate": 2.8908043976903065e-06,
+      "loss": 0.5907905101776123,
+      "step": 8500
+    },
+    {
+      "epoch": 1.512455516014235,
+      "eval_loss": 0.6111557483673096,
+      "eval_runtime": 408.1089,
+      "eval_samples_per_second": 12.323,
+      "eval_steps_per_second": 6.163,
+      "step": 8500
+    },
+    {
+      "epoch": 1.5142348754448398,
+      "grad_norm": 0.6754533648490906,
+      "learning_rate": 2.8708205426915858e-06,
+      "loss": 0.5735606670379638,
+      "step": 8510
+    },
+    {
+      "epoch": 1.5160142348754448,
+      "grad_norm": 0.9227625727653503,
+      "learning_rate": 2.850894417599154e-06,
+      "loss": 0.586240005493164,
+      "step": 8520
+    },
+    {
+      "epoch": 1.5177935943060499,
+      "grad_norm": 1.1501375436782837,
+      "learning_rate": 2.8310261837686594e-06,
+      "loss": 0.5986839771270752,
+      "step": 8530
+    },
+    {
+      "epoch": 1.5195729537366547,
+      "grad_norm": 0.8052434325218201,
+      "learning_rate": 2.811216002086954e-06,
+      "loss": 0.6387444972991944,
+      "step": 8540
+    },
+    {
+      "epoch": 1.52135231316726,
+      "grad_norm": 0.9873703718185425,
+      "learning_rate": 2.791464032970812e-06,
+      "loss": 0.6114506244659423,
+      "step": 8550
+    },
+    {
+      "epoch": 1.5231316725978647,
+      "grad_norm": 0.8305763006210327,
+      "learning_rate": 2.771770436365612e-06,
+      "loss": 0.65102219581604,
+      "step": 8560
+    },
+    {
+      "epoch": 1.5249110320284698,
+      "grad_norm": 0.9540684223175049,
+      "learning_rate": 2.7521353717440523e-06,
+      "loss": 0.6242643833160401,
+      "step": 8570
+    },
+    {
+      "epoch": 1.5266903914590748,
+      "grad_norm": 0.764707624912262,
+      "learning_rate": 2.732558998104855e-06,
+      "loss": 0.6340816020965576,
+      "step": 8580
+    },
+    {
+      "epoch": 1.5284697508896796,
+      "grad_norm": 0.8800462484359741,
+      "learning_rate": 2.7130414739714884e-06,
+      "loss": 0.6269487857818603,
+      "step": 8590
+    },
+    {
+      "epoch": 1.5302491103202847,
+      "grad_norm": 0.7068182826042175,
+      "learning_rate": 2.6935829573908645e-06,
+      "loss": 0.6376915454864502,
+      "step": 8600
+    },
+    {
+      "epoch": 1.5320284697508897,
+      "grad_norm": 0.9243912100791931,
+      "learning_rate": 2.6741836059320813e-06,
+      "loss": 0.6650017261505127,
+      "step": 8610
+    },
+    {
+      "epoch": 1.5338078291814945,
+      "grad_norm": 0.8034710884094238,
+      "learning_rate": 2.654843576685129e-06,
+      "loss": 0.6140787601470947,
+      "step": 8620
+    },
+    {
+      "epoch": 1.5355871886120998,
+      "grad_norm": 0.6924305558204651,
+      "learning_rate": 2.635563026259622e-06,
+      "loss": 0.592177438735962,
+      "step": 8630
+    },
+    {
+      "epoch": 1.5373665480427046,
+      "grad_norm": 0.9960291385650635,
+      "learning_rate": 2.616342110783544e-06,
+      "loss": 0.620767879486084,
+      "step": 8640
+    },
+    {
+      "epoch": 1.5391459074733096,
+      "grad_norm": 0.9700925350189209,
+      "learning_rate": 2.5971809859019616e-06,
+      "loss": 0.6059544563293457,
+      "step": 8650
+    },
+    {
+      "epoch": 1.5409252669039146,
+      "grad_norm": 0.8979527354240417,
+      "learning_rate": 2.578079806775786e-06,
+      "loss": 0.6165768146514893,
+      "step": 8660
+    },
+    {
+      "epoch": 1.5427046263345194,
+      "grad_norm": 0.7694889903068542,
+      "learning_rate": 2.559038728080495e-06,
+      "loss": 0.6269434452056885,
+      "step": 8670
+    },
+    {
+      "epoch": 1.5444839857651247,
+      "grad_norm": 0.7812192440032959,
+      "learning_rate": 2.5400579040049045e-06,
+      "loss": 0.6156674385070801,
+      "step": 8680
+    },
+    {
+      "epoch": 1.5462633451957295,
+      "grad_norm": 0.8593633770942688,
+      "learning_rate": 2.521137488249892e-06,
+      "loss": 0.5995303630828858,
+      "step": 8690
+    },
+    {
+      "epoch": 1.5480427046263345,
+      "grad_norm": 0.7315478920936584,
+      "learning_rate": 2.5022776340271827e-06,
+      "loss": 0.6255430698394775,
+      "step": 8700
+    },
+    {
+      "epoch": 1.5498220640569396,
+      "grad_norm": 0.7979263663291931,
+      "learning_rate": 2.483478494058081e-06,
+      "loss": 0.617695426940918,
+      "step": 8710
+    },
+    {
+      "epoch": 1.5516014234875444,
+      "grad_norm": 0.791521430015564,
+      "learning_rate": 2.4647402205722513e-06,
+      "loss": 0.5985545158386231,
+      "step": 8720
+    },
+    {
+      "epoch": 1.5533807829181496,
+      "grad_norm": 0.8115281462669373,
+      "learning_rate": 2.446062965306476e-06,
+      "loss": 0.6137414932250976,
+      "step": 8730
+    },
+    {
+      "epoch": 1.5551601423487544,
+      "grad_norm": 0.73805832862854,
+      "learning_rate": 2.427446879503439e-06,
+      "loss": 0.6292818069458008,
+      "step": 8740
+    },
+    {
+      "epoch": 1.5569395017793595,
+      "grad_norm": 0.7681689262390137,
+      "learning_rate": 2.408892113910485e-06,
+      "loss": 0.6184986114501954,
+      "step": 8750
+    },
+    {
+      "epoch": 1.5587188612099645,
+      "grad_norm": 0.8392589688301086,
+      "learning_rate": 2.390398818778403e-06,
+      "loss": 0.6432001113891601,
+      "step": 8760
+    },
+    {
+      "epoch": 1.5604982206405693,
+      "grad_norm": 0.7140183448791504,
+      "learning_rate": 2.3719671438602287e-06,
+      "loss": 0.5931034088134766,
+      "step": 8770
+    },
+    {
+      "epoch": 1.5622775800711743,
+      "grad_norm": 1.0340561866760254,
+      "learning_rate": 2.353597238409997e-06,
+      "loss": 0.6184981346130372,
+      "step": 8780
+    },
+    {
+      "epoch": 1.5640569395017794,
+      "grad_norm": 1.3788442611694336,
+      "learning_rate": 2.3352892511815695e-06,
+      "loss": 0.6254217147827148,
+      "step": 8790
+    },
+    {
+      "epoch": 1.5658362989323842,
+      "grad_norm": 0.8685413599014282,
+      "learning_rate": 2.3170433304274e-06,
+      "loss": 0.6440535545349121,
+      "step": 8800
+    },
+    {
+      "epoch": 1.5676156583629894,
+      "grad_norm": 0.8243607878684998,
+      "learning_rate": 2.298859623897357e-06,
+      "loss": 0.6024580001831055,
+      "step": 8810
+    },
+    {
+      "epoch": 1.5693950177935942,
+      "grad_norm": 0.9618055820465088,
+      "learning_rate": 2.280738278837508e-06,
+      "loss": 0.6213048934936524,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5711743772241993,
+      "grad_norm": 0.8591263294219971,
+      "learning_rate": 2.2626794419889463e-06,
+      "loss": 0.6279497146606445,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5729537366548043,
+      "grad_norm": 0.8621962070465088,
+      "learning_rate": 2.2446832595865833e-06,
+      "loss": 0.5932358741760254,
+      "step": 8840
+    },
+    {
+      "epoch": 1.5747330960854091,
+      "grad_norm": 0.9073888659477234,
+      "learning_rate": 2.226749877357983e-06,
+      "loss": 0.6451629161834717,
+      "step": 8850
+    },
+    {
+      "epoch": 1.5765124555160144,
+      "grad_norm": 0.9106960892677307,
+      "learning_rate": 2.208879440522167e-06,
+      "loss": 0.6058315753936767,
+      "step": 8860
+    },
+    {
+      "epoch": 1.5782918149466192,
+      "grad_norm": 0.7805183529853821,
+      "learning_rate": 2.1910720937884432e-06,
+      "loss": 0.6308177947998047,
+      "step": 8870
+    },
+    {
+      "epoch": 1.5800711743772242,
+      "grad_norm": 0.8798645734786987,
+      "learning_rate": 2.1733279813552386e-06,
+      "loss": 0.6120688438415527,
+      "step": 8880
+    },
+    {
+      "epoch": 1.5818505338078293,
+      "grad_norm": 0.7852058410644531,
+      "learning_rate": 2.1556472469089305e-06,
+      "loss": 0.6294644832611084,
+      "step": 8890
+    },
+    {
+      "epoch": 1.583629893238434,
+      "grad_norm": 1.0219141244888306,
+      "learning_rate": 2.1380300336226756e-06,
+      "loss": 0.6379860877990723,
+      "step": 8900
+    },
+    {
+      "epoch": 1.585409252669039,
+      "grad_norm": 0.8301982879638672,
+      "learning_rate": 2.120476484155255e-06,
+      "loss": 0.5950196743011474,
+      "step": 8910
+    },
+    {
+      "epoch": 1.5871886120996441,
+      "grad_norm": 0.781911313533783,
+      "learning_rate": 2.102986740649928e-06,
+      "loss": 0.6155390739440918,
+      "step": 8920
+    },
+    {
+      "epoch": 1.5889679715302492,
+      "grad_norm": 0.8553647994995117,
+      "learning_rate": 2.0855609447332635e-06,
+      "loss": 0.6030520439147949,
+      "step": 8930
+    },
+    {
+      "epoch": 1.5907473309608542,
+      "grad_norm": 0.8010233640670776,
+      "learning_rate": 2.06819923751401e-06,
+      "loss": 0.5782717227935791,
+      "step": 8940
+    },
+    {
+      "epoch": 1.592526690391459,
+      "grad_norm": 0.7767868041992188,
+      "learning_rate": 2.050901759581937e-06,
+      "loss": 0.5862733364105225,
+      "step": 8950
+    },
+    {
+      "epoch": 1.594306049822064,
+      "grad_norm": 0.894256591796875,
+      "learning_rate": 2.033668651006715e-06,
+      "loss": 0.6170249938964844,
+      "step": 8960
+    },
+    {
+      "epoch": 1.596085409252669,
+      "grad_norm": 0.8816112875938416,
+      "learning_rate": 2.0165000513367604e-06,
+      "loss": 0.5960803508758545,
+      "step": 8970
+    },
+    {
+      "epoch": 1.5978647686832739,
+      "grad_norm": 0.7646706700325012,
+      "learning_rate": 1.9993960995981287e-06,
+      "loss": 0.609787130355835,
+      "step": 8980
+    },
+    {
+      "epoch": 1.5996441281138791,
+      "grad_norm": 0.8380435109138489,
+      "learning_rate": 1.9823569342933624e-06,
+      "loss": 0.5809425354003906,
+      "step": 8990
+    },
+    {
+      "epoch": 1.601423487544484,
+      "grad_norm": 0.7763661742210388,
+      "learning_rate": 1.965382693400396e-06,
+      "loss": 0.6282608032226562,
+      "step": 9000
+    },
+    {
+      "epoch": 1.601423487544484,
+      "eval_loss": 0.6098406314849854,
+      "eval_runtime": 407.9599,
+      "eval_samples_per_second": 12.327,
+      "eval_steps_per_second": 6.165,
+      "step": 9000
+    },
+    {
+      "epoch": 1.603202846975089,
+      "grad_norm": 0.7980552315711975,
+      "learning_rate": 1.9484735143714184e-06,
+      "loss": 0.5892675876617431,
+      "step": 9010
+    },
+    {
+      "epoch": 1.604982206405694,
+      "grad_norm": 0.7350971102714539,
+      "learning_rate": 1.931629534131769e-06,
+      "loss": 0.615026330947876,
+      "step": 9020
+    },
+    {
+      "epoch": 1.6067615658362988,
+      "grad_norm": 0.9350213408470154,
+      "learning_rate": 1.9148508890788263e-06,
+      "loss": 0.5945257663726806,
+      "step": 9030
+    },
+    {
+      "epoch": 1.608540925266904,
+      "grad_norm": 0.8562370538711548,
+      "learning_rate": 1.8981377150809111e-06,
+      "loss": 0.6524335384368897,
+      "step": 9040
+    },
+    {
+      "epoch": 1.6103202846975089,
+      "grad_norm": 0.8481477499008179,
+      "learning_rate": 1.8814901474761704e-06,
+      "loss": 0.572668981552124,
+      "step": 9050
+    },
+    {
+      "epoch": 1.612099644128114,
+      "grad_norm": 0.8405110239982605,
+      "learning_rate": 1.8649083210714946e-06,
+      "loss": 0.6025730609893799,
+      "step": 9060
+    },
+    {
+      "epoch": 1.613879003558719,
+      "grad_norm": 0.6442246437072754,
+      "learning_rate": 1.8483923701414274e-06,
+      "loss": 0.619115161895752,
+      "step": 9070
+    },
+    {
+      "epoch": 1.6156583629893237,
+      "grad_norm": 0.9033611416816711,
+      "learning_rate": 1.8319424284270638e-06,
+      "loss": 0.6313360214233399,
+      "step": 9080
+    },
+    {
+      "epoch": 1.6174377224199288,
+      "grad_norm": 0.8982630968093872,
+      "learning_rate": 1.8155586291349836e-06,
+      "loss": 0.6707229137420654,
+      "step": 9090
+    },
+    {
+      "epoch": 1.6192170818505338,
+      "grad_norm": 0.8548517823219299,
+      "learning_rate": 1.7992411049361612e-06,
+      "loss": 0.5913959503173828,
+      "step": 9100
+    },
+    {
+      "epoch": 1.6209964412811388,
+      "grad_norm": 0.8158797025680542,
+      "learning_rate": 1.7829899879649005e-06,
+      "loss": 0.6455716133117676,
+      "step": 9110
+    },
+    {
+      "epoch": 1.6227758007117439,
+      "grad_norm": 1.0375958681106567,
+      "learning_rate": 1.7668054098177512e-06,
+      "loss": 0.6378812789916992,
+      "step": 9120
+    },
+    {
+      "epoch": 1.6245551601423487,
+      "grad_norm": 0.9441247582435608,
+      "learning_rate": 1.7506875015524649e-06,
+      "loss": 0.6197398662567138,
+      "step": 9130
+    },
+    {
+      "epoch": 1.6263345195729537,
+      "grad_norm": 0.7956865429878235,
+      "learning_rate": 1.7346363936869082e-06,
+      "loss": 0.6085912704467773,
+      "step": 9140
+    },
+    {
+      "epoch": 1.6281138790035588,
+      "grad_norm": 0.7922634482383728,
+      "learning_rate": 1.7186522161980279e-06,
+      "loss": 0.6055495738983154,
+      "step": 9150
+    },
+    {
+      "epoch": 1.6298932384341636,
+      "grad_norm": 0.9479956030845642,
+      "learning_rate": 1.7027350985207847e-06,
+      "loss": 0.5977864265441895,
+      "step": 9160
+    },
+    {
+      "epoch": 1.6316725978647688,
+      "grad_norm": 0.8339085578918457,
+      "learning_rate": 1.68688516954711e-06,
+      "loss": 0.6118191242218017,
+      "step": 9170
+    },
+    {
+      "epoch": 1.6334519572953736,
+      "grad_norm": 0.8879317045211792,
+      "learning_rate": 1.6711025576248586e-06,
+      "loss": 0.607643461227417,
+      "step": 9180
+    },
+    {
+      "epoch": 1.6352313167259787,
+      "grad_norm": 0.7563620805740356,
+      "learning_rate": 1.655387390556782e-06,
+      "loss": 0.566606092453003,
+      "step": 9190
+    },
+    {
+      "epoch": 1.6370106761565837,
+      "grad_norm": 0.8990421295166016,
+      "learning_rate": 1.6397397955994742e-06,
+      "loss": 0.6038555145263672,
+      "step": 9200
+    },
+    {
+      "epoch": 1.6387900355871885,
+      "grad_norm": 0.8692913055419922,
+      "learning_rate": 1.624159899462353e-06,
+      "loss": 0.6045926570892334,
+      "step": 9210
+    },
+    {
+      "epoch": 1.6405693950177938,
+      "grad_norm": 0.7604990005493164,
+      "learning_rate": 1.6086478283066386e-06,
+      "loss": 0.6289362907409668,
+      "step": 9220
+    },
+    {
+      "epoch": 1.6423487544483986,
+      "grad_norm": 0.8552067875862122,
+      "learning_rate": 1.5932037077443163e-06,
+      "loss": 0.5995774269104004,
+      "step": 9230
+    },
+    {
+      "epoch": 1.6441281138790036,
+      "grad_norm": 0.9507073760032654,
+      "learning_rate": 1.577827662837136e-06,
+      "loss": 0.6592823028564453,
+      "step": 9240
+    },
+    {
+      "epoch": 1.6459074733096086,
+      "grad_norm": 0.7741204500198364,
+      "learning_rate": 1.5625198180955836e-06,
+      "loss": 0.5817698001861572,
+      "step": 9250
+    },
+    {
+      "epoch": 1.6476868327402134,
+      "grad_norm": 0.906514048576355,
+      "learning_rate": 1.5472802974778911e-06,
+      "loss": 0.6105581760406494,
+      "step": 9260
+    },
+    {
+      "epoch": 1.6494661921708185,
+      "grad_norm": 1.1253999471664429,
+      "learning_rate": 1.5321092243890112e-06,
+      "loss": 0.6198730945587159,
+      "step": 9270
+    },
+    {
+      "epoch": 1.6512455516014235,
+      "grad_norm": 0.7494739294052124,
+      "learning_rate": 1.5170067216796425e-06,
+      "loss": 0.5921574115753174,
+      "step": 9280
+    },
+    {
+      "epoch": 1.6530249110320283,
+      "grad_norm": 0.674996018409729,
+      "learning_rate": 1.5019729116452086e-06,
+      "loss": 0.6231479644775391,
+      "step": 9290
+    },
+    {
+      "epoch": 1.6548042704626336,
+      "grad_norm": 0.8674888610839844,
+      "learning_rate": 1.4870079160248918e-06,
+      "loss": 0.5954656600952148,
+      "step": 9300
+    },
+    {
+      "epoch": 1.6565836298932384,
+      "grad_norm": 0.8911182284355164,
+      "learning_rate": 1.4721118560006298e-06,
+      "loss": 0.5914658069610595,
+      "step": 9310
+    },
+    {
+      "epoch": 1.6583629893238434,
+      "grad_norm": 0.9357933402061462,
+      "learning_rate": 1.4572848521961414e-06,
+      "loss": 0.6310736656188964,
+      "step": 9320
+    },
+    {
+      "epoch": 1.6601423487544484,
+      "grad_norm": 0.936082124710083,
+      "learning_rate": 1.4425270246759549e-06,
+      "loss": 0.5978594303131104,
+      "step": 9330
+    },
+    {
+      "epoch": 1.6619217081850532,
+      "grad_norm": 0.8404752016067505,
+      "learning_rate": 1.4278384929444233e-06,
+      "loss": 0.5916398048400879,
+      "step": 9340
+    },
+    {
+      "epoch": 1.6637010676156585,
+      "grad_norm": 0.6900134086608887,
+      "learning_rate": 1.4132193759447665e-06,
+      "loss": 0.6098292350769043,
+      "step": 9350
+    },
+    {
+      "epoch": 1.6654804270462633,
+      "grad_norm": 0.8345168828964233,
+      "learning_rate": 1.398669792058105e-06,
+      "loss": 0.6169853687286377,
+      "step": 9360
+    },
+    {
+      "epoch": 1.6672597864768683,
+      "grad_norm": 0.8186418414115906,
+      "learning_rate": 1.384189859102505e-06,
+      "loss": 0.6554735660552978,
+      "step": 9370
+    },
+    {
+      "epoch": 1.6690391459074734,
+      "grad_norm": 0.8094502687454224,
+      "learning_rate": 1.369779694332012e-06,
+      "loss": 0.57339506149292,
+      "step": 9380
+    },
+    {
+      "epoch": 1.6708185053380782,
+      "grad_norm": 0.8150178790092468,
+      "learning_rate": 1.3554394144357218e-06,
+      "loss": 0.5624317169189453,
+      "step": 9390
+    },
+    {
+      "epoch": 1.6725978647686834,
+      "grad_norm": 0.7492024898529053,
+      "learning_rate": 1.341169135536814e-06,
+      "loss": 0.619264554977417,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6743772241992882,
+      "grad_norm": 0.7049137949943542,
+      "learning_rate": 1.326968973191628e-06,
+      "loss": 0.5918323040008545,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6761565836298933,
+      "grad_norm": 0.8279508948326111,
+      "learning_rate": 1.312839042388716e-06,
+      "loss": 0.6143953800201416,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6779359430604983,
+      "grad_norm": 1.0093584060668945,
+      "learning_rate": 1.2987794575479218e-06,
+      "loss": 0.6140541076660156,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6797153024911031,
+      "grad_norm": 0.9923004508018494,
+      "learning_rate": 1.2847903325194455e-06,
+      "loss": 0.6000154495239258,
+      "step": 9440
+    },
+    {
+      "epoch": 1.6814946619217082,
+      "grad_norm": 0.9289199113845825,
+      "learning_rate": 1.270871780582924e-06,
+      "loss": 0.6113490104675293,
+      "step": 9450
+    },
+    {
+      "epoch": 1.6832740213523132,
+      "grad_norm": 0.6759699583053589,
+      "learning_rate": 1.2570239144465212e-06,
+      "loss": 0.6019508838653564,
+      "step": 9460
+    },
+    {
+      "epoch": 1.685053380782918,
+      "grad_norm": 0.8999039530754089,
+      "learning_rate": 1.2432468462460024e-06,
+      "loss": 0.6267662048339844,
+      "step": 9470
+    },
+    {
+      "epoch": 1.6868327402135233,
+      "grad_norm": 0.9530540704727173,
+      "learning_rate": 1.2295406875438377e-06,
+      "loss": 0.5737581729888916,
+      "step": 9480
+    },
+    {
+      "epoch": 1.688612099644128,
+      "grad_norm": 0.9010872840881348,
+      "learning_rate": 1.2159055493282911e-06,
+      "loss": 0.6142421722412109,
+      "step": 9490
+    },
+    {
+      "epoch": 1.690391459074733,
+      "grad_norm": 0.9528436660766602,
+      "learning_rate": 1.2023415420125262e-06,
+      "loss": 0.5961336612701416,
+      "step": 9500
+    },
+    {
+      "epoch": 1.690391459074733,
+      "eval_loss": 0.6090449094772339,
+      "eval_runtime": 409.4467,
+      "eval_samples_per_second": 12.282,
+      "eval_steps_per_second": 6.142,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6921708185053381,
+      "grad_norm": 0.863777220249176,
+      "learning_rate": 1.1888487754337052e-06,
+      "loss": 0.6345890045166016,
+      "step": 9510
+    },
+    {
+      "epoch": 1.693950177935943,
+      "grad_norm": 1.0439646244049072,
+      "learning_rate": 1.1754273588521149e-06,
+      "loss": 0.6450316429138183,
+      "step": 9520
+    },
+    {
+      "epoch": 1.6957295373665482,
+      "grad_norm": 0.8258497714996338,
+      "learning_rate": 1.162077400950261e-06,
+      "loss": 0.6218976497650146,
+      "step": 9530
+    },
+    {
+      "epoch": 1.697508896797153,
+      "grad_norm": 0.8564954996109009,
+      "learning_rate": 1.148799009832009e-06,
+      "loss": 0.6231951713562012,
+      "step": 9540
+    },
+    {
+      "epoch": 1.699288256227758,
+      "grad_norm": 0.7925037741661072,
+      "learning_rate": 1.1355922930216867e-06,
+      "loss": 0.5838134765625,
+      "step": 9550
+    },
+    {
+      "epoch": 1.701067615658363,
+      "grad_norm": 0.8268294930458069,
+      "learning_rate": 1.122457357463238e-06,
+      "loss": 0.6027390480041503,
+      "step": 9560
+    },
+    {
+      "epoch": 1.7028469750889679,
+      "grad_norm": 0.9756558537483215,
+      "learning_rate": 1.1093943095193328e-06,
+      "loss": 0.6181758403778076,
+      "step": 9570
+    },
+    {
+      "epoch": 1.704626334519573,
+      "grad_norm": 0.8576071262359619,
+      "learning_rate": 1.096403254970526e-06,
+      "loss": 0.6374680519104003,
+      "step": 9580
+    },
+    {
+      "epoch": 1.706405693950178,
+      "grad_norm": 0.9196457266807556,
+      "learning_rate": 1.083484299014389e-06,
+      "loss": 0.6039985179901123,
+      "step": 9590
+    },
+    {
+      "epoch": 1.708185053380783,
+      "grad_norm": 0.9652523994445801,
+      "learning_rate": 1.0706375462646557e-06,
+      "loss": 0.6314795970916748,
+      "step": 9600
+    },
+    {
+      "epoch": 1.709964412811388,
+      "grad_norm": 0.9042948484420776,
+      "learning_rate": 1.0578631007503914e-06,
+      "loss": 0.6144063472747803,
+      "step": 9610
+    },
+    {
+      "epoch": 1.7117437722419928,
+      "grad_norm": 1.0348018407821655,
+      "learning_rate": 1.045161065915129e-06,
+      "loss": 0.6448927879333496,
+      "step": 9620
+    },
+    {
+      "epoch": 1.7135231316725978,
+      "grad_norm": 0.9222016334533691,
+      "learning_rate": 1.03253154461605e-06,
+      "loss": 0.6264961719512939,
+      "step": 9630
+    },
+    {
+      "epoch": 1.7153024911032029,
+      "grad_norm": 0.8399068117141724,
+      "learning_rate": 1.019974639123138e-06,
+      "loss": 0.6070163726806641,
+      "step": 9640
+    },
+    {
+      "epoch": 1.7170818505338077,
+      "grad_norm": 0.8186811208724976,
+      "learning_rate": 1.0074904511183592e-06,
+      "loss": 0.6530916690826416,
+      "step": 9650
+    },
+    {
+      "epoch": 1.718861209964413,
+      "grad_norm": 0.8317114114761353,
+      "learning_rate": 9.95079081694832e-07,
+      "loss": 0.6258822917938233,
+      "step": 9660
+    },
+    {
+      "epoch": 1.7206405693950177,
+      "grad_norm": 0.8472376465797424,
+      "learning_rate": 9.827406313560194e-07,
+      "loss": 0.6343019485473633,
+      "step": 9670
+    },
+    {
+      "epoch": 1.7224199288256228,
+      "grad_norm": 0.7907516956329346,
+      "learning_rate": 9.704752000148998e-07,
+      "loss": 0.5924992561340332,
+      "step": 9680
+    },
+    {
+      "epoch": 1.7241992882562278,
+      "grad_norm": 0.8266814351081848,
+      "learning_rate": 9.582828869931749e-07,
+      "loss": 0.63353590965271,
+      "step": 9690
+    },
+    {
+      "epoch": 1.7259786476868326,
+      "grad_norm": 0.8828052878379822,
+      "learning_rate": 9.461637910204468e-07,
+      "loss": 0.5913454532623291,
+      "step": 9700
+    },
+    {
+      "epoch": 1.7277580071174379,
+      "grad_norm": 0.795305609703064,
+      "learning_rate": 9.341180102334391e-07,
+      "loss": 0.625618314743042,
+      "step": 9710
+    },
+    {
+      "epoch": 1.7295373665480427,
+      "grad_norm": 0.8134214878082275,
+      "learning_rate": 9.221456421751828e-07,
+      "loss": 0.6241415023803711,
+      "step": 9720
+    },
+    {
+      "epoch": 1.7313167259786477,
+      "grad_norm": 0.8993518948554993,
+      "learning_rate": 9.102467837942364e-07,
+      "loss": 0.6320163249969483,
+      "step": 9730
+    },
+    {
+      "epoch": 1.7330960854092528,
+      "grad_norm": 0.8220566511154175,
+      "learning_rate": 8.984215314439071e-07,
+      "loss": 0.5945559501647949,
+      "step": 9740
+    },
+    {
+      "epoch": 1.7348754448398576,
+      "grad_norm": 0.8112704753875732,
+      "learning_rate": 8.866699808814516e-07,
+      "loss": 0.5934808731079102,
+      "step": 9750
+    },
+    {
+      "epoch": 1.7366548042704626,
+      "grad_norm": 0.9084371328353882,
+      "learning_rate": 8.749922272673228e-07,
+      "loss": 0.614768123626709,
+      "step": 9760
+    },
+    {
+      "epoch": 1.7384341637010676,
+      "grad_norm": 1.0116448402404785,
+      "learning_rate": 8.633883651643815e-07,
+      "loss": 0.6284510612487793,
+      "step": 9770
+    },
+    {
+      "epoch": 1.7402135231316724,
+      "grad_norm": 0.7356001734733582,
+      "learning_rate": 8.518584885371417e-07,
+      "loss": 0.5880512237548828,
+      "step": 9780
+    },
+    {
+      "epoch": 1.7419928825622777,
+      "grad_norm": 1.018543004989624,
+      "learning_rate": 8.404026907510043e-07,
+      "loss": 0.6093903064727784,
+      "step": 9790
+    },
+    {
+      "epoch": 1.7437722419928825,
+      "grad_norm": 0.7327563166618347,
+      "learning_rate": 8.290210645715002e-07,
+      "loss": 0.6004554748535156,
+      "step": 9800
+    },
+    {
+      "epoch": 1.7455516014234875,
+      "grad_norm": 0.8729904294013977,
+      "learning_rate": 8.177137021635439e-07,
+      "loss": 0.6066318035125733,
+      "step": 9810
+    },
+    {
+      "epoch": 1.7473309608540926,
+      "grad_norm": 0.9604983925819397,
+      "learning_rate": 8.064806950906867e-07,
+      "loss": 0.617271900177002,
+      "step": 9820
+    },
+    {
+      "epoch": 1.7491103202846974,
+      "grad_norm": 0.8511675000190735,
+      "learning_rate": 7.95322134314368e-07,
+      "loss": 0.6143304824829101,
+      "step": 9830
+    },
+    {
+      "epoch": 1.7508896797153026,
+      "grad_norm": 1.240357518196106,
+      "learning_rate": 7.842381101931895e-07,
+      "loss": 0.5943353652954102,
+      "step": 9840
+    },
+    {
+      "epoch": 1.7526690391459074,
+      "grad_norm": 0.7550404667854309,
+      "learning_rate": 7.732287124821747e-07,
+      "loss": 0.6033238887786865,
+      "step": 9850
+    },
+    {
+      "epoch": 1.7544483985765125,
+      "grad_norm": 0.8071495294570923,
+      "learning_rate": 7.622940303320426e-07,
+      "loss": 0.6051639080047607,
+      "step": 9860
+    },
+    {
+      "epoch": 1.7562277580071175,
+      "grad_norm": 1.0281059741973877,
+      "learning_rate": 7.514341522884971e-07,
+      "loss": 0.6654407501220703,
+      "step": 9870
+    },
+    {
+      "epoch": 1.7580071174377223,
+      "grad_norm": 0.8573912978172302,
+      "learning_rate": 7.406491662914917e-07,
+      "loss": 0.5752899646759033,
+      "step": 9880
+    },
+    {
+      "epoch": 1.7597864768683276,
+      "grad_norm": 0.9136864542961121,
+      "learning_rate": 7.299391596745342e-07,
+      "loss": 0.5918805599212646,
+      "step": 9890
+    },
+    {
+      "epoch": 1.7615658362989324,
+      "grad_norm": 0.7273184657096863,
+      "learning_rate": 7.193042191639677e-07,
+      "loss": 0.6129786491394043,
+      "step": 9900
+    },
+    {
+      "epoch": 1.7633451957295374,
+      "grad_norm": 0.8380533456802368,
+      "learning_rate": 7.087444308782787e-07,
+      "loss": 0.6275459289550781,
+      "step": 9910
+    },
+    {
+      "epoch": 1.7651245551601424,
+      "grad_norm": 0.8067348003387451,
+      "learning_rate": 6.982598803273854e-07,
+      "loss": 0.6274401187896729,
+      "step": 9920
+    },
+    {
+      "epoch": 1.7669039145907472,
+      "grad_norm": 0.9172580242156982,
+      "learning_rate": 6.878506524119644e-07,
+      "loss": 0.6016909122467041,
+      "step": 9930
+    },
+    {
+      "epoch": 1.7686832740213523,
+      "grad_norm": 0.8141267895698547,
+      "learning_rate": 6.775168314227442e-07,
+      "loss": 0.5785222053527832,
+      "step": 9940
+    },
+    {
+      "epoch": 1.7704626334519573,
+      "grad_norm": 0.8267232179641724,
+      "learning_rate": 6.672585010398347e-07,
+      "loss": 0.59950852394104,
+      "step": 9950
+    },
+    {
+      "epoch": 1.7722419928825621,
+      "grad_norm": 1.015816569328308,
+      "learning_rate": 6.570757443320441e-07,
+      "loss": 0.5815055847167969,
+      "step": 9960
+    },
+    {
+      "epoch": 1.7740213523131674,
+      "grad_norm": 0.9419827461242676,
+      "learning_rate": 6.469686437562117e-07,
+      "loss": 0.6218266963958741,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7758007117437722,
+      "grad_norm": 0.8901230692863464,
+      "learning_rate": 6.369372811565333e-07,
+      "loss": 0.6288963317871094,
+      "step": 9980
+    },
+    {
+      "epoch": 1.7775800711743772,
+      "grad_norm": 0.8378164172172546,
+      "learning_rate": 6.269817377639054e-07,
+      "loss": 0.5956534385681153,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7793594306049823,
+      "grad_norm": 0.6490439772605896,
+      "learning_rate": 6.171020941952611e-07,
+      "loss": 0.6209733486175537,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7793594306049823,
+      "eval_loss": 0.6085147261619568,
+      "eval_runtime": 413.8517,
+      "eval_samples_per_second": 12.152,
+      "eval_steps_per_second": 6.077,
+      "step": 10000
+    },
+    {
+      "epoch": 1.781138790035587,
+      "grad_norm": 0.7866795063018799,
+      "learning_rate": 6.072984304529172e-07,
+      "loss": 0.620451545715332,
+      "step": 10010
+    },
+    {
+      "epoch": 1.7829181494661923,
+      "grad_norm": 0.7238859534263611,
+      "learning_rate": 5.975708259239366e-07,
+      "loss": 0.6792274951934815,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7846975088967971,
+      "grad_norm": 0.7588909268379211,
+      "learning_rate": 5.879193593794708e-07,
+      "loss": 0.5726376056671143,
+      "step": 10030
+    },
+    {
+      "epoch": 1.7864768683274022,
+      "grad_norm": 0.7037184834480286,
+      "learning_rate": 5.783441089741359e-07,
+      "loss": 0.617862319946289,
+      "step": 10040
+    },
+    {
+      "epoch": 1.7882562277580072,
+      "grad_norm": 0.9185910224914551,
+      "learning_rate": 5.688451522453664e-07,
+      "loss": 0.6003572463989257,
+      "step": 10050
+    },
+    {
+      "epoch": 1.790035587188612,
+      "grad_norm": 0.8440210223197937,
+      "learning_rate": 5.594225661128005e-07,
+      "loss": 0.6071921825408936,
+      "step": 10060
+    },
+    {
+      "epoch": 1.791814946619217,
+      "grad_norm": 0.8852623105049133,
+      "learning_rate": 5.500764268776437e-07,
+      "loss": 0.5917786121368408,
+      "step": 10070
+    },
+    {
+      "epoch": 1.793594306049822,
+      "grad_norm": 0.9420585632324219,
+      "learning_rate": 5.408068102220665e-07,
+      "loss": 0.5636492252349854,
+      "step": 10080
+    },
+    {
+      "epoch": 1.795373665480427,
+      "grad_norm": 0.8379149436950684,
+      "learning_rate": 5.316137912085761e-07,
+      "loss": 0.5862648963928223,
+      "step": 10090
+    },
+    {
+      "epoch": 1.7971530249110321,
+      "grad_norm": 0.9554212689399719,
+      "learning_rate": 5.224974442794162e-07,
+      "loss": 0.5997903347015381,
+      "step": 10100
+    },
+    {
+      "epoch": 1.798932384341637,
+      "grad_norm": 0.784550666809082,
+      "learning_rate": 5.134578432559667e-07,
+      "loss": 0.6203852653503418,
+      "step": 10110
+    },
+    {
+      "epoch": 1.800711743772242,
+      "grad_norm": 0.9477099180221558,
+      "learning_rate": 5.044950613381417e-07,
+      "loss": 0.6295520782470703,
+      "step": 10120
+    },
+    {
+      "epoch": 1.802491103202847,
+      "grad_norm": 0.8329566717147827,
+      "learning_rate": 4.956091711037969e-07,
+      "loss": 0.6353004932403564,
+      "step": 10130
+    },
+    {
+      "epoch": 1.8042704626334518,
+      "grad_norm": 0.8794506788253784,
+      "learning_rate": 4.868002445081421e-07,
+      "loss": 0.6215813636779786,
+      "step": 10140
+    },
+    {
+      "epoch": 1.806049822064057,
+      "grad_norm": 0.9077814221382141,
+      "learning_rate": 4.780683528831609e-07,
+      "loss": 0.6175337314605713,
+      "step": 10150
+    },
+    {
+      "epoch": 1.8078291814946619,
+      "grad_norm": 0.8299528956413269,
+      "learning_rate": 4.6941356693703034e-07,
+      "loss": 0.6105076789855957,
+      "step": 10160
+    },
+    {
+      "epoch": 1.809608540925267,
+      "grad_norm": 0.8642259240150452,
+      "learning_rate": 4.608359567535503e-07,
+      "loss": 0.6132296085357666,
+      "step": 10170
+    },
+    {
+      "epoch": 1.811387900355872,
+      "grad_norm": 0.7558045387268066,
+      "learning_rate": 4.523355917915739e-07,
+      "loss": 0.5516393661499024,
+      "step": 10180
+    },
+    {
+      "epoch": 1.8131672597864767,
+      "grad_norm": 1.0017114877700806,
+      "learning_rate": 4.4391254088444693e-07,
+      "loss": 0.6016965866088867,
+      "step": 10190
+    },
+    {
+      "epoch": 1.814946619217082,
+      "grad_norm": 0.8784421682357788,
+      "learning_rate": 4.35566872239449e-07,
+      "loss": 0.6034076690673829,
+      "step": 10200
+    },
+    {
+      "epoch": 1.8167259786476868,
+      "grad_norm": 0.9458897113800049,
+      "learning_rate": 4.2729865343724254e-07,
+      "loss": 0.6276726722717285,
+      "step": 10210
+    },
+    {
+      "epoch": 1.8185053380782918,
+      "grad_norm": 0.7058622241020203,
+      "learning_rate": 4.1910795143132456e-07,
+      "loss": 0.6476336002349854,
+      "step": 10220
+    },
+    {
+      "epoch": 1.8202846975088969,
+      "grad_norm": 0.8122621774673462,
+      "learning_rate": 4.109948325474855e-07,
+      "loss": 0.6218099594116211,
+      "step": 10230
+    },
+    {
+      "epoch": 1.8220640569395017,
+      "grad_norm": 0.7891798615455627,
+      "learning_rate": 4.029593624832706e-07,
+      "loss": 0.6232967853546143,
+      "step": 10240
+    },
+    {
+      "epoch": 1.8238434163701067,
+      "grad_norm": 0.9203746914863586,
+      "learning_rate": 3.9500160630745023e-07,
+      "loss": 0.6128015518188477,
+      "step": 10250
+    },
+    {
+      "epoch": 1.8256227758007118,
+      "grad_norm": 0.8149620294570923,
+      "learning_rate": 3.871216284594881e-07,
+      "loss": 0.6312131881713867,
+      "step": 10260
+    },
+    {
+      "epoch": 1.8274021352313166,
+      "grad_norm": 0.8598204255104065,
+      "learning_rate": 3.7931949274902737e-07,
+      "loss": 0.5871327877044678,
+      "step": 10270
+    },
+    {
+      "epoch": 1.8291814946619218,
+      "grad_norm": 0.8834412693977356,
+      "learning_rate": 3.715952623553676e-07,
+      "loss": 0.5824168682098388,
+      "step": 10280
+    },
+    {
+      "epoch": 1.8309608540925266,
+      "grad_norm": 1.0543700456619263,
+      "learning_rate": 3.6394899982695297e-07,
+      "loss": 0.6174266338348389,
+      "step": 10290
+    },
+    {
+      "epoch": 1.8327402135231317,
+      "grad_norm": 0.8103786706924438,
+      "learning_rate": 3.563807670808717e-07,
+      "loss": 0.6298078536987305,
+      "step": 10300
+    },
+    {
+      "epoch": 1.8345195729537367,
+      "grad_norm": 0.9107673764228821,
+      "learning_rate": 3.488906254023472e-07,
+      "loss": 0.6039840698242187,
+      "step": 10310
+    },
+    {
+      "epoch": 1.8362989323843415,
+      "grad_norm": 0.8083145022392273,
+      "learning_rate": 3.4147863544424896e-07,
+      "loss": 0.6395132064819335,
+      "step": 10320
+    },
+    {
+      "epoch": 1.8380782918149468,
+      "grad_norm": 0.9567161202430725,
+      "learning_rate": 3.3414485722659374e-07,
+      "loss": 0.6278172969818115,
+      "step": 10330
+    },
+    {
+      "epoch": 1.8398576512455516,
+      "grad_norm": 0.8990734219551086,
+      "learning_rate": 3.2688935013606706e-07,
+      "loss": 0.6095623016357422,
+      "step": 10340
+    },
+    {
+      "epoch": 1.8416370106761566,
+      "grad_norm": 0.8392049670219421,
+      "learning_rate": 3.19712172925537e-07,
+      "loss": 0.5831768989562989,
+      "step": 10350
+    },
+    {
+      "epoch": 1.8434163701067616,
+      "grad_norm": 0.8746337890625,
+      "learning_rate": 3.126133837135825e-07,
+      "loss": 0.6429958820343018,
+      "step": 10360
+    },
+    {
+      "epoch": 1.8451957295373664,
+      "grad_norm": 0.842336893081665,
+      "learning_rate": 3.055930399840168e-07,
+      "loss": 0.6127379417419434,
+      "step": 10370
+    },
+    {
+      "epoch": 1.8469750889679717,
+      "grad_norm": 0.8559570908546448,
+      "learning_rate": 2.9865119858542904e-07,
+      "loss": 0.5989023208618164,
+      "step": 10380
+    },
+    {
+      "epoch": 1.8487544483985765,
+      "grad_norm": 0.9535056352615356,
+      "learning_rate": 2.9178791573071907e-07,
+      "loss": 0.6665677070617676,
+      "step": 10390
+    },
+    {
+      "epoch": 1.8505338078291815,
+      "grad_norm": 0.781708300113678,
+      "learning_rate": 2.8500324699664575e-07,
+      "loss": 0.6113322257995606,
+      "step": 10400
+    },
+    {
+      "epoch": 1.8523131672597866,
+      "grad_norm": 0.6821208000183105,
+      "learning_rate": 2.7829724732336914e-07,
+      "loss": 0.6359220027923584,
+      "step": 10410
+    },
+    {
+      "epoch": 1.8540925266903914,
+      "grad_norm": 0.8178049921989441,
+      "learning_rate": 2.7166997101402025e-07,
+      "loss": 0.605013656616211,
+      "step": 10420
+    },
+    {
+      "epoch": 1.8558718861209964,
+      "grad_norm": 0.8356890678405762,
+      "learning_rate": 2.651214717342454e-07,
+      "loss": 0.5625947475433349,
+      "step": 10430
+    },
+    {
+      "epoch": 1.8576512455516014,
+      "grad_norm": 0.9983499646186829,
+      "learning_rate": 2.5865180251178014e-07,
+      "loss": 0.623445463180542,
+      "step": 10440
+    },
+    {
+      "epoch": 1.8594306049822062,
+      "grad_norm": 0.7427167296409607,
+      "learning_rate": 2.522610157360228e-07,
+      "loss": 0.6319187641143799,
+      "step": 10450
+    },
+    {
+      "epoch": 1.8612099644128115,
+      "grad_norm": 0.7787860631942749,
+      "learning_rate": 2.459491631575983e-07,
+      "loss": 0.616677713394165,
+      "step": 10460
+    },
+    {
+      "epoch": 1.8629893238434163,
+      "grad_norm": 0.9523722529411316,
+      "learning_rate": 2.397162958879551e-07,
+      "loss": 0.6324071884155273,
+      "step": 10470
+    },
+    {
+      "epoch": 1.8647686832740213,
+      "grad_norm": 0.9004319310188293,
+      "learning_rate": 2.3356246439893425e-07,
+      "loss": 0.6087874412536621,
+      "step": 10480
+    },
+    {
+      "epoch": 1.8665480427046264,
+      "grad_norm": 0.8960275053977966,
+      "learning_rate": 2.2748771852237672e-07,
+      "loss": 0.5879695415496826,
+      "step": 10490
+    },
+    {
+      "epoch": 1.8683274021352312,
+      "grad_norm": 0.9242817759513855,
+      "learning_rate": 2.2149210744970896e-07,
+      "loss": 0.6239465713500977,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8683274021352312,
+      "eval_loss": 0.6083104014396667,
+      "eval_runtime": 425.8922,
+      "eval_samples_per_second": 11.808,
+      "eval_steps_per_second": 5.905,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8701067615658364,
+      "grad_norm": 0.7920718789100647,
+      "learning_rate": 2.155756797315478e-07,
+      "loss": 0.5853692531585694,
+      "step": 10510
+    },
+    {
+      "epoch": 1.8718861209964412,
+      "grad_norm": 0.8980622291564941,
+      "learning_rate": 2.097384832773075e-07,
+      "loss": 0.6284151077270508,
+      "step": 10520
+    },
+    {
+      "epoch": 1.8736654804270463,
+      "grad_norm": 0.8127864599227905,
+      "learning_rate": 2.0398056535481547e-07,
+      "loss": 0.6136427402496338,
+      "step": 10530
+    },
+    {
+      "epoch": 1.8754448398576513,
+      "grad_norm": 0.7731335759162903,
+      "learning_rate": 1.9830197258992158e-07,
+      "loss": 0.6308210849761963,
+      "step": 10540
+    },
+    {
+      "epoch": 1.8772241992882561,
+      "grad_norm": 0.7324168086051941,
+      "learning_rate": 1.9270275096612612e-07,
+      "loss": 0.6134955406188964,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8790035587188612,
+      "grad_norm": 0.8311629891395569,
+      "learning_rate": 1.871829458242058e-07,
+      "loss": 0.6565740585327149,
+      "step": 10560
+    },
+    {
+      "epoch": 1.8807829181494662,
+      "grad_norm": 0.9405824542045593,
+      "learning_rate": 1.8174260186184956e-07,
+      "loss": 0.5937122344970703,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8825622775800712,
+      "grad_norm": 0.977630615234375,
+      "learning_rate": 1.7638176313329203e-07,
+      "loss": 0.5619380474090576,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8843416370106763,
+      "grad_norm": 0.8144553303718567,
+      "learning_rate": 1.7110047304895738e-07,
+      "loss": 0.5915993213653564,
+      "step": 10590
+    },
+    {
+      "epoch": 1.886120996441281,
+      "grad_norm": 0.7680257558822632,
+      "learning_rate": 1.6589877437511502e-07,
+      "loss": 0.6170778751373291,
+      "step": 10600
+    },
+    {
+      "epoch": 1.887900355871886,
+      "grad_norm": 0.7631038427352905,
+      "learning_rate": 1.6077670923351884e-07,
+      "loss": 0.6260135173797607,
+      "step": 10610
+    },
+    {
+      "epoch": 1.8896797153024911,
+      "grad_norm": 0.8002687692642212,
+      "learning_rate": 1.5573431910108404e-07,
+      "loss": 0.6207282543182373,
+      "step": 10620
+    },
+    {
+      "epoch": 1.891459074733096,
+      "grad_norm": 0.8984543681144714,
+      "learning_rate": 1.5077164480953422e-07,
+      "loss": 0.6157234668731689,
+      "step": 10630
+    },
+    {
+      "epoch": 1.8932384341637012,
+      "grad_norm": 0.8822320103645325,
+      "learning_rate": 1.458887265450859e-07,
+      "loss": 0.6343502044677735,
+      "step": 10640
+    },
+    {
+      "epoch": 1.895017793594306,
+      "grad_norm": 0.8803090453147888,
+      "learning_rate": 1.4108560384810787e-07,
+      "loss": 0.6439435005187988,
+      "step": 10650
+    },
+    {
+      "epoch": 1.896797153024911,
+      "grad_norm": 0.9255337715148926,
+      "learning_rate": 1.3636231561281464e-07,
+      "loss": 0.614884328842163,
+      "step": 10660
+    },
+    {
+      "epoch": 1.898576512455516,
+      "grad_norm": 0.9993085861206055,
+      "learning_rate": 1.3171890008694454e-07,
+      "loss": 0.5709932327270508,
+      "step": 10670
+    },
+    {
+      "epoch": 1.9003558718861209,
+      "grad_norm": 0.7870404720306396,
+      "learning_rate": 1.2715539487145212e-07,
+      "loss": 0.6171539783477783,
+      "step": 10680
+    },
+    {
+      "epoch": 1.9021352313167261,
+      "grad_norm": 0.7632233500480652,
+      "learning_rate": 1.2267183692019845e-07,
+      "loss": 0.6215362548828125,
+      "step": 10690
+    },
+    {
+      "epoch": 1.903914590747331,
+      "grad_norm": 0.8149640560150146,
+      "learning_rate": 1.1826826253966029e-07,
+      "loss": 0.6112423419952393,
+      "step": 10700
+    },
+    {
+      "epoch": 1.905693950177936,
+      "grad_norm": 0.875933825969696,
+      "learning_rate": 1.1394470738863128e-07,
+      "loss": 0.6336851119995117,
+      "step": 10710
+    },
+    {
+      "epoch": 1.907473309608541,
+      "grad_norm": 0.8472994565963745,
+      "learning_rate": 1.0970120647793347e-07,
+      "loss": 0.6054853916168212,
+      "step": 10720
+    },
+    {
+      "epoch": 1.9092526690391458,
+      "grad_norm": 0.8871169090270996,
+      "learning_rate": 1.055377941701341e-07,
+      "loss": 0.6998202323913574,
+      "step": 10730
+    },
+    {
+      "epoch": 1.9110320284697508,
+      "grad_norm": 0.9006623029708862,
+      "learning_rate": 1.0145450417926806e-07,
+      "loss": 0.6072861194610596,
+      "step": 10740
+    },
+    {
+      "epoch": 1.9128113879003559,
+      "grad_norm": 0.781328558921814,
+      "learning_rate": 9.745136957056368e-08,
+      "loss": 0.6054211616516113,
+      "step": 10750
+    },
+    {
+      "epoch": 1.914590747330961,
+      "grad_norm": 0.8939800262451172,
+      "learning_rate": 9.352842276017516e-08,
+      "loss": 0.5997004985809327,
+      "step": 10760
+    },
+    {
+      "epoch": 1.916370106761566,
+      "grad_norm": 0.8280763030052185,
+      "learning_rate": 8.968569551492278e-08,
+      "loss": 0.6360623836517334,
+      "step": 10770
+    },
+    {
+      "epoch": 1.9181494661921707,
+      "grad_norm": 0.9543731808662415,
+      "learning_rate": 8.592321895203093e-08,
+      "loss": 0.5910046100616455,
+      "step": 10780
+    },
+    {
+      "epoch": 1.9199288256227758,
+      "grad_norm": 0.85121750831604,
+      "learning_rate": 8.224102353887931e-08,
+      "loss": 0.6036299228668213,
+      "step": 10790
+    },
+    {
+      "epoch": 1.9217081850533808,
+      "grad_norm": 0.8640086054801941,
+      "learning_rate": 7.863913909275656e-08,
+      "loss": 0.63557448387146,
+      "step": 10800
+    },
+    {
+      "epoch": 1.9234875444839856,
+      "grad_norm": 0.8530787825584412,
+      "learning_rate": 7.511759478061709e-08,
+      "loss": 0.5753629684448243,
+      "step": 10810
+    },
+    {
+      "epoch": 1.9252669039145909,
+      "grad_norm": 0.799323320388794,
+      "learning_rate": 7.16764191188446e-08,
+      "loss": 0.6003875732421875,
+      "step": 10820
+    },
+    {
+      "epoch": 1.9270462633451957,
+      "grad_norm": 0.8886220455169678,
+      "learning_rate": 6.831563997302337e-08,
+      "loss": 0.6101795196533203,
+      "step": 10830
+    },
+    {
+      "epoch": 1.9288256227758007,
+      "grad_norm": 0.9187102913856506,
+      "learning_rate": 6.50352845577118e-08,
+      "loss": 0.5787545680999756,
+      "step": 10840
+    },
+    {
+      "epoch": 1.9306049822064058,
+      "grad_norm": 0.8966771960258484,
+      "learning_rate": 6.18353794362192e-08,
+      "loss": 0.6280184745788574,
+      "step": 10850
+    },
+    {
+      "epoch": 1.9323843416370106,
+      "grad_norm": 1.0109564065933228,
+      "learning_rate": 5.8715950520396024e-08,
+      "loss": 0.5829953193664551,
+      "step": 10860
+    },
+    {
+      "epoch": 1.9341637010676158,
+      "grad_norm": 0.8951236009597778,
+      "learning_rate": 5.5677023070420665e-08,
+      "loss": 0.6363350868225097,
+      "step": 10870
+    },
+    {
+      "epoch": 1.9359430604982206,
+      "grad_norm": 1.040250539779663,
+      "learning_rate": 5.271862169459296e-08,
+      "loss": 0.6194582939147949,
+      "step": 10880
+    },
+    {
+      "epoch": 1.9377224199288257,
+      "grad_norm": 0.7401016354560852,
+      "learning_rate": 4.98407703491377e-08,
+      "loss": 0.6160064220428467,
+      "step": 10890
+    },
+    {
+      "epoch": 1.9395017793594307,
+      "grad_norm": 0.9136718511581421,
+      "learning_rate": 4.704349233801475e-08,
+      "loss": 0.6500599384307861,
+      "step": 10900
+    },
+    {
+      "epoch": 1.9412811387900355,
+      "grad_norm": 0.9346413612365723,
+      "learning_rate": 4.4326810312720345e-08,
+      "loss": 0.5980278968811035,
+      "step": 10910
+    },
+    {
+      "epoch": 1.9430604982206405,
+      "grad_norm": 0.8895809054374695,
+      "learning_rate": 4.169074627211278e-08,
+      "loss": 0.6100972652435303,
+      "step": 10920
+    },
+    {
+      "epoch": 1.9448398576512456,
+      "grad_norm": 0.8696462512016296,
+      "learning_rate": 3.913532156222699e-08,
+      "loss": 0.6204128742218018,
+      "step": 10930
+    },
+    {
+      "epoch": 1.9466192170818504,
+      "grad_norm": 0.8176571130752563,
+      "learning_rate": 3.6660556876110256e-08,
+      "loss": 0.6417241096496582,
+      "step": 10940
+    },
+    {
+      "epoch": 1.9483985765124556,
+      "grad_norm": 0.6966422200202942,
+      "learning_rate": 3.426647225364565e-08,
+      "loss": 0.5693438053131104,
+      "step": 10950
+    },
+    {
+      "epoch": 1.9501779359430604,
+      "grad_norm": 0.7873883247375488,
+      "learning_rate": 3.195308708139555e-08,
+      "loss": 0.5997710704803467,
+      "step": 10960
+    },
+    {
+      "epoch": 1.9519572953736655,
+      "grad_norm": 0.8650481700897217,
+      "learning_rate": 2.9720420092442803e-08,
+      "loss": 0.6134130477905273,
+      "step": 10970
+    },
+    {
+      "epoch": 1.9537366548042705,
+      "grad_norm": 0.7547999620437622,
+      "learning_rate": 2.756848936623979e-08,
+      "loss": 0.5843913555145264,
+      "step": 10980
+    },
+    {
+      "epoch": 1.9555160142348753,
+      "grad_norm": 1.011824369430542,
+      "learning_rate": 2.5497312328460756e-08,
+      "loss": 0.5818317413330079,
+      "step": 10990
+    },
+    {
+      "epoch": 1.9572953736654806,
+      "grad_norm": 0.9581059217453003,
+      "learning_rate": 2.3506905750859678e-08,
+      "loss": 0.5937665939331055,
+      "step": 11000
+    },
+    {
+      "epoch": 1.9572953736654806,
+      "eval_loss": 0.6082284450531006,
+      "eval_runtime": 725.2,
+      "eval_samples_per_second": 6.935,
+      "eval_steps_per_second": 3.468,
+      "step": 11000
+    },
+    {
+      "epoch": 1.9590747330960854,
+      "grad_norm": 0.9106253385543823,
+      "learning_rate": 2.1597285751140397e-08,
+      "loss": 0.6484842777252198,
+      "step": 11010
+    },
+    {
+      "epoch": 1.9608540925266904,
+      "grad_norm": 0.8921802639961243,
+      "learning_rate": 1.9768467792817825e-08,
+      "loss": 0.5928434371948242,
+      "step": 11020
+    },
+    {
+      "epoch": 1.9626334519572954,
+      "grad_norm": 0.8079352378845215,
+      "learning_rate": 1.802046668509916e-08,
+      "loss": 0.6446592807769775,
+      "step": 11030
+    },
+    {
+      "epoch": 1.9644128113879002,
+      "grad_norm": 0.9401485323905945,
+      "learning_rate": 1.6353296582758416e-08,
+      "loss": 0.6148817539215088,
+      "step": 11040
+    },
+    {
+      "epoch": 1.9661921708185055,
+      "grad_norm": 0.89174485206604,
+      "learning_rate": 1.476697098602986e-08,
+      "loss": 0.6317745685577393,
+      "step": 11050
+    },
+    {
+      "epoch": 1.9679715302491103,
+      "grad_norm": 0.8186066150665283,
+      "learning_rate": 1.3261502740490317e-08,
+      "loss": 0.604107666015625,
+      "step": 11060
+    },
+    {
+      "epoch": 1.9697508896797153,
+      "grad_norm": 1.0631054639816284,
+      "learning_rate": 1.1836904036959252e-08,
+      "loss": 0.6059531688690185,
+      "step": 11070
+    },
+    {
+      "epoch": 1.9715302491103204,
+      "grad_norm": 0.7882722020149231,
+      "learning_rate": 1.0493186411398848e-08,
+      "loss": 0.5915180683135987,
+      "step": 11080
+    },
+    {
+      "epoch": 1.9733096085409252,
+      "grad_norm": 0.9640932679176331,
+      "learning_rate": 9.230360744822975e-09,
+      "loss": 0.6303177356719971,
+      "step": 11090
+    },
+    {
+      "epoch": 1.9750889679715302,
+      "grad_norm": 0.8806138634681702,
+      "learning_rate": 8.048437263205033e-09,
+      "loss": 0.6188820362091064,
+      "step": 11100
+    },
+    {
+      "epoch": 1.9768683274021353,
+      "grad_norm": 0.7664757370948792,
+      "learning_rate": 6.947425537400243e-09,
+      "loss": 0.5683889865875245,
+      "step": 11110
+    },
+    {
+      "epoch": 1.97864768683274,
+      "grad_norm": 1.0602748394012451,
+      "learning_rate": 5.927334483061264e-09,
+      "loss": 0.5750191688537598,
+      "step": 11120
+    },
+    {
+      "epoch": 1.9804270462633453,
+      "grad_norm": 1.0477946996688843,
+      "learning_rate": 4.988172360573806e-09,
+      "loss": 0.6328764438629151,
+      "step": 11130
+    },
+    {
+      "epoch": 1.9822064056939501,
+      "grad_norm": 0.7248848080635071,
+      "learning_rate": 4.129946774984461e-09,
+      "loss": 0.6349906921386719,
+      "step": 11140
+    },
+    {
+      "epoch": 1.9839857651245552,
+      "grad_norm": 0.8907039761543274,
+      "learning_rate": 3.352664675938533e-09,
+      "loss": 0.6227742195129394,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9857651245551602,
+      "grad_norm": 0.7807096242904663,
+      "learning_rate": 2.656332357631186e-09,
+      "loss": 0.6313164234161377,
+      "step": 11160
+    },
+    {
+      "epoch": 1.987544483985765,
+      "grad_norm": 0.6477228403091431,
+      "learning_rate": 2.040955458744165e-09,
+      "loss": 0.6183161258697509,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9893238434163703,
+      "grad_norm": 0.7875910997390747,
+      "learning_rate": 1.5065389624124848e-09,
+      "loss": 0.6141678810119628,
+      "step": 11180
+    },
+    {
+      "epoch": 1.991103202846975,
+      "grad_norm": 0.8616359829902649,
+      "learning_rate": 1.0530871961778044e-09,
+      "loss": 0.6099072933197022,
+      "step": 11190
+    },
+    {
+      "epoch": 1.99288256227758,
+      "grad_norm": 0.9508731961250305,
+      "learning_rate": 6.806038319506769e-10,
+      "loss": 0.6085052967071534,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9946619217081851,
+      "grad_norm": 0.7681756019592285,
+      "learning_rate": 3.8909188598945656e-10,
+      "loss": 0.6388015270233154,
+      "step": 11210
+    },
+    {
+      "epoch": 1.99644128113879,
+      "grad_norm": 0.7130521535873413,
+      "learning_rate": 1.7855371886699169e-10,
+      "loss": 0.643953514099121,
+      "step": 11220
+    },
+    {
+      "epoch": 1.998220640569395,
+      "grad_norm": 0.8050049543380737,
+      "learning_rate": 4.8991035457301994e-11,
+      "loss": 0.6025295257568359,
+      "step": 11230
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8600704669952393,
+      "learning_rate": 4.04884918925319e-13,
+      "loss": 0.6060295104980469,
+      "step": 11240
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 11240,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.298854192009953e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}