diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4177 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.49016360866397296,
+  "eval_steps": 500,
+  "global_step": 59200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008279790686891435,
+      "grad_norm": 0.2026122659444809,
+      "learning_rate": 0.00019994481795043316,
+      "loss": 1.9308,
+      "step": 100
+    },
+    {
+      "epoch": 0.001655958137378287,
+      "grad_norm": 0.189473494887352,
+      "learning_rate": 0.00019988961934128355,
+      "loss": 1.796,
+      "step": 200
+    },
+    {
+      "epoch": 0.0024839372060674308,
+      "grad_norm": 0.21440361440181732,
+      "learning_rate": 0.00019983442073213394,
+      "loss": 1.7292,
+      "step": 300
+    },
+    {
+      "epoch": 0.003311916274756574,
+      "grad_norm": 0.4124305844306946,
+      "learning_rate": 0.00019977922212298434,
+      "loss": 1.7479,
+      "step": 400
+    },
+    {
+      "epoch": 0.004139895343445717,
+      "grad_norm": 0.326767235994339,
+      "learning_rate": 0.00019972402351383473,
+      "loss": 1.7445,
+      "step": 500
+    },
+    {
+      "epoch": 0.0049678744121348616,
+      "grad_norm": 0.23368559777736664,
+      "learning_rate": 0.00019966882490468513,
+      "loss": 1.7146,
+      "step": 600
+    },
+    {
+      "epoch": 0.005795853480824005,
+      "grad_norm": 0.2080787867307663,
+      "learning_rate": 0.00019961362629553552,
+      "loss": 1.7009,
+      "step": 700
+    },
+    {
+      "epoch": 0.006623832549513148,
+      "grad_norm": 0.23906764388084412,
+      "learning_rate": 0.00019955842768638592,
+      "loss": 1.7346,
+      "step": 800
+    },
+    {
+      "epoch": 0.0074518116182022915,
+      "grad_norm": 0.21370458602905273,
+      "learning_rate": 0.00019950322907723634,
+      "loss": 1.7161,
+      "step": 900
+    },
+    {
+      "epoch": 0.008279790686891435,
+      "grad_norm": 0.2232082337141037,
+      "learning_rate": 0.0001994480304680867,
+      "loss": 1.6845,
+      "step": 1000
+    },
+    {
+      "epoch": 0.009107769755580579,
+      "grad_norm": 0.28137433528900146,
+      "learning_rate": 0.00019939283185893713,
+      "loss": 1.7064,
+      "step": 1100
+    },
+    {
+      "epoch": 0.009935748824269723,
+      "grad_norm": 0.27577510476112366,
+      "learning_rate": 0.00019933763324978752,
+      "loss": 1.7216,
+      "step": 1200
+    },
+    {
+      "epoch": 0.010763727892958866,
+      "grad_norm": 0.19732429087162018,
+      "learning_rate": 0.0001992824346406379,
+      "loss": 1.6785,
+      "step": 1300
+    },
+    {
+      "epoch": 0.01159170696164801,
+      "grad_norm": 0.2802606523036957,
+      "learning_rate": 0.0001992272360314883,
+      "loss": 1.6987,
+      "step": 1400
+    },
+    {
+      "epoch": 0.012419686030337154,
+      "grad_norm": 0.2615717053413391,
+      "learning_rate": 0.0001991720374223387,
+      "loss": 1.7157,
+      "step": 1500
+    },
+    {
+      "epoch": 0.013247665099026296,
+      "grad_norm": 0.21463246643543243,
+      "learning_rate": 0.00019911683881318907,
+      "loss": 1.709,
+      "step": 1600
+    },
+    {
+      "epoch": 0.01407564416771544,
+      "grad_norm": 0.23596777021884918,
+      "learning_rate": 0.0001990616402040395,
+      "loss": 1.7288,
+      "step": 1700
+    },
+    {
+      "epoch": 0.014903623236404583,
+      "grad_norm": 0.26049914956092834,
+      "learning_rate": 0.00019900644159488989,
+      "loss": 1.7007,
+      "step": 1800
+    },
+    {
+      "epoch": 0.01573160230509373,
+      "grad_norm": 0.1828412264585495,
+      "learning_rate": 0.00019895124298574025,
+      "loss": 1.6872,
+      "step": 1900
+    },
+    {
+      "epoch": 0.01655958137378287,
+      "grad_norm": 0.30434760451316833,
+      "learning_rate": 0.00019889604437659067,
+      "loss": 1.7084,
+      "step": 2000
+    },
+    {
+      "epoch": 0.017387560442472014,
+      "grad_norm": 0.19470958411693573,
+      "learning_rate": 0.00019884084576744107,
+      "loss": 1.6804,
+      "step": 2100
+    },
+    {
+      "epoch": 0.018215539511161158,
+      "grad_norm": 0.23792260885238647,
+      "learning_rate": 0.00019878564715829144,
+      "loss": 1.6873,
+      "step": 2200
+    },
+    {
+      "epoch": 0.019043518579850302,
+      "grad_norm": 0.2714627683162689,
+      "learning_rate": 0.00019873044854914186,
+      "loss": 1.6803,
+      "step": 2300
+    },
+    {
+      "epoch": 0.019871497648539446,
+      "grad_norm": 0.2310653179883957,
+      "learning_rate": 0.00019867524993999225,
+      "loss": 1.7056,
+      "step": 2400
+    },
+    {
+      "epoch": 0.020699476717228587,
+      "grad_norm": 0.21665313839912415,
+      "learning_rate": 0.00019862005133084262,
+      "loss": 1.6636,
+      "step": 2500
+    },
+    {
+      "epoch": 0.02152745578591773,
+      "grad_norm": 0.26358917355537415,
+      "learning_rate": 0.00019856485272169304,
+      "loss": 1.6567,
+      "step": 2600
+    },
+    {
+      "epoch": 0.022355434854606875,
+      "grad_norm": 0.2334413081407547,
+      "learning_rate": 0.00019850965411254343,
+      "loss": 1.6666,
+      "step": 2700
+    },
+    {
+      "epoch": 0.02318341392329602,
+      "grad_norm": 0.20429347455501556,
+      "learning_rate": 0.0001984544555033938,
+      "loss": 1.7123,
+      "step": 2800
+    },
+    {
+      "epoch": 0.024011392991985164,
+      "grad_norm": 0.22018109261989594,
+      "learning_rate": 0.00019839925689424422,
+      "loss": 1.6405,
+      "step": 2900
+    },
+    {
+      "epoch": 0.024839372060674308,
+      "grad_norm": 0.2289331555366516,
+      "learning_rate": 0.00019834405828509462,
+      "loss": 1.6769,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02566735112936345,
+      "grad_norm": 0.21042221784591675,
+      "learning_rate": 0.00019828885967594498,
+      "loss": 1.6559,
+      "step": 3100
+    },
+    {
+      "epoch": 0.026495330198052593,
+      "grad_norm": 0.24456587433815002,
+      "learning_rate": 0.0001982336610667954,
+      "loss": 1.6889,
+      "step": 3200
+    },
+    {
+      "epoch": 0.027323309266741737,
+      "grad_norm": 0.19650936126708984,
+      "learning_rate": 0.0001981784624576458,
+      "loss": 1.6577,
+      "step": 3300
+    },
+    {
+      "epoch": 0.02815128833543088,
+      "grad_norm": 0.2078315168619156,
+      "learning_rate": 0.00019812326384849617,
+      "loss": 1.6968,
+      "step": 3400
+    },
+    {
+      "epoch": 0.028979267404120025,
+      "grad_norm": 0.27083057165145874,
+      "learning_rate": 0.0001980680652393466,
+      "loss": 1.6851,
+      "step": 3500
+    },
+    {
+      "epoch": 0.029807246472809166,
+      "grad_norm": 0.19873012602329254,
+      "learning_rate": 0.00019801286663019698,
+      "loss": 1.6642,
+      "step": 3600
+    },
+    {
+      "epoch": 0.03063522554149831,
+      "grad_norm": 0.25127413868904114,
+      "learning_rate": 0.00019795766802104735,
+      "loss": 1.6925,
+      "step": 3700
+    },
+    {
+      "epoch": 0.03146320461018746,
+      "grad_norm": 0.28270578384399414,
+      "learning_rate": 0.00019790246941189777,
+      "loss": 1.6823,
+      "step": 3800
+    },
+    {
+      "epoch": 0.032291183678876595,
+      "grad_norm": 0.22762413322925568,
+      "learning_rate": 0.00019784727080274816,
+      "loss": 1.6802,
+      "step": 3900
+    },
+    {
+      "epoch": 0.03311916274756574,
+      "grad_norm": 0.24451415240764618,
+      "learning_rate": 0.00019779207219359853,
+      "loss": 1.6855,
+      "step": 4000
+    },
+    {
+      "epoch": 0.03394714181625488,
+      "grad_norm": 0.18539206683635712,
+      "learning_rate": 0.00019773687358444895,
+      "loss": 1.6753,
+      "step": 4100
+    },
+    {
+      "epoch": 0.03477512088494403,
+      "grad_norm": 0.20416654646396637,
+      "learning_rate": 0.00019768167497529935,
+      "loss": 1.6892,
+      "step": 4200
+    },
+    {
+      "epoch": 0.03560309995363317,
+      "grad_norm": 0.22939777374267578,
+      "learning_rate": 0.00019762647636614971,
+      "loss": 1.6953,
+      "step": 4300
+    },
+    {
+      "epoch": 0.036431079022322316,
+      "grad_norm": 0.21451617777347565,
+      "learning_rate": 0.00019757127775700014,
+      "loss": 1.6609,
+      "step": 4400
+    },
+    {
+      "epoch": 0.03725905809101146,
+      "grad_norm": 0.20271216332912445,
+      "learning_rate": 0.00019751607914785053,
+      "loss": 1.6868,
+      "step": 4500
+    },
+    {
+      "epoch": 0.038087037159700604,
+      "grad_norm": 0.23506902158260345,
+      "learning_rate": 0.0001974608805387009,
+      "loss": 1.7012,
+      "step": 4600
+    },
+    {
+      "epoch": 0.03891501622838975,
+      "grad_norm": 0.19889087975025177,
+      "learning_rate": 0.00019740568192955132,
+      "loss": 1.6867,
+      "step": 4700
+    },
+    {
+      "epoch": 0.03974299529707889,
+      "grad_norm": 0.21165409684181213,
+      "learning_rate": 0.0001973504833204017,
+      "loss": 1.6666,
+      "step": 4800
+    },
+    {
+      "epoch": 0.04057097436576804,
+      "grad_norm": 0.25541654229164124,
+      "learning_rate": 0.0001972952847112521,
+      "loss": 1.7012,
+      "step": 4900
+    },
+    {
+      "epoch": 0.041398953434457174,
+      "grad_norm": 0.2289835512638092,
+      "learning_rate": 0.0001972400861021025,
+      "loss": 1.6351,
+      "step": 5000
+    },
+    {
+      "epoch": 0.04222693250314632,
+      "grad_norm": 0.20699161291122437,
+      "learning_rate": 0.0001971848874929529,
+      "loss": 1.6834,
+      "step": 5100
+    },
+    {
+      "epoch": 0.04305491157183546,
+      "grad_norm": 0.21547134220600128,
+      "learning_rate": 0.0001971296888838033,
+      "loss": 1.6605,
+      "step": 5200
+    },
+    {
+      "epoch": 0.043882890640524606,
+      "grad_norm": 0.21996253728866577,
+      "learning_rate": 0.00019707449027465368,
+      "loss": 1.6747,
+      "step": 5300
+    },
+    {
+      "epoch": 0.04471086970921375,
+      "grad_norm": 0.19425031542778015,
+      "learning_rate": 0.00019701929166550408,
+      "loss": 1.6677,
+      "step": 5400
+    },
+    {
+      "epoch": 0.045538848777902895,
+      "grad_norm": 0.21223612129688263,
+      "learning_rate": 0.00019696409305635447,
+      "loss": 1.6401,
+      "step": 5500
+    },
+    {
+      "epoch": 0.04636682784659204,
+      "grad_norm": 0.21771861612796783,
+      "learning_rate": 0.00019690889444720487,
+      "loss": 1.6655,
+      "step": 5600
+    },
+    {
+      "epoch": 0.04719480691528118,
+      "grad_norm": 0.17102624475955963,
+      "learning_rate": 0.00019685369583805526,
+      "loss": 1.6373,
+      "step": 5700
+    },
+    {
+      "epoch": 0.04802278598397033,
+      "grad_norm": 1.002648115158081,
+      "learning_rate": 0.00019679849722890565,
+      "loss": 1.6578,
+      "step": 5800
+    },
+    {
+      "epoch": 0.04885076505265947,
+      "grad_norm": 0.209450364112854,
+      "learning_rate": 0.00019674329861975605,
+      "loss": 1.7114,
+      "step": 5900
+    },
+    {
+      "epoch": 0.049678744121348616,
+      "grad_norm": 0.21328583359718323,
+      "learning_rate": 0.00019668810001060644,
+      "loss": 1.6816,
+      "step": 6000
+    },
+    {
+      "epoch": 0.05050672319003775,
+      "grad_norm": 0.2184838354587555,
+      "learning_rate": 0.00019663290140145684,
+      "loss": 1.6596,
+      "step": 6100
+    },
+    {
+      "epoch": 0.0513347022587269,
+      "grad_norm": 0.23963460326194763,
+      "learning_rate": 0.00019657825477839872,
+      "loss": 1.6627,
+      "step": 6200
+    },
+    {
+      "epoch": 0.05216268132741604,
+      "grad_norm": 0.2165103405714035,
+      "learning_rate": 0.00019652305616924912,
+      "loss": 1.6808,
+      "step": 6300
+    },
+    {
+      "epoch": 0.052990660396105185,
+      "grad_norm": 0.19727864861488342,
+      "learning_rate": 0.0001964678575600995,
+      "loss": 1.7164,
+      "step": 6400
+    },
+    {
+      "epoch": 0.05381863946479433,
+      "grad_norm": 0.23533885180950165,
+      "learning_rate": 0.0001964126589509499,
+      "loss": 1.6575,
+      "step": 6500
+    },
+    {
+      "epoch": 0.054646618533483474,
+      "grad_norm": 0.23083338141441345,
+      "learning_rate": 0.0001963574603418003,
+      "loss": 1.6481,
+      "step": 6600
+    },
+    {
+      "epoch": 0.05547459760217262,
+      "grad_norm": 0.18879123032093048,
+      "learning_rate": 0.0001963022617326507,
+      "loss": 1.6798,
+      "step": 6700
+    },
+    {
+      "epoch": 0.05630257667086176,
+      "grad_norm": 0.2472197264432907,
+      "learning_rate": 0.0001962470631235011,
+      "loss": 1.6739,
+      "step": 6800
+    },
+    {
+      "epoch": 0.057130555739550906,
+      "grad_norm": 0.19961301982402802,
+      "learning_rate": 0.00019619186451435148,
+      "loss": 1.6307,
+      "step": 6900
+    },
+    {
+      "epoch": 0.05795853480824005,
+      "grad_norm": 0.1921355426311493,
+      "learning_rate": 0.00019613666590520188,
+      "loss": 1.6432,
+      "step": 7000
+    },
+    {
+      "epoch": 0.058786513876929194,
+      "grad_norm": 0.29221248626708984,
+      "learning_rate": 0.00019608146729605227,
+      "loss": 1.6591,
+      "step": 7100
+    },
+    {
+      "epoch": 0.05961449294561833,
+      "grad_norm": 0.2565169632434845,
+      "learning_rate": 0.00019602626868690267,
+      "loss": 1.6314,
+      "step": 7200
+    },
+    {
+      "epoch": 0.060442472014307476,
+      "grad_norm": 0.19964313507080078,
+      "learning_rate": 0.00019597107007775306,
+      "loss": 1.6571,
+      "step": 7300
+    },
+    {
+      "epoch": 0.06127045108299662,
+      "grad_norm": 0.19322757422924042,
+      "learning_rate": 0.00019591587146860345,
+      "loss": 1.6553,
+      "step": 7400
+    },
+    {
+      "epoch": 0.062098430151685764,
+      "grad_norm": 0.2548861801624298,
+      "learning_rate": 0.00019586067285945385,
+      "loss": 1.6791,
+      "step": 7500
+    },
+    {
+      "epoch": 0.06292640922037492,
+      "grad_norm": 0.2616177797317505,
+      "learning_rate": 0.00019580547425030424,
+      "loss": 1.6208,
+      "step": 7600
+    },
+    {
+      "epoch": 0.06375438828906405,
+      "grad_norm": 0.2758786976337433,
+      "learning_rate": 0.00019575027564115464,
+      "loss": 1.6838,
+      "step": 7700
+    },
+    {
+      "epoch": 0.06458236735775319,
+      "grad_norm": 0.1881067007780075,
+      "learning_rate": 0.00019569507703200503,
+      "loss": 1.6577,
+      "step": 7800
+    },
+    {
+      "epoch": 0.06541034642644233,
+      "grad_norm": 0.21222031116485596,
+      "learning_rate": 0.00019563987842285543,
+      "loss": 1.6952,
+      "step": 7900
+    },
+    {
+      "epoch": 0.06623832549513148,
+      "grad_norm": 0.23496469855308533,
+      "learning_rate": 0.00019558467981370582,
+      "loss": 1.6604,
+      "step": 8000
+    },
+    {
+      "epoch": 0.06706630456382062,
+      "grad_norm": 0.21739377081394196,
+      "learning_rate": 0.00019552948120455621,
+      "loss": 1.6493,
+      "step": 8100
+    },
+    {
+      "epoch": 0.06789428363250977,
+      "grad_norm": 0.2083568572998047,
+      "learning_rate": 0.0001954742825954066,
+      "loss": 1.6507,
+      "step": 8200
+    },
+    {
+      "epoch": 0.06872226270119891,
+      "grad_norm": 0.2494797557592392,
+      "learning_rate": 0.000195419083986257,
+      "loss": 1.6718,
+      "step": 8300
+    },
+    {
+      "epoch": 0.06955024176988805,
+      "grad_norm": 0.22432386875152588,
+      "learning_rate": 0.0001953638853771074,
+      "loss": 1.6758,
+      "step": 8400
+    },
+    {
+      "epoch": 0.0703782208385772,
+      "grad_norm": 0.21674951910972595,
+      "learning_rate": 0.00019530923875404928,
+      "loss": 1.6437,
+      "step": 8500
+    },
+    {
+      "epoch": 0.07120619990726634,
+      "grad_norm": 0.188558429479599,
+      "learning_rate": 0.00019525404014489968,
+      "loss": 1.6345,
+      "step": 8600
+    },
+    {
+      "epoch": 0.07203417897595549,
+      "grad_norm": 0.22968587279319763,
+      "learning_rate": 0.00019519884153575007,
+      "loss": 1.6609,
+      "step": 8700
+    },
+    {
+      "epoch": 0.07286215804464463,
+      "grad_norm": 0.2717007100582123,
+      "learning_rate": 0.00019514364292660047,
+      "loss": 1.6514,
+      "step": 8800
+    },
+    {
+      "epoch": 0.07369013711333378,
+      "grad_norm": 0.2857931852340698,
+      "learning_rate": 0.00019508844431745086,
+      "loss": 1.6669,
+      "step": 8900
+    },
+    {
+      "epoch": 0.07451811618202292,
+      "grad_norm": 0.23388004302978516,
+      "learning_rate": 0.00019503324570830125,
+      "loss": 1.6785,
+      "step": 9000
+    },
+    {
+      "epoch": 0.07534609525071206,
+      "grad_norm": 0.20342163741588593,
+      "learning_rate": 0.00019497804709915165,
+      "loss": 1.6777,
+      "step": 9100
+    },
+    {
+      "epoch": 0.07617407431940121,
+      "grad_norm": 0.2345508486032486,
+      "learning_rate": 0.00019492340047609353,
+      "loss": 1.6672,
+      "step": 9200
+    },
+    {
+      "epoch": 0.07700205338809035,
+      "grad_norm": 0.20348191261291504,
+      "learning_rate": 0.00019486820186694396,
+      "loss": 1.6851,
+      "step": 9300
+    },
+    {
+      "epoch": 0.0778300324567795,
+      "grad_norm": 0.2147115170955658,
+      "learning_rate": 0.00019481300325779435,
+      "loss": 1.6768,
+      "step": 9400
+    },
+    {
+      "epoch": 0.07865801152546864,
+      "grad_norm": 0.24411408603191376,
+      "learning_rate": 0.00019475780464864472,
+      "loss": 1.6488,
+      "step": 9500
+    },
+    {
+      "epoch": 0.07948599059415778,
+      "grad_norm": 0.2360156625509262,
+      "learning_rate": 0.00019470260603949514,
+      "loss": 1.6248,
+      "step": 9600
+    },
+    {
+      "epoch": 0.08031396966284693,
+      "grad_norm": 0.24534212052822113,
+      "learning_rate": 0.00019464740743034553,
+      "loss": 1.6425,
+      "step": 9700
+    },
+    {
+      "epoch": 0.08114194873153607,
+      "grad_norm": 0.26849544048309326,
+      "learning_rate": 0.0001945922088211959,
+      "loss": 1.668,
+      "step": 9800
+    },
+    {
+      "epoch": 0.0819699278002252,
+      "grad_norm": 0.20883604884147644,
+      "learning_rate": 0.00019453701021204632,
+      "loss": 1.6492,
+      "step": 9900
+    },
+    {
+      "epoch": 0.08279790686891435,
+      "grad_norm": 0.22967711091041565,
+      "learning_rate": 0.00019448181160289672,
+      "loss": 1.6667,
+      "step": 10000
+    },
+    {
+      "epoch": 0.08362588593760349,
+      "grad_norm": 0.24589480459690094,
+      "learning_rate": 0.00019442661299374708,
+      "loss": 1.62,
+      "step": 10100
+    },
+    {
+      "epoch": 0.08445386500629264,
+      "grad_norm": 0.26812034845352173,
+      "learning_rate": 0.0001943714143845975,
+      "loss": 1.6351,
+      "step": 10200
+    },
+    {
+      "epoch": 0.08528184407498178,
+      "grad_norm": 0.24378615617752075,
+      "learning_rate": 0.0001943162157754479,
+      "loss": 1.6097,
+      "step": 10300
+    },
+    {
+      "epoch": 0.08610982314367092,
+      "grad_norm": 0.31285640597343445,
+      "learning_rate": 0.00019426101716629827,
+      "loss": 1.7014,
+      "step": 10400
+    },
+    {
+      "epoch": 0.08693780221236007,
+      "grad_norm": 0.277204692363739,
+      "learning_rate": 0.0001942058185571487,
+      "loss": 1.6482,
+      "step": 10500
+    },
+    {
+      "epoch": 0.08776578128104921,
+      "grad_norm": 0.22598931193351746,
+      "learning_rate": 0.00019415061994799908,
+      "loss": 1.6459,
+      "step": 10600
+    },
+    {
+      "epoch": 0.08859376034973836,
+      "grad_norm": 0.3453767001628876,
+      "learning_rate": 0.00019409542133884945,
+      "loss": 1.6549,
+      "step": 10700
+    },
+    {
+      "epoch": 0.0894217394184275,
+      "grad_norm": 0.2226220667362213,
+      "learning_rate": 0.00019404022272969987,
+      "loss": 1.6481,
+      "step": 10800
+    },
+    {
+      "epoch": 0.09024971848711665,
+      "grad_norm": 0.2280457317829132,
+      "learning_rate": 0.00019398502412055026,
+      "loss": 1.6587,
+      "step": 10900
+    },
+    {
+      "epoch": 0.09107769755580579,
+      "grad_norm": 0.2163500040769577,
+      "learning_rate": 0.00019392982551140063,
+      "loss": 1.6693,
+      "step": 11000
+    },
+    {
+      "epoch": 0.09190567662449493,
+      "grad_norm": 0.19597041606903076,
+      "learning_rate": 0.00019387462690225105,
+      "loss": 1.6313,
+      "step": 11100
+    },
+    {
+      "epoch": 0.09273365569318408,
+      "grad_norm": 0.2143152803182602,
+      "learning_rate": 0.00019381942829310145,
+      "loss": 1.6543,
+      "step": 11200
+    },
+    {
+      "epoch": 0.09356163476187322,
+      "grad_norm": 0.2231709212064743,
+      "learning_rate": 0.0001937642296839518,
+      "loss": 1.6534,
+      "step": 11300
+    },
+    {
+      "epoch": 0.09438961383056237,
+      "grad_norm": 0.21003256738185883,
+      "learning_rate": 0.00019370903107480223,
+      "loss": 1.644,
+      "step": 11400
+    },
+    {
+      "epoch": 0.09521759289925151,
+      "grad_norm": 0.23484201729297638,
+      "learning_rate": 0.00019365383246565263,
+      "loss": 1.67,
+      "step": 11500
+    },
+    {
+      "epoch": 0.09604557196794065,
+      "grad_norm": 0.23727314174175262,
+      "learning_rate": 0.000193598633856503,
+      "loss": 1.6078,
+      "step": 11600
+    },
+    {
+      "epoch": 0.0968735510366298,
+      "grad_norm": 0.2734909653663635,
+      "learning_rate": 0.00019354343524735342,
+      "loss": 1.6389,
+      "step": 11700
+    },
+    {
+      "epoch": 0.09770153010531894,
+      "grad_norm": 0.266926646232605,
+      "learning_rate": 0.0001934882366382038,
+      "loss": 1.6522,
+      "step": 11800
+    },
+    {
+      "epoch": 0.09852950917400809,
+      "grad_norm": 0.22209756076335907,
+      "learning_rate": 0.00019343303802905418,
+      "loss": 1.679,
+      "step": 11900
+    },
+    {
+      "epoch": 0.09935748824269723,
+      "grad_norm": 0.20740656554698944,
+      "learning_rate": 0.0001933778394199046,
+      "loss": 1.6494,
+      "step": 12000
+    },
+    {
+      "epoch": 0.10018546731138636,
+      "grad_norm": 0.19304044544696808,
+      "learning_rate": 0.000193322640810755,
+      "loss": 1.644,
+      "step": 12100
+    },
+    {
+      "epoch": 0.1010134463800755,
+      "grad_norm": 0.2168605774641037,
+      "learning_rate": 0.0001932674422016054,
+      "loss": 1.673,
+      "step": 12200
+    },
+    {
+      "epoch": 0.10184142544876465,
+      "grad_norm": 0.2477482557296753,
+      "learning_rate": 0.00019321224359245578,
+      "loss": 1.6912,
+      "step": 12300
+    },
+    {
+      "epoch": 0.1026694045174538,
+      "grad_norm": 0.220897376537323,
+      "learning_rate": 0.00019315704498330618,
+      "loss": 1.6553,
+      "step": 12400
+    },
+    {
+      "epoch": 0.10349738358614294,
+      "grad_norm": 0.22782349586486816,
+      "learning_rate": 0.00019310184637415657,
+      "loss": 1.6585,
+      "step": 12500
+    },
+    {
+      "epoch": 0.10432536265483208,
+      "grad_norm": 0.22199232876300812,
+      "learning_rate": 0.00019304664776500697,
+      "loss": 1.6617,
+      "step": 12600
+    },
+    {
+      "epoch": 0.10515334172352123,
+      "grad_norm": 0.19177618622779846,
+      "learning_rate": 0.00019299144915585736,
+      "loss": 1.6788,
+      "step": 12700
+    },
+    {
+      "epoch": 0.10598132079221037,
+      "grad_norm": 0.20250776410102844,
+      "learning_rate": 0.00019293625054670775,
+      "loss": 1.6329,
+      "step": 12800
+    },
+    {
+      "epoch": 0.10680929986089951,
+      "grad_norm": 0.23451390862464905,
+      "learning_rate": 0.00019288105193755815,
+      "loss": 1.6336,
+      "step": 12900
+    },
+    {
+      "epoch": 0.10763727892958866,
+      "grad_norm": 0.20917341113090515,
+      "learning_rate": 0.00019282585332840854,
+      "loss": 1.6436,
+      "step": 13000
+    },
+    {
+      "epoch": 0.1084652579982778,
+      "grad_norm": 0.24702110886573792,
+      "learning_rate": 0.00019277065471925894,
+      "loss": 1.6569,
+      "step": 13100
+    },
+    {
+      "epoch": 0.10929323706696695,
+      "grad_norm": 0.2046515792608261,
+      "learning_rate": 0.00019271545611010933,
+      "loss": 1.6728,
+      "step": 13200
+    },
+    {
+      "epoch": 0.11012121613565609,
+      "grad_norm": 0.23590846359729767,
+      "learning_rate": 0.00019266025750095973,
+      "loss": 1.6808,
+      "step": 13300
+    },
+    {
+      "epoch": 0.11094919520434524,
+      "grad_norm": 0.2132890671491623,
+      "learning_rate": 0.00019260505889181012,
+      "loss": 1.6428,
+      "step": 13400
+    },
+    {
+      "epoch": 0.11177717427303438,
+      "grad_norm": 0.22132565081119537,
+      "learning_rate": 0.000192550412268752,
+      "loss": 1.688,
+      "step": 13500
+    },
+    {
+      "epoch": 0.11260515334172352,
+      "grad_norm": 0.2686227262020111,
+      "learning_rate": 0.0001924952136596024,
+      "loss": 1.6523,
+      "step": 13600
+    },
+    {
+      "epoch": 0.11343313241041267,
+      "grad_norm": 0.2114957571029663,
+      "learning_rate": 0.0001924400150504528,
+      "loss": 1.643,
+      "step": 13700
+    },
+    {
+      "epoch": 0.11426111147910181,
+      "grad_norm": 0.22338928282260895,
+      "learning_rate": 0.0001923848164413032,
+      "loss": 1.6371,
+      "step": 13800
+    },
+    {
+      "epoch": 0.11508909054779096,
+      "grad_norm": 0.27738282084465027,
+      "learning_rate": 0.00019232961783215358,
+      "loss": 1.664,
+      "step": 13900
+    },
+    {
+      "epoch": 0.1159170696164801,
+      "grad_norm": 0.211566761136055,
+      "learning_rate": 0.00019227441922300398,
+      "loss": 1.7005,
+      "step": 14000
+    },
+    {
+      "epoch": 0.11674504868516924,
+      "grad_norm": 0.23775742948055267,
+      "learning_rate": 0.00019221922061385437,
+      "loss": 1.6538,
+      "step": 14100
+    },
+    {
+      "epoch": 0.11757302775385839,
+      "grad_norm": 0.24485789239406586,
+      "learning_rate": 0.00019216402200470477,
+      "loss": 1.6969,
+      "step": 14200
+    },
+    {
+      "epoch": 0.11840100682254752,
+      "grad_norm": 0.2573772966861725,
+      "learning_rate": 0.00019210882339555516,
+      "loss": 1.6268,
+      "step": 14300
+    },
+    {
+      "epoch": 0.11922898589123666,
+      "grad_norm": 0.24128000438213348,
+      "learning_rate": 0.00019205362478640555,
+      "loss": 1.6765,
+      "step": 14400
+    },
+    {
+      "epoch": 0.12005696495992581,
+      "grad_norm": 0.2213265597820282,
+      "learning_rate": 0.00019199842617725595,
+      "loss": 1.6483,
+      "step": 14500
+    },
+    {
+      "epoch": 0.12088494402861495,
+      "grad_norm": 0.3523562252521515,
+      "learning_rate": 0.00019194322756810634,
+      "loss": 1.6394,
+      "step": 14600
+    },
+    {
+      "epoch": 0.1217129230973041,
+      "grad_norm": 0.26578399538993835,
+      "learning_rate": 0.00019188802895895674,
+      "loss": 1.7057,
+      "step": 14700
+    },
+    {
+      "epoch": 0.12254090216599324,
+      "grad_norm": 0.2317681610584259,
+      "learning_rate": 0.00019183283034980713,
+      "loss": 1.6244,
+      "step": 14800
+    },
+    {
+      "epoch": 0.12336888123468238,
+      "grad_norm": 0.2221396416425705,
+      "learning_rate": 0.00019177763174065752,
+      "loss": 1.6442,
+      "step": 14900
+    },
+    {
+      "epoch": 0.12419686030337153,
+      "grad_norm": 0.2175658941268921,
+      "learning_rate": 0.00019172243313150792,
+      "loss": 1.6643,
+      "step": 15000
+    },
+    {
+      "epoch": 0.1250248393720607,
+      "grad_norm": 0.21838437020778656,
+      "learning_rate": 0.0001916672345223583,
+      "loss": 1.6629,
+      "step": 15100
+    },
+    {
+      "epoch": 0.12585281844074983,
+      "grad_norm": 0.19501908123493195,
+      "learning_rate": 0.0001916120359132087,
+      "loss": 1.6466,
+      "step": 15200
+    },
+    {
+      "epoch": 0.12668079750943897,
+      "grad_norm": 0.24163535237312317,
+      "learning_rate": 0.0001915568373040591,
+      "loss": 1.6448,
+      "step": 15300
+    },
+    {
+      "epoch": 0.1275087765781281,
+      "grad_norm": 0.2042425274848938,
+      "learning_rate": 0.0001915016386949095,
+      "loss": 1.6639,
+      "step": 15400
+    },
+    {
+      "epoch": 0.12833675564681724,
+      "grad_norm": 0.22937864065170288,
+      "learning_rate": 0.0001914464400857599,
+      "loss": 1.635,
+      "step": 15500
+    },
+    {
+      "epoch": 0.12916473471550638,
+      "grad_norm": 0.2340078055858612,
+      "learning_rate": 0.00019139124147661028,
+      "loss": 1.6335,
+      "step": 15600
+    },
+    {
+      "epoch": 0.12999271378419552,
+      "grad_norm": 0.24161306023597717,
+      "learning_rate": 0.00019133604286746068,
+      "loss": 1.634,
+      "step": 15700
+    },
+    {
+      "epoch": 0.13082069285288467,
+      "grad_norm": 0.25775766372680664,
+      "learning_rate": 0.00019128084425831107,
+      "loss": 1.6331,
+      "step": 15800
+    },
+    {
+      "epoch": 0.1316486719215738,
+      "grad_norm": 0.23609529435634613,
+      "learning_rate": 0.00019122564564916147,
+      "loss": 1.6431,
+      "step": 15900
+    },
+    {
+      "epoch": 0.13247665099026296,
+      "grad_norm": 0.24818001687526703,
+      "learning_rate": 0.00019117044704001186,
+      "loss": 1.6355,
+      "step": 16000
+    },
+    {
+      "epoch": 0.1333046300589521,
+      "grad_norm": 0.2805030941963196,
+      "learning_rate": 0.00019111580041695375,
+      "loss": 1.6459,
+      "step": 16100
+    },
+    {
+      "epoch": 0.13413260912764124,
+      "grad_norm": 0.2721966505050659,
+      "learning_rate": 0.00019106060180780417,
+      "loss": 1.6384,
+      "step": 16200
+    },
+    {
+      "epoch": 0.1349605881963304,
+      "grad_norm": 0.23994600772857666,
+      "learning_rate": 0.00019100540319865454,
+      "loss": 1.627,
+      "step": 16300
+    },
+    {
+      "epoch": 0.13578856726501953,
+      "grad_norm": 0.20296916365623474,
+      "learning_rate": 0.00019095020458950496,
+      "loss": 1.6528,
+      "step": 16400
+    },
+    {
+      "epoch": 0.13661654633370868,
+      "grad_norm": 0.21805709600448608,
+      "learning_rate": 0.00019089500598035535,
+      "loss": 1.64,
+      "step": 16500
+    },
+    {
+      "epoch": 0.13744452540239782,
+      "grad_norm": 0.25123125314712524,
+      "learning_rate": 0.00019083980737120572,
+      "loss": 1.6889,
+      "step": 16600
+    },
+    {
+      "epoch": 0.13827250447108697,
+      "grad_norm": 0.2638598680496216,
+      "learning_rate": 0.00019078460876205614,
+      "loss": 1.6507,
+      "step": 16700
+    },
+    {
+      "epoch": 0.1391004835397761,
+      "grad_norm": 0.21334075927734375,
+      "learning_rate": 0.00019072941015290653,
+      "loss": 1.6428,
+      "step": 16800
+    },
+    {
+      "epoch": 0.13992846260846525,
+      "grad_norm": 0.2150932401418686,
+      "learning_rate": 0.0001906742115437569,
+      "loss": 1.6277,
+      "step": 16900
+    },
+    {
+      "epoch": 0.1407564416771544,
+      "grad_norm": 0.28504666686058044,
+      "learning_rate": 0.00019061901293460732,
+      "loss": 1.6609,
+      "step": 17000
+    },
+    {
+      "epoch": 0.14158442074584354,
+      "grad_norm": 0.22296425700187683,
+      "learning_rate": 0.00019056381432545772,
+      "loss": 1.6359,
+      "step": 17100
+    },
+    {
+      "epoch": 0.1424123998145327,
+      "grad_norm": 0.23885318636894226,
+      "learning_rate": 0.00019050861571630808,
+      "loss": 1.6262,
+      "step": 17200
+    },
+    {
+      "epoch": 0.14324037888322183,
+      "grad_norm": 0.22231832146644592,
+      "learning_rate": 0.0001904534171071585,
+      "loss": 1.6369,
+      "step": 17300
+    },
+    {
+      "epoch": 0.14406835795191097,
+      "grad_norm": 0.25279131531715393,
+      "learning_rate": 0.0001903982184980089,
+      "loss": 1.6376,
+      "step": 17400
+    },
+    {
+      "epoch": 0.14489633702060012,
+      "grad_norm": 0.2754824459552765,
+      "learning_rate": 0.00019034301988885927,
+      "loss": 1.6579,
+      "step": 17500
+    },
+    {
+      "epoch": 0.14572431608928926,
+      "grad_norm": 0.22775475680828094,
+      "learning_rate": 0.0001902878212797097,
+      "loss": 1.674,
+      "step": 17600
+    },
+    {
+      "epoch": 0.1465522951579784,
+      "grad_norm": 0.21519581973552704,
+      "learning_rate": 0.00019023262267056008,
+      "loss": 1.654,
+      "step": 17700
+    },
+    {
+      "epoch": 0.14738027422666755,
+      "grad_norm": 0.2484564632177353,
+      "learning_rate": 0.00019017742406141045,
+      "loss": 1.6046,
+      "step": 17800
+    },
+    {
+      "epoch": 0.1482082532953567,
+      "grad_norm": 0.21091294288635254,
+      "learning_rate": 0.00019012222545226087,
+      "loss": 1.616,
+      "step": 17900
+    },
+    {
+      "epoch": 0.14903623236404584,
+      "grad_norm": 0.27948471903800964,
+      "learning_rate": 0.00019006702684311127,
+      "loss": 1.6866,
+      "step": 18000
+    },
+    {
+      "epoch": 0.14986421143273498,
+      "grad_norm": 0.28303253650665283,
+      "learning_rate": 0.00019001182823396163,
+      "loss": 1.6569,
+      "step": 18100
+    },
+    {
+      "epoch": 0.15069219050142413,
+      "grad_norm": 0.2028988003730774,
+      "learning_rate": 0.00018995662962481205,
+      "loss": 1.6752,
+      "step": 18200
+    },
+    {
+      "epoch": 0.15152016957011327,
+      "grad_norm": 0.237883523106575,
+      "learning_rate": 0.0001899019830017539,
+      "loss": 1.6618,
+      "step": 18300
+    },
+    {
+      "epoch": 0.15234814863880242,
+      "grad_norm": 0.21337077021598816,
+      "learning_rate": 0.00018984678439260433,
+      "loss": 1.6397,
+      "step": 18400
+    },
+    {
+      "epoch": 0.15317612770749156,
+      "grad_norm": 0.20675146579742432,
+      "learning_rate": 0.00018979158578345473,
+      "loss": 1.6382,
+      "step": 18500
+    },
+    {
+      "epoch": 0.1540041067761807,
+      "grad_norm": 0.23405767977237701,
+      "learning_rate": 0.0001897363871743051,
+      "loss": 1.6112,
+      "step": 18600
+    },
+    {
+      "epoch": 0.15483208584486985,
+      "grad_norm": 0.22801779210567474,
+      "learning_rate": 0.00018968118856515552,
+      "loss": 1.6332,
+      "step": 18700
+    },
+    {
+      "epoch": 0.155660064913559,
+      "grad_norm": 0.28877681493759155,
+      "learning_rate": 0.0001896259899560059,
+      "loss": 1.618,
+      "step": 18800
+    },
+    {
+      "epoch": 0.15648804398224814,
+      "grad_norm": 0.2644197940826416,
+      "learning_rate": 0.00018957079134685628,
+      "loss": 1.6617,
+      "step": 18900
+    },
+    {
+      "epoch": 0.15731602305093728,
+      "grad_norm": 0.20182272791862488,
+      "learning_rate": 0.0001895155927377067,
+      "loss": 1.6273,
+      "step": 19000
+    },
+    {
+      "epoch": 0.15814400211962643,
+      "grad_norm": 0.2702679932117462,
+      "learning_rate": 0.0001894603941285571,
+      "loss": 1.6573,
+      "step": 19100
+    },
+    {
+      "epoch": 0.15897198118831557,
+      "grad_norm": 0.24677981436252594,
+      "learning_rate": 0.00018940519551940746,
+      "loss": 1.645,
+      "step": 19200
+    },
+    {
+      "epoch": 0.15979996025700471,
+      "grad_norm": 0.22584903240203857,
+      "learning_rate": 0.00018934999691025788,
+      "loss": 1.6496,
+      "step": 19300
+    },
+    {
+      "epoch": 0.16062793932569386,
+      "grad_norm": 0.25998055934906006,
+      "learning_rate": 0.00018929479830110828,
+      "loss": 1.6406,
+      "step": 19400
+    },
+    {
+      "epoch": 0.161455918394383,
+      "grad_norm": 0.24679061770439148,
+      "learning_rate": 0.00018923959969195867,
+      "loss": 1.664,
+      "step": 19500
+    },
+    {
+      "epoch": 0.16228389746307215,
+      "grad_norm": 0.21543079614639282,
+      "learning_rate": 0.00018918440108280907,
+      "loss": 1.6316,
+      "step": 19600
+    },
+    {
+      "epoch": 0.1631118765317613,
+      "grad_norm": 0.20581622421741486,
+      "learning_rate": 0.00018912920247365946,
+      "loss": 1.6417,
+      "step": 19700
+    },
+    {
+      "epoch": 0.1639398556004504,
+      "grad_norm": 0.24385277926921844,
+      "learning_rate": 0.00018907400386450985,
+      "loss": 1.6462,
+      "step": 19800
+    },
+    {
+      "epoch": 0.16476783466913955,
+      "grad_norm": 0.21104101836681366,
+      "learning_rate": 0.00018901880525536025,
+      "loss": 1.6255,
+      "step": 19900
+    },
+    {
+      "epoch": 0.1655958137378287,
+      "grad_norm": 0.2160460352897644,
+      "learning_rate": 0.00018896360664621064,
+      "loss": 1.6602,
+      "step": 20000
+    },
+    {
+      "epoch": 0.16642379280651784,
+      "grad_norm": 0.27930399775505066,
+      "learning_rate": 0.00018890840803706104,
+      "loss": 1.676,
+      "step": 20100
+    },
+    {
+      "epoch": 0.16725177187520698,
+      "grad_norm": 0.21541720628738403,
+      "learning_rate": 0.00018885320942791143,
+      "loss": 1.627,
+      "step": 20200
+    },
+    {
+      "epoch": 0.16807975094389613,
+      "grad_norm": 0.21303394436836243,
+      "learning_rate": 0.00018879801081876182,
+      "loss": 1.6332,
+      "step": 20300
+    },
+    {
+      "epoch": 0.16890773001258527,
+      "grad_norm": 0.22445669770240784,
+      "learning_rate": 0.0001887433641957037,
+      "loss": 1.6466,
+      "step": 20400
+    },
+    {
+      "epoch": 0.16973570908127442,
+      "grad_norm": 0.21410098671913147,
+      "learning_rate": 0.0001886881655865541,
+      "loss": 1.6392,
+      "step": 20500
+    },
+    {
+      "epoch": 0.17056368814996356,
+      "grad_norm": 0.19233547151088715,
+      "learning_rate": 0.0001886329669774045,
+      "loss": 1.6396,
+      "step": 20600
+    },
+    {
+      "epoch": 0.1713916672186527,
+      "grad_norm": 0.20589858293533325,
+      "learning_rate": 0.0001885777683682549,
+      "loss": 1.6295,
+      "step": 20700
+    },
+    {
+      "epoch": 0.17221964628734185,
+      "grad_norm": 0.25907692313194275,
+      "learning_rate": 0.0001885225697591053,
+      "loss": 1.6329,
+      "step": 20800
+    },
+    {
+      "epoch": 0.173047625356031,
+      "grad_norm": 0.2765398919582367,
+      "learning_rate": 0.00018846737114995568,
+      "loss": 1.6868,
+      "step": 20900
+    },
+    {
+      "epoch": 0.17387560442472014,
+      "grad_norm": 0.23948393762111664,
+      "learning_rate": 0.00018841217254080608,
+      "loss": 1.6664,
+      "step": 21000
+    },
+    {
+      "epoch": 0.17470358349340928,
+      "grad_norm": 0.2371809184551239,
+      "learning_rate": 0.00018835697393165647,
+      "loss": 1.6843,
+      "step": 21100
+    },
+    {
+      "epoch": 0.17553156256209843,
+      "grad_norm": 0.19910277426242828,
+      "learning_rate": 0.00018830177532250686,
+      "loss": 1.619,
+      "step": 21200
+    },
+    {
+      "epoch": 0.17635954163078757,
+      "grad_norm": 0.24008843302726746,
+      "learning_rate": 0.00018824657671335726,
+      "loss": 1.6254,
+      "step": 21300
+    },
+    {
+      "epoch": 0.1771875206994767,
+      "grad_norm": 0.22527045011520386,
+      "learning_rate": 0.00018819137810420765,
+      "loss": 1.656,
+      "step": 21400
+    },
+    {
+      "epoch": 0.17801549976816586,
+      "grad_norm": 0.21581338346004486,
+      "learning_rate": 0.00018813617949505805,
+      "loss": 1.6297,
+      "step": 21500
+    },
+    {
+      "epoch": 0.178843478836855,
+      "grad_norm": 0.23672404885292053,
+      "learning_rate": 0.00018808098088590844,
+      "loss": 1.6493,
+      "step": 21600
+    },
+    {
+      "epoch": 0.17967145790554415,
+      "grad_norm": 0.22824163734912872,
+      "learning_rate": 0.00018802578227675884,
+      "loss": 1.6428,
+      "step": 21700
+    },
+    {
+      "epoch": 0.1804994369742333,
+      "grad_norm": 0.21679380536079407,
+      "learning_rate": 0.00018797058366760923,
+      "loss": 1.6373,
+      "step": 21800
+    },
+    {
+      "epoch": 0.18132741604292243,
+      "grad_norm": 0.23232513666152954,
+      "learning_rate": 0.00018791538505845962,
+      "loss": 1.6408,
+      "step": 21900
+    },
+    {
+      "epoch": 0.18215539511161158,
+      "grad_norm": 0.21634866297245026,
+      "learning_rate": 0.00018786018644931002,
+      "loss": 1.6428,
+      "step": 22000
+    },
+    {
+      "epoch": 0.18298337418030072,
+      "grad_norm": 0.24199073016643524,
+      "learning_rate": 0.0001878049878401604,
+      "loss": 1.6912,
+      "step": 22100
+    },
+    {
+      "epoch": 0.18381135324898987,
+      "grad_norm": 0.24073319137096405,
+      "learning_rate": 0.0001877497892310108,
+      "loss": 1.6697,
+      "step": 22200
+    },
+    {
+      "epoch": 0.184639332317679,
+      "grad_norm": 0.20666244626045227,
+      "learning_rate": 0.0001876945906218612,
+      "loss": 1.6372,
+      "step": 22300
+    },
+    {
+      "epoch": 0.18546731138636816,
+      "grad_norm": 0.24439455568790436,
+      "learning_rate": 0.0001876393920127116,
+      "loss": 1.6562,
+      "step": 22400
+    },
+    {
+      "epoch": 0.1862952904550573,
+      "grad_norm": 0.20771746337413788,
+      "learning_rate": 0.000187584193403562,
+      "loss": 1.6342,
+      "step": 22500
+    },
+    {
+      "epoch": 0.18712326952374644,
+      "grad_norm": 0.23821604251861572,
+      "learning_rate": 0.00018752899479441238,
+      "loss": 1.6266,
+      "step": 22600
+    },
+    {
+      "epoch": 0.1879512485924356,
+      "grad_norm": 0.20013689994812012,
+      "learning_rate": 0.00018747379618526278,
+      "loss": 1.6394,
+      "step": 22700
+    },
+    {
+      "epoch": 0.18877922766112473,
+      "grad_norm": 0.24145907163619995,
+      "learning_rate": 0.00018741859757611317,
+      "loss": 1.639,
+      "step": 22800
+    },
+    {
+      "epoch": 0.18960720672981388,
+      "grad_norm": 0.227446511387825,
+      "learning_rate": 0.00018736395095305509,
+      "loss": 1.6245,
+      "step": 22900
+    },
+    {
+      "epoch": 0.19043518579850302,
+      "grad_norm": 0.23569269478321075,
+      "learning_rate": 0.00018730875234390545,
+      "loss": 1.6536,
+      "step": 23000
+    },
+    {
+      "epoch": 0.19126316486719216,
+      "grad_norm": 0.26718559861183167,
+      "learning_rate": 0.00018725355373475585,
+      "loss": 1.6152,
+      "step": 23100
+    },
+    {
+      "epoch": 0.1920911439358813,
+      "grad_norm": 0.2448846399784088,
+      "learning_rate": 0.00018719835512560627,
+      "loss": 1.6462,
+      "step": 23200
+    },
+    {
+      "epoch": 0.19291912300457045,
+      "grad_norm": 0.22957266867160797,
+      "learning_rate": 0.00018714315651645664,
+      "loss": 1.6493,
+      "step": 23300
+    },
+    {
+      "epoch": 0.1937471020732596,
+      "grad_norm": 0.24881067872047424,
+      "learning_rate": 0.00018708795790730703,
+      "loss": 1.6509,
+      "step": 23400
+    },
+    {
+      "epoch": 0.19457508114194874,
+      "grad_norm": 0.26392731070518494,
+      "learning_rate": 0.00018703275929815745,
+      "loss": 1.6436,
+      "step": 23500
+    },
+    {
+      "epoch": 0.19540306021063789,
+      "grad_norm": 0.23469410836696625,
+      "learning_rate": 0.00018697756068900782,
+      "loss": 1.6494,
+      "step": 23600
+    },
+    {
+      "epoch": 0.19623103927932703,
+      "grad_norm": 0.191993847489357,
+      "learning_rate": 0.00018692236207985824,
+      "loss": 1.6118,
+      "step": 23700
+    },
+    {
+      "epoch": 0.19705901834801617,
+      "grad_norm": 0.24076974391937256,
+      "learning_rate": 0.00018686716347070863,
+      "loss": 1.6414,
+      "step": 23800
+    },
+    {
+      "epoch": 0.19788699741670532,
+      "grad_norm": 0.22676746547222137,
+      "learning_rate": 0.000186811964861559,
+      "loss": 1.6392,
+      "step": 23900
+    },
+    {
+      "epoch": 0.19871497648539446,
+      "grad_norm": 0.2693268060684204,
+      "learning_rate": 0.00018675676625240942,
+      "loss": 1.6254,
+      "step": 24000
+    },
+    {
+      "epoch": 0.1995429555540836,
+      "grad_norm": 0.23930659890174866,
+      "learning_rate": 0.00018670156764325982,
+      "loss": 1.6451,
+      "step": 24100
+    },
+    {
+      "epoch": 0.20037093462277272,
+      "grad_norm": 0.23256762325763702,
+      "learning_rate": 0.00018664636903411018,
+      "loss": 1.6424,
+      "step": 24200
+    },
+    {
+      "epoch": 0.20119891369146187,
+      "grad_norm": 0.2577391564846039,
+      "learning_rate": 0.0001865911704249606,
+      "loss": 1.6212,
+      "step": 24300
+    },
+    {
+      "epoch": 0.202026892760151,
+      "grad_norm": 0.26071691513061523,
+      "learning_rate": 0.000186535971815811,
+      "loss": 1.6314,
+      "step": 24400
+    },
+    {
+      "epoch": 0.20285487182884016,
+      "grad_norm": 0.22462689876556396,
+      "learning_rate": 0.00018648077320666137,
+      "loss": 1.6237,
+      "step": 24500
+    },
+    {
+      "epoch": 0.2036828508975293,
+      "grad_norm": 0.24064601957798004,
+      "learning_rate": 0.0001864255745975118,
+      "loss": 1.6391,
+      "step": 24600
+    },
+    {
+      "epoch": 0.20451082996621844,
+      "grad_norm": 0.22928814589977264,
+      "learning_rate": 0.00018637037598836218,
+      "loss": 1.6147,
+      "step": 24700
+    },
+    {
+      "epoch": 0.2053388090349076,
+      "grad_norm": 0.22983402013778687,
+      "learning_rate": 0.00018631517737921255,
+      "loss": 1.6427,
+      "step": 24800
+    },
+    {
+      "epoch": 0.20616678810359673,
+      "grad_norm": 0.22635537385940552,
+      "learning_rate": 0.00018625997877006297,
+      "loss": 1.6421,
+      "step": 24900
+    },
+    {
+      "epoch": 0.20699476717228588,
+      "grad_norm": 0.23368728160858154,
+      "learning_rate": 0.00018620478016091336,
+      "loss": 1.6472,
+      "step": 25000
+    },
+    {
+      "epoch": 0.20782274624097502,
+      "grad_norm": 0.24677674472332,
+      "learning_rate": 0.00018614958155176373,
+      "loss": 1.6697,
+      "step": 25100
+    },
+    {
+      "epoch": 0.20865072530966416,
+      "grad_norm": 0.2160155028104782,
+      "learning_rate": 0.00018609438294261415,
+      "loss": 1.6327,
+      "step": 25200
+    },
+    {
+      "epoch": 0.2094787043783533,
+      "grad_norm": 0.20933164656162262,
+      "learning_rate": 0.00018603918433346455,
+      "loss": 1.6464,
+      "step": 25300
+    },
+    {
+      "epoch": 0.21030668344704245,
+      "grad_norm": 0.2508947253227234,
+      "learning_rate": 0.00018598398572431491,
+      "loss": 1.6531,
+      "step": 25400
+    },
+    {
+      "epoch": 0.2111346625157316,
+      "grad_norm": 0.264946311712265,
+      "learning_rate": 0.00018592878711516534,
+      "loss": 1.6285,
+      "step": 25500
+    },
+    {
+      "epoch": 0.21196264158442074,
+      "grad_norm": 0.23839199542999268,
+      "learning_rate": 0.00018587358850601573,
+      "loss": 1.6601,
+      "step": 25600
+    },
+    {
+      "epoch": 0.21279062065310989,
+      "grad_norm": 0.20937936007976532,
+      "learning_rate": 0.0001858183898968661,
+      "loss": 1.6199,
+      "step": 25700
+    },
+    {
+      "epoch": 0.21361859972179903,
+      "grad_norm": 0.2978517413139343,
+      "learning_rate": 0.00018576319128771652,
+      "loss": 1.6233,
+      "step": 25800
+    },
+    {
+      "epoch": 0.21444657879048817,
+      "grad_norm": 0.22715617716312408,
+      "learning_rate": 0.0001857079926785669,
+      "loss": 1.6616,
+      "step": 25900
+    },
+    {
+      "epoch": 0.21527455785917732,
+      "grad_norm": 0.21679271757602692,
+      "learning_rate": 0.00018565279406941728,
+      "loss": 1.6562,
+      "step": 26000
+    },
+    {
+      "epoch": 0.21610253692786646,
+      "grad_norm": 0.2540909945964813,
+      "learning_rate": 0.0001855975954602677,
+      "loss": 1.6292,
+      "step": 26100
+    },
+    {
+      "epoch": 0.2169305159965556,
+      "grad_norm": 0.23456817865371704,
+      "learning_rate": 0.0001855423968511181,
+      "loss": 1.6117,
+      "step": 26200
+    },
+    {
+      "epoch": 0.21775849506524475,
+      "grad_norm": 0.2706565856933594,
+      "learning_rate": 0.00018548719824196846,
+      "loss": 1.6119,
+      "step": 26300
+    },
+    {
+      "epoch": 0.2185864741339339,
+      "grad_norm": 0.24881504476070404,
+      "learning_rate": 0.00018543199963281888,
+      "loss": 1.6263,
+      "step": 26400
+    },
+    {
+      "epoch": 0.21941445320262304,
+      "grad_norm": 0.3882363736629486,
+      "learning_rate": 0.00018537680102366928,
+      "loss": 1.6537,
+      "step": 26500
+    },
+    {
+      "epoch": 0.22024243227131218,
+      "grad_norm": 0.19953063130378723,
+      "learning_rate": 0.00018532160241451965,
+      "loss": 1.6786,
+      "step": 26600
+    },
+    {
+      "epoch": 0.22107041134000133,
+      "grad_norm": 0.21415996551513672,
+      "learning_rate": 0.00018526640380537007,
+      "loss": 1.6582,
+      "step": 26700
+    },
+    {
+      "epoch": 0.22189839040869047,
+      "grad_norm": 0.22917871177196503,
+      "learning_rate": 0.00018521120519622046,
+      "loss": 1.5913,
+      "step": 26800
+    },
+    {
+      "epoch": 0.22272636947737962,
+      "grad_norm": 0.23988696932792664,
+      "learning_rate": 0.00018515600658707083,
+      "loss": 1.6284,
+      "step": 26900
+    },
+    {
+      "epoch": 0.22355434854606876,
+      "grad_norm": 0.21336591243743896,
+      "learning_rate": 0.00018510080797792125,
+      "loss": 1.6594,
+      "step": 27000
+    },
+    {
+      "epoch": 0.2243823276147579,
+      "grad_norm": 0.22978056967258453,
+      "learning_rate": 0.00018504560936877164,
+      "loss": 1.653,
+      "step": 27100
+    },
+    {
+      "epoch": 0.22521030668344705,
+      "grad_norm": 0.27317842841148376,
+      "learning_rate": 0.00018499096274571353,
+      "loss": 1.6267,
+      "step": 27200
+    },
+    {
+      "epoch": 0.2260382857521362,
+      "grad_norm": 0.2396378219127655,
+      "learning_rate": 0.00018493576413656392,
+      "loss": 1.6452,
+      "step": 27300
+    },
+    {
+      "epoch": 0.22686626482082534,
+      "grad_norm": 0.2800294756889343,
+      "learning_rate": 0.00018488056552741432,
+      "loss": 1.6722,
+      "step": 27400
+    },
+    {
+      "epoch": 0.22769424388951448,
+      "grad_norm": 0.24665579199790955,
+      "learning_rate": 0.0001848253669182647,
+      "loss": 1.6473,
+      "step": 27500
+    },
+    {
+      "epoch": 0.22852222295820362,
+      "grad_norm": 0.23755255341529846,
+      "learning_rate": 0.0001847701683091151,
+      "loss": 1.6072,
+      "step": 27600
+    },
+    {
+      "epoch": 0.22935020202689277,
+      "grad_norm": 0.23274143040180206,
+      "learning_rate": 0.0001847149696999655,
+      "loss": 1.6547,
+      "step": 27700
+    },
+    {
+      "epoch": 0.2301781810955819,
+      "grad_norm": 0.20112739503383636,
+      "learning_rate": 0.0001846597710908159,
+      "loss": 1.6307,
+      "step": 27800
+    },
+    {
+      "epoch": 0.23100616016427106,
+      "grad_norm": 0.29942500591278076,
+      "learning_rate": 0.0001846045724816663,
+      "loss": 1.6384,
+      "step": 27900
+    },
+    {
+      "epoch": 0.2318341392329602,
+      "grad_norm": 0.2471247911453247,
+      "learning_rate": 0.00018454937387251668,
+      "loss": 1.6568,
+      "step": 28000
+    },
+    {
+      "epoch": 0.23266211830164935,
+      "grad_norm": 0.2586652934551239,
+      "learning_rate": 0.00018449417526336708,
+      "loss": 1.6167,
+      "step": 28100
+    },
+    {
+      "epoch": 0.2334900973703385,
+      "grad_norm": 0.23116792738437653,
+      "learning_rate": 0.00018443897665421747,
+      "loss": 1.6222,
+      "step": 28200
+    },
+    {
+      "epoch": 0.23431807643902763,
+      "grad_norm": 0.22467024624347687,
+      "learning_rate": 0.00018438433003115936,
+      "loss": 1.6343,
+      "step": 28300
+    },
+    {
+      "epoch": 0.23514605550771678,
+      "grad_norm": 0.28815844655036926,
+      "learning_rate": 0.00018432913142200975,
+      "loss": 1.6388,
+      "step": 28400
+    },
+    {
+      "epoch": 0.23597403457640592,
+      "grad_norm": 0.22312502562999725,
+      "learning_rate": 0.00018427393281286017,
+      "loss": 1.639,
+      "step": 28500
+    },
+    {
+      "epoch": 0.23680201364509504,
+      "grad_norm": 0.2542140781879425,
+      "learning_rate": 0.00018421873420371054,
+      "loss": 1.6019,
+      "step": 28600
+    },
+    {
+      "epoch": 0.23762999271378418,
+      "grad_norm": 0.2540210783481598,
+      "learning_rate": 0.00018416353559456094,
+      "loss": 1.6217,
+      "step": 28700
+    },
+    {
+      "epoch": 0.23845797178247333,
+      "grad_norm": 0.2396947294473648,
+      "learning_rate": 0.00018410833698541136,
+      "loss": 1.6186,
+      "step": 28800
+    },
+    {
+      "epoch": 0.23928595085116247,
+      "grad_norm": 0.2223241776227951,
+      "learning_rate": 0.00018405313837626172,
+      "loss": 1.6385,
+      "step": 28900
+    },
+    {
+      "epoch": 0.24011392991985162,
+      "grad_norm": 0.21971747279167175,
+      "learning_rate": 0.00018399793976711212,
+      "loss": 1.6357,
+      "step": 29000
+    },
+    {
+      "epoch": 0.24094190898854076,
+      "grad_norm": 0.24293990433216095,
+      "learning_rate": 0.00018394274115796254,
+      "loss": 1.6677,
+      "step": 29100
+    },
+    {
+      "epoch": 0.2417698880572299,
+      "grad_norm": 0.2490108758211136,
+      "learning_rate": 0.0001838875425488129,
+      "loss": 1.6423,
+      "step": 29200
+    },
+    {
+      "epoch": 0.24259786712591905,
+      "grad_norm": 0.2689921259880066,
+      "learning_rate": 0.0001838323439396633,
+      "loss": 1.6273,
+      "step": 29300
+    },
+    {
+      "epoch": 0.2434258461946082,
+      "grad_norm": 0.21055381000041962,
+      "learning_rate": 0.0001837776973166052,
+      "loss": 1.5876,
+      "step": 29400
+    },
+    {
+      "epoch": 0.24425382526329734,
+      "grad_norm": 0.24339735507965088,
+      "learning_rate": 0.00018372249870745558,
+      "loss": 1.5942,
+      "step": 29500
+    },
+    {
+      "epoch": 0.24508180433198648,
+      "grad_norm": 0.22557999193668365,
+      "learning_rate": 0.000183667300098306,
+      "loss": 1.6407,
+      "step": 29600
+    },
+    {
+      "epoch": 0.24590978340067562,
+      "grad_norm": 0.27617794275283813,
+      "learning_rate": 0.00018361210148915637,
+      "loss": 1.626,
+      "step": 29700
+    },
+    {
+      "epoch": 0.24673776246936477,
+      "grad_norm": 0.2763831615447998,
+      "learning_rate": 0.00018355690288000676,
+      "loss": 1.6035,
+      "step": 29800
+    },
+    {
+      "epoch": 0.2475657415380539,
+      "grad_norm": 0.23689210414886475,
+      "learning_rate": 0.00018350170427085719,
+      "loss": 1.6772,
+      "step": 29900
+    },
+    {
+      "epoch": 0.24839372060674306,
+      "grad_norm": 0.22767847776412964,
+      "learning_rate": 0.00018344650566170755,
+      "loss": 1.6315,
+      "step": 30000
+    },
+    {
+      "epoch": 0.2492216996754322,
+      "grad_norm": 0.23645687103271484,
+      "learning_rate": 0.00018339130705255795,
+      "loss": 1.6424,
+      "step": 30100
+    },
+    {
+      "epoch": 0.2500496787441214,
+      "grad_norm": 0.29076236486434937,
+      "learning_rate": 0.00018333610844340837,
+      "loss": 1.6175,
+      "step": 30200
+    },
+    {
+      "epoch": 0.2508776578128105,
+      "grad_norm": 0.22849859297275543,
+      "learning_rate": 0.00018328090983425874,
+      "loss": 1.6264,
+      "step": 30300
+    },
+    {
+      "epoch": 0.25170563688149966,
+      "grad_norm": 0.19252796471118927,
+      "learning_rate": 0.00018322571122510913,
+      "loss": 1.6386,
+      "step": 30400
+    },
+    {
+      "epoch": 0.2525336159501888,
+      "grad_norm": 0.22710925340652466,
+      "learning_rate": 0.00018317051261595955,
+      "loss": 1.6098,
+      "step": 30500
+    },
+    {
+      "epoch": 0.25336159501887795,
+      "grad_norm": 0.22989875078201294,
+      "learning_rate": 0.00018311531400680992,
+      "loss": 1.6339,
+      "step": 30600
+    },
+    {
+      "epoch": 0.2541895740875671,
+      "grad_norm": 0.25290095806121826,
+      "learning_rate": 0.0001830601153976603,
+      "loss": 1.634,
+      "step": 30700
+    },
+    {
+      "epoch": 0.2550175531562562,
+      "grad_norm": 0.24599549174308777,
+      "learning_rate": 0.00018300491678851073,
+      "loss": 1.6577,
+      "step": 30800
+    },
+    {
+      "epoch": 0.2558455322249453,
+      "grad_norm": 0.37027794122695923,
+      "learning_rate": 0.0001829497181793611,
+      "loss": 1.644,
+      "step": 30900
+    },
+    {
+      "epoch": 0.25667351129363447,
+      "grad_norm": 0.2629932463169098,
+      "learning_rate": 0.00018289451957021152,
+      "loss": 1.622,
+      "step": 31000
+    },
+    {
+      "epoch": 0.2575014903623236,
+      "grad_norm": 0.33217158913612366,
+      "learning_rate": 0.00018283932096106192,
+      "loss": 1.6334,
+      "step": 31100
+    },
+    {
+      "epoch": 0.25832946943101276,
+      "grad_norm": 0.23134025931358337,
+      "learning_rate": 0.00018278412235191228,
+      "loss": 1.6411,
+      "step": 31200
+    },
+    {
+      "epoch": 0.2591574484997019,
+      "grad_norm": 0.2514587342739105,
+      "learning_rate": 0.0001827289237427627,
+      "loss": 1.6125,
+      "step": 31300
+    },
+    {
+      "epoch": 0.25998542756839105,
+      "grad_norm": 0.2340904325246811,
+      "learning_rate": 0.0001826737251336131,
+      "loss": 1.6567,
+      "step": 31400
+    },
+    {
+      "epoch": 0.2608134066370802,
+      "grad_norm": 0.27328845858573914,
+      "learning_rate": 0.00018261852652446347,
+      "loss": 1.6479,
+      "step": 31500
+    },
+    {
+      "epoch": 0.26164138570576934,
+      "grad_norm": 0.2700946033000946,
+      "learning_rate": 0.0001825633279153139,
+      "loss": 1.6262,
+      "step": 31600
+    },
+    {
+      "epoch": 0.2624693647744585,
+      "grad_norm": 0.25395748019218445,
+      "learning_rate": 0.00018250812930616428,
+      "loss": 1.6166,
+      "step": 31700
+    },
+    {
+      "epoch": 0.2632973438431476,
+      "grad_norm": 0.23837615549564362,
+      "learning_rate": 0.00018245293069701465,
+      "loss": 1.6378,
+      "step": 31800
+    },
+    {
+      "epoch": 0.26412532291183677,
+      "grad_norm": 0.2575034201145172,
+      "learning_rate": 0.00018239773208786507,
+      "loss": 1.6448,
+      "step": 31900
+    },
+    {
+      "epoch": 0.2649533019805259,
+      "grad_norm": 0.25627613067626953,
+      "learning_rate": 0.00018234253347871546,
+      "loss": 1.6674,
+      "step": 32000
+    },
+    {
+      "epoch": 0.26578128104921506,
+      "grad_norm": 0.24171142280101776,
+      "learning_rate": 0.00018228733486956583,
+      "loss": 1.6584,
+      "step": 32100
+    },
+    {
+      "epoch": 0.2666092601179042,
+      "grad_norm": 0.2696898877620697,
+      "learning_rate": 0.00018223213626041625,
+      "loss": 1.6346,
+      "step": 32200
+    },
+    {
+      "epoch": 0.26743723918659335,
+      "grad_norm": 0.2296200692653656,
+      "learning_rate": 0.00018217693765126665,
+      "loss": 1.6376,
+      "step": 32300
+    },
+    {
+      "epoch": 0.2682652182552825,
+      "grad_norm": 0.22668029367923737,
+      "learning_rate": 0.00018212173904211701,
+      "loss": 1.6568,
+      "step": 32400
+    },
+    {
+      "epoch": 0.26909319732397163,
+      "grad_norm": 0.23633837699890137,
+      "learning_rate": 0.00018206654043296744,
+      "loss": 1.6355,
+      "step": 32500
+    },
+    {
+      "epoch": 0.2699211763926608,
+      "grad_norm": 0.3272385597229004,
+      "learning_rate": 0.00018201134182381783,
+      "loss": 1.6333,
+      "step": 32600
+    },
+    {
+      "epoch": 0.2707491554613499,
+      "grad_norm": 0.22284284234046936,
+      "learning_rate": 0.0001819561432146682,
+      "loss": 1.6482,
+      "step": 32700
+    },
+    {
+      "epoch": 0.27157713453003907,
+      "grad_norm": 0.22121630609035492,
+      "learning_rate": 0.00018190094460551862,
+      "loss": 1.6496,
+      "step": 32800
+    },
+    {
+      "epoch": 0.2724051135987282,
+      "grad_norm": 0.2997874319553375,
+      "learning_rate": 0.000181845745996369,
+      "loss": 1.6416,
+      "step": 32900
+    },
+    {
+      "epoch": 0.27323309266741735,
+      "grad_norm": 0.26545023918151855,
+      "learning_rate": 0.00018179054738721938,
+      "loss": 1.6452,
+      "step": 33000
+    },
+    {
+      "epoch": 0.2740610717361065,
+      "grad_norm": 0.2760174870491028,
+      "learning_rate": 0.0001817353487780698,
+      "loss": 1.6617,
+      "step": 33100
+    },
+    {
+      "epoch": 0.27488905080479564,
+      "grad_norm": 0.2267056554555893,
+      "learning_rate": 0.0001816801501689202,
+      "loss": 1.6201,
+      "step": 33200
+    },
+    {
+      "epoch": 0.2757170298734848,
+      "grad_norm": 0.2418147623538971,
+      "learning_rate": 0.00018162495155977056,
+      "loss": 1.6369,
+      "step": 33300
+    },
+    {
+      "epoch": 0.27654500894217393,
+      "grad_norm": 0.2712232172489166,
+      "learning_rate": 0.00018156975295062098,
+      "loss": 1.61,
+      "step": 33400
+    },
+    {
+      "epoch": 0.2773729880108631,
+      "grad_norm": 0.21989206969738007,
+      "learning_rate": 0.00018151455434147138,
+      "loss": 1.6385,
+      "step": 33500
+    },
+    {
+      "epoch": 0.2782009670795522,
+      "grad_norm": 0.23285841941833496,
+      "learning_rate": 0.00018145935573232174,
+      "loss": 1.6234,
+      "step": 33600
+    },
+    {
+      "epoch": 0.27902894614824136,
+      "grad_norm": 0.2723333537578583,
+      "learning_rate": 0.00018140415712317217,
+      "loss": 1.6265,
+      "step": 33700
+    },
+    {
+      "epoch": 0.2798569252169305,
+      "grad_norm": 0.29798418283462524,
+      "learning_rate": 0.00018134895851402256,
+      "loss": 1.6278,
+      "step": 33800
+    },
+    {
+      "epoch": 0.28068490428561965,
+      "grad_norm": 0.22149847447872162,
+      "learning_rate": 0.00018129375990487293,
+      "loss": 1.6347,
+      "step": 33900
+    },
+    {
+      "epoch": 0.2815128833543088,
+      "grad_norm": 0.2652026116847992,
+      "learning_rate": 0.00018123856129572335,
+      "loss": 1.6388,
+      "step": 34000
+    },
+    {
+      "epoch": 0.28234086242299794,
+      "grad_norm": 0.21799111366271973,
+      "learning_rate": 0.00018118336268657374,
+      "loss": 1.6031,
+      "step": 34100
+    },
+    {
+      "epoch": 0.2831688414916871,
+      "grad_norm": 0.23749162256717682,
+      "learning_rate": 0.0001811281640774241,
+      "loss": 1.6268,
+      "step": 34200
+    },
+    {
+      "epoch": 0.28399682056037623,
+      "grad_norm": 0.20693838596343994,
+      "learning_rate": 0.00018107351745436602,
+      "loss": 1.6406,
+      "step": 34300
+    },
+    {
+      "epoch": 0.2848247996290654,
+      "grad_norm": 0.27509090304374695,
+      "learning_rate": 0.00018101831884521642,
+      "loss": 1.6291,
+      "step": 34400
+    },
+    {
+      "epoch": 0.2856527786977545,
+      "grad_norm": 0.24260659515857697,
+      "learning_rate": 0.0001809631202360668,
+      "loss": 1.6418,
+      "step": 34500
+    },
+    {
+      "epoch": 0.28648075776644366,
+      "grad_norm": 0.22240006923675537,
+      "learning_rate": 0.0001809079216269172,
+      "loss": 1.589,
+      "step": 34600
+    },
+    {
+      "epoch": 0.2873087368351328,
+      "grad_norm": 0.21402128040790558,
+      "learning_rate": 0.0001808527230177676,
+      "loss": 1.6004,
+      "step": 34700
+    },
+    {
+      "epoch": 0.28813671590382195,
+      "grad_norm": 0.23821817338466644,
+      "learning_rate": 0.000180797524408618,
+      "loss": 1.6203,
+      "step": 34800
+    },
+    {
+      "epoch": 0.2889646949725111,
+      "grad_norm": 0.22152476012706757,
+      "learning_rate": 0.0001807423257994684,
+      "loss": 1.6351,
+      "step": 34900
+    },
+    {
+      "epoch": 0.28979267404120024,
+      "grad_norm": 0.269619882106781,
+      "learning_rate": 0.00018068712719031878,
+      "loss": 1.6269,
+      "step": 35000
+    },
+    {
+      "epoch": 0.2906206531098894,
+      "grad_norm": 0.2366069257259369,
+      "learning_rate": 0.00018063192858116918,
+      "loss": 1.6363,
+      "step": 35100
+    },
+    {
+      "epoch": 0.2914486321785785,
+      "grad_norm": 0.2468317300081253,
+      "learning_rate": 0.00018057672997201957,
+      "loss": 1.6265,
+      "step": 35200
+    },
+    {
+      "epoch": 0.29227661124726767,
+      "grad_norm": 0.2998073697090149,
+      "learning_rate": 0.00018052153136286997,
+      "loss": 1.6213,
+      "step": 35300
+    },
+    {
+      "epoch": 0.2931045903159568,
+      "grad_norm": 0.2570279836654663,
+      "learning_rate": 0.00018046633275372036,
+      "loss": 1.6229,
+      "step": 35400
+    },
+    {
+      "epoch": 0.29393256938464596,
+      "grad_norm": 0.29977649450302124,
+      "learning_rate": 0.00018041113414457075,
+      "loss": 1.5778,
+      "step": 35500
+    },
+    {
+      "epoch": 0.2947605484533351,
+      "grad_norm": 0.23830650746822357,
+      "learning_rate": 0.00018035593553542115,
+      "loss": 1.643,
+      "step": 35600
+    },
+    {
+      "epoch": 0.29558852752202425,
+      "grad_norm": 0.2635740339756012,
+      "learning_rate": 0.00018030073692627154,
+      "loss": 1.6369,
+      "step": 35700
+    },
+    {
+      "epoch": 0.2964165065907134,
+      "grad_norm": 0.24356907606124878,
+      "learning_rate": 0.00018024553831712194,
+      "loss": 1.6231,
+      "step": 35800
+    },
+    {
+      "epoch": 0.29724448565940254,
+      "grad_norm": 0.19405247271060944,
+      "learning_rate": 0.00018019033970797233,
+      "loss": 1.6504,
+      "step": 35900
+    },
+    {
+      "epoch": 0.2980724647280917,
+      "grad_norm": 0.22447918355464935,
+      "learning_rate": 0.00018013514109882273,
+      "loss": 1.6385,
+      "step": 36000
+    },
+    {
+      "epoch": 0.2989004437967808,
+      "grad_norm": 0.2469175159931183,
+      "learning_rate": 0.00018007994248967312,
+      "loss": 1.6285,
+      "step": 36100
+    },
+    {
+      "epoch": 0.29972842286546997,
+      "grad_norm": 0.2642788290977478,
+      "learning_rate": 0.00018002474388052351,
+      "loss": 1.6119,
+      "step": 36200
+    },
+    {
+      "epoch": 0.3005564019341591,
+      "grad_norm": 0.2765555679798126,
+      "learning_rate": 0.0001799695452713739,
+      "loss": 1.6178,
+      "step": 36300
+    },
+    {
+      "epoch": 0.30138438100284826,
+      "grad_norm": 0.3768673837184906,
+      "learning_rate": 0.0001799143466622243,
+      "loss": 1.6424,
+      "step": 36400
+    },
+    {
+      "epoch": 0.3022123600715374,
+      "grad_norm": 0.26609617471694946,
+      "learning_rate": 0.0001798591480530747,
+      "loss": 1.6599,
+      "step": 36500
+    },
+    {
+      "epoch": 0.30304033914022654,
+      "grad_norm": 0.24030715227127075,
+      "learning_rate": 0.0001798039494439251,
+      "loss": 1.64,
+      "step": 36600
+    },
+    {
+      "epoch": 0.3038683182089157,
+      "grad_norm": 0.24733193218708038,
+      "learning_rate": 0.00017974875083477549,
+      "loss": 1.6097,
+      "step": 36700
+    },
+    {
+      "epoch": 0.30469629727760483,
+      "grad_norm": 0.24855640530586243,
+      "learning_rate": 0.00017969355222562588,
+      "loss": 1.6355,
+      "step": 36800
+    },
+    {
+      "epoch": 0.305524276346294,
+      "grad_norm": 0.2558852434158325,
+      "learning_rate": 0.00017963835361647627,
+      "loss": 1.6395,
+      "step": 36900
+    },
+    {
+      "epoch": 0.3063522554149831,
+      "grad_norm": 0.252812922000885,
+      "learning_rate": 0.00017958315500732667,
+      "loss": 1.6019,
+      "step": 37000
+    },
+    {
+      "epoch": 0.30718023448367227,
+      "grad_norm": 0.2745685279369354,
+      "learning_rate": 0.0001795279563981771,
+      "loss": 1.6438,
+      "step": 37100
+    },
+    {
+      "epoch": 0.3080082135523614,
+      "grad_norm": 0.24301989376544952,
+      "learning_rate": 0.00017947330977511895,
+      "loss": 1.5943,
+      "step": 37200
+    },
+    {
+      "epoch": 0.30883619262105055,
+      "grad_norm": 0.25345346331596375,
+      "learning_rate": 0.00017941811116596937,
+      "loss": 1.6255,
+      "step": 37300
+    },
+    {
+      "epoch": 0.3096641716897397,
+      "grad_norm": 0.2757332921028137,
+      "learning_rate": 0.00017936291255681974,
+      "loss": 1.622,
+      "step": 37400
+    },
+    {
+      "epoch": 0.31049215075842884,
+      "grad_norm": 0.2548620402812958,
+      "learning_rate": 0.00017930771394767013,
+      "loss": 1.6315,
+      "step": 37500
+    },
+    {
+      "epoch": 0.311320129827118,
+      "grad_norm": 0.2474357783794403,
+      "learning_rate": 0.00017925251533852055,
+      "loss": 1.6391,
+      "step": 37600
+    },
+    {
+      "epoch": 0.31214810889580713,
+      "grad_norm": 0.23792117834091187,
+      "learning_rate": 0.00017919731672937092,
+      "loss": 1.6349,
+      "step": 37700
+    },
+    {
+      "epoch": 0.3129760879644963,
+      "grad_norm": 0.22419889271259308,
+      "learning_rate": 0.00017914211812022131,
+      "loss": 1.6439,
+      "step": 37800
+    },
+    {
+      "epoch": 0.3138040670331854,
+      "grad_norm": 0.2642965316772461,
+      "learning_rate": 0.00017908691951107174,
+      "loss": 1.6347,
+      "step": 37900
+    },
+    {
+      "epoch": 0.31463204610187456,
+      "grad_norm": 0.2979467511177063,
+      "learning_rate": 0.0001790317209019221,
+      "loss": 1.6478,
+      "step": 38000
+    },
+    {
+      "epoch": 0.3154600251705637,
+      "grad_norm": 0.33403944969177246,
+      "learning_rate": 0.0001789765222927725,
+      "loss": 1.5852,
+      "step": 38100
+    },
+    {
+      "epoch": 0.31628800423925285,
+      "grad_norm": 0.2425394505262375,
+      "learning_rate": 0.00017892132368362292,
+      "loss": 1.596,
+      "step": 38200
+    },
+    {
+      "epoch": 0.317115983307942,
+      "grad_norm": 0.24871741235256195,
+      "learning_rate": 0.00017886612507447328,
+      "loss": 1.6358,
+      "step": 38300
+    },
+    {
+      "epoch": 0.31794396237663114,
+      "grad_norm": 0.2535829246044159,
+      "learning_rate": 0.00017881092646532368,
+      "loss": 1.6175,
+      "step": 38400
+    },
+    {
+      "epoch": 0.3187719414453203,
+      "grad_norm": 0.3197729289531708,
+      "learning_rate": 0.0001787557278561741,
+      "loss": 1.6211,
+      "step": 38500
+    },
+    {
+      "epoch": 0.31959992051400943,
+      "grad_norm": 0.23783159255981445,
+      "learning_rate": 0.00017870052924702447,
+      "loss": 1.6677,
+      "step": 38600
+    },
+    {
+      "epoch": 0.32042789958269857,
+      "grad_norm": 0.2013082504272461,
+      "learning_rate": 0.00017864533063787486,
+      "loss": 1.5916,
+      "step": 38700
+    },
+    {
+      "epoch": 0.3212558786513877,
+      "grad_norm": 0.2517942190170288,
+      "learning_rate": 0.00017859013202872528,
+      "loss": 1.61,
+      "step": 38800
+    },
+    {
+      "epoch": 0.32208385772007686,
+      "grad_norm": 0.22963842749595642,
+      "learning_rate": 0.00017853493341957565,
+      "loss": 1.5888,
+      "step": 38900
+    },
+    {
+      "epoch": 0.322911836788766,
+      "grad_norm": 0.27445054054260254,
+      "learning_rate": 0.00017847973481042604,
+      "loss": 1.6021,
+      "step": 39000
+    },
+    {
+      "epoch": 0.32373981585745515,
+      "grad_norm": 0.21975164115428925,
+      "learning_rate": 0.00017842453620127647,
+      "loss": 1.6261,
+      "step": 39100
+    },
+    {
+      "epoch": 0.3245677949261443,
+      "grad_norm": 0.27244439721107483,
+      "learning_rate": 0.00017836988957821835,
+      "loss": 1.6478,
+      "step": 39200
+    },
+    {
+      "epoch": 0.32539577399483344,
+      "grad_norm": 0.2596853971481323,
+      "learning_rate": 0.00017831469096906875,
+      "loss": 1.5833,
+      "step": 39300
+    },
+    {
+      "epoch": 0.3262237530635226,
+      "grad_norm": 0.22705727815628052,
+      "learning_rate": 0.0001782594923599191,
+      "loss": 1.64,
+      "step": 39400
+    },
+    {
+      "epoch": 0.32705173213221167,
+      "grad_norm": 0.24524278938770294,
+      "learning_rate": 0.00017820429375076953,
+      "loss": 1.6177,
+      "step": 39500
+    },
+    {
+      "epoch": 0.3278797112009008,
+      "grad_norm": 0.2465396374464035,
+      "learning_rate": 0.00017814909514161993,
+      "loss": 1.6241,
+      "step": 39600
+    },
+    {
+      "epoch": 0.32870769026958996,
+      "grad_norm": 0.24681350588798523,
+      "learning_rate": 0.0001780938965324703,
+      "loss": 1.6366,
+      "step": 39700
+    },
+    {
+      "epoch": 0.3295356693382791,
+      "grad_norm": 0.2725008726119995,
+      "learning_rate": 0.00017803869792332072,
+      "loss": 1.6168,
+      "step": 39800
+    },
+    {
+      "epoch": 0.33036364840696825,
+      "grad_norm": 0.24282367527484894,
+      "learning_rate": 0.0001779834993141711,
+      "loss": 1.6427,
+      "step": 39900
+    },
+    {
+      "epoch": 0.3311916274756574,
+      "grad_norm": 0.22297324240207672,
+      "learning_rate": 0.00017792830070502148,
+      "loss": 1.6094,
+      "step": 40000
+    },
+    {
+      "epoch": 0.33201960654434653,
+      "grad_norm": 0.23385342955589294,
+      "learning_rate": 0.0001778736540819634,
+      "loss": 1.6268,
+      "step": 40100
+    },
+    {
+      "epoch": 0.3328475856130357,
+      "grad_norm": 0.24977736175060272,
+      "learning_rate": 0.00017781845547281376,
+      "loss": 1.6237,
+      "step": 40200
+    },
+    {
+      "epoch": 0.3336755646817248,
+      "grad_norm": 0.24268653988838196,
+      "learning_rate": 0.00017776325686366418,
+      "loss": 1.6471,
+      "step": 40300
+    },
+    {
+      "epoch": 0.33450354375041397,
+      "grad_norm": 0.1996404081583023,
+      "learning_rate": 0.00017770805825451457,
+      "loss": 1.6132,
+      "step": 40400
+    },
+    {
+      "epoch": 0.3353315228191031,
+      "grad_norm": 0.26735562086105347,
+      "learning_rate": 0.00017765285964536494,
+      "loss": 1.6451,
+      "step": 40500
+    },
+    {
+      "epoch": 0.33615950188779226,
+      "grad_norm": 0.24058261513710022,
+      "learning_rate": 0.00017759766103621536,
+      "loss": 1.6199,
+      "step": 40600
+    },
+    {
+      "epoch": 0.3369874809564814,
+      "grad_norm": 0.23054581880569458,
+      "learning_rate": 0.00017754246242706576,
+      "loss": 1.6227,
+      "step": 40700
+    },
+    {
+      "epoch": 0.33781546002517054,
+      "grad_norm": 0.245001882314682,
+      "learning_rate": 0.00017748726381791615,
+      "loss": 1.6422,
+      "step": 40800
+    },
+    {
+      "epoch": 0.3386434390938597,
+      "grad_norm": 0.40328505635261536,
+      "learning_rate": 0.00017743206520876655,
+      "loss": 1.6102,
+      "step": 40900
+    },
+    {
+      "epoch": 0.33947141816254883,
+      "grad_norm": 0.21955619752407074,
+      "learning_rate": 0.00017737686659961694,
+      "loss": 1.6439,
+      "step": 41000
+    },
+    {
+      "epoch": 0.340299397231238,
+      "grad_norm": 0.2642820477485657,
+      "learning_rate": 0.00017732166799046733,
+      "loss": 1.6576,
+      "step": 41100
+    },
+    {
+      "epoch": 0.3411273762999271,
+      "grad_norm": 0.22586101293563843,
+      "learning_rate": 0.00017726646938131773,
+      "loss": 1.6331,
+      "step": 41200
+    },
+    {
+      "epoch": 0.34195535536861627,
+      "grad_norm": 0.22351595759391785,
+      "learning_rate": 0.00017721127077216812,
+      "loss": 1.6218,
+      "step": 41300
+    },
+    {
+      "epoch": 0.3427833344373054,
+      "grad_norm": 0.5699514150619507,
+      "learning_rate": 0.00017715607216301852,
+      "loss": 1.6398,
+      "step": 41400
+    },
+    {
+      "epoch": 0.34361131350599455,
+      "grad_norm": 0.3038058280944824,
+      "learning_rate": 0.0001771008735538689,
+      "loss": 1.6364,
+      "step": 41500
+    },
+    {
+      "epoch": 0.3444392925746837,
+      "grad_norm": 0.25284209847450256,
+      "learning_rate": 0.0001770456749447193,
+      "loss": 1.6345,
+      "step": 41600
+    },
+    {
+      "epoch": 0.34526727164337284,
+      "grad_norm": 0.5244471430778503,
+      "learning_rate": 0.0001769904763355697,
+      "loss": 1.6348,
+      "step": 41700
+    },
+    {
+      "epoch": 0.346095250712062,
+      "grad_norm": 0.240703284740448,
+      "learning_rate": 0.0001769352777264201,
+      "loss": 1.6312,
+      "step": 41800
+    },
+    {
+      "epoch": 0.34692322978075113,
+      "grad_norm": 0.23771056532859802,
+      "learning_rate": 0.0001768800791172705,
+      "loss": 1.6526,
+      "step": 41900
+    },
+    {
+      "epoch": 0.3477512088494403,
+      "grad_norm": 0.25365206599235535,
+      "learning_rate": 0.00017682488050812088,
+      "loss": 1.6376,
+      "step": 42000
+    },
+    {
+      "epoch": 0.3485791879181294,
+      "grad_norm": 0.27795979380607605,
+      "learning_rate": 0.00017676968189897128,
+      "loss": 1.6621,
+      "step": 42100
+    },
+    {
+      "epoch": 0.34940716698681856,
+      "grad_norm": 0.24921217560768127,
+      "learning_rate": 0.00017671448328982167,
+      "loss": 1.622,
+      "step": 42200
+    },
+    {
+      "epoch": 0.3502351460555077,
+      "grad_norm": 0.24285615980625153,
+      "learning_rate": 0.00017665928468067207,
+      "loss": 1.6524,
+      "step": 42300
+    },
+    {
+      "epoch": 0.35106312512419685,
+      "grad_norm": 0.23597076535224915,
+      "learning_rate": 0.00017660408607152246,
+      "loss": 1.6257,
+      "step": 42400
+    },
+    {
+      "epoch": 0.351891104192886,
+      "grad_norm": 0.2752334177494049,
+      "learning_rate": 0.00017654888746237285,
+      "loss": 1.6223,
+      "step": 42500
+    },
+    {
+      "epoch": 0.35271908326157514,
+      "grad_norm": 0.24339468777179718,
+      "learning_rate": 0.00017649424083931474,
+      "loss": 1.6307,
+      "step": 42600
+    },
+    {
+      "epoch": 0.3535470623302643,
+      "grad_norm": 0.22084777057170868,
+      "learning_rate": 0.00017643904223016513,
+      "loss": 1.6245,
+      "step": 42700
+    },
+    {
+      "epoch": 0.3543750413989534,
+      "grad_norm": 0.2476748824119568,
+      "learning_rate": 0.00017638384362101556,
+      "loss": 1.6262,
+      "step": 42800
+    },
+    {
+      "epoch": 0.35520302046764257,
+      "grad_norm": 0.2630603611469269,
+      "learning_rate": 0.00017632864501186592,
+      "loss": 1.6396,
+      "step": 42900
+    },
+    {
+      "epoch": 0.3560309995363317,
+      "grad_norm": 0.27675458788871765,
+      "learning_rate": 0.00017627344640271632,
+      "loss": 1.6497,
+      "step": 43000
+    },
+    {
+      "epoch": 0.35685897860502086,
+      "grad_norm": 0.23261144757270813,
+      "learning_rate": 0.00017621824779356674,
+      "loss": 1.6097,
+      "step": 43100
+    },
+    {
+      "epoch": 0.35768695767371,
+      "grad_norm": 0.24998947978019714,
+      "learning_rate": 0.0001761630491844171,
+      "loss": 1.6306,
+      "step": 43200
+    },
+    {
+      "epoch": 0.35851493674239915,
+      "grad_norm": 0.2858506441116333,
+      "learning_rate": 0.0001761078505752675,
+      "loss": 1.6167,
+      "step": 43300
+    },
+    {
+      "epoch": 0.3593429158110883,
+      "grad_norm": 0.24741750955581665,
+      "learning_rate": 0.00017605265196611792,
+      "loss": 1.6434,
+      "step": 43400
+    },
+    {
+      "epoch": 0.36017089487977744,
+      "grad_norm": 0.2413889467716217,
+      "learning_rate": 0.0001759974533569683,
+      "loss": 1.6175,
+      "step": 43500
+    },
+    {
+      "epoch": 0.3609988739484666,
+      "grad_norm": 0.23276059329509735,
+      "learning_rate": 0.00017594225474781868,
+      "loss": 1.6506,
+      "step": 43600
+    },
+    {
+      "epoch": 0.3618268530171557,
+      "grad_norm": 0.7301183938980103,
+      "learning_rate": 0.0001758870561386691,
+      "loss": 1.6602,
+      "step": 43700
+    },
+    {
+      "epoch": 0.36265483208584487,
+      "grad_norm": 0.26085367798805237,
+      "learning_rate": 0.00017583185752951947,
+      "loss": 1.6361,
+      "step": 43800
+    },
+    {
+      "epoch": 0.363482811154534,
+      "grad_norm": 0.2507553994655609,
+      "learning_rate": 0.00017577665892036987,
+      "loss": 1.6207,
+      "step": 43900
+    },
+    {
+      "epoch": 0.36431079022322316,
+      "grad_norm": 0.2663898766040802,
+      "learning_rate": 0.00017572146031122029,
+      "loss": 1.6443,
+      "step": 44000
+    },
+    {
+      "epoch": 0.3651387692919123,
+      "grad_norm": 0.34235909581184387,
+      "learning_rate": 0.00017566626170207065,
+      "loss": 1.668,
+      "step": 44100
+    },
+    {
+      "epoch": 0.36596674836060145,
+      "grad_norm": 0.26196587085723877,
+      "learning_rate": 0.00017561106309292105,
+      "loss": 1.6544,
+      "step": 44200
+    },
+    {
+      "epoch": 0.3667947274292906,
+      "grad_norm": 0.23575757443904877,
+      "learning_rate": 0.00017555586448377147,
+      "loss": 1.6252,
+      "step": 44300
+    },
+    {
+      "epoch": 0.36762270649797973,
+      "grad_norm": 0.2926133871078491,
+      "learning_rate": 0.00017550066587462184,
+      "loss": 1.6282,
+      "step": 44400
+    },
+    {
+      "epoch": 0.3684506855666689,
+      "grad_norm": 0.21387304365634918,
+      "learning_rate": 0.00017544546726547223,
+      "loss": 1.6542,
+      "step": 44500
+    },
+    {
+      "epoch": 0.369278664635358,
+      "grad_norm": 0.24483337998390198,
+      "learning_rate": 0.00017539026865632265,
+      "loss": 1.626,
+      "step": 44600
+    },
+    {
+      "epoch": 0.37010664370404717,
+      "grad_norm": 0.3214597702026367,
+      "learning_rate": 0.00017533507004717302,
+      "loss": 1.5933,
+      "step": 44700
+    },
+    {
+      "epoch": 0.3709346227727363,
+      "grad_norm": 0.23716263473033905,
+      "learning_rate": 0.0001752798714380234,
+      "loss": 1.6189,
+      "step": 44800
+    },
+    {
+      "epoch": 0.37176260184142546,
+      "grad_norm": 0.22332042455673218,
+      "learning_rate": 0.00017522467282887383,
+      "loss": 1.6488,
+      "step": 44900
+    },
+    {
+      "epoch": 0.3725905809101146,
+      "grad_norm": 0.2706310749053955,
+      "learning_rate": 0.0001751694742197242,
+      "loss": 1.6269,
+      "step": 45000
+    },
+    {
+      "epoch": 0.37341855997880374,
+      "grad_norm": 0.2572775185108185,
+      "learning_rate": 0.0001751142756105746,
+      "loss": 1.6432,
+      "step": 45100
+    },
+    {
+      "epoch": 0.3742465390474929,
+      "grad_norm": 0.23288114368915558,
+      "learning_rate": 0.00017505907700142502,
+      "loss": 1.6531,
+      "step": 45200
+    },
+    {
+      "epoch": 0.37507451811618203,
+      "grad_norm": 0.2673160433769226,
+      "learning_rate": 0.00017500387839227538,
+      "loss": 1.6047,
+      "step": 45300
+    },
+    {
+      "epoch": 0.3759024971848712,
+      "grad_norm": 0.29170718789100647,
+      "learning_rate": 0.00017494867978312578,
+      "loss": 1.6119,
+      "step": 45400
+    },
+    {
+      "epoch": 0.3767304762535603,
+      "grad_norm": 0.2626097500324249,
+      "learning_rate": 0.0001748934811739762,
+      "loss": 1.6329,
+      "step": 45500
+    },
+    {
+      "epoch": 0.37755845532224946,
+      "grad_norm": 0.22249017655849457,
+      "learning_rate": 0.00017483828256482657,
+      "loss": 1.6284,
+      "step": 45600
+    },
+    {
+      "epoch": 0.3783864343909386,
+      "grad_norm": 0.36833006143569946,
+      "learning_rate": 0.00017478308395567696,
+      "loss": 1.6079,
+      "step": 45700
+    },
+    {
+      "epoch": 0.37921441345962775,
+      "grad_norm": 0.23582060635089874,
+      "learning_rate": 0.00017472788534652738,
+      "loss": 1.6243,
+      "step": 45800
+    },
+    {
+      "epoch": 0.3800423925283169,
+      "grad_norm": 0.20202980935573578,
+      "learning_rate": 0.00017467268673737775,
+      "loss": 1.6312,
+      "step": 45900
+    },
+    {
+      "epoch": 0.38087037159700604,
+      "grad_norm": 0.3692006468772888,
+      "learning_rate": 0.00017461748812822814,
+      "loss": 1.6048,
+      "step": 46000
+    },
+    {
+      "epoch": 0.3816983506656952,
+      "grad_norm": 0.27726104855537415,
+      "learning_rate": 0.00017456228951907857,
+      "loss": 1.6396,
+      "step": 46100
+    },
+    {
+      "epoch": 0.38252632973438433,
+      "grad_norm": 0.2585281431674957,
+      "learning_rate": 0.00017450709090992893,
+      "loss": 1.6185,
+      "step": 46200
+    },
+    {
+      "epoch": 0.3833543088030735,
+      "grad_norm": 0.24427704513072968,
+      "learning_rate": 0.00017445189230077935,
+      "loss": 1.5996,
+      "step": 46300
+    },
+    {
+      "epoch": 0.3841822878717626,
+      "grad_norm": 0.2784598469734192,
+      "learning_rate": 0.00017439669369162975,
+      "loss": 1.6012,
+      "step": 46400
+    },
+    {
+      "epoch": 0.38501026694045176,
+      "grad_norm": 0.25094759464263916,
+      "learning_rate": 0.00017434149508248012,
+      "loss": 1.6043,
+      "step": 46500
+    },
+    {
+      "epoch": 0.3858382460091409,
+      "grad_norm": 0.2598397731781006,
+      "learning_rate": 0.00017428629647333054,
+      "loss": 1.6278,
+      "step": 46600
+    },
+    {
+      "epoch": 0.38666622507783005,
+      "grad_norm": 0.22646105289459229,
+      "learning_rate": 0.00017423109786418093,
+      "loss": 1.6297,
+      "step": 46700
+    },
+    {
+      "epoch": 0.3874942041465192,
+      "grad_norm": 0.29078444838523865,
+      "learning_rate": 0.0001741758992550313,
+      "loss": 1.6509,
+      "step": 46800
+    },
+    {
+      "epoch": 0.38832218321520834,
+      "grad_norm": 0.2686766982078552,
+      "learning_rate": 0.00017412070064588172,
+      "loss": 1.6145,
+      "step": 46900
+    },
+    {
+      "epoch": 0.3891501622838975,
+      "grad_norm": 0.2642682194709778,
+      "learning_rate": 0.0001740655020367321,
+      "loss": 1.6421,
+      "step": 47000
+    },
+    {
+      "epoch": 0.3899781413525866,
+      "grad_norm": 0.2528562545776367,
+      "learning_rate": 0.000174010855413674,
+      "loss": 1.6607,
+      "step": 47100
+    },
+    {
+      "epoch": 0.39080612042127577,
+      "grad_norm": 0.2299865484237671,
+      "learning_rate": 0.00017395620879061589,
+      "loss": 1.6112,
+      "step": 47200
+    },
+    {
+      "epoch": 0.3916340994899649,
+      "grad_norm": 0.24297955632209778,
+      "learning_rate": 0.00017390101018146628,
+      "loss": 1.6081,
+      "step": 47300
+    },
+    {
+      "epoch": 0.39246207855865406,
+      "grad_norm": 0.243024542927742,
+      "learning_rate": 0.00017384581157231667,
+      "loss": 1.6703,
+      "step": 47400
+    },
+    {
+      "epoch": 0.3932900576273432,
+      "grad_norm": 0.24055872857570648,
+      "learning_rate": 0.00017379061296316707,
+      "loss": 1.5895,
+      "step": 47500
+    },
+    {
+      "epoch": 0.39411803669603235,
+      "grad_norm": 0.2664213478565216,
+      "learning_rate": 0.00017373541435401746,
+      "loss": 1.6173,
+      "step": 47600
+    },
+    {
+      "epoch": 0.3949460157647215,
+      "grad_norm": 0.23474732041358948,
+      "learning_rate": 0.00017368021574486786,
+      "loss": 1.6347,
+      "step": 47700
+    },
+    {
+      "epoch": 0.39577399483341064,
+      "grad_norm": 0.24745678901672363,
+      "learning_rate": 0.00017362501713571825,
+      "loss": 1.6112,
+      "step": 47800
+    },
+    {
+      "epoch": 0.3966019739020998,
+      "grad_norm": 0.2576342523097992,
+      "learning_rate": 0.00017356981852656865,
+      "loss": 1.6181,
+      "step": 47900
+    },
+    {
+      "epoch": 0.3974299529707889,
+      "grad_norm": 0.27308520674705505,
+      "learning_rate": 0.00017351461991741904,
+      "loss": 1.6037,
+      "step": 48000
+    },
+    {
+      "epoch": 0.39825793203947807,
+      "grad_norm": 0.2533150315284729,
+      "learning_rate": 0.00017345942130826943,
+      "loss": 1.6561,
+      "step": 48100
+    },
+    {
+      "epoch": 0.3990859111081672,
+      "grad_norm": 0.24679023027420044,
+      "learning_rate": 0.00017340422269911983,
+      "loss": 1.6129,
+      "step": 48200
+    },
+    {
+      "epoch": 0.3999138901768563,
+      "grad_norm": 0.28216132521629333,
+      "learning_rate": 0.00017334902408997022,
+      "loss": 1.6374,
+      "step": 48300
+    },
+    {
+      "epoch": 0.40074186924554545,
+      "grad_norm": 0.26568201184272766,
+      "learning_rate": 0.00017329382548082062,
+      "loss": 1.6395,
+      "step": 48400
+    },
+    {
+      "epoch": 0.4015698483142346,
+      "grad_norm": 0.2306041121482849,
+      "learning_rate": 0.000173238626871671,
+      "loss": 1.6055,
+      "step": 48500
+    },
+    {
+      "epoch": 0.40239782738292373,
+      "grad_norm": 0.38825201988220215,
+      "learning_rate": 0.0001731834282625214,
+      "loss": 1.6479,
+      "step": 48600
+    },
+    {
+      "epoch": 0.4032258064516129,
+      "grad_norm": 0.2115429788827896,
+      "learning_rate": 0.0001731282296533718,
+      "loss": 1.616,
+      "step": 48700
+    },
+    {
+      "epoch": 0.404053785520302,
+      "grad_norm": 0.2399633675813675,
+      "learning_rate": 0.0001730730310442222,
+      "loss": 1.644,
+      "step": 48800
+    },
+    {
+      "epoch": 0.40488176458899117,
+      "grad_norm": 0.2738541066646576,
+      "learning_rate": 0.0001730178324350726,
+      "loss": 1.6346,
+      "step": 48900
+    },
+    {
+      "epoch": 0.4057097436576803,
+      "grad_norm": 0.24475158751010895,
+      "learning_rate": 0.00017296263382592298,
+      "loss": 1.6143,
+      "step": 49000
+    },
+    {
+      "epoch": 0.40653772272636945,
+      "grad_norm": 0.2846825122833252,
+      "learning_rate": 0.00017290743521677338,
+      "loss": 1.627,
+      "step": 49100
+    },
+    {
+      "epoch": 0.4073657017950586,
+      "grad_norm": 0.24657665193080902,
+      "learning_rate": 0.00017285223660762377,
+      "loss": 1.6223,
+      "step": 49200
+    },
+    {
+      "epoch": 0.40819368086374774,
+      "grad_norm": 0.24977359175682068,
+      "learning_rate": 0.00017279703799847416,
+      "loss": 1.6454,
+      "step": 49300
+    },
+    {
+      "epoch": 0.4090216599324369,
+      "grad_norm": 0.24544388055801392,
+      "learning_rate": 0.00017274183938932456,
+      "loss": 1.6557,
+      "step": 49400
+    },
+    {
+      "epoch": 0.40984963900112603,
+      "grad_norm": 0.23884369432926178,
+      "learning_rate": 0.00017268664078017495,
+      "loss": 1.6128,
+      "step": 49500
+    },
+    {
+      "epoch": 0.4106776180698152,
+      "grad_norm": 0.36807382106781006,
+      "learning_rate": 0.00017263144217102535,
+      "loss": 1.6237,
+      "step": 49600
+    },
+    {
+      "epoch": 0.4115055971385043,
+      "grad_norm": 0.291007936000824,
+      "learning_rate": 0.00017257624356187574,
+      "loss": 1.6439,
+      "step": 49700
+    },
+    {
+      "epoch": 0.41233357620719346,
+      "grad_norm": 0.22296257317066193,
+      "learning_rate": 0.00017252104495272614,
+      "loss": 1.636,
+      "step": 49800
+    },
+    {
+      "epoch": 0.4131615552758826,
+      "grad_norm": 0.281239777803421,
+      "learning_rate": 0.00017246584634357653,
+      "loss": 1.6008,
+      "step": 49900
+    },
+    {
+      "epoch": 0.41398953434457175,
+      "grad_norm": 0.30486106872558594,
+      "learning_rate": 0.00017241064773442692,
+      "loss": 1.5891,
+      "step": 50000
+    },
+    {
+      "epoch": 0.4148175134132609,
+      "grad_norm": 0.24123990535736084,
+      "learning_rate": 0.00017235544912527732,
+      "loss": 1.6299,
+      "step": 50100
+    },
+    {
+      "epoch": 0.41564549248195004,
+      "grad_norm": 0.3727368116378784,
+      "learning_rate": 0.0001723002505161277,
+      "loss": 1.6166,
+      "step": 50200
+    },
+    {
+      "epoch": 0.4164734715506392,
+      "grad_norm": 0.24565471708774567,
+      "learning_rate": 0.0001722450519069781,
+      "loss": 1.6129,
+      "step": 50300
+    },
+    {
+      "epoch": 0.41730145061932833,
+      "grad_norm": 0.2552422285079956,
+      "learning_rate": 0.00017219040528392002,
+      "loss": 1.6117,
+      "step": 50400
+    },
+    {
+      "epoch": 0.4181294296880175,
+      "grad_norm": 0.21545369923114777,
+      "learning_rate": 0.0001721352066747704,
+      "loss": 1.6433,
+      "step": 50500
+    },
+    {
+      "epoch": 0.4189574087567066,
+      "grad_norm": 0.23316791653633118,
+      "learning_rate": 0.00017208000806562078,
+      "loss": 1.6806,
+      "step": 50600
+    },
+    {
+      "epoch": 0.41978538782539576,
+      "grad_norm": 0.2567419409751892,
+      "learning_rate": 0.0001720248094564712,
+      "loss": 1.6253,
+      "step": 50700
+    },
+    {
+      "epoch": 0.4206133668940849,
+      "grad_norm": 0.31419897079467773,
+      "learning_rate": 0.00017196961084732157,
+      "loss": 1.611,
+      "step": 50800
+    },
+    {
+      "epoch": 0.42144134596277405,
+      "grad_norm": 0.26649272441864014,
+      "learning_rate": 0.00017191441223817196,
+      "loss": 1.6269,
+      "step": 50900
+    },
+    {
+      "epoch": 0.4222693250314632,
+      "grad_norm": 0.24049483239650726,
+      "learning_rate": 0.00017185921362902239,
+      "loss": 1.6257,
+      "step": 51000
+    },
+    {
+      "epoch": 0.42309730410015234,
+      "grad_norm": 0.22467993199825287,
+      "learning_rate": 0.00017180401501987275,
+      "loss": 1.5821,
+      "step": 51100
+    },
+    {
+      "epoch": 0.4239252831688415,
+      "grad_norm": 0.27348917722702026,
+      "learning_rate": 0.00017174881641072315,
+      "loss": 1.6243,
+      "step": 51200
+    },
+    {
+      "epoch": 0.4247532622375306,
+      "grad_norm": 0.23247596621513367,
+      "learning_rate": 0.00017169361780157357,
+      "loss": 1.6333,
+      "step": 51300
+    },
+    {
+      "epoch": 0.42558124130621977,
+      "grad_norm": 0.25708910822868347,
+      "learning_rate": 0.00017163841919242394,
+      "loss": 1.6168,
+      "step": 51400
+    },
+    {
+      "epoch": 0.4264092203749089,
+      "grad_norm": 0.2495090365409851,
+      "learning_rate": 0.00017158322058327433,
+      "loss": 1.6105,
+      "step": 51500
+    },
+    {
+      "epoch": 0.42723719944359806,
+      "grad_norm": 0.2512340545654297,
+      "learning_rate": 0.00017152802197412475,
+      "loss": 1.631,
+      "step": 51600
+    },
+    {
+      "epoch": 0.4280651785122872,
+      "grad_norm": 0.27026236057281494,
+      "learning_rate": 0.00017147282336497512,
+      "loss": 1.5874,
+      "step": 51700
+    },
+    {
+      "epoch": 0.42889315758097635,
+      "grad_norm": 0.23251605033874512,
+      "learning_rate": 0.0001714176247558255,
+      "loss": 1.6383,
+      "step": 51800
+    },
+    {
+      "epoch": 0.4297211366496655,
+      "grad_norm": 0.3068407475948334,
+      "learning_rate": 0.00017136242614667593,
+      "loss": 1.607,
+      "step": 51900
+    },
+    {
+      "epoch": 0.43054911571835464,
+      "grad_norm": 0.22653043270111084,
+      "learning_rate": 0.0001713072275375263,
+      "loss": 1.5913,
+      "step": 52000
+    },
+    {
+      "epoch": 0.4313770947870438,
+      "grad_norm": 0.22893564403057098,
+      "learning_rate": 0.0001712520289283767,
+      "loss": 1.6218,
+      "step": 52100
+    },
+    {
+      "epoch": 0.4322050738557329,
+      "grad_norm": 0.23899468779563904,
+      "learning_rate": 0.00017119683031922712,
+      "loss": 1.6178,
+      "step": 52200
+    },
+    {
+      "epoch": 0.43303305292442207,
+      "grad_norm": 0.26594260334968567,
+      "learning_rate": 0.00017114163171007748,
+      "loss": 1.6118,
+      "step": 52300
+    },
+    {
+      "epoch": 0.4338610319931112,
+      "grad_norm": 0.29181089997291565,
+      "learning_rate": 0.00017108643310092788,
+      "loss": 1.6391,
+      "step": 52400
+    },
+    {
+      "epoch": 0.43468901106180036,
+      "grad_norm": 0.35307735204696655,
+      "learning_rate": 0.0001710312344917783,
+      "loss": 1.601,
+      "step": 52500
+    },
+    {
+      "epoch": 0.4355169901304895,
+      "grad_norm": 0.25828322768211365,
+      "learning_rate": 0.00017097603588262867,
+      "loss": 1.6421,
+      "step": 52600
+    },
+    {
+      "epoch": 0.43634496919917864,
+      "grad_norm": 0.30803248286247253,
+      "learning_rate": 0.00017092083727347906,
+      "loss": 1.655,
+      "step": 52700
+    },
+    {
+      "epoch": 0.4371729482678678,
+      "grad_norm": 0.23458120226860046,
+      "learning_rate": 0.00017086563866432948,
+      "loss": 1.6099,
+      "step": 52800
+    },
+    {
+      "epoch": 0.43800092733655693,
+      "grad_norm": 0.23281985521316528,
+      "learning_rate": 0.00017081044005517985,
+      "loss": 1.6388,
+      "step": 52900
+    },
+    {
+      "epoch": 0.4388289064052461,
+      "grad_norm": 0.23414187133312225,
+      "learning_rate": 0.00017075524144603024,
+      "loss": 1.5956,
+      "step": 53000
+    },
+    {
+      "epoch": 0.4396568854739352,
+      "grad_norm": 0.25793951749801636,
+      "learning_rate": 0.00017070004283688066,
+      "loss": 1.6734,
+      "step": 53100
+    },
+    {
+      "epoch": 0.44048486454262437,
+      "grad_norm": 0.23932301998138428,
+      "learning_rate": 0.00017064484422773103,
+      "loss": 1.6106,
+      "step": 53200
+    },
+    {
+      "epoch": 0.4413128436113135,
+      "grad_norm": 0.24835747480392456,
+      "learning_rate": 0.00017058964561858143,
+      "loss": 1.6077,
+      "step": 53300
+    },
+    {
+      "epoch": 0.44214082268000265,
+      "grad_norm": 0.26732298731803894,
+      "learning_rate": 0.00017053444700943185,
+      "loss": 1.5729,
+      "step": 53400
+    },
+    {
+      "epoch": 0.4429688017486918,
+      "grad_norm": 0.2826359272003174,
+      "learning_rate": 0.00017047924840028221,
+      "loss": 1.6623,
+      "step": 53500
+    },
+    {
+      "epoch": 0.44379678081738094,
+      "grad_norm": 0.24312160909175873,
+      "learning_rate": 0.00017042404979113264,
+      "loss": 1.6075,
+      "step": 53600
+    },
+    {
+      "epoch": 0.4446247598860701,
+      "grad_norm": 0.293720006942749,
+      "learning_rate": 0.00017036885118198303,
+      "loss": 1.6293,
+      "step": 53700
+    },
+    {
+      "epoch": 0.44545273895475923,
+      "grad_norm": 0.2534725069999695,
+      "learning_rate": 0.0001703136525728334,
+      "loss": 1.5869,
+      "step": 53800
+    },
+    {
+      "epoch": 0.4462807180234484,
+      "grad_norm": 0.25833797454833984,
+      "learning_rate": 0.00017025845396368382,
+      "loss": 1.6359,
+      "step": 53900
+    },
+    {
+      "epoch": 0.4471086970921375,
+      "grad_norm": 0.2595597207546234,
+      "learning_rate": 0.0001702032553545342,
+      "loss": 1.598,
+      "step": 54000
+    },
+    {
+      "epoch": 0.44793667616082666,
+      "grad_norm": 0.2243652492761612,
+      "learning_rate": 0.00017014805674538458,
+      "loss": 1.6555,
+      "step": 54100
+    },
+    {
+      "epoch": 0.4487646552295158,
+      "grad_norm": 0.2626785337924957,
+      "learning_rate": 0.000170092858136235,
+      "loss": 1.6327,
+      "step": 54200
+    },
+    {
+      "epoch": 0.44959263429820495,
+      "grad_norm": 0.28202512860298157,
+      "learning_rate": 0.0001700376595270854,
+      "loss": 1.5877,
+      "step": 54300
+    },
+    {
+      "epoch": 0.4504206133668941,
+      "grad_norm": 0.23521679639816284,
+      "learning_rate": 0.00016998246091793576,
+      "loss": 1.6357,
+      "step": 54400
+    },
+    {
+      "epoch": 0.45124859243558324,
+      "grad_norm": 0.24466179311275482,
+      "learning_rate": 0.00016992726230878618,
+      "loss": 1.6089,
+      "step": 54500
+    },
+    {
+      "epoch": 0.4520765715042724,
+      "grad_norm": 0.3313222825527191,
+      "learning_rate": 0.00016987261568572804,
+      "loss": 1.634,
+      "step": 54600
+    },
+    {
+      "epoch": 0.45290455057296153,
+      "grad_norm": 0.21911990642547607,
+      "learning_rate": 0.00016981741707657846,
+      "loss": 1.6241,
+      "step": 54700
+    },
+    {
+      "epoch": 0.4537325296416507,
+      "grad_norm": 0.24226447939872742,
+      "learning_rate": 0.00016976221846742886,
+      "loss": 1.5985,
+      "step": 54800
+    },
+    {
+      "epoch": 0.4545605087103398,
+      "grad_norm": 0.25372591614723206,
+      "learning_rate": 0.00016970701985827925,
+      "loss": 1.5914,
+      "step": 54900
+    },
+    {
+      "epoch": 0.45538848777902896,
+      "grad_norm": 0.27368178963661194,
+      "learning_rate": 0.00016965182124912965,
+      "loss": 1.652,
+      "step": 55000
+    },
+    {
+      "epoch": 0.4562164668477181,
+      "grad_norm": 0.305602103471756,
+      "learning_rate": 0.00016959662263998004,
+      "loss": 1.651,
+      "step": 55100
+    },
+    {
+      "epoch": 0.45704444591640725,
+      "grad_norm": 0.2739979326725006,
+      "learning_rate": 0.00016954142403083044,
+      "loss": 1.6583,
+      "step": 55200
+    },
+    {
+      "epoch": 0.4578724249850964,
+      "grad_norm": 0.2888374924659729,
+      "learning_rate": 0.00016948622542168083,
+      "loss": 1.6283,
+      "step": 55300
+    },
+    {
+      "epoch": 0.45870040405378554,
+      "grad_norm": 0.26104286313056946,
+      "learning_rate": 0.00016943102681253122,
+      "loss": 1.6222,
+      "step": 55400
+    },
+    {
+      "epoch": 0.4595283831224747,
+      "grad_norm": 0.2540399134159088,
+      "learning_rate": 0.00016937582820338162,
+      "loss": 1.62,
+      "step": 55500
+    },
+    {
+      "epoch": 0.4603563621911638,
+      "grad_norm": 0.2812783718109131,
+      "learning_rate": 0.000169320629594232,
+      "loss": 1.6215,
+      "step": 55600
+    },
+    {
+      "epoch": 0.46118434125985297,
+      "grad_norm": 0.28794047236442566,
+      "learning_rate": 0.0001692654309850824,
+      "loss": 1.6361,
+      "step": 55700
+    },
+    {
+      "epoch": 0.4620123203285421,
+      "grad_norm": 0.24678725004196167,
+      "learning_rate": 0.0001692102323759328,
+      "loss": 1.6053,
+      "step": 55800
+    },
+    {
+      "epoch": 0.46284029939723126,
+      "grad_norm": 0.2598378360271454,
+      "learning_rate": 0.0001691550337667832,
+      "loss": 1.606,
+      "step": 55900
+    },
+    {
+      "epoch": 0.4636682784659204,
+      "grad_norm": 0.3878665566444397,
+      "learning_rate": 0.0001690998351576336,
+      "loss": 1.6494,
+      "step": 56000
+    },
+    {
+      "epoch": 0.46449625753460955,
+      "grad_norm": 0.24927765130996704,
+      "learning_rate": 0.00016904463654848398,
+      "loss": 1.6219,
+      "step": 56100
+    },
+    {
+      "epoch": 0.4653242366032987,
+      "grad_norm": 0.2416529506444931,
+      "learning_rate": 0.00016898943793933438,
+      "loss": 1.6164,
+      "step": 56200
+    },
+    {
+      "epoch": 0.46615221567198784,
+      "grad_norm": 0.3543456792831421,
+      "learning_rate": 0.00016893423933018477,
+      "loss": 1.604,
+      "step": 56300
+    },
+    {
+      "epoch": 0.466980194740677,
+      "grad_norm": 0.3229213058948517,
+      "learning_rate": 0.00016887904072103517,
+      "loss": 1.6126,
+      "step": 56400
+    },
+    {
+      "epoch": 0.4678081738093661,
+      "grad_norm": 0.24769534170627594,
+      "learning_rate": 0.00016882384211188556,
+      "loss": 1.6298,
+      "step": 56500
+    },
+    {
+      "epoch": 0.46863615287805527,
+      "grad_norm": 0.25397947430610657,
+      "learning_rate": 0.00016876864350273595,
+      "loss": 1.6389,
+      "step": 56600
+    },
+    {
+      "epoch": 0.4694641319467444,
+      "grad_norm": 0.2763853371143341,
+      "learning_rate": 0.00016871399687967784,
+      "loss": 1.602,
+      "step": 56700
+    },
+    {
+      "epoch": 0.47029211101543356,
+      "grad_norm": 0.24106226861476898,
+      "learning_rate": 0.00016865879827052824,
+      "loss": 1.6317,
+      "step": 56800
+    },
+    {
+      "epoch": 0.4711200900841227,
+      "grad_norm": 0.31650885939598083,
+      "learning_rate": 0.00016860359966137863,
+      "loss": 1.6287,
+      "step": 56900
+    },
+    {
+      "epoch": 0.47194806915281184,
+      "grad_norm": 0.22699052095413208,
+      "learning_rate": 0.00016854840105222902,
+      "loss": 1.6059,
+      "step": 57000
+    },
+    {
+      "epoch": 0.47277604822150093,
+      "grad_norm": 0.23599205911159515,
+      "learning_rate": 0.00016849320244307942,
+      "loss": 1.6359,
+      "step": 57100
+    },
+    {
+      "epoch": 0.4736040272901901,
+      "grad_norm": 0.2733438014984131,
+      "learning_rate": 0.0001684380038339298,
+      "loss": 1.6221,
+      "step": 57200
+    },
+    {
+      "epoch": 0.4744320063588792,
+      "grad_norm": 0.22067943215370178,
+      "learning_rate": 0.0001683828052247802,
+      "loss": 1.603,
+      "step": 57300
+    },
+    {
+      "epoch": 0.47525998542756837,
+      "grad_norm": 0.25246864557266235,
+      "learning_rate": 0.0001683276066156306,
+      "loss": 1.5875,
+      "step": 57400
+    },
+    {
+      "epoch": 0.4760879644962575,
+      "grad_norm": 0.2747368812561035,
+      "learning_rate": 0.000168272408006481,
+      "loss": 1.6522,
+      "step": 57500
+    },
+    {
+      "epoch": 0.47691594356494665,
+      "grad_norm": 0.352892130613327,
+      "learning_rate": 0.0001682172093973314,
+      "loss": 1.609,
+      "step": 57600
+    },
+    {
+      "epoch": 0.4777439226336358,
+      "grad_norm": 0.26428744196891785,
+      "learning_rate": 0.00016816201078818178,
+      "loss": 1.6191,
+      "step": 57700
+    },
+    {
+      "epoch": 0.47857190170232494,
+      "grad_norm": 0.2937561571598053,
+      "learning_rate": 0.0001681068121790322,
+      "loss": 1.6204,
+      "step": 57800
+    },
+    {
+      "epoch": 0.4793998807710141,
+      "grad_norm": 0.23604629933834076,
+      "learning_rate": 0.00016805161356988257,
+      "loss": 1.6416,
+      "step": 57900
+    },
+    {
+      "epoch": 0.48022785983970323,
+      "grad_norm": 0.2349170446395874,
+      "learning_rate": 0.00016799641496073297,
+      "loss": 1.6116,
+      "step": 58000
+    },
+    {
+      "epoch": 0.4810558389083924,
+      "grad_norm": 0.28929609060287476,
+      "learning_rate": 0.0001679412163515834,
+      "loss": 1.5956,
+      "step": 58100
+    },
+    {
+      "epoch": 0.4818838179770815,
+      "grad_norm": 0.2797716557979584,
+      "learning_rate": 0.00016788601774243375,
+      "loss": 1.598,
+      "step": 58200
+    },
+    {
+      "epoch": 0.48271179704577066,
+      "grad_norm": 0.30529841780662537,
+      "learning_rate": 0.00016783081913328415,
+      "loss": 1.6324,
+      "step": 58300
+    },
+    {
+      "epoch": 0.4835397761144598,
+      "grad_norm": 0.35713014006614685,
+      "learning_rate": 0.00016777562052413457,
+      "loss": 1.624,
+      "step": 58400
+    },
+    {
+      "epoch": 0.48436775518314895,
+      "grad_norm": 0.29055917263031006,
+      "learning_rate": 0.00016772042191498494,
+      "loss": 1.618,
+      "step": 58500
+    },
+    {
+      "epoch": 0.4851957342518381,
+      "grad_norm": 0.24039526283740997,
+      "learning_rate": 0.00016766522330583533,
+      "loss": 1.6056,
+      "step": 58600
+    },
+    {
+      "epoch": 0.48602371332052724,
+      "grad_norm": 0.26216012239456177,
+      "learning_rate": 0.00016761002469668575,
+      "loss": 1.6113,
+      "step": 58700
+    },
+    {
+      "epoch": 0.4868516923892164,
+      "grad_norm": 0.2159920334815979,
+      "learning_rate": 0.00016755482608753612,
+      "loss": 1.6335,
+      "step": 58800
+    },
+    {
+      "epoch": 0.48767967145790553,
+      "grad_norm": 0.26959407329559326,
+      "learning_rate": 0.00016749962747838651,
+      "loss": 1.6376,
+      "step": 58900
+    },
+    {
+      "epoch": 0.4885076505265947,
+      "grad_norm": 0.2333938479423523,
+      "learning_rate": 0.00016744442886923694,
+      "loss": 1.6516,
+      "step": 59000
+    },
+    {
+      "epoch": 0.4893356295952838,
+      "grad_norm": 0.3124794065952301,
+      "learning_rate": 0.0001673892302600873,
+      "loss": 1.6144,
+      "step": 59100
+    },
+    {
+      "epoch": 0.49016360866397296,
+      "grad_norm": 0.26165392994880676,
+      "learning_rate": 0.0001673340316509377,
+      "loss": 1.6102,
+      "step": 59200
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 362328,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1600,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0468457732150804e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}