diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6491355898522886,
+  "eval_steps": 500,
+  "global_step": 78400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008279790686891435,
+      "grad_norm": 0.1965542435646057,
+      "learning_rate": 0.00019994481795043316,
+      "loss": 1.926,
+      "step": 100
+    },
+    {
+      "epoch": 0.001655958137378287,
+      "grad_norm": 0.17061847448349,
+      "learning_rate": 0.00019988961934128355,
+      "loss": 1.7922,
+      "step": 200
+    },
+    {
+      "epoch": 0.0024839372060674308,
+      "grad_norm": 0.19415344297885895,
+      "learning_rate": 0.00019983442073213394,
+      "loss": 1.7264,
+      "step": 300
+    },
+    {
+      "epoch": 0.003311916274756574,
+      "grad_norm": 0.33725494146347046,
+      "learning_rate": 0.00019977922212298434,
+      "loss": 1.7463,
+      "step": 400
+    },
+    {
+      "epoch": 0.004139895343445717,
+      "grad_norm": 0.24688084423542023,
+      "learning_rate": 0.00019972402351383473,
+      "loss": 1.7424,
+      "step": 500
+    },
+    {
+      "epoch": 0.0049678744121348616,
+      "grad_norm": 0.235696479678154,
+      "learning_rate": 0.00019966882490468513,
+      "loss": 1.7127,
+      "step": 600
+    },
+    {
+      "epoch": 0.005795853480824005,
+      "grad_norm": 0.20345468819141388,
+      "learning_rate": 0.00019961362629553552,
+      "loss": 1.6991,
+      "step": 700
+    },
+    {
+      "epoch": 0.006623832549513148,
+      "grad_norm": 0.2328668087720871,
+      "learning_rate": 0.00019955842768638592,
+      "loss": 1.7335,
+      "step": 800
+    },
+    {
+      "epoch": 0.0074518116182022915,
+      "grad_norm": 0.21569831669330597,
+      "learning_rate": 0.00019950322907723634,
+      "loss": 1.7145,
+      "step": 900
+    },
+    {
+      "epoch": 0.008279790686891435,
+      "grad_norm": 0.2174810767173767,
+      "learning_rate": 0.0001994480304680867,
+      "loss": 1.6832,
+      "step": 1000
+    },
+    {
+      "epoch": 0.009107769755580579,
+      "grad_norm": 0.24059951305389404,
+      "learning_rate": 0.00019939283185893713,
+      "loss": 1.7052,
+      "step": 1100
+    },
+    {
+      "epoch": 0.009935748824269723,
+      "grad_norm": 0.2508637011051178,
+      "learning_rate": 0.00019933763324978752,
+      "loss": 1.7207,
+      "step": 1200
+    },
+    {
+      "epoch": 0.010763727892958866,
+      "grad_norm": 0.20193152129650116,
+      "learning_rate": 0.0001992824346406379,
+      "loss": 1.6772,
+      "step": 1300
+    },
+    {
+      "epoch": 0.01159170696164801,
+      "grad_norm": 0.23993732035160065,
+      "learning_rate": 0.0001992272360314883,
+      "loss": 1.6975,
+      "step": 1400
+    },
+    {
+      "epoch": 0.012419686030337154,
+      "grad_norm": 0.2582526206970215,
+      "learning_rate": 0.0001991720374223387,
+      "loss": 1.7142,
+      "step": 1500
+    },
+    {
+      "epoch": 0.013247665099026296,
+      "grad_norm": 0.20497630536556244,
+      "learning_rate": 0.00019911683881318907,
+      "loss": 1.7078,
+      "step": 1600
+    },
+    {
+      "epoch": 0.01407564416771544,
+      "grad_norm": 0.22760722041130066,
+      "learning_rate": 0.0001990616402040395,
+      "loss": 1.7277,
+      "step": 1700
+    },
+    {
+      "epoch": 0.014903623236404583,
+      "grad_norm": 0.22834879159927368,
+      "learning_rate": 0.00019900644159488989,
+      "loss": 1.6996,
+      "step": 1800
+    },
+    {
+      "epoch": 0.01573160230509373,
+      "grad_norm": 0.1762934923171997,
+      "learning_rate": 0.00019895124298574025,
+      "loss": 1.6856,
+      "step": 1900
+    },
+    {
+      "epoch": 0.01655958137378287,
+      "grad_norm": 0.2605816125869751,
+      "learning_rate": 0.00019889604437659067,
+      "loss": 1.707,
+      "step": 2000
+    },
+    {
+      "epoch": 0.017387560442472014,
+      "grad_norm": 0.190592423081398,
+      "learning_rate": 0.00019884084576744107,
+      "loss": 1.6793,
+      "step": 2100
+    },
+    {
+      "epoch": 0.018215539511161158,
+      "grad_norm": 0.22705131769180298,
+      "learning_rate": 0.00019878564715829144,
+      "loss": 1.6856,
+      "step": 2200
+    },
+    {
+      "epoch": 0.019043518579850302,
+      "grad_norm": 0.26750609278678894,
+      "learning_rate": 0.00019873044854914186,
+      "loss": 1.6791,
+      "step": 2300
+    },
+    {
+      "epoch": 0.019871497648539446,
+      "grad_norm": 0.20754960179328918,
+      "learning_rate": 0.00019867524993999225,
+      "loss": 1.7049,
+      "step": 2400
+    },
+    {
+      "epoch": 0.020699476717228587,
+      "grad_norm": 0.2278178185224533,
+      "learning_rate": 0.00019862005133084262,
+      "loss": 1.6614,
+      "step": 2500
+    },
+    {
+      "epoch": 0.02152745578591773,
+      "grad_norm": 0.2631435990333557,
+      "learning_rate": 0.00019856485272169304,
+      "loss": 1.6555,
+      "step": 2600
+    },
+    {
+      "epoch": 0.022355434854606875,
+      "grad_norm": 0.22711312770843506,
+      "learning_rate": 0.00019850965411254343,
+      "loss": 1.6652,
+      "step": 2700
+    },
+    {
+      "epoch": 0.02318341392329602,
+      "grad_norm": 0.204128697514534,
+      "learning_rate": 0.0001984544555033938,
+      "loss": 1.7111,
+      "step": 2800
+    },
+    {
+      "epoch": 0.024011392991985164,
+      "grad_norm": 0.2155551165342331,
+      "learning_rate": 0.00019839925689424422,
+      "loss": 1.6391,
+      "step": 2900
+    },
+    {
+      "epoch": 0.024839372060674308,
+      "grad_norm": 0.222556933760643,
+      "learning_rate": 0.00019834405828509462,
+      "loss": 1.6754,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02566735112936345,
+      "grad_norm": 0.21225596964359283,
+      "learning_rate": 0.00019828885967594498,
+      "loss": 1.6551,
+      "step": 3100
+    },
+    {
+      "epoch": 0.026495330198052593,
+      "grad_norm": 0.2524341642856598,
+      "learning_rate": 0.0001982336610667954,
+      "loss": 1.6876,
+      "step": 3200
+    },
+    {
+      "epoch": 0.027323309266741737,
+      "grad_norm": 0.19305479526519775,
+      "learning_rate": 0.0001981784624576458,
+      "loss": 1.6574,
+      "step": 3300
+    },
+    {
+      "epoch": 0.02815128833543088,
+      "grad_norm": 0.20817072689533234,
+      "learning_rate": 0.00019812326384849617,
+      "loss": 1.6962,
+      "step": 3400
+    },
+    {
+      "epoch": 0.028979267404120025,
+      "grad_norm": 0.2652389705181122,
+      "learning_rate": 0.0001980680652393466,
+      "loss": 1.6845,
+      "step": 3500
+    },
+    {
+      "epoch": 0.029807246472809166,
+      "grad_norm": 0.19599197804927826,
+      "learning_rate": 0.00019801286663019698,
+      "loss": 1.6631,
+      "step": 3600
+    },
+    {
+      "epoch": 0.03063522554149831,
+      "grad_norm": 0.23823082447052002,
+      "learning_rate": 0.00019795766802104735,
+      "loss": 1.6919,
+      "step": 3700
+    },
+    {
+      "epoch": 0.03146320461018746,
+      "grad_norm": 0.2747247517108917,
+      "learning_rate": 0.00019790246941189777,
+      "loss": 1.6821,
+      "step": 3800
+    },
+    {
+      "epoch": 0.032291183678876595,
+      "grad_norm": 0.2241993397474289,
+      "learning_rate": 0.00019784727080274816,
+      "loss": 1.6776,
+      "step": 3900
+    },
+    {
+      "epoch": 0.03311916274756574,
+      "grad_norm": 0.246478870511055,
+      "learning_rate": 0.00019779207219359853,
+      "loss": 1.6837,
+      "step": 4000
+    },
+    {
+      "epoch": 0.03394714181625488,
+      "grad_norm": 0.18319016695022583,
+      "learning_rate": 0.00019773687358444895,
+      "loss": 1.6739,
+      "step": 4100
+    },
+    {
+      "epoch": 0.03477512088494403,
+      "grad_norm": 0.19710256159305573,
+      "learning_rate": 0.00019768167497529935,
+      "loss": 1.6885,
+      "step": 4200
+    },
+    {
+      "epoch": 0.03560309995363317,
+      "grad_norm": 0.21982429921627045,
+      "learning_rate": 0.00019762647636614971,
+      "loss": 1.6945,
+      "step": 4300
+    },
+    {
+      "epoch": 0.036431079022322316,
+      "grad_norm": 0.20753520727157593,
+      "learning_rate": 0.00019757127775700014,
+      "loss": 1.6601,
+      "step": 4400
+    },
+    {
+      "epoch": 0.03725905809101146,
+      "grad_norm": 0.2044544219970703,
+      "learning_rate": 0.00019751607914785053,
+      "loss": 1.6857,
+      "step": 4500
+    },
+    {
+      "epoch": 0.038087037159700604,
+      "grad_norm": 0.22137849032878876,
+      "learning_rate": 0.0001974608805387009,
+      "loss": 1.6997,
+      "step": 4600
+    },
+    {
+      "epoch": 0.03891501622838975,
+      "grad_norm": 0.1963178962469101,
+      "learning_rate": 0.00019740568192955132,
+      "loss": 1.6851,
+      "step": 4700
+    },
+    {
+      "epoch": 0.03974299529707889,
+      "grad_norm": 0.20963279902935028,
+      "learning_rate": 0.0001973504833204017,
+      "loss": 1.6656,
+      "step": 4800
+    },
+    {
+      "epoch": 0.04057097436576804,
+      "grad_norm": 0.25534650683403015,
+      "learning_rate": 0.0001972952847112521,
+      "loss": 1.6999,
+      "step": 4900
+    },
+    {
+      "epoch": 0.041398953434457174,
+      "grad_norm": 0.22999747097492218,
+      "learning_rate": 0.0001972400861021025,
+      "loss": 1.634,
+      "step": 5000
+    },
+    {
+      "epoch": 0.04222693250314632,
+      "grad_norm": 0.21249458193778992,
+      "learning_rate": 0.0001971848874929529,
+      "loss": 1.6824,
+      "step": 5100
+    },
+    {
+      "epoch": 0.04305491157183546,
+      "grad_norm": 0.20595486462116241,
+      "learning_rate": 0.0001971296888838033,
+      "loss": 1.6587,
+      "step": 5200
+    },
+    {
+      "epoch": 0.043882890640524606,
+      "grad_norm": 0.21987958252429962,
+      "learning_rate": 0.00019707449027465368,
+      "loss": 1.6727,
+      "step": 5300
+    },
+    {
+      "epoch": 0.04471086970921375,
+      "grad_norm": 0.1945623904466629,
+      "learning_rate": 0.00019701929166550408,
+      "loss": 1.665,
+      "step": 5400
+    },
+    {
+      "epoch": 0.045538848777902895,
+      "grad_norm": 0.20507337152957916,
+      "learning_rate": 0.00019696409305635447,
+      "loss": 1.6388,
+      "step": 5500
+    },
+    {
+      "epoch": 0.04636682784659204,
+      "grad_norm": 0.21563030779361725,
+      "learning_rate": 0.00019690889444720487,
+      "loss": 1.6643,
+      "step": 5600
+    },
+    {
+      "epoch": 0.04719480691528118,
+      "grad_norm": 0.15810565650463104,
+      "learning_rate": 0.00019685369583805526,
+      "loss": 1.6367,
+      "step": 5700
+    },
+    {
+      "epoch": 0.04802278598397033,
+      "grad_norm": 0.3873724937438965,
+      "learning_rate": 0.00019679849722890565,
+      "loss": 1.6568,
+      "step": 5800
+    },
+    {
+      "epoch": 0.04885076505265947,
+      "grad_norm": 0.1988883763551712,
+      "learning_rate": 0.00019674329861975605,
+      "loss": 1.7088,
+      "step": 5900
+    },
+    {
+      "epoch": 0.049678744121348616,
+      "grad_norm": 0.20232902467250824,
+      "learning_rate": 0.00019668810001060644,
+      "loss": 1.6803,
+      "step": 6000
+    },
+    {
+      "epoch": 0.05050672319003775,
+      "grad_norm": 0.2232154905796051,
+      "learning_rate": 0.00019663290140145684,
+      "loss": 1.6575,
+      "step": 6100
+    },
+    {
+      "epoch": 0.0513347022587269,
+      "grad_norm": 0.23148497939109802,
+      "learning_rate": 0.00019657770279230723,
+      "loss": 1.6604,
+      "step": 6200
+    },
+    {
+      "epoch": 0.05216268132741604,
+      "grad_norm": 0.21519899368286133,
+      "learning_rate": 0.00019652250418315763,
+      "loss": 1.6789,
+      "step": 6300
+    },
+    {
+      "epoch": 0.052990660396105185,
+      "grad_norm": 0.1947392374277115,
+      "learning_rate": 0.00019646730557400802,
+      "loss": 1.7142,
+      "step": 6400
+    },
+    {
+      "epoch": 0.05381863946479433,
+      "grad_norm": 0.22812333703041077,
+      "learning_rate": 0.00019641210696485841,
+      "loss": 1.6561,
+      "step": 6500
+    },
+    {
+      "epoch": 0.054646618533483474,
+      "grad_norm": 0.22744423151016235,
+      "learning_rate": 0.0001963569083557088,
+      "loss": 1.6466,
+      "step": 6600
+    },
+    {
+      "epoch": 0.05547459760217262,
+      "grad_norm": 0.18678049743175507,
+      "learning_rate": 0.0001963017097465592,
+      "loss": 1.6779,
+      "step": 6700
+    },
+    {
+      "epoch": 0.05630257667086176,
+      "grad_norm": 0.23794223368167877,
+      "learning_rate": 0.0001962465111374096,
+      "loss": 1.6715,
+      "step": 6800
+    },
+    {
+      "epoch": 0.057130555739550906,
+      "grad_norm": 0.2025059461593628,
+      "learning_rate": 0.00019619131252826,
+      "loss": 1.6296,
+      "step": 6900
+    },
+    {
+      "epoch": 0.05795853480824005,
+      "grad_norm": 0.18825982511043549,
+      "learning_rate": 0.00019613611391911039,
+      "loss": 1.6415,
+      "step": 7000
+    },
+    {
+      "epoch": 0.058786513876929194,
+      "grad_norm": 0.28439417481422424,
+      "learning_rate": 0.00019608091530996078,
+      "loss": 1.6574,
+      "step": 7100
+    },
+    {
+      "epoch": 0.05961449294561833,
+      "grad_norm": 0.25905317068099976,
+      "learning_rate": 0.00019602571670081117,
+      "loss": 1.6296,
+      "step": 7200
+    },
+    {
+      "epoch": 0.060442472014307476,
+      "grad_norm": 0.19414661824703217,
+      "learning_rate": 0.00019597051809166157,
+      "loss": 1.6556,
+      "step": 7300
+    },
+    {
+      "epoch": 0.06127045108299662,
+      "grad_norm": 0.19103066623210907,
+      "learning_rate": 0.00019591531948251196,
+      "loss": 1.6536,
+      "step": 7400
+    },
+    {
+      "epoch": 0.062098430151685764,
+      "grad_norm": 0.2532936632633209,
+      "learning_rate": 0.00019586012087336236,
+      "loss": 1.6773,
+      "step": 7500
+    },
+    {
+      "epoch": 0.06292640922037492,
+      "grad_norm": 0.2522614300251007,
+      "learning_rate": 0.00019580492226421275,
+      "loss": 1.6192,
+      "step": 7600
+    },
+    {
+      "epoch": 0.06375438828906405,
+      "grad_norm": 0.2658845782279968,
+      "learning_rate": 0.00019574972365506315,
+      "loss": 1.6819,
+      "step": 7700
+    },
+    {
+      "epoch": 0.06458236735775319,
+      "grad_norm": 0.18649327754974365,
+      "learning_rate": 0.00019569452504591354,
+      "loss": 1.656,
+      "step": 7800
+    },
+    {
+      "epoch": 0.06541034642644233,
+      "grad_norm": 0.20903462171554565,
+      "learning_rate": 0.00019563932643676393,
+      "loss": 1.6931,
+      "step": 7900
+    },
+    {
+      "epoch": 0.06623832549513148,
+      "grad_norm": 0.23540228605270386,
+      "learning_rate": 0.00019558412782761433,
+      "loss": 1.6575,
+      "step": 8000
+    },
+    {
+      "epoch": 0.06706630456382062,
+      "grad_norm": 0.21330572664737701,
+      "learning_rate": 0.00019552892921846472,
+      "loss": 1.6479,
+      "step": 8100
+    },
+    {
+      "epoch": 0.06789428363250977,
+      "grad_norm": 0.19642361998558044,
+      "learning_rate": 0.00019547373060931512,
+      "loss": 1.649,
+      "step": 8200
+    },
+    {
+      "epoch": 0.06872226270119891,
+      "grad_norm": 0.24467919766902924,
+      "learning_rate": 0.0001954185320001655,
+      "loss": 1.6701,
+      "step": 8300
+    },
+    {
+      "epoch": 0.06955024176988805,
+      "grad_norm": 0.21666431427001953,
+      "learning_rate": 0.0001953633333910159,
+      "loss": 1.6745,
+      "step": 8400
+    },
+    {
+      "epoch": 0.0703782208385772,
+      "grad_norm": 0.20939518511295319,
+      "learning_rate": 0.0001953081347818663,
+      "loss": 1.642,
+      "step": 8500
+    },
+    {
+      "epoch": 0.07120619990726634,
+      "grad_norm": 0.1840800642967224,
+      "learning_rate": 0.0001952529361727167,
+      "loss": 1.6324,
+      "step": 8600
+    },
+    {
+      "epoch": 0.07203417897595549,
+      "grad_norm": 0.22671596705913544,
+      "learning_rate": 0.0001951977375635671,
+      "loss": 1.6594,
+      "step": 8700
+    },
+    {
+      "epoch": 0.07286215804464463,
+      "grad_norm": 0.2556142210960388,
+      "learning_rate": 0.00019514253895441748,
+      "loss": 1.6491,
+      "step": 8800
+    },
+    {
+      "epoch": 0.07369013711333378,
+      "grad_norm": 0.2819201648235321,
+      "learning_rate": 0.00019508734034526788,
+      "loss": 1.665,
+      "step": 8900
+    },
+    {
+      "epoch": 0.07451811618202292,
+      "grad_norm": 0.22931043803691864,
+      "learning_rate": 0.00019503214173611827,
+      "loss": 1.6766,
+      "step": 9000
+    },
+    {
+      "epoch": 0.07534609525071206,
+      "grad_norm": 0.2069370597600937,
+      "learning_rate": 0.00019497749511306016,
+      "loss": 1.6753,
+      "step": 9100
+    },
+    {
+      "epoch": 0.07617407431940121,
+      "grad_norm": 0.23254962265491486,
+      "learning_rate": 0.00019492340047609353,
+      "loss": 1.6642,
+      "step": 9200
+    },
+    {
+      "epoch": 0.07700205338809035,
+      "grad_norm": 0.19698099792003632,
+      "learning_rate": 0.00019486820186694396,
+      "loss": 1.6838,
+      "step": 9300
+    },
+    {
+      "epoch": 0.0778300324567795,
+      "grad_norm": 0.21683326363563538,
+      "learning_rate": 0.00019481300325779435,
+      "loss": 1.6741,
+      "step": 9400
+    },
+    {
+      "epoch": 0.07865801152546864,
+      "grad_norm": 0.2396763414144516,
+      "learning_rate": 0.00019475780464864472,
+      "loss": 1.6475,
+      "step": 9500
+    },
+    {
+      "epoch": 0.07948599059415778,
+      "grad_norm": 0.224142923951149,
+      "learning_rate": 0.00019470260603949514,
+      "loss": 1.6221,
+      "step": 9600
+    },
+    {
+      "epoch": 0.08031396966284693,
+      "grad_norm": 0.24930523335933685,
+      "learning_rate": 0.00019464740743034553,
+      "loss": 1.6401,
+      "step": 9700
+    },
+    {
+      "epoch": 0.08114194873153607,
+      "grad_norm": 0.26143988966941833,
+      "learning_rate": 0.0001945922088211959,
+      "loss": 1.6655,
+      "step": 9800
+    },
+    {
+      "epoch": 0.0819699278002252,
+      "grad_norm": 0.19812200963497162,
+      "learning_rate": 0.00019453701021204632,
+      "loss": 1.6469,
+      "step": 9900
+    },
+    {
+      "epoch": 0.08279790686891435,
+      "grad_norm": 0.23136496543884277,
+      "learning_rate": 0.00019448181160289672,
+      "loss": 1.6643,
+      "step": 10000
+    },
+    {
+      "epoch": 0.08362588593760349,
+      "grad_norm": 0.24188588559627533,
+      "learning_rate": 0.00019442661299374708,
+      "loss": 1.6182,
+      "step": 10100
+    },
+    {
+      "epoch": 0.08445386500629264,
+      "grad_norm": 0.2655825912952423,
+      "learning_rate": 0.0001943714143845975,
+      "loss": 1.6339,
+      "step": 10200
+    },
+    {
+      "epoch": 0.08528184407498178,
+      "grad_norm": 0.2417413890361786,
+      "learning_rate": 0.0001943162157754479,
+      "loss": 1.6076,
+      "step": 10300
+    },
+    {
+      "epoch": 0.08610982314367092,
+      "grad_norm": 0.2977447807788849,
+      "learning_rate": 0.00019426101716629827,
+      "loss": 1.6999,
+      "step": 10400
+    },
+    {
+      "epoch": 0.08693780221236007,
+      "grad_norm": 0.2565799057483673,
+      "learning_rate": 0.0001942058185571487,
+      "loss": 1.6461,
+      "step": 10500
+    },
+    {
+      "epoch": 0.08776578128104921,
+      "grad_norm": 0.2196560502052307,
+      "learning_rate": 0.00019415061994799908,
+      "loss": 1.6441,
+      "step": 10600
+    },
+    {
+      "epoch": 0.08859376034973836,
+      "grad_norm": 0.35161322355270386,
+      "learning_rate": 0.00019409542133884945,
+      "loss": 1.6521,
+      "step": 10700
+    },
+    {
+      "epoch": 0.0894217394184275,
+      "grad_norm": 0.21615535020828247,
+      "learning_rate": 0.00019404022272969987,
+      "loss": 1.6467,
+      "step": 10800
+    },
+    {
+      "epoch": 0.09024971848711665,
+      "grad_norm": 0.22406460344791412,
+      "learning_rate": 0.00019398502412055026,
+      "loss": 1.6562,
+      "step": 10900
+    },
+    {
+      "epoch": 0.09107769755580579,
+      "grad_norm": 0.21635471284389496,
+      "learning_rate": 0.00019392982551140063,
+      "loss": 1.6678,
+      "step": 11000
+    },
+    {
+      "epoch": 0.09190567662449493,
+      "grad_norm": 0.19095148146152496,
+      "learning_rate": 0.00019387462690225105,
+      "loss": 1.6288,
+      "step": 11100
+    },
+    {
+      "epoch": 0.09273365569318408,
+      "grad_norm": 0.2104116678237915,
+      "learning_rate": 0.00019381942829310145,
+      "loss": 1.6518,
+      "step": 11200
+    },
+    {
+      "epoch": 0.09356163476187322,
+      "grad_norm": 0.217295840382576,
+      "learning_rate": 0.0001937642296839518,
+      "loss": 1.6517,
+      "step": 11300
+    },
+    {
+      "epoch": 0.09438961383056237,
+      "grad_norm": 0.2044863998889923,
+      "learning_rate": 0.00019370903107480223,
+      "loss": 1.6424,
+      "step": 11400
+    },
+    {
+      "epoch": 0.09521759289925151,
+      "grad_norm": 0.257137656211853,
+      "learning_rate": 0.00019365383246565263,
+      "loss": 1.6686,
+      "step": 11500
+    },
+    {
+      "epoch": 0.09604557196794065,
+      "grad_norm": 0.19957780838012695,
+      "learning_rate": 0.000193598633856503,
+      "loss": 1.6052,
+      "step": 11600
+    },
+    {
+      "epoch": 0.0968735510366298,
+      "grad_norm": 0.2578992545604706,
+      "learning_rate": 0.00019354343524735342,
+      "loss": 1.637,
+      "step": 11700
+    },
+    {
+      "epoch": 0.09770153010531894,
+      "grad_norm": 0.25783538818359375,
+      "learning_rate": 0.0001934882366382038,
+      "loss": 1.6513,
+      "step": 11800
+    },
+    {
+      "epoch": 0.09852950917400809,
+      "grad_norm": 0.23085756599903107,
+      "learning_rate": 0.00019343303802905418,
+      "loss": 1.6775,
+      "step": 11900
+    },
+    {
+      "epoch": 0.09935748824269723,
+      "grad_norm": 0.20355592668056488,
+      "learning_rate": 0.0001933778394199046,
+      "loss": 1.6475,
+      "step": 12000
+    },
+    {
+      "epoch": 0.10018546731138636,
+      "grad_norm": 0.1904350370168686,
+      "learning_rate": 0.000193322640810755,
+      "loss": 1.6432,
+      "step": 12100
+    },
+    {
+      "epoch": 0.1010134463800755,
+      "grad_norm": 0.2179509401321411,
+      "learning_rate": 0.0001932674422016054,
+      "loss": 1.6708,
+      "step": 12200
+    },
+    {
+      "epoch": 0.10184142544876465,
+      "grad_norm": 0.23058493435382843,
+      "learning_rate": 0.00019321224359245578,
+      "loss": 1.6893,
+      "step": 12300
+    },
+    {
+      "epoch": 0.1026694045174538,
+      "grad_norm": 0.21472018957138062,
+      "learning_rate": 0.00019315704498330618,
+      "loss": 1.6534,
+      "step": 12400
+    },
+    {
+      "epoch": 0.10349738358614294,
+      "grad_norm": 0.22223271429538727,
+      "learning_rate": 0.00019310184637415657,
+      "loss": 1.6569,
+      "step": 12500
+    },
+    {
+      "epoch": 0.10432536265483208,
+      "grad_norm": 0.218324676156044,
+      "learning_rate": 0.00019304664776500697,
+      "loss": 1.6592,
+      "step": 12600
+    },
+    {
+      "epoch": 0.10515334172352123,
+      "grad_norm": 0.18915393948554993,
+      "learning_rate": 0.00019299144915585736,
+      "loss": 1.6757,
+      "step": 12700
+    },
+    {
+      "epoch": 0.10598132079221037,
+      "grad_norm": 0.19763177633285522,
+      "learning_rate": 0.00019293625054670775,
+      "loss": 1.6327,
+      "step": 12800
+    },
+    {
+      "epoch": 0.10680929986089951,
+      "grad_norm": 0.20737317204475403,
+      "learning_rate": 0.00019288105193755815,
+      "loss": 1.6319,
+      "step": 12900
+    },
+    {
+      "epoch": 0.10763727892958866,
+      "grad_norm": 0.20598194003105164,
+      "learning_rate": 0.00019282585332840854,
+      "loss": 1.6431,
+      "step": 13000
+    },
+    {
+      "epoch": 0.1084652579982778,
+      "grad_norm": 0.24075473845005035,
+      "learning_rate": 0.00019277065471925894,
+      "loss": 1.6557,
+      "step": 13100
+    },
+    {
+      "epoch": 0.10929323706696695,
+      "grad_norm": 0.19969555735588074,
+      "learning_rate": 0.00019271545611010933,
+      "loss": 1.6702,
+      "step": 13200
+    },
+    {
+      "epoch": 0.11012121613565609,
+      "grad_norm": 0.2330610454082489,
+      "learning_rate": 0.00019266025750095973,
+      "loss": 1.6787,
+      "step": 13300
+    },
+    {
+      "epoch": 0.11094919520434524,
+      "grad_norm": 0.20984840393066406,
+      "learning_rate": 0.00019260505889181012,
+      "loss": 1.6416,
+      "step": 13400
+    },
+    {
+      "epoch": 0.11177717427303438,
+      "grad_norm": 0.2110757976770401,
+      "learning_rate": 0.00019254986028266051,
+      "loss": 1.686,
+      "step": 13500
+    },
+    {
+      "epoch": 0.11260515334172352,
+      "grad_norm": 0.2638145685195923,
+      "learning_rate": 0.0001924946616735109,
+      "loss": 1.6507,
+      "step": 13600
+    },
+    {
+      "epoch": 0.11343313241041267,
+      "grad_norm": 0.20532745122909546,
+      "learning_rate": 0.0001924394630643613,
+      "loss": 1.6409,
+      "step": 13700
+    },
+    {
+      "epoch": 0.11426111147910181,
+      "grad_norm": 0.21913674473762512,
+      "learning_rate": 0.0001923842644552117,
+      "loss": 1.6341,
+      "step": 13800
+    },
+    {
+      "epoch": 0.11508909054779096,
+      "grad_norm": 0.29349780082702637,
+      "learning_rate": 0.0001923290658460621,
+      "loss": 1.6626,
+      "step": 13900
+    },
+    {
+      "epoch": 0.1159170696164801,
+      "grad_norm": 0.20620688796043396,
+      "learning_rate": 0.00019227386723691248,
+      "loss": 1.6993,
+      "step": 14000
+    },
+    {
+      "epoch": 0.11674504868516924,
+      "grad_norm": 0.2426968812942505,
+      "learning_rate": 0.00019221866862776288,
+      "loss": 1.6519,
+      "step": 14100
+    },
+    {
+      "epoch": 0.11757302775385839,
+      "grad_norm": 0.24563075602054596,
+      "learning_rate": 0.00019216347001861327,
+      "loss": 1.6951,
+      "step": 14200
+    },
+    {
+      "epoch": 0.11840100682254752,
+      "grad_norm": 0.2335137128829956,
+      "learning_rate": 0.00019210827140946367,
+      "loss": 1.6252,
+      "step": 14300
+    },
+    {
+      "epoch": 0.11922898589123666,
+      "grad_norm": 0.2369394600391388,
+      "learning_rate": 0.00019205307280031406,
+      "loss": 1.6747,
+      "step": 14400
+    },
+    {
+      "epoch": 0.12005696495992581,
+      "grad_norm": 0.22131401300430298,
+      "learning_rate": 0.00019199787419116446,
+      "loss": 1.6462,
+      "step": 14500
+    },
+    {
+      "epoch": 0.12088494402861495,
+      "grad_norm": 0.3465203046798706,
+      "learning_rate": 0.00019194267558201485,
+      "loss": 1.6379,
+      "step": 14600
+    },
+    {
+      "epoch": 0.1217129230973041,
+      "grad_norm": 0.26189711689949036,
+      "learning_rate": 0.00019188747697286524,
+      "loss": 1.7039,
+      "step": 14700
+    },
+    {
+      "epoch": 0.12254090216599324,
+      "grad_norm": 0.23303379118442535,
+      "learning_rate": 0.00019183227836371564,
+      "loss": 1.623,
+      "step": 14800
+    },
+    {
+      "epoch": 0.12336888123468238,
+      "grad_norm": 0.21591255068778992,
+      "learning_rate": 0.00019177707975456603,
+      "loss": 1.6421,
+      "step": 14900
+    },
+    {
+      "epoch": 0.12419686030337153,
+      "grad_norm": 0.21272686123847961,
+      "learning_rate": 0.00019172188114541643,
+      "loss": 1.6627,
+      "step": 15000
+    },
+    {
+      "epoch": 0.1250248393720607,
+      "grad_norm": 0.21305076777935028,
+      "learning_rate": 0.00019166668253626682,
+      "loss": 1.6609,
+      "step": 15100
+    },
+    {
+      "epoch": 0.12585281844074983,
+      "grad_norm": 0.19881877303123474,
+      "learning_rate": 0.00019161148392711722,
+      "loss": 1.6446,
+      "step": 15200
+    },
+    {
+      "epoch": 0.12668079750943897,
+      "grad_norm": 0.23909015953540802,
+      "learning_rate": 0.0001915568373040591,
+      "loss": 1.6429,
+      "step": 15300
+    },
+    {
+      "epoch": 0.1275087765781281,
+      "grad_norm": 0.2043364942073822,
+      "learning_rate": 0.0001915016386949095,
+      "loss": 1.6618,
+      "step": 15400
+    },
+    {
+      "epoch": 0.12833675564681724,
+      "grad_norm": 0.2276351898908615,
+      "learning_rate": 0.0001914464400857599,
+      "loss": 1.6341,
+      "step": 15500
+    },
+    {
+      "epoch": 0.12916473471550638,
+      "grad_norm": 0.23034417629241943,
+      "learning_rate": 0.00019139124147661028,
+      "loss": 1.6323,
+      "step": 15600
+    },
+    {
+      "epoch": 0.12999271378419552,
+      "grad_norm": 0.24165049195289612,
+      "learning_rate": 0.00019133604286746068,
+      "loss": 1.6327,
+      "step": 15700
+    },
+    {
+      "epoch": 0.13082069285288467,
+      "grad_norm": 0.2675997018814087,
+      "learning_rate": 0.00019128084425831107,
+      "loss": 1.6315,
+      "step": 15800
+    },
+    {
+      "epoch": 0.1316486719215738,
+      "grad_norm": 0.228688582777977,
+      "learning_rate": 0.00019122564564916147,
+      "loss": 1.6417,
+      "step": 15900
+    },
+    {
+      "epoch": 0.13247665099026296,
+      "grad_norm": 0.23833218216896057,
+      "learning_rate": 0.00019117044704001186,
+      "loss": 1.6341,
+      "step": 16000
+    },
+    {
+      "epoch": 0.1333046300589521,
+      "grad_norm": 0.26954251527786255,
+      "learning_rate": 0.00019111580041695375,
+      "loss": 1.6444,
+      "step": 16100
+    },
+    {
+      "epoch": 0.13413260912764124,
+      "grad_norm": 0.2615845203399658,
+      "learning_rate": 0.00019106060180780417,
+      "loss": 1.6373,
+      "step": 16200
+    },
+    {
+      "epoch": 0.1349605881963304,
+      "grad_norm": 0.24307669699192047,
+      "learning_rate": 0.00019100540319865454,
+      "loss": 1.6249,
+      "step": 16300
+    },
+    {
+      "epoch": 0.13578856726501953,
+      "grad_norm": 0.2018076479434967,
+      "learning_rate": 0.00019095020458950496,
+      "loss": 1.6502,
+      "step": 16400
+    },
+    {
+      "epoch": 0.13661654633370868,
+      "grad_norm": 0.21472395956516266,
+      "learning_rate": 0.00019089500598035535,
+      "loss": 1.6392,
+      "step": 16500
+    },
+    {
+      "epoch": 0.13744452540239782,
+      "grad_norm": 0.24748454988002777,
+      "learning_rate": 0.00019083980737120572,
+      "loss": 1.687,
+      "step": 16600
+    },
+    {
+      "epoch": 0.13827250447108697,
+      "grad_norm": 0.24712525308132172,
+      "learning_rate": 0.00019078460876205614,
+      "loss": 1.6488,
+      "step": 16700
+    },
+    {
+      "epoch": 0.1391004835397761,
+      "grad_norm": 0.2053895890712738,
+      "learning_rate": 0.00019072941015290653,
+      "loss": 1.6407,
+      "step": 16800
+    },
+    {
+      "epoch": 0.13992846260846525,
+      "grad_norm": 0.2149922251701355,
+      "learning_rate": 0.0001906742115437569,
+      "loss": 1.6265,
+      "step": 16900
+    },
+    {
+      "epoch": 0.1407564416771544,
+      "grad_norm": 0.26773732900619507,
+      "learning_rate": 0.00019061901293460732,
+      "loss": 1.6593,
+      "step": 17000
+    },
+    {
+      "epoch": 0.14158442074584354,
+      "grad_norm": 0.20975753664970398,
+      "learning_rate": 0.00019056381432545772,
+      "loss": 1.6345,
+      "step": 17100
+    },
+    {
+      "epoch": 0.1424123998145327,
+      "grad_norm": 0.2390817403793335,
+      "learning_rate": 0.00019050861571630808,
+      "loss": 1.6241,
+      "step": 17200
+    },
+    {
+      "epoch": 0.14324037888322183,
+      "grad_norm": 0.2248607873916626,
+      "learning_rate": 0.0001904534171071585,
+      "loss": 1.6352,
+      "step": 17300
+    },
+    {
+      "epoch": 0.14406835795191097,
+      "grad_norm": 0.25773122906684875,
+      "learning_rate": 0.0001903982184980089,
+      "loss": 1.6365,
+      "step": 17400
+    },
+    {
+      "epoch": 0.14489633702060012,
+      "grad_norm": 0.2824070155620575,
+      "learning_rate": 0.00019034301988885927,
+      "loss": 1.6565,
+      "step": 17500
+    },
+    {
+      "epoch": 0.14572431608928926,
+      "grad_norm": 0.21992579102516174,
+      "learning_rate": 0.0001902878212797097,
+      "loss": 1.6707,
+      "step": 17600
+    },
+    {
+      "epoch": 0.1465522951579784,
+      "grad_norm": 0.2076718509197235,
+      "learning_rate": 0.00019023262267056008,
+      "loss": 1.6515,
+      "step": 17700
+    },
+    {
+      "epoch": 0.14738027422666755,
+      "grad_norm": 0.2828584313392639,
+      "learning_rate": 0.00019017742406141045,
+      "loss": 1.6021,
+      "step": 17800
+    },
+    {
+      "epoch": 0.1482082532953567,
+      "grad_norm": 0.20166254043579102,
+      "learning_rate": 0.00019012222545226087,
+      "loss": 1.6144,
+      "step": 17900
+    },
+    {
+      "epoch": 0.14903623236404584,
+      "grad_norm": 0.2705940008163452,
+      "learning_rate": 0.00019006702684311127,
+      "loss": 1.6853,
+      "step": 18000
+    },
+    {
+      "epoch": 0.14986421143273498,
+      "grad_norm": 0.2661746144294739,
+      "learning_rate": 0.00019001182823396163,
+      "loss": 1.6554,
+      "step": 18100
+    },
+    {
+      "epoch": 0.15069219050142413,
+      "grad_norm": 0.19586409628391266,
+      "learning_rate": 0.00018995662962481205,
+      "loss": 1.6739,
+      "step": 18200
+    },
+    {
+      "epoch": 0.15152016957011327,
+      "grad_norm": 0.22720645368099213,
+      "learning_rate": 0.00018990143101566245,
+      "loss": 1.6599,
+      "step": 18300
+    },
+    {
+      "epoch": 0.15234814863880242,
+      "grad_norm": 0.20618566870689392,
+      "learning_rate": 0.00018984623240651282,
+      "loss": 1.6389,
+      "step": 18400
+    },
+    {
+      "epoch": 0.15317612770749156,
+      "grad_norm": 0.20746736228466034,
+      "learning_rate": 0.00018979103379736324,
+      "loss": 1.6367,
+      "step": 18500
+    },
+    {
+      "epoch": 0.1540041067761807,
+      "grad_norm": 0.2366802841424942,
+      "learning_rate": 0.00018973583518821363,
+      "loss": 1.6096,
+      "step": 18600
+    },
+    {
+      "epoch": 0.15483208584486985,
+      "grad_norm": 0.22609083354473114,
+      "learning_rate": 0.000189680636579064,
+      "loss": 1.6324,
+      "step": 18700
+    },
+    {
+      "epoch": 0.155660064913559,
+      "grad_norm": 0.27429673075675964,
+      "learning_rate": 0.00018962543796991442,
+      "loss": 1.617,
+      "step": 18800
+    },
+    {
+      "epoch": 0.15648804398224814,
+      "grad_norm": 0.24921450018882751,
+      "learning_rate": 0.0001895702393607648,
+      "loss": 1.6605,
+      "step": 18900
+    },
+    {
+      "epoch": 0.15731602305093728,
+      "grad_norm": 0.20069612562656403,
+      "learning_rate": 0.00018951504075161518,
+      "loss": 1.6253,
+      "step": 19000
+    },
+    {
+      "epoch": 0.15814400211962643,
+      "grad_norm": 0.26492151618003845,
+      "learning_rate": 0.0001894598421424656,
+      "loss": 1.6561,
+      "step": 19100
+    },
+    {
+      "epoch": 0.15897198118831557,
+      "grad_norm": 0.24309319257736206,
+      "learning_rate": 0.000189404643533316,
+      "loss": 1.6432,
+      "step": 19200
+    },
+    {
+      "epoch": 0.15979996025700471,
+      "grad_norm": 0.22135823965072632,
+      "learning_rate": 0.00018934944492416636,
+      "loss": 1.6476,
+      "step": 19300
+    },
+    {
+      "epoch": 0.16062793932569386,
+      "grad_norm": 0.2506055533885956,
+      "learning_rate": 0.00018929424631501678,
+      "loss": 1.6391,
+      "step": 19400
+    },
+    {
+      "epoch": 0.161455918394383,
+      "grad_norm": 0.24688918888568878,
+      "learning_rate": 0.00018923904770586718,
+      "loss": 1.662,
+      "step": 19500
+    },
+    {
+      "epoch": 0.16228389746307215,
+      "grad_norm": 0.21143893897533417,
+      "learning_rate": 0.00018918384909671755,
+      "loss": 1.6302,
+      "step": 19600
+    },
+    {
+      "epoch": 0.1631118765317613,
+      "grad_norm": 0.20364534854888916,
+      "learning_rate": 0.00018912865048756797,
+      "loss": 1.6402,
+      "step": 19700
+    },
+    {
+      "epoch": 0.1639398556004504,
+      "grad_norm": 0.24464745819568634,
+      "learning_rate": 0.00018907345187841836,
+      "loss": 1.6452,
+      "step": 19800
+    },
+    {
+      "epoch": 0.16476783466913955,
+      "grad_norm": 0.20947644114494324,
+      "learning_rate": 0.00018901825326926873,
+      "loss": 1.624,
+      "step": 19900
+    },
+    {
+      "epoch": 0.1655958137378287,
+      "grad_norm": 0.20495034754276276,
+      "learning_rate": 0.00018896305466011915,
+      "loss": 1.6585,
+      "step": 20000
+    },
+    {
+      "epoch": 0.16642379280651784,
+      "grad_norm": 0.26043763756752014,
+      "learning_rate": 0.00018890785605096954,
+      "loss": 1.6746,
+      "step": 20100
+    },
+    {
+      "epoch": 0.16725177187520698,
+      "grad_norm": 0.20851582288742065,
+      "learning_rate": 0.00018885320942791143,
+      "loss": 1.625,
+      "step": 20200
+    },
+    {
+      "epoch": 0.16807975094389613,
+      "grad_norm": 0.2077847421169281,
+      "learning_rate": 0.00018879801081876182,
+      "loss": 1.6315,
+      "step": 20300
+    },
+    {
+      "epoch": 0.16890773001258527,
+      "grad_norm": 0.21309372782707214,
+      "learning_rate": 0.0001887433641957037,
+      "loss": 1.6452,
+      "step": 20400
+    },
+    {
+      "epoch": 0.16973570908127442,
+      "grad_norm": 0.205887570977211,
+      "learning_rate": 0.0001886881655865541,
+      "loss": 1.637,
+      "step": 20500
+    },
+    {
+      "epoch": 0.17056368814996356,
+      "grad_norm": 0.18793857097625732,
+      "learning_rate": 0.0001886329669774045,
+      "loss": 1.6381,
+      "step": 20600
+    },
+    {
+      "epoch": 0.1713916672186527,
+      "grad_norm": 0.20410719513893127,
+      "learning_rate": 0.0001885777683682549,
+      "loss": 1.6282,
+      "step": 20700
+    },
+    {
+      "epoch": 0.17221964628734185,
+      "grad_norm": 0.2512223422527313,
+      "learning_rate": 0.0001885225697591053,
+      "loss": 1.6311,
+      "step": 20800
+    },
+    {
+      "epoch": 0.173047625356031,
+      "grad_norm": 0.2699196934700012,
+      "learning_rate": 0.00018846737114995568,
+      "loss": 1.6852,
+      "step": 20900
+    },
+    {
+      "epoch": 0.17387560442472014,
+      "grad_norm": 0.24233217537403107,
+      "learning_rate": 0.00018841217254080608,
+      "loss": 1.6642,
+      "step": 21000
+    },
+    {
+      "epoch": 0.17470358349340928,
+      "grad_norm": 0.23415082693099976,
+      "learning_rate": 0.00018835697393165647,
+      "loss": 1.6825,
+      "step": 21100
+    },
+    {
+      "epoch": 0.17553156256209843,
+      "grad_norm": 0.20289281010627747,
+      "learning_rate": 0.00018830177532250686,
+      "loss": 1.6164,
+      "step": 21200
+    },
+    {
+      "epoch": 0.17635954163078757,
+      "grad_norm": 0.2364041954278946,
+      "learning_rate": 0.00018824657671335726,
+      "loss": 1.6243,
+      "step": 21300
+    },
+    {
+      "epoch": 0.1771875206994767,
+      "grad_norm": 0.21854723989963531,
+      "learning_rate": 0.00018819137810420765,
+      "loss": 1.6549,
+      "step": 21400
+    },
+    {
+      "epoch": 0.17801549976816586,
+      "grad_norm": 0.2151113748550415,
+      "learning_rate": 0.00018813617949505805,
+      "loss": 1.6278,
+      "step": 21500
+    },
+    {
+      "epoch": 0.178843478836855,
+      "grad_norm": 0.2160249501466751,
+      "learning_rate": 0.00018808098088590844,
+      "loss": 1.6475,
+      "step": 21600
+    },
+    {
+      "epoch": 0.17967145790554415,
+      "grad_norm": 0.22971031069755554,
+      "learning_rate": 0.00018802578227675884,
+      "loss": 1.6414,
+      "step": 21700
+    },
+    {
+      "epoch": 0.1804994369742333,
+      "grad_norm": 0.2136770784854889,
+      "learning_rate": 0.00018797058366760923,
+      "loss": 1.6364,
+      "step": 21800
+    },
+    {
+      "epoch": 0.18132741604292243,
+      "grad_norm": 0.22959814965724945,
+      "learning_rate": 0.00018791538505845962,
+      "loss": 1.639,
+      "step": 21900
+    },
+    {
+      "epoch": 0.18215539511161158,
+      "grad_norm": 0.21293580532073975,
+      "learning_rate": 0.00018786018644931002,
+      "loss": 1.6419,
+      "step": 22000
+    },
+    {
+      "epoch": 0.18298337418030072,
+      "grad_norm": 0.241853266954422,
+      "learning_rate": 0.0001878049878401604,
+      "loss": 1.6887,
+      "step": 22100
+    },
+    {
+      "epoch": 0.18381135324898987,
+      "grad_norm": 0.23669332265853882,
+      "learning_rate": 0.0001877497892310108,
+      "loss": 1.6685,
+      "step": 22200
+    },
+    {
+      "epoch": 0.184639332317679,
+      "grad_norm": 0.20697368681430817,
+      "learning_rate": 0.0001876945906218612,
+      "loss": 1.6351,
+      "step": 22300
+    },
+    {
+      "epoch": 0.18546731138636816,
+      "grad_norm": 0.2443273812532425,
+      "learning_rate": 0.0001876393920127116,
+      "loss": 1.6545,
+      "step": 22400
+    },
+    {
+      "epoch": 0.1862952904550573,
+      "grad_norm": 0.20272250473499298,
+      "learning_rate": 0.000187584193403562,
+      "loss": 1.6325,
+      "step": 22500
+    },
+    {
+      "epoch": 0.18712326952374644,
+      "grad_norm": 0.23750340938568115,
+      "learning_rate": 0.00018752899479441238,
+      "loss": 1.6248,
+      "step": 22600
+    },
+    {
+      "epoch": 0.1879512485924356,
+      "grad_norm": 0.20021654665470123,
+      "learning_rate": 0.00018747379618526278,
+      "loss": 1.6383,
+      "step": 22700
+    },
+    {
+      "epoch": 0.18877922766112473,
+      "grad_norm": 0.23722049593925476,
+      "learning_rate": 0.00018741859757611317,
+      "loss": 1.6377,
+      "step": 22800
+    },
+    {
+      "epoch": 0.18960720672981388,
+      "grad_norm": 0.22547928988933563,
+      "learning_rate": 0.00018736395095305509,
+      "loss": 1.6223,
+      "step": 22900
+    },
+    {
+      "epoch": 0.19043518579850302,
+      "grad_norm": 0.23537902534008026,
+      "learning_rate": 0.00018730875234390545,
+      "loss": 1.6527,
+      "step": 23000
+    },
+    {
+      "epoch": 0.19126316486719216,
+      "grad_norm": 0.25864705443382263,
+      "learning_rate": 0.00018725355373475585,
+      "loss": 1.6138,
+      "step": 23100
+    },
+    {
+      "epoch": 0.1920911439358813,
+      "grad_norm": 0.24374592304229736,
+      "learning_rate": 0.00018719835512560627,
+      "loss": 1.6443,
+      "step": 23200
+    },
+    {
+      "epoch": 0.19291912300457045,
+      "grad_norm": 0.22655454277992249,
+      "learning_rate": 0.00018714315651645664,
+      "loss": 1.6472,
+      "step": 23300
+    },
+    {
+      "epoch": 0.1937471020732596,
+      "grad_norm": 0.23370279371738434,
+      "learning_rate": 0.00018708795790730703,
+      "loss": 1.6491,
+      "step": 23400
+    },
+    {
+      "epoch": 0.19457508114194874,
+      "grad_norm": 0.2537304162979126,
+      "learning_rate": 0.00018703275929815745,
+      "loss": 1.6426,
+      "step": 23500
+    },
+    {
+      "epoch": 0.19540306021063789,
+      "grad_norm": 0.23084019124507904,
+      "learning_rate": 0.00018697756068900782,
+      "loss": 1.6487,
+      "step": 23600
+    },
+    {
+      "epoch": 0.19623103927932703,
+      "grad_norm": 0.19103647768497467,
+      "learning_rate": 0.00018692236207985824,
+      "loss": 1.6103,
+      "step": 23700
+    },
+    {
+      "epoch": 0.19705901834801617,
+      "grad_norm": 0.23761576414108276,
+      "learning_rate": 0.00018686716347070863,
+      "loss": 1.6402,
+      "step": 23800
+    },
+    {
+      "epoch": 0.19788699741670532,
+      "grad_norm": 0.2252289056777954,
+      "learning_rate": 0.000186811964861559,
+      "loss": 1.6376,
+      "step": 23900
+    },
+    {
+      "epoch": 0.19871497648539446,
+      "grad_norm": 0.25417011976242065,
+      "learning_rate": 0.00018675676625240942,
+      "loss": 1.6236,
+      "step": 24000
+    },
+    {
+      "epoch": 0.1995429555540836,
+      "grad_norm": 0.2344467043876648,
+      "learning_rate": 0.00018670156764325982,
+      "loss": 1.6432,
+      "step": 24100
+    },
+    {
+      "epoch": 0.20037093462277272,
+      "grad_norm": 0.22560527920722961,
+      "learning_rate": 0.00018664636903411018,
+      "loss": 1.6404,
+      "step": 24200
+    },
+    {
+      "epoch": 0.20119891369146187,
+      "grad_norm": 0.24285413324832916,
+      "learning_rate": 0.0001865911704249606,
+      "loss": 1.6199,
+      "step": 24300
+    },
+    {
+      "epoch": 0.202026892760151,
+      "grad_norm": 0.26057812571525574,
+      "learning_rate": 0.000186535971815811,
+      "loss": 1.6302,
+      "step": 24400
+    },
+    {
+      "epoch": 0.20285487182884016,
+      "grad_norm": 0.22331953048706055,
+      "learning_rate": 0.00018648077320666137,
+      "loss": 1.623,
+      "step": 24500
+    },
+    {
+      "epoch": 0.2036828508975293,
+      "grad_norm": 0.24363332986831665,
+      "learning_rate": 0.0001864255745975118,
+      "loss": 1.6365,
+      "step": 24600
+    },
+    {
+      "epoch": 0.20451082996621844,
+      "grad_norm": 0.22553332149982452,
+      "learning_rate": 0.00018637037598836218,
+      "loss": 1.6135,
+      "step": 24700
+    },
+    {
+      "epoch": 0.2053388090349076,
+      "grad_norm": 0.23822011053562164,
+      "learning_rate": 0.00018631517737921255,
+      "loss": 1.6404,
+      "step": 24800
+    },
+    {
+      "epoch": 0.20616678810359673,
+      "grad_norm": 0.22308780252933502,
+      "learning_rate": 0.00018625997877006297,
+      "loss": 1.6401,
+      "step": 24900
+    },
+    {
+      "epoch": 0.20699476717228588,
+      "grad_norm": 0.232479065656662,
+      "learning_rate": 0.00018620478016091336,
+      "loss": 1.6457,
+      "step": 25000
+    },
+    {
+      "epoch": 0.20782274624097502,
+      "grad_norm": 0.23883666098117828,
+      "learning_rate": 0.00018614958155176373,
+      "loss": 1.6687,
+      "step": 25100
+    },
+    {
+      "epoch": 0.20865072530966416,
+      "grad_norm": 0.215969979763031,
+      "learning_rate": 0.00018609438294261415,
+      "loss": 1.6314,
+      "step": 25200
+    },
+    {
+      "epoch": 0.2094787043783533,
+      "grad_norm": 0.2068208009004593,
+      "learning_rate": 0.00018603918433346455,
+      "loss": 1.6458,
+      "step": 25300
+    },
+    {
+      "epoch": 0.21030668344704245,
+      "grad_norm": 0.2438194751739502,
+      "learning_rate": 0.00018598398572431491,
+      "loss": 1.6519,
+      "step": 25400
+    },
+    {
+      "epoch": 0.2111346625157316,
+      "grad_norm": 0.26404204964637756,
+      "learning_rate": 0.00018592878711516534,
+      "loss": 1.6276,
+      "step": 25500
+    },
+    {
+      "epoch": 0.21196264158442074,
+      "grad_norm": 0.2229745090007782,
+      "learning_rate": 0.0001858741404921072,
+      "loss": 1.6585,
+      "step": 25600
+    },
+    {
+      "epoch": 0.21279062065310989,
+      "grad_norm": 0.2071082442998886,
+      "learning_rate": 0.00018581894188295762,
+      "loss": 1.6187,
+      "step": 25700
+    },
+    {
+      "epoch": 0.21361859972179903,
+      "grad_norm": 0.41804060339927673,
+      "learning_rate": 0.000185763743273808,
+      "loss": 1.6219,
+      "step": 25800
+    },
+    {
+      "epoch": 0.21444657879048817,
+      "grad_norm": 0.22351470589637756,
+      "learning_rate": 0.00018570854466465838,
+      "loss": 1.6616,
+      "step": 25900
+    },
+    {
+      "epoch": 0.21527455785917732,
+      "grad_norm": 0.219690203666687,
+      "learning_rate": 0.0001856533460555088,
+      "loss": 1.6551,
+      "step": 26000
+    },
+    {
+      "epoch": 0.21610253692786646,
+      "grad_norm": 0.25187328457832336,
+      "learning_rate": 0.0001855981474463592,
+      "loss": 1.6272,
+      "step": 26100
+    },
+    {
+      "epoch": 0.2169305159965556,
+      "grad_norm": 0.22479000687599182,
+      "learning_rate": 0.00018554294883720956,
+      "loss": 1.6096,
+      "step": 26200
+    },
+    {
+      "epoch": 0.21775849506524475,
+      "grad_norm": 0.2545852065086365,
+      "learning_rate": 0.00018548775022805998,
+      "loss": 1.6099,
+      "step": 26300
+    },
+    {
+      "epoch": 0.2185864741339339,
+      "grad_norm": 0.24930129945278168,
+      "learning_rate": 0.00018543255161891038,
+      "loss": 1.6247,
+      "step": 26400
+    },
+    {
+      "epoch": 0.21941445320262304,
+      "grad_norm": 0.3432662785053253,
+      "learning_rate": 0.00018537735300976077,
+      "loss": 1.6523,
+      "step": 26500
+    },
+    {
+      "epoch": 0.22024243227131218,
+      "grad_norm": 0.19649013876914978,
+      "learning_rate": 0.00018532215440061116,
+      "loss": 1.6761,
+      "step": 26600
+    },
+    {
+      "epoch": 0.22107041134000133,
+      "grad_norm": 0.22068659961223602,
+      "learning_rate": 0.00018526695579146156,
+      "loss": 1.655,
+      "step": 26700
+    },
+    {
+      "epoch": 0.22189839040869047,
+      "grad_norm": 0.2267031967639923,
+      "learning_rate": 0.00018521175718231195,
+      "loss": 1.5902,
+      "step": 26800
+    },
+    {
+      "epoch": 0.22272636947737962,
+      "grad_norm": 0.2132362276315689,
+      "learning_rate": 0.00018515655857316235,
+      "loss": 1.6265,
+      "step": 26900
+    },
+    {
+      "epoch": 0.22355434854606876,
+      "grad_norm": 0.21281544864177704,
+      "learning_rate": 0.00018510135996401274,
+      "loss": 1.6584,
+      "step": 27000
+    },
+    {
+      "epoch": 0.2243823276147579,
+      "grad_norm": 0.22479993104934692,
+      "learning_rate": 0.00018504616135486314,
+      "loss": 1.6518,
+      "step": 27100
+    },
+    {
+      "epoch": 0.22521030668344705,
+      "grad_norm": 0.27082720398902893,
+      "learning_rate": 0.00018499096274571353,
+      "loss": 1.6251,
+      "step": 27200
+    },
+    {
+      "epoch": 0.2260382857521362,
+      "grad_norm": 0.22913210093975067,
+      "learning_rate": 0.00018493576413656392,
+      "loss": 1.6431,
+      "step": 27300
+    },
+    {
+      "epoch": 0.22686626482082534,
+      "grad_norm": 0.2798856496810913,
+      "learning_rate": 0.00018488056552741432,
+      "loss": 1.6707,
+      "step": 27400
+    },
+    {
+      "epoch": 0.22769424388951448,
+      "grad_norm": 0.2440473437309265,
+      "learning_rate": 0.0001848253669182647,
+      "loss": 1.6457,
+      "step": 27500
+    },
+    {
+      "epoch": 0.22852222295820362,
+      "grad_norm": 0.22927863895893097,
+      "learning_rate": 0.0001847701683091151,
+      "loss": 1.6066,
+      "step": 27600
+    },
+    {
+      "epoch": 0.22935020202689277,
+      "grad_norm": 0.2303207516670227,
+      "learning_rate": 0.0001847149696999655,
+      "loss": 1.6533,
+      "step": 27700
+    },
+    {
+      "epoch": 0.2301781810955819,
+      "grad_norm": 0.1961013227701187,
+      "learning_rate": 0.0001846597710908159,
+      "loss": 1.6287,
+      "step": 27800
+    },
+    {
+      "epoch": 0.23100616016427106,
+      "grad_norm": 0.29377955198287964,
+      "learning_rate": 0.0001846045724816663,
+      "loss": 1.6367,
+      "step": 27900
+    },
+    {
+      "epoch": 0.2318341392329602,
+      "grad_norm": 0.22385641932487488,
+      "learning_rate": 0.00018454937387251668,
+      "loss": 1.6553,
+      "step": 28000
+    },
+    {
+      "epoch": 0.23266211830164935,
+      "grad_norm": 0.23252160847187042,
+      "learning_rate": 0.00018449417526336708,
+      "loss": 1.6148,
+      "step": 28100
+    },
+    {
+      "epoch": 0.2334900973703385,
+      "grad_norm": 0.22680921852588654,
+      "learning_rate": 0.00018443897665421747,
+      "loss": 1.6208,
+      "step": 28200
+    },
+    {
+      "epoch": 0.23431807643902763,
+      "grad_norm": 0.2204786092042923,
+      "learning_rate": 0.00018438433003115936,
+      "loss": 1.6322,
+      "step": 28300
+    },
+    {
+      "epoch": 0.23514605550771678,
+      "grad_norm": 0.2748723030090332,
+      "learning_rate": 0.00018432913142200975,
+      "loss": 1.6382,
+      "step": 28400
+    },
+    {
+      "epoch": 0.23597403457640592,
+      "grad_norm": 0.21922816336154938,
+      "learning_rate": 0.00018427393281286017,
+      "loss": 1.6371,
+      "step": 28500
+    },
+    {
+      "epoch": 0.23680201364509504,
+      "grad_norm": 0.2469870150089264,
+      "learning_rate": 0.00018421873420371054,
+      "loss": 1.5986,
+      "step": 28600
+    },
+    {
+      "epoch": 0.23762999271378418,
+      "grad_norm": 0.24729487299919128,
+      "learning_rate": 0.00018416353559456094,
+      "loss": 1.6186,
+      "step": 28700
+    },
+    {
+      "epoch": 0.23845797178247333,
+      "grad_norm": 0.2277776002883911,
+      "learning_rate": 0.00018410833698541136,
+      "loss": 1.617,
+      "step": 28800
+    },
+    {
+      "epoch": 0.23928595085116247,
+      "grad_norm": 0.23853392899036407,
+      "learning_rate": 0.00018405313837626172,
+      "loss": 1.6358,
+      "step": 28900
+    },
+    {
+      "epoch": 0.24011392991985162,
+      "grad_norm": 0.22271224856376648,
+      "learning_rate": 0.00018399793976711212,
+      "loss": 1.6349,
+      "step": 29000
+    },
+    {
+      "epoch": 0.24094190898854076,
+      "grad_norm": 0.2435714304447174,
+      "learning_rate": 0.00018394274115796254,
+      "loss": 1.6673,
+      "step": 29100
+    },
+    {
+      "epoch": 0.2417698880572299,
+      "grad_norm": 0.2396841198205948,
+      "learning_rate": 0.0001838875425488129,
+      "loss": 1.641,
+      "step": 29200
+    },
+    {
+      "epoch": 0.24259786712591905,
+      "grad_norm": 0.26187509298324585,
+      "learning_rate": 0.0001838323439396633,
+      "loss": 1.6256,
+      "step": 29300
+    },
+    {
+      "epoch": 0.2434258461946082,
+      "grad_norm": 0.21034131944179535,
+      "learning_rate": 0.00018377714533051372,
+      "loss": 1.5857,
+      "step": 29400
+    },
+    {
+      "epoch": 0.24425382526329734,
+      "grad_norm": 0.23442912101745605,
+      "learning_rate": 0.0001837219467213641,
+      "loss": 1.5935,
+      "step": 29500
+    },
+    {
+      "epoch": 0.24508180433198648,
+      "grad_norm": 0.21707892417907715,
+      "learning_rate": 0.00018366674811221448,
+      "loss": 1.6391,
+      "step": 29600
+    },
+    {
+      "epoch": 0.24590978340067562,
+      "grad_norm": 0.2780430316925049,
+      "learning_rate": 0.0001836115495030649,
+      "loss": 1.6243,
+      "step": 29700
+    },
+    {
+      "epoch": 0.24673776246936477,
+      "grad_norm": 0.2786915898323059,
+      "learning_rate": 0.00018355635089391527,
+      "loss": 1.6023,
+      "step": 29800
+    },
+    {
+      "epoch": 0.2475657415380539,
+      "grad_norm": 0.23516952991485596,
+      "learning_rate": 0.00018350115228476567,
+      "loss": 1.6745,
+      "step": 29900
+    },
+    {
+      "epoch": 0.24839372060674306,
+      "grad_norm": 0.22414027154445648,
+      "learning_rate": 0.0001834459536756161,
+      "loss": 1.6293,
+      "step": 30000
+    },
+    {
+      "epoch": 0.2492216996754322,
+      "grad_norm": 0.2355353832244873,
+      "learning_rate": 0.00018339075506646645,
+      "loss": 1.6405,
+      "step": 30100
+    },
+    {
+      "epoch": 0.2500496787441214,
+      "grad_norm": 0.284574031829834,
+      "learning_rate": 0.00018333555645731685,
+      "loss": 1.6156,
+      "step": 30200
+    },
+    {
+      "epoch": 0.2508776578128105,
+      "grad_norm": 0.22318606078624725,
+      "learning_rate": 0.00018328035784816727,
+      "loss": 1.6259,
+      "step": 30300
+    },
+    {
+      "epoch": 0.25170563688149966,
+      "grad_norm": 0.18916834890842438,
+      "learning_rate": 0.00018322515923901764,
+      "loss": 1.6371,
+      "step": 30400
+    },
+    {
+      "epoch": 0.2525336159501888,
+      "grad_norm": 0.21935884654521942,
+      "learning_rate": 0.00018316996062986803,
+      "loss": 1.608,
+      "step": 30500
+    },
+    {
+      "epoch": 0.25336159501887795,
+      "grad_norm": 0.22668689489364624,
+      "learning_rate": 0.00018311476202071845,
+      "loss": 1.6319,
+      "step": 30600
+    },
+    {
+      "epoch": 0.2541895740875671,
+      "grad_norm": 0.2440369874238968,
+      "learning_rate": 0.00018305956341156882,
+      "loss": 1.6317,
+      "step": 30700
+    },
+    {
+      "epoch": 0.2550175531562562,
+      "grad_norm": 0.24555020034313202,
+      "learning_rate": 0.00018300436480241921,
+      "loss": 1.6545,
+      "step": 30800
+    },
+    {
+      "epoch": 0.2558455322249453,
+      "grad_norm": 0.24021874368190765,
+      "learning_rate": 0.00018294916619326964,
+      "loss": 1.6426,
+      "step": 30900
+    },
+    {
+      "epoch": 0.25667351129363447,
+      "grad_norm": 0.24666373431682587,
+      "learning_rate": 0.00018289451957021152,
+      "loss": 1.6185,
+      "step": 31000
+    },
+    {
+      "epoch": 0.2575014903623236,
+      "grad_norm": 0.33455362915992737,
+      "learning_rate": 0.00018283932096106192,
+      "loss": 1.63,
+      "step": 31100
+    },
+    {
+      "epoch": 0.25832946943101276,
+      "grad_norm": 0.22457076609134674,
+      "learning_rate": 0.00018278412235191228,
+      "loss": 1.6384,
+      "step": 31200
+    },
+    {
+      "epoch": 0.2591574484997019,
+      "grad_norm": 0.23992955684661865,
+      "learning_rate": 0.0001827289237427627,
+      "loss": 1.6101,
+      "step": 31300
+    },
+    {
+      "epoch": 0.25998542756839105,
+      "grad_norm": 0.23438423871994019,
+      "learning_rate": 0.0001826737251336131,
+      "loss": 1.6541,
+      "step": 31400
+    },
+    {
+      "epoch": 0.2608134066370802,
+      "grad_norm": 0.2723533511161804,
+      "learning_rate": 0.00018261852652446347,
+      "loss": 1.6466,
+      "step": 31500
+    },
+    {
+      "epoch": 0.26164138570576934,
+      "grad_norm": 0.284927099943161,
+      "learning_rate": 0.0001825633279153139,
+      "loss": 1.6252,
+      "step": 31600
+    },
+    {
+      "epoch": 0.2624693647744585,
+      "grad_norm": 0.2600990831851959,
+      "learning_rate": 0.00018250812930616428,
+      "loss": 1.6156,
+      "step": 31700
+    },
+    {
+      "epoch": 0.2632973438431476,
+      "grad_norm": 0.2504718005657196,
+      "learning_rate": 0.00018245293069701465,
+      "loss": 1.6356,
+      "step": 31800
+    },
+    {
+      "epoch": 0.26412532291183677,
+      "grad_norm": 0.2585621476173401,
+      "learning_rate": 0.00018239773208786507,
+      "loss": 1.6444,
+      "step": 31900
+    },
+    {
+      "epoch": 0.2649533019805259,
+      "grad_norm": 0.2442982792854309,
+      "learning_rate": 0.00018234253347871546,
+      "loss": 1.6647,
+      "step": 32000
+    },
+    {
+      "epoch": 0.26578128104921506,
+      "grad_norm": 0.23902744054794312,
+      "learning_rate": 0.00018228733486956583,
+      "loss": 1.6553,
+      "step": 32100
+    },
+    {
+      "epoch": 0.2666092601179042,
+      "grad_norm": 0.2639334499835968,
+      "learning_rate": 0.00018223213626041625,
+      "loss": 1.6324,
+      "step": 32200
+    },
+    {
+      "epoch": 0.26743723918659335,
+      "grad_norm": 0.21691977977752686,
+      "learning_rate": 0.00018217693765126665,
+      "loss": 1.6354,
+      "step": 32300
+    },
+    {
+      "epoch": 0.2682652182552825,
+      "grad_norm": 0.22446656227111816,
+      "learning_rate": 0.00018212173904211701,
+      "loss": 1.6549,
+      "step": 32400
+    },
+    {
+      "epoch": 0.26909319732397163,
+      "grad_norm": 0.23378610610961914,
+      "learning_rate": 0.00018206654043296744,
+      "loss": 1.6331,
+      "step": 32500
+    },
+    {
+      "epoch": 0.2699211763926608,
+      "grad_norm": 0.31450459361076355,
+      "learning_rate": 0.00018201134182381783,
+      "loss": 1.6313,
+      "step": 32600
+    },
+    {
+      "epoch": 0.2707491554613499,
+      "grad_norm": 0.22139056026935577,
+      "learning_rate": 0.0001819561432146682,
+      "loss": 1.6464,
+      "step": 32700
+    },
+    {
+      "epoch": 0.27157713453003907,
+      "grad_norm": 0.2225482016801834,
+      "learning_rate": 0.00018190094460551862,
+      "loss": 1.648,
+      "step": 32800
+    },
+    {
+      "epoch": 0.2724051135987282,
+      "grad_norm": 0.26663103699684143,
+      "learning_rate": 0.000181845745996369,
+      "loss": 1.6397,
+      "step": 32900
+    },
+    {
+      "epoch": 0.27323309266741735,
+      "grad_norm": 0.2755584716796875,
+      "learning_rate": 0.0001817910993733109,
+      "loss": 1.6449,
+      "step": 33000
+    },
+    {
+      "epoch": 0.2740610717361065,
+      "grad_norm": 0.25144413113594055,
+      "learning_rate": 0.0001817359007641613,
+      "loss": 1.6593,
+      "step": 33100
+    },
+    {
+      "epoch": 0.27488905080479564,
+      "grad_norm": 0.23322436213493347,
+      "learning_rate": 0.00018168070215501166,
+      "loss": 1.6178,
+      "step": 33200
+    },
+    {
+      "epoch": 0.2757170298734848,
+      "grad_norm": 0.23854686319828033,
+      "learning_rate": 0.00018162550354586208,
+      "loss": 1.6349,
+      "step": 33300
+    },
+    {
+      "epoch": 0.27654500894217393,
+      "grad_norm": 0.27168381214141846,
+      "learning_rate": 0.00018157030493671248,
+      "loss": 1.6077,
+      "step": 33400
+    },
+    {
+      "epoch": 0.2773729880108631,
+      "grad_norm": 0.2243831604719162,
+      "learning_rate": 0.00018151510632756284,
+      "loss": 1.6365,
+      "step": 33500
+    },
+    {
+      "epoch": 0.2782009670795522,
+      "grad_norm": 0.2275819331407547,
+      "learning_rate": 0.00018145990771841326,
+      "loss": 1.6231,
+      "step": 33600
+    },
+    {
+      "epoch": 0.27902894614824136,
+      "grad_norm": 0.2547686994075775,
+      "learning_rate": 0.00018140470910926366,
+      "loss": 1.6246,
+      "step": 33700
+    },
+    {
+      "epoch": 0.2798569252169305,
+      "grad_norm": 0.299843430519104,
+      "learning_rate": 0.00018134951050011405,
+      "loss": 1.6258,
+      "step": 33800
+    },
+    {
+      "epoch": 0.28068490428561965,
+      "grad_norm": 0.21927513182163239,
+      "learning_rate": 0.00018129431189096445,
+      "loss": 1.6332,
+      "step": 33900
+    },
+    {
+      "epoch": 0.2815128833543088,
+      "grad_norm": 0.25713977217674255,
+      "learning_rate": 0.00018123911328181484,
+      "loss": 1.6369,
+      "step": 34000
+    },
+    {
+      "epoch": 0.28234086242299794,
+      "grad_norm": 0.20918399095535278,
+      "learning_rate": 0.00018118391467266524,
+      "loss": 1.6009,
+      "step": 34100
+    },
+    {
+      "epoch": 0.2831688414916871,
+      "grad_norm": 0.2372296303510666,
+      "learning_rate": 0.00018112871606351563,
+      "loss": 1.6264,
+      "step": 34200
+    },
+    {
+      "epoch": 0.28399682056037623,
+      "grad_norm": 0.20098775625228882,
+      "learning_rate": 0.00018107351745436602,
+      "loss": 1.6386,
+      "step": 34300
+    },
+    {
+      "epoch": 0.2848247996290654,
+      "grad_norm": 0.26937851309776306,
+      "learning_rate": 0.00018101831884521642,
+      "loss": 1.6269,
+      "step": 34400
+    },
+    {
+      "epoch": 0.2856527786977545,
+      "grad_norm": 0.2214965969324112,
+      "learning_rate": 0.0001809631202360668,
+      "loss": 1.6395,
+      "step": 34500
+    },
+    {
+      "epoch": 0.28648075776644366,
+      "grad_norm": 0.21897448599338531,
+      "learning_rate": 0.0001809079216269172,
+      "loss": 1.5864,
+      "step": 34600
+    },
+    {
+      "epoch": 0.2873087368351328,
+      "grad_norm": 0.21076589822769165,
+      "learning_rate": 0.0001808527230177676,
+      "loss": 1.598,
+      "step": 34700
+    },
+    {
+      "epoch": 0.28813671590382195,
+      "grad_norm": 0.2242691069841385,
+      "learning_rate": 0.000180797524408618,
+      "loss": 1.6202,
+      "step": 34800
+    },
+    {
+      "epoch": 0.2889646949725111,
+      "grad_norm": 0.21810419857501984,
+      "learning_rate": 0.0001807423257994684,
+      "loss": 1.6345,
+      "step": 34900
+    },
+    {
+      "epoch": 0.28979267404120024,
+      "grad_norm": 0.274824857711792,
+      "learning_rate": 0.00018068712719031878,
+      "loss": 1.626,
+      "step": 35000
+    },
+    {
+      "epoch": 0.2906206531098894,
+      "grad_norm": 0.23478242754936218,
+      "learning_rate": 0.00018063192858116918,
+      "loss": 1.6346,
+      "step": 35100
+    },
+    {
+      "epoch": 0.2914486321785785,
+      "grad_norm": 0.24421899020671844,
+      "learning_rate": 0.00018057672997201957,
+      "loss": 1.6243,
+      "step": 35200
+    },
+    {
+      "epoch": 0.29227661124726767,
+      "grad_norm": 0.28101858496665955,
+      "learning_rate": 0.00018052153136286997,
+      "loss": 1.619,
+      "step": 35300
+    },
+    {
+      "epoch": 0.2931045903159568,
+      "grad_norm": 0.24983233213424683,
+      "learning_rate": 0.00018046633275372036,
+      "loss": 1.621,
+      "step": 35400
+    },
+    {
+      "epoch": 0.29393256938464596,
+      "grad_norm": 0.3039966821670532,
+      "learning_rate": 0.00018041113414457075,
+      "loss": 1.5755,
+      "step": 35500
+    },
+    {
+      "epoch": 0.2947605484533351,
+      "grad_norm": 0.23376306891441345,
+      "learning_rate": 0.00018035593553542115,
+      "loss": 1.6419,
+      "step": 35600
+    },
+    {
+      "epoch": 0.29558852752202425,
+      "grad_norm": 0.2499234825372696,
+      "learning_rate": 0.00018030073692627154,
+      "loss": 1.634,
+      "step": 35700
+    },
+    {
+      "epoch": 0.2964165065907134,
+      "grad_norm": 0.24096137285232544,
+      "learning_rate": 0.00018024553831712194,
+      "loss": 1.621,
+      "step": 35800
+    },
+    {
+      "epoch": 0.29724448565940254,
+      "grad_norm": 0.19286566972732544,
+      "learning_rate": 0.00018019033970797233,
+      "loss": 1.6477,
+      "step": 35900
+    },
+    {
+      "epoch": 0.2980724647280917,
+      "grad_norm": 0.22015613317489624,
+      "learning_rate": 0.00018013514109882273,
+      "loss": 1.6364,
+      "step": 36000
+    },
+    {
+      "epoch": 0.2989004437967808,
+      "grad_norm": 0.2441215068101883,
+      "learning_rate": 0.00018008049447576464,
+      "loss": 1.6261,
+      "step": 36100
+    },
+    {
+      "epoch": 0.29972842286546997,
+      "grad_norm": 0.24828499555587769,
+      "learning_rate": 0.000180025295866615,
+      "loss": 1.6095,
+      "step": 36200
+    },
+    {
+      "epoch": 0.3005564019341591,
+      "grad_norm": 0.27119019627571106,
+      "learning_rate": 0.0001799700972574654,
+      "loss": 1.617,
+      "step": 36300
+    },
+    {
+      "epoch": 0.30138438100284826,
+      "grad_norm": 0.33090922236442566,
+      "learning_rate": 0.00017991489864831582,
+      "loss": 1.6395,
+      "step": 36400
+    },
+    {
+      "epoch": 0.3022123600715374,
+      "grad_norm": 0.25472140312194824,
+      "learning_rate": 0.0001798597000391662,
+      "loss": 1.6586,
+      "step": 36500
+    },
+    {
+      "epoch": 0.30304033914022654,
+      "grad_norm": 0.24048671126365662,
+      "learning_rate": 0.00017980450143001658,
+      "loss": 1.6378,
+      "step": 36600
+    },
+    {
+      "epoch": 0.3038683182089157,
+      "grad_norm": 0.25020936131477356,
+      "learning_rate": 0.000179749302820867,
+      "loss": 1.6086,
+      "step": 36700
+    },
+    {
+      "epoch": 0.30469629727760483,
+      "grad_norm": 0.2425384819507599,
+      "learning_rate": 0.00017969410421171737,
+      "loss": 1.6332,
+      "step": 36800
+    },
+    {
+      "epoch": 0.305524276346294,
+      "grad_norm": 0.24413828551769257,
+      "learning_rate": 0.00017963890560256777,
+      "loss": 1.637,
+      "step": 36900
+    },
+    {
+      "epoch": 0.3063522554149831,
+      "grad_norm": 0.24942953884601593,
+      "learning_rate": 0.0001795837069934182,
+      "loss": 1.6016,
+      "step": 37000
+    },
+    {
+      "epoch": 0.30718023448367227,
+      "grad_norm": 0.2701806426048279,
+      "learning_rate": 0.00017952850838426855,
+      "loss": 1.6413,
+      "step": 37100
+    },
+    {
+      "epoch": 0.3080082135523614,
+      "grad_norm": 0.23709261417388916,
+      "learning_rate": 0.00017947330977511895,
+      "loss": 1.5924,
+      "step": 37200
+    },
+    {
+      "epoch": 0.30883619262105055,
+      "grad_norm": 0.252654105424881,
+      "learning_rate": 0.00017941811116596937,
+      "loss": 1.6229,
+      "step": 37300
+    },
+    {
+      "epoch": 0.3096641716897397,
+      "grad_norm": 0.27790337800979614,
+      "learning_rate": 0.00017936291255681974,
+      "loss": 1.62,
+      "step": 37400
+    },
+    {
+      "epoch": 0.31049215075842884,
+      "grad_norm": 0.25286659598350525,
+      "learning_rate": 0.00017930771394767013,
+      "loss": 1.6305,
+      "step": 37500
+    },
+    {
+      "epoch": 0.311320129827118,
+      "grad_norm": 0.24629120528697968,
+      "learning_rate": 0.00017925251533852055,
+      "loss": 1.637,
+      "step": 37600
+    },
+    {
+      "epoch": 0.31214810889580713,
+      "grad_norm": 0.26566171646118164,
+      "learning_rate": 0.00017919731672937092,
+      "loss": 1.6333,
+      "step": 37700
+    },
+    {
+      "epoch": 0.3129760879644963,
+      "grad_norm": 0.22510451078414917,
+      "learning_rate": 0.00017914211812022131,
+      "loss": 1.6408,
+      "step": 37800
+    },
+    {
+      "epoch": 0.3138040670331854,
+      "grad_norm": 0.27187344431877136,
+      "learning_rate": 0.00017908691951107174,
+      "loss": 1.633,
+      "step": 37900
+    },
+    {
+      "epoch": 0.31463204610187456,
+      "grad_norm": 0.29819074273109436,
+      "learning_rate": 0.0001790317209019221,
+      "loss": 1.6448,
+      "step": 38000
+    },
+    {
+      "epoch": 0.3154600251705637,
+      "grad_norm": 0.33122682571411133,
+      "learning_rate": 0.0001789765222927725,
+      "loss": 1.583,
+      "step": 38100
+    },
+    {
+      "epoch": 0.31628800423925285,
+      "grad_norm": 0.24350272119045258,
+      "learning_rate": 0.00017892132368362292,
+      "loss": 1.5941,
+      "step": 38200
+    },
+    {
+      "epoch": 0.317115983307942,
+      "grad_norm": 0.2509153485298157,
+      "learning_rate": 0.00017886612507447328,
+      "loss": 1.6332,
+      "step": 38300
+    },
+    {
+      "epoch": 0.31794396237663114,
+      "grad_norm": 0.25258669257164,
+      "learning_rate": 0.00017881092646532368,
+      "loss": 1.6157,
+      "step": 38400
+    },
+    {
+      "epoch": 0.3187719414453203,
+      "grad_norm": 0.31328025460243225,
+      "learning_rate": 0.0001787557278561741,
+      "loss": 1.619,
+      "step": 38500
+    },
+    {
+      "epoch": 0.31959992051400943,
+      "grad_norm": 0.24107754230499268,
+      "learning_rate": 0.00017870052924702447,
+      "loss": 1.666,
+      "step": 38600
+    },
+    {
+      "epoch": 0.32042789958269857,
+      "grad_norm": 0.19994518160820007,
+      "learning_rate": 0.00017864533063787486,
+      "loss": 1.5895,
+      "step": 38700
+    },
+    {
+      "epoch": 0.3212558786513877,
+      "grad_norm": 0.25987759232521057,
+      "learning_rate": 0.00017859013202872528,
+      "loss": 1.6075,
+      "step": 38800
+    },
+    {
+      "epoch": 0.32208385772007686,
+      "grad_norm": 0.23311209678649902,
+      "learning_rate": 0.00017853493341957565,
+      "loss": 1.5872,
+      "step": 38900
+    },
+    {
+      "epoch": 0.322911836788766,
+      "grad_norm": 0.2703259289264679,
+      "learning_rate": 0.00017847973481042604,
+      "loss": 1.6011,
+      "step": 39000
+    },
+    {
+      "epoch": 0.32373981585745515,
+      "grad_norm": 0.21432837843894958,
+      "learning_rate": 0.00017842453620127647,
+      "loss": 1.6239,
+      "step": 39100
+    },
+    {
+      "epoch": 0.3245677949261443,
+      "grad_norm": 0.2730528712272644,
+      "learning_rate": 0.00017836988957821835,
+      "loss": 1.6453,
+      "step": 39200
+    },
+    {
+      "epoch": 0.32539577399483344,
+      "grad_norm": 0.25443777441978455,
+      "learning_rate": 0.00017831469096906875,
+      "loss": 1.5818,
+      "step": 39300
+    },
+    {
+      "epoch": 0.3262237530635226,
+      "grad_norm": 0.23115037381649017,
+      "learning_rate": 0.0001782594923599191,
+      "loss": 1.6373,
+      "step": 39400
+    },
+    {
+      "epoch": 0.32705173213221167,
+      "grad_norm": 0.232584148645401,
+      "learning_rate": 0.00017820429375076953,
+      "loss": 1.6154,
+      "step": 39500
+    },
+    {
+      "epoch": 0.3278797112009008,
+      "grad_norm": 0.2370355725288391,
+      "learning_rate": 0.00017814909514161993,
+      "loss": 1.6214,
+      "step": 39600
+    },
+    {
+      "epoch": 0.32870769026958996,
+      "grad_norm": 0.2408798784017563,
+      "learning_rate": 0.0001780938965324703,
+      "loss": 1.6338,
+      "step": 39700
+    },
+    {
+      "epoch": 0.3295356693382791,
+      "grad_norm": 0.2674672603607178,
+      "learning_rate": 0.00017803869792332072,
+      "loss": 1.6144,
+      "step": 39800
+    },
+    {
+      "epoch": 0.33036364840696825,
+      "grad_norm": 0.2323506474494934,
+      "learning_rate": 0.0001779834993141711,
+      "loss": 1.642,
+      "step": 39900
+    },
+    {
+      "epoch": 0.3311916274756574,
+      "grad_norm": 0.22446438670158386,
+      "learning_rate": 0.00017792830070502148,
+      "loss": 1.6089,
+      "step": 40000
+    },
+    {
+      "epoch": 0.33201960654434653,
+      "grad_norm": 0.22976459562778473,
+      "learning_rate": 0.0001778731020958719,
+      "loss": 1.6249,
+      "step": 40100
+    },
+    {
+      "epoch": 0.3328475856130357,
+      "grad_norm": 0.2459111511707306,
+      "learning_rate": 0.0001778179034867223,
+      "loss": 1.6212,
+      "step": 40200
+    },
+    {
+      "epoch": 0.3336755646817248,
+      "grad_norm": 0.2351800501346588,
+      "learning_rate": 0.00017776270487757266,
+      "loss": 1.645,
+      "step": 40300
+    },
+    {
+      "epoch": 0.33450354375041397,
+      "grad_norm": 0.1990719586610794,
+      "learning_rate": 0.00017770750626842308,
+      "loss": 1.6108,
+      "step": 40400
+    },
+    {
+      "epoch": 0.3353315228191031,
+      "grad_norm": 0.26577112078666687,
+      "learning_rate": 0.00017765230765927348,
+      "loss": 1.6433,
+      "step": 40500
+    },
+    {
+      "epoch": 0.33615950188779226,
+      "grad_norm": 0.24102246761322021,
+      "learning_rate": 0.00017759710905012384,
+      "loss": 1.6177,
+      "step": 40600
+    },
+    {
+      "epoch": 0.3369874809564814,
+      "grad_norm": 0.22635315358638763,
+      "learning_rate": 0.00017754191044097427,
+      "loss": 1.6226,
+      "step": 40700
+    },
+    {
+      "epoch": 0.33781546002517054,
+      "grad_norm": 0.24064379930496216,
+      "learning_rate": 0.00017748671183182466,
+      "loss": 1.6408,
+      "step": 40800
+    },
+    {
+      "epoch": 0.3386434390938597,
+      "grad_norm": 0.2781156897544861,
+      "learning_rate": 0.00017743151322267503,
+      "loss": 1.6081,
+      "step": 40900
+    },
+    {
+      "epoch": 0.33947141816254883,
+      "grad_norm": 0.21876955032348633,
+      "learning_rate": 0.00017737631461352545,
+      "loss": 1.6409,
+      "step": 41000
+    },
+    {
+      "epoch": 0.340299397231238,
+      "grad_norm": 0.2646994888782501,
+      "learning_rate": 0.00017732111600437584,
+      "loss": 1.6566,
+      "step": 41100
+    },
+    {
+      "epoch": 0.3411273762999271,
+      "grad_norm": 0.22703330218791962,
+      "learning_rate": 0.0001772659173952262,
+      "loss": 1.6322,
+      "step": 41200
+    },
+    {
+      "epoch": 0.34195535536861627,
+      "grad_norm": 0.22421209514141083,
+      "learning_rate": 0.00017721071878607663,
+      "loss": 1.62,
+      "step": 41300
+    },
+    {
+      "epoch": 0.3427833344373054,
+      "grad_norm": 0.6235564947128296,
+      "learning_rate": 0.00017715552017692703,
+      "loss": 1.6388,
+      "step": 41400
+    },
+    {
+      "epoch": 0.34361131350599455,
+      "grad_norm": 0.2891361117362976,
+      "learning_rate": 0.0001771003215677774,
+      "loss": 1.6348,
+      "step": 41500
+    },
+    {
+      "epoch": 0.3444392925746837,
+      "grad_norm": 0.2454879879951477,
+      "learning_rate": 0.00017704512295862781,
+      "loss": 1.6338,
+      "step": 41600
+    },
+    {
+      "epoch": 0.34526727164337284,
+      "grad_norm": 0.39638543128967285,
+      "learning_rate": 0.0001769899243494782,
+      "loss": 1.6329,
+      "step": 41700
+    },
+    {
+      "epoch": 0.346095250712062,
+      "grad_norm": 0.24618402123451233,
+      "learning_rate": 0.00017693472574032858,
+      "loss": 1.6304,
+      "step": 41800
+    },
+    {
+      "epoch": 0.34692322978075113,
+      "grad_norm": 0.22741757333278656,
+      "learning_rate": 0.000176879527131179,
+      "loss": 1.6514,
+      "step": 41900
+    },
+    {
+      "epoch": 0.3477512088494403,
+      "grad_norm": 0.2486487478017807,
+      "learning_rate": 0.00017682488050812088,
+      "loss": 1.6365,
+      "step": 42000
+    },
+    {
+      "epoch": 0.3485791879181294,
+      "grad_norm": 0.26512429118156433,
+      "learning_rate": 0.00017676968189897128,
+      "loss": 1.66,
+      "step": 42100
+    },
+    {
+      "epoch": 0.34940716698681856,
+      "grad_norm": 0.24919961392879486,
+      "learning_rate": 0.00017671448328982167,
+      "loss": 1.6216,
+      "step": 42200
+    },
+    {
+      "epoch": 0.3502351460555077,
+      "grad_norm": 0.24012117087841034,
+      "learning_rate": 0.00017665928468067207,
+      "loss": 1.6522,
+      "step": 42300
+    },
+    {
+      "epoch": 0.35106312512419685,
+      "grad_norm": 0.231460839509964,
+      "learning_rate": 0.00017660408607152246,
+      "loss": 1.6249,
+      "step": 42400
+    },
+    {
+      "epoch": 0.351891104192886,
+      "grad_norm": 0.28340089321136475,
+      "learning_rate": 0.00017654888746237285,
+      "loss": 1.6198,
+      "step": 42500
+    },
+    {
+      "epoch": 0.35271908326157514,
+      "grad_norm": 0.2706835865974426,
+      "learning_rate": 0.00017649368885322325,
+      "loss": 1.6292,
+      "step": 42600
+    },
+    {
+      "epoch": 0.3535470623302643,
+      "grad_norm": 0.22307871282100677,
+      "learning_rate": 0.00017643849024407364,
+      "loss": 1.6236,
+      "step": 42700
+    },
+    {
+      "epoch": 0.3543750413989534,
+      "grad_norm": 0.2504919767379761,
+      "learning_rate": 0.00017638384362101556,
+      "loss": 1.625,
+      "step": 42800
+    },
+    {
+      "epoch": 0.35520302046764257,
+      "grad_norm": 0.2657232880592346,
+      "learning_rate": 0.00017632864501186592,
+      "loss": 1.6394,
+      "step": 42900
+    },
+    {
+      "epoch": 0.3560309995363317,
+      "grad_norm": 0.2763049304485321,
+      "learning_rate": 0.00017627344640271632,
+      "loss": 1.6491,
+      "step": 43000
+    },
+    {
+      "epoch": 0.35685897860502086,
+      "grad_norm": 0.2219325453042984,
+      "learning_rate": 0.00017621824779356674,
+      "loss": 1.6078,
+      "step": 43100
+    },
+    {
+      "epoch": 0.35768695767371,
+      "grad_norm": 0.24751123785972595,
+      "learning_rate": 0.0001761630491844171,
+      "loss": 1.6286,
+      "step": 43200
+    },
+    {
+      "epoch": 0.35851493674239915,
+      "grad_norm": 0.2798210382461548,
+      "learning_rate": 0.0001761078505752675,
+      "loss": 1.6157,
+      "step": 43300
+    },
+    {
+      "epoch": 0.3593429158110883,
+      "grad_norm": 0.2589050829410553,
+      "learning_rate": 0.00017605265196611792,
+      "loss": 1.6419,
+      "step": 43400
+    },
+    {
+      "epoch": 0.36017089487977744,
+      "grad_norm": 0.24586708843708038,
+      "learning_rate": 0.0001759974533569683,
+      "loss": 1.6156,
+      "step": 43500
+    },
+    {
+      "epoch": 0.3609988739484666,
+      "grad_norm": 0.2211756557226181,
+      "learning_rate": 0.00017594225474781868,
+      "loss": 1.6491,
+      "step": 43600
+    },
+    {
+      "epoch": 0.3618268530171557,
+      "grad_norm": 0.47307145595550537,
+      "learning_rate": 0.0001758870561386691,
+      "loss": 1.66,
+      "step": 43700
+    },
+    {
+      "epoch": 0.36265483208584487,
+      "grad_norm": 0.260375440120697,
+      "learning_rate": 0.00017583185752951947,
+      "loss": 1.6352,
+      "step": 43800
+    },
+    {
+      "epoch": 0.363482811154534,
+      "grad_norm": 0.28375470638275146,
+      "learning_rate": 0.00017577665892036987,
+      "loss": 1.6187,
+      "step": 43900
+    },
+    {
+      "epoch": 0.36431079022322316,
+      "grad_norm": 0.27511486411094666,
+      "learning_rate": 0.00017572146031122029,
+      "loss": 1.6423,
+      "step": 44000
+    },
+    {
+      "epoch": 0.3651387692919123,
+      "grad_norm": 0.32292941212654114,
+      "learning_rate": 0.00017566626170207065,
+      "loss": 1.6655,
+      "step": 44100
+    },
+    {
+      "epoch": 0.36596674836060145,
+      "grad_norm": 0.2551015019416809,
+      "learning_rate": 0.00017561106309292105,
+      "loss": 1.6541,
+      "step": 44200
+    },
+    {
+      "epoch": 0.3667947274292906,
+      "grad_norm": 0.24115487933158875,
+      "learning_rate": 0.00017555586448377147,
+      "loss": 1.6233,
+      "step": 44300
+    },
+    {
+      "epoch": 0.36762270649797973,
+      "grad_norm": 0.2830246388912201,
+      "learning_rate": 0.00017550066587462184,
+      "loss": 1.6267,
+      "step": 44400
+    },
+    {
+      "epoch": 0.3684506855666689,
+      "grad_norm": 0.20589886605739594,
+      "learning_rate": 0.00017544546726547223,
+      "loss": 1.6542,
+      "step": 44500
+    },
+    {
+      "epoch": 0.369278664635358,
+      "grad_norm": 0.24481356143951416,
+      "learning_rate": 0.00017539026865632265,
+      "loss": 1.625,
+      "step": 44600
+    },
+    {
+      "epoch": 0.37010664370404717,
+      "grad_norm": 0.31838709115982056,
+      "learning_rate": 0.00017533507004717302,
+      "loss": 1.593,
+      "step": 44700
+    },
+    {
+      "epoch": 0.3709346227727363,
+      "grad_norm": 0.23912674188613892,
+      "learning_rate": 0.0001752798714380234,
+      "loss": 1.6169,
+      "step": 44800
+    },
+    {
+      "epoch": 0.37176260184142546,
+      "grad_norm": 0.22365044057369232,
+      "learning_rate": 0.00017522467282887383,
+      "loss": 1.6472,
+      "step": 44900
+    },
+    {
+      "epoch": 0.3725905809101146,
+      "grad_norm": 0.26311957836151123,
+      "learning_rate": 0.0001751694742197242,
+      "loss": 1.6257,
+      "step": 45000
+    },
+    {
+      "epoch": 0.37341855997880374,
+      "grad_norm": 0.25842225551605225,
+      "learning_rate": 0.0001751142756105746,
+      "loss": 1.6422,
+      "step": 45100
+    },
+    {
+      "epoch": 0.3742465390474929,
+      "grad_norm": 0.233975350856781,
+      "learning_rate": 0.00017505907700142502,
+      "loss": 1.6528,
+      "step": 45200
+    },
+    {
+      "epoch": 0.37507451811618203,
+      "grad_norm": 0.2946456968784332,
+      "learning_rate": 0.00017500387839227538,
+      "loss": 1.6038,
+      "step": 45300
+    },
+    {
+      "epoch": 0.3759024971848712,
+      "grad_norm": 0.28840962052345276,
+      "learning_rate": 0.00017494867978312578,
+      "loss": 1.6088,
+      "step": 45400
+    },
+    {
+      "epoch": 0.3767304762535603,
+      "grad_norm": 0.2456226348876953,
+      "learning_rate": 0.0001748934811739762,
+      "loss": 1.6323,
+      "step": 45500
+    },
+    {
+      "epoch": 0.37755845532224946,
+      "grad_norm": 0.22499750554561615,
+      "learning_rate": 0.00017483828256482657,
+      "loss": 1.6261,
+      "step": 45600
+    },
+    {
+      "epoch": 0.3783864343909386,
+      "grad_norm": 0.2697817087173462,
+      "learning_rate": 0.00017478308395567696,
+      "loss": 1.6062,
+      "step": 45700
+    },
+    {
+      "epoch": 0.37921441345962775,
+      "grad_norm": 0.21315142512321472,
+      "learning_rate": 0.00017472788534652738,
+      "loss": 1.6245,
+      "step": 45800
+    },
+    {
+      "epoch": 0.3800423925283169,
+      "grad_norm": 0.20391778647899628,
+      "learning_rate": 0.00017467268673737775,
+      "loss": 1.6293,
+      "step": 45900
+    },
+    {
+      "epoch": 0.38087037159700604,
+      "grad_norm": 0.30484962463378906,
+      "learning_rate": 0.00017461748812822814,
+      "loss": 1.6034,
+      "step": 46000
+    },
+    {
+      "epoch": 0.3816983506656952,
+      "grad_norm": 0.2746950089931488,
+      "learning_rate": 0.00017456228951907857,
+      "loss": 1.6382,
+      "step": 46100
+    },
+    {
+      "epoch": 0.38252632973438433,
+      "grad_norm": 0.2682898938655853,
+      "learning_rate": 0.00017450709090992893,
+      "loss": 1.6189,
+      "step": 46200
+    },
+    {
+      "epoch": 0.3833543088030735,
+      "grad_norm": 0.23377636075019836,
+      "learning_rate": 0.00017445189230077935,
+      "loss": 1.5986,
+      "step": 46300
+    },
+    {
+      "epoch": 0.3841822878717626,
+      "grad_norm": 0.285535603761673,
+      "learning_rate": 0.00017439669369162975,
+      "loss": 1.6005,
+      "step": 46400
+    },
+    {
+      "epoch": 0.38501026694045176,
+      "grad_norm": 0.25021353363990784,
+      "learning_rate": 0.00017434149508248012,
+      "loss": 1.6028,
+      "step": 46500
+    },
+    {
+      "epoch": 0.3858382460091409,
+      "grad_norm": 0.23996229469776154,
+      "learning_rate": 0.00017428629647333054,
+      "loss": 1.6265,
+      "step": 46600
+    },
+    {
+      "epoch": 0.38666622507783005,
+      "grad_norm": 0.22236377000808716,
+      "learning_rate": 0.00017423109786418093,
+      "loss": 1.6285,
+      "step": 46700
+    },
+    {
+      "epoch": 0.3874942041465192,
+      "grad_norm": 0.3048860728740692,
+      "learning_rate": 0.0001741758992550313,
+      "loss": 1.6503,
+      "step": 46800
+    },
+    {
+      "epoch": 0.38832218321520834,
+      "grad_norm": 0.2737395167350769,
+      "learning_rate": 0.00017412070064588172,
+      "loss": 1.6136,
+      "step": 46900
+    },
+    {
+      "epoch": 0.3891501622838975,
+      "grad_norm": 0.26264122128486633,
+      "learning_rate": 0.0001740655020367321,
+      "loss": 1.6407,
+      "step": 47000
+    },
+    {
+      "epoch": 0.3899781413525866,
+      "grad_norm": 0.2572570741176605,
+      "learning_rate": 0.00017401030342758248,
+      "loss": 1.659,
+      "step": 47100
+    },
+    {
+      "epoch": 0.39080612042127577,
+      "grad_norm": 0.23144544661045074,
+      "learning_rate": 0.0001739556568045244,
+      "loss": 1.6108,
+      "step": 47200
+    },
+    {
+      "epoch": 0.3916340994899649,
+      "grad_norm": 0.2429717630147934,
+      "learning_rate": 0.00017390045819537476,
+      "loss": 1.6062,
+      "step": 47300
+    },
+    {
+      "epoch": 0.39246207855865406,
+      "grad_norm": 0.23754027485847473,
+      "learning_rate": 0.00017384525958622518,
+      "loss": 1.6677,
+      "step": 47400
+    },
+    {
+      "epoch": 0.3932900576273432,
+      "grad_norm": 0.23894472420215607,
+      "learning_rate": 0.00017379006097707558,
+      "loss": 1.5879,
+      "step": 47500
+    },
+    {
+      "epoch": 0.39411803669603235,
+      "grad_norm": 0.2626464366912842,
+      "learning_rate": 0.00017373486236792594,
+      "loss": 1.6149,
+      "step": 47600
+    },
+    {
+      "epoch": 0.3949460157647215,
+      "grad_norm": 0.22693473100662231,
+      "learning_rate": 0.00017367966375877637,
+      "loss": 1.6335,
+      "step": 47700
+    },
+    {
+      "epoch": 0.39577399483341064,
+      "grad_norm": 0.250009149312973,
+      "learning_rate": 0.00017362446514962676,
+      "loss": 1.61,
+      "step": 47800
+    },
+    {
+      "epoch": 0.3966019739020998,
+      "grad_norm": 0.2577686309814453,
+      "learning_rate": 0.00017356926654047713,
+      "loss": 1.6171,
+      "step": 47900
+    },
+    {
+      "epoch": 0.3974299529707889,
+      "grad_norm": 0.2702758312225342,
+      "learning_rate": 0.00017351406793132755,
+      "loss": 1.6016,
+      "step": 48000
+    },
+    {
+      "epoch": 0.39825793203947807,
+      "grad_norm": 0.24689382314682007,
+      "learning_rate": 0.00017345886932217794,
+      "loss": 1.6545,
+      "step": 48100
+    },
+    {
+      "epoch": 0.3990859111081672,
+      "grad_norm": 0.29933419823646545,
+      "learning_rate": 0.0001734036707130283,
+      "loss": 1.612,
+      "step": 48200
+    },
+    {
+      "epoch": 0.3999138901768563,
+      "grad_norm": 0.2971803545951843,
+      "learning_rate": 0.00017334847210387873,
+      "loss": 1.6364,
+      "step": 48300
+    },
+    {
+      "epoch": 0.40074186924554545,
+      "grad_norm": 0.2572464048862457,
+      "learning_rate": 0.00017329327349472912,
+      "loss": 1.6386,
+      "step": 48400
+    },
+    {
+      "epoch": 0.4015698483142346,
+      "grad_norm": 0.23949342966079712,
+      "learning_rate": 0.0001732380748855795,
+      "loss": 1.6038,
+      "step": 48500
+    },
+    {
+      "epoch": 0.40239782738292373,
+      "grad_norm": 0.3574795424938202,
+      "learning_rate": 0.0001731828762764299,
+      "loss": 1.6472,
+      "step": 48600
+    },
+    {
+      "epoch": 0.4032258064516129,
+      "grad_norm": 0.21243545413017273,
+      "learning_rate": 0.0001731276776672803,
+      "loss": 1.6144,
+      "step": 48700
+    },
+    {
+      "epoch": 0.404053785520302,
+      "grad_norm": 0.237880140542984,
+      "learning_rate": 0.00017307247905813067,
+      "loss": 1.6422,
+      "step": 48800
+    },
+    {
+      "epoch": 0.40488176458899117,
+      "grad_norm": 0.2700289189815521,
+      "learning_rate": 0.0001730172804489811,
+      "loss": 1.6329,
+      "step": 48900
+    },
+    {
+      "epoch": 0.4057097436576803,
+      "grad_norm": 0.24220387637615204,
+      "learning_rate": 0.0001729620818398315,
+      "loss": 1.6129,
+      "step": 49000
+    },
+    {
+      "epoch": 0.40653772272636945,
+      "grad_norm": 0.2865298390388489,
+      "learning_rate": 0.00017290688323068188,
+      "loss": 1.6253,
+      "step": 49100
+    },
+    {
+      "epoch": 0.4073657017950586,
+      "grad_norm": 0.24625296890735626,
+      "learning_rate": 0.00017285168462153228,
+      "loss": 1.6206,
+      "step": 49200
+    },
+    {
+      "epoch": 0.40819368086374774,
+      "grad_norm": 0.24253369867801666,
+      "learning_rate": 0.00017279648601238267,
+      "loss": 1.6423,
+      "step": 49300
+    },
+    {
+      "epoch": 0.4090216599324369,
+      "grad_norm": 0.2467760145664215,
+      "learning_rate": 0.00017274183938932456,
+      "loss": 1.655,
+      "step": 49400
+    },
+    {
+      "epoch": 0.40984963900112603,
+      "grad_norm": 0.2473566234111786,
+      "learning_rate": 0.00017268664078017495,
+      "loss": 1.6121,
+      "step": 49500
+    },
+    {
+      "epoch": 0.4106776180698152,
+      "grad_norm": 0.2393588274717331,
+      "learning_rate": 0.00017263144217102535,
+      "loss": 1.6231,
+      "step": 49600
+    },
+    {
+      "epoch": 0.4115055971385043,
+      "grad_norm": 0.28404486179351807,
+      "learning_rate": 0.00017257624356187574,
+      "loss": 1.6423,
+      "step": 49700
+    },
+    {
+      "epoch": 0.41233357620719346,
+      "grad_norm": 0.22529253363609314,
+      "learning_rate": 0.00017252104495272614,
+      "loss": 1.6341,
+      "step": 49800
+    },
+    {
+      "epoch": 0.4131615552758826,
+      "grad_norm": 0.27375975251197815,
+      "learning_rate": 0.00017246584634357653,
+      "loss": 1.5997,
+      "step": 49900
+    },
+    {
+      "epoch": 0.41398953434457175,
+      "grad_norm": 0.2889430820941925,
+      "learning_rate": 0.00017241064773442692,
+      "loss": 1.5884,
+      "step": 50000
+    },
+    {
+      "epoch": 0.4148175134132609,
+      "grad_norm": 0.2392152100801468,
+      "learning_rate": 0.00017235544912527732,
+      "loss": 1.6298,
+      "step": 50100
+    },
+    {
+      "epoch": 0.41564549248195004,
+      "grad_norm": 0.22931554913520813,
+      "learning_rate": 0.0001723002505161277,
+      "loss": 1.6139,
+      "step": 50200
+    },
+    {
+      "epoch": 0.4164734715506392,
+      "grad_norm": 0.24450622498989105,
+      "learning_rate": 0.0001722450519069781,
+      "loss": 1.6112,
+      "step": 50300
+    },
+    {
+      "epoch": 0.41730145061932833,
+      "grad_norm": 0.25789907574653625,
+      "learning_rate": 0.00017219040528392002,
+      "loss": 1.6098,
+      "step": 50400
+    },
+    {
+      "epoch": 0.4181294296880175,
+      "grad_norm": 0.21035891771316528,
+      "learning_rate": 0.0001721352066747704,
+      "loss": 1.6414,
+      "step": 50500
+    },
+    {
+      "epoch": 0.4189574087567066,
+      "grad_norm": 0.21786290407180786,
+      "learning_rate": 0.00017208000806562078,
+      "loss": 1.6782,
+      "step": 50600
+    },
+    {
+      "epoch": 0.41978538782539576,
+      "grad_norm": 0.25614434480667114,
+      "learning_rate": 0.0001720248094564712,
+      "loss": 1.6238,
+      "step": 50700
+    },
+    {
+      "epoch": 0.4206133668940849,
+      "grad_norm": 0.3049182891845703,
+      "learning_rate": 0.00017196961084732157,
+      "loss": 1.6096,
+      "step": 50800
+    },
+    {
+      "epoch": 0.42144134596277405,
+      "grad_norm": 0.2816081643104553,
+      "learning_rate": 0.00017191441223817196,
+      "loss": 1.6256,
+      "step": 50900
+    },
+    {
+      "epoch": 0.4222693250314632,
+      "grad_norm": 0.23279207944869995,
+      "learning_rate": 0.00017185921362902239,
+      "loss": 1.6243,
+      "step": 51000
+    },
+    {
+      "epoch": 0.42309730410015234,
+      "grad_norm": 0.23084907233715057,
+      "learning_rate": 0.00017180401501987275,
+      "loss": 1.5797,
+      "step": 51100
+    },
+    {
+      "epoch": 0.4239252831688415,
+      "grad_norm": 0.26271852850914,
+      "learning_rate": 0.00017174881641072315,
+      "loss": 1.6228,
+      "step": 51200
+    },
+    {
+      "epoch": 0.4247532622375306,
+      "grad_norm": 0.22446395456790924,
+      "learning_rate": 0.00017169361780157357,
+      "loss": 1.632,
+      "step": 51300
+    },
+    {
+      "epoch": 0.42558124130621977,
+      "grad_norm": 0.2501899302005768,
+      "learning_rate": 0.00017163841919242394,
+      "loss": 1.6155,
+      "step": 51400
+    },
+    {
+      "epoch": 0.4264092203749089,
+      "grad_norm": 0.24416933953762054,
+      "learning_rate": 0.00017158322058327433,
+      "loss": 1.6104,
+      "step": 51500
+    },
+    {
+      "epoch": 0.42723719944359806,
+      "grad_norm": 0.24799175560474396,
+      "learning_rate": 0.00017152802197412475,
+      "loss": 1.629,
+      "step": 51600
+    },
+    {
+      "epoch": 0.4280651785122872,
+      "grad_norm": 0.2706332504749298,
+      "learning_rate": 0.00017147282336497512,
+      "loss": 1.5854,
+      "step": 51700
+    },
+    {
+      "epoch": 0.42889315758097635,
+      "grad_norm": 0.2325417846441269,
+      "learning_rate": 0.0001714176247558255,
+      "loss": 1.6373,
+      "step": 51800
+    },
+    {
+      "epoch": 0.4297211366496655,
+      "grad_norm": 0.3152548372745514,
+      "learning_rate": 0.00017136242614667593,
+      "loss": 1.6051,
+      "step": 51900
+    },
+    {
+      "epoch": 0.43054911571835464,
+      "grad_norm": 0.22651983797550201,
+      "learning_rate": 0.0001713072275375263,
+      "loss": 1.588,
+      "step": 52000
+    },
+    {
+      "epoch": 0.4313770947870438,
+      "grad_norm": 0.22875835001468658,
+      "learning_rate": 0.0001712520289283767,
+      "loss": 1.6193,
+      "step": 52100
+    },
+    {
+      "epoch": 0.4322050738557329,
+      "grad_norm": 0.244579017162323,
+      "learning_rate": 0.00017119683031922712,
+      "loss": 1.6165,
+      "step": 52200
+    },
+    {
+      "epoch": 0.43303305292442207,
+      "grad_norm": 0.2727806270122528,
+      "learning_rate": 0.00017114163171007748,
+      "loss": 1.6106,
+      "step": 52300
+    },
+    {
+      "epoch": 0.4338610319931112,
+      "grad_norm": 0.2944915294647217,
+      "learning_rate": 0.00017108643310092788,
+      "loss": 1.6379,
+      "step": 52400
+    },
+    {
+      "epoch": 0.43468901106180036,
+      "grad_norm": 0.32742735743522644,
+      "learning_rate": 0.0001710312344917783,
+      "loss": 1.5989,
+      "step": 52500
+    },
+    {
+      "epoch": 0.4355169901304895,
+      "grad_norm": 0.2617054879665375,
+      "learning_rate": 0.00017097603588262867,
+      "loss": 1.6405,
+      "step": 52600
+    },
+    {
+      "epoch": 0.43634496919917864,
+      "grad_norm": 0.2704140841960907,
+      "learning_rate": 0.00017092083727347906,
+      "loss": 1.654,
+      "step": 52700
+    },
+    {
+      "epoch": 0.4371729482678678,
+      "grad_norm": 0.22568093240261078,
+      "learning_rate": 0.00017086563866432948,
+      "loss": 1.6086,
+      "step": 52800
+    },
+    {
+      "epoch": 0.43800092733655693,
+      "grad_norm": 0.2364298552274704,
+      "learning_rate": 0.00017081044005517985,
+      "loss": 1.637,
+      "step": 52900
+    },
+    {
+      "epoch": 0.4388289064052461,
+      "grad_norm": 0.24392692744731903,
+      "learning_rate": 0.00017075524144603024,
+      "loss": 1.5939,
+      "step": 53000
+    },
+    {
+      "epoch": 0.4396568854739352,
+      "grad_norm": 0.2571870684623718,
+      "learning_rate": 0.00017070004283688066,
+      "loss": 1.6708,
+      "step": 53100
+    },
+    {
+      "epoch": 0.44048486454262437,
+      "grad_norm": 0.23539692163467407,
+      "learning_rate": 0.00017064484422773103,
+      "loss": 1.6085,
+      "step": 53200
+    },
+    {
+      "epoch": 0.4413128436113135,
+      "grad_norm": 0.25460028648376465,
+      "learning_rate": 0.00017058964561858143,
+      "loss": 1.6051,
+      "step": 53300
+    },
+    {
+      "epoch": 0.44214082268000265,
+      "grad_norm": 0.3254430294036865,
+      "learning_rate": 0.00017053444700943185,
+      "loss": 1.5707,
+      "step": 53400
+    },
+    {
+      "epoch": 0.4429688017486918,
+      "grad_norm": 0.2832888960838318,
+      "learning_rate": 0.00017047924840028221,
+      "loss": 1.6606,
+      "step": 53500
+    },
+    {
+      "epoch": 0.44379678081738094,
+      "grad_norm": 0.2448771893978119,
+      "learning_rate": 0.00017042404979113264,
+      "loss": 1.6052,
+      "step": 53600
+    },
+    {
+      "epoch": 0.4446247598860701,
+      "grad_norm": 0.28213441371917725,
+      "learning_rate": 0.00017036885118198303,
+      "loss": 1.6286,
+      "step": 53700
+    },
+    {
+      "epoch": 0.44545273895475923,
+      "grad_norm": 0.24302446842193604,
+      "learning_rate": 0.0001703136525728334,
+      "loss": 1.5793,
+      "step": 53800
+    },
+    {
+      "epoch": 0.4462807180234484,
+      "grad_norm": 0.25916993618011475,
+      "learning_rate": 0.0001702590059497753,
+      "loss": 1.6339,
+      "step": 53900
+    },
+    {
+      "epoch": 0.4471086970921375,
+      "grad_norm": 0.24786342680454254,
+      "learning_rate": 0.00017020380734062568,
+      "loss": 1.5977,
+      "step": 54000
+    },
+    {
+      "epoch": 0.44793667616082666,
+      "grad_norm": 0.22815145552158356,
+      "learning_rate": 0.0001701486087314761,
+      "loss": 1.6541,
+      "step": 54100
+    },
+    {
+      "epoch": 0.4487646552295158,
+      "grad_norm": 0.26695653796195984,
+      "learning_rate": 0.0001700934101223265,
+      "loss": 1.6313,
+      "step": 54200
+    },
+    {
+      "epoch": 0.44959263429820495,
+      "grad_norm": 0.2767154276371002,
+      "learning_rate": 0.00017003821151317686,
+      "loss": 1.5868,
+      "step": 54300
+    },
+    {
+      "epoch": 0.4504206133668941,
+      "grad_norm": 0.23107494413852692,
+      "learning_rate": 0.00016998301290402728,
+      "loss": 1.6353,
+      "step": 54400
+    },
+    {
+      "epoch": 0.45124859243558324,
+      "grad_norm": 0.24376466870307922,
+      "learning_rate": 0.00016992781429487768,
+      "loss": 1.6075,
+      "step": 54500
+    },
+    {
+      "epoch": 0.4520765715042724,
+      "grad_norm": 0.3318908214569092,
+      "learning_rate": 0.00016987261568572804,
+      "loss": 1.6337,
+      "step": 54600
+    },
+    {
+      "epoch": 0.45290455057296153,
+      "grad_norm": 0.2159205824136734,
+      "learning_rate": 0.00016981741707657846,
+      "loss": 1.6237,
+      "step": 54700
+    },
+    {
+      "epoch": 0.4537325296416507,
+      "grad_norm": 0.24641750752925873,
+      "learning_rate": 0.00016976221846742886,
+      "loss": 1.5984,
+      "step": 54800
+    },
+    {
+      "epoch": 0.4545605087103398,
+      "grad_norm": 0.25615862011909485,
+      "learning_rate": 0.00016970701985827925,
+      "loss": 1.5901,
+      "step": 54900
+    },
+    {
+      "epoch": 0.45538848777902896,
+      "grad_norm": 0.27317163348197937,
+      "learning_rate": 0.00016965182124912965,
+      "loss": 1.651,
+      "step": 55000
+    },
+    {
+      "epoch": 0.4562164668477181,
+      "grad_norm": 0.2555563747882843,
+      "learning_rate": 0.00016959662263998004,
+      "loss": 1.6497,
+      "step": 55100
+    },
+    {
+      "epoch": 0.45704444591640725,
+      "grad_norm": 0.2796775698661804,
+      "learning_rate": 0.00016954142403083044,
+      "loss": 1.6579,
+      "step": 55200
+    },
+    {
+      "epoch": 0.4578724249850964,
+      "grad_norm": 0.26606446504592896,
+      "learning_rate": 0.00016948622542168083,
+      "loss": 1.6271,
+      "step": 55300
+    },
+    {
+      "epoch": 0.45870040405378554,
+      "grad_norm": 0.2654423117637634,
+      "learning_rate": 0.00016943102681253122,
+      "loss": 1.6223,
+      "step": 55400
+    },
+    {
+      "epoch": 0.4595283831224747,
+      "grad_norm": 0.24428655207157135,
+      "learning_rate": 0.00016937582820338162,
+      "loss": 1.6188,
+      "step": 55500
+    },
+    {
+      "epoch": 0.4603563621911638,
+      "grad_norm": 0.28939154744148254,
+      "learning_rate": 0.000169320629594232,
+      "loss": 1.619,
+      "step": 55600
+    },
+    {
+      "epoch": 0.46118434125985297,
+      "grad_norm": 0.28437426686286926,
+      "learning_rate": 0.0001692654309850824,
+      "loss": 1.6357,
+      "step": 55700
+    },
+    {
+      "epoch": 0.4620123203285421,
+      "grad_norm": 0.2547878921031952,
+      "learning_rate": 0.0001692102323759328,
+      "loss": 1.6044,
+      "step": 55800
+    },
+    {
+      "epoch": 0.46284029939723126,
+      "grad_norm": 0.25692984461784363,
+      "learning_rate": 0.0001691550337667832,
+      "loss": 1.6059,
+      "step": 55900
+    },
+    {
+      "epoch": 0.4636682784659204,
+      "grad_norm": 0.3365028500556946,
+      "learning_rate": 0.0001690998351576336,
+      "loss": 1.6484,
+      "step": 56000
+    },
+    {
+      "epoch": 0.46449625753460955,
+      "grad_norm": 0.2529739737510681,
+      "learning_rate": 0.00016904463654848398,
+      "loss": 1.6212,
+      "step": 56100
+    },
+    {
+      "epoch": 0.4653242366032987,
+      "grad_norm": 0.23951299488544464,
+      "learning_rate": 0.00016898943793933438,
+      "loss": 1.6151,
+      "step": 56200
+    },
+    {
+      "epoch": 0.46615221567198784,
+      "grad_norm": 0.3531137704849243,
+      "learning_rate": 0.00016893423933018477,
+      "loss": 1.6013,
+      "step": 56300
+    },
+    {
+      "epoch": 0.466980194740677,
+      "grad_norm": 0.30527693033218384,
+      "learning_rate": 0.00016887904072103517,
+      "loss": 1.6112,
+      "step": 56400
+    },
+    {
+      "epoch": 0.4678081738093661,
+      "grad_norm": 0.25314369797706604,
+      "learning_rate": 0.00016882384211188556,
+      "loss": 1.6281,
+      "step": 56500
+    },
+    {
+      "epoch": 0.46863615287805527,
+      "grad_norm": 0.25864383578300476,
+      "learning_rate": 0.00016876864350273595,
+      "loss": 1.6376,
+      "step": 56600
+    },
+    {
+      "epoch": 0.4694641319467444,
+      "grad_norm": 0.27216991782188416,
+      "learning_rate": 0.00016871344489358635,
+      "loss": 1.5999,
+      "step": 56700
+    },
+    {
+      "epoch": 0.47029211101543356,
+      "grad_norm": 0.23203931748867035,
+      "learning_rate": 0.00016865824628443674,
+      "loss": 1.6304,
+      "step": 56800
+    },
+    {
+      "epoch": 0.4711200900841227,
+      "grad_norm": 0.3125530183315277,
+      "learning_rate": 0.00016860304767528714,
+      "loss": 1.6264,
+      "step": 56900
+    },
+    {
+      "epoch": 0.47194806915281184,
+      "grad_norm": 0.22787316143512726,
+      "learning_rate": 0.00016854784906613753,
+      "loss": 1.6044,
+      "step": 57000
+    },
+    {
+      "epoch": 0.47277604822150093,
+      "grad_norm": 0.24813027679920197,
+      "learning_rate": 0.00016849265045698793,
+      "loss": 1.6347,
+      "step": 57100
+    },
+    {
+      "epoch": 0.4736040272901901,
+      "grad_norm": 0.2771335542201996,
+      "learning_rate": 0.00016843745184783832,
+      "loss": 1.6186,
+      "step": 57200
+    },
+    {
+      "epoch": 0.4744320063588792,
+      "grad_norm": 0.2221546620130539,
+      "learning_rate": 0.00016838225323868871,
+      "loss": 1.6019,
+      "step": 57300
+    },
+    {
+      "epoch": 0.47525998542756837,
+      "grad_norm": 0.24322104454040527,
+      "learning_rate": 0.0001683270546295391,
+      "loss": 1.5865,
+      "step": 57400
+    },
+    {
+      "epoch": 0.4760879644962575,
+      "grad_norm": 0.26411768794059753,
+      "learning_rate": 0.0001682718560203895,
+      "loss": 1.6504,
+      "step": 57500
+    },
+    {
+      "epoch": 0.47691594356494665,
+      "grad_norm": 0.319624662399292,
+      "learning_rate": 0.0001682166574112399,
+      "loss": 1.6068,
+      "step": 57600
+    },
+    {
+      "epoch": 0.4777439226336358,
+      "grad_norm": 0.2550843358039856,
+      "learning_rate": 0.0001681614588020903,
+      "loss": 1.6177,
+      "step": 57700
+    },
+    {
+      "epoch": 0.47857190170232494,
+      "grad_norm": 0.28622475266456604,
+      "learning_rate": 0.00016810626019294069,
+      "loss": 1.619,
+      "step": 57800
+    },
+    {
+      "epoch": 0.4793998807710141,
+      "grad_norm": 0.2016662210226059,
+      "learning_rate": 0.00016805106158379108,
+      "loss": 1.6413,
+      "step": 57900
+    },
+    {
+      "epoch": 0.48022785983970323,
+      "grad_norm": 0.2323019951581955,
+      "learning_rate": 0.00016799586297464147,
+      "loss": 1.6104,
+      "step": 58000
+    },
+    {
+      "epoch": 0.4810558389083924,
+      "grad_norm": 0.2892739176750183,
+      "learning_rate": 0.00016794066436549187,
+      "loss": 1.5928,
+      "step": 58100
+    },
+    {
+      "epoch": 0.4818838179770815,
+      "grad_norm": 0.2968466877937317,
+      "learning_rate": 0.00016788546575634226,
+      "loss": 1.5965,
+      "step": 58200
+    },
+    {
+      "epoch": 0.48271179704577066,
+      "grad_norm": 0.29501309990882874,
+      "learning_rate": 0.00016783026714719266,
+      "loss": 1.6309,
+      "step": 58300
+    },
+    {
+      "epoch": 0.4835397761144598,
+      "grad_norm": 0.3167082369327545,
+      "learning_rate": 0.00016777506853804305,
+      "loss": 1.622,
+      "step": 58400
+    },
+    {
+      "epoch": 0.48436775518314895,
+      "grad_norm": 0.2594543397426605,
+      "learning_rate": 0.00016771986992889345,
+      "loss": 1.6183,
+      "step": 58500
+    },
+    {
+      "epoch": 0.4851957342518381,
+      "grad_norm": 0.23582147061824799,
+      "learning_rate": 0.00016766467131974384,
+      "loss": 1.6039,
+      "step": 58600
+    },
+    {
+      "epoch": 0.48602371332052724,
+      "grad_norm": 0.24872687458992004,
+      "learning_rate": 0.00016760947271059423,
+      "loss": 1.6092,
+      "step": 58700
+    },
+    {
+      "epoch": 0.4868516923892164,
+      "grad_norm": 0.21512049436569214,
+      "learning_rate": 0.00016755427410144463,
+      "loss": 1.6323,
+      "step": 58800
+    },
+    {
+      "epoch": 0.48767967145790553,
+      "grad_norm": 0.25735777616500854,
+      "learning_rate": 0.00016749907549229502,
+      "loss": 1.6367,
+      "step": 58900
+    },
+    {
+      "epoch": 0.4885076505265947,
+      "grad_norm": 0.23260653018951416,
+      "learning_rate": 0.00016744387688314542,
+      "loss": 1.6497,
+      "step": 59000
+    },
+    {
+      "epoch": 0.4893356295952838,
+      "grad_norm": 0.31416621804237366,
+      "learning_rate": 0.00016738867827399584,
+      "loss": 1.6128,
+      "step": 59100
+    },
+    {
+      "epoch": 0.49016360866397296,
+      "grad_norm": 0.2689904570579529,
+      "learning_rate": 0.00016733458363702922,
+      "loss": 1.6084,
+      "step": 59200
+    },
+    {
+      "epoch": 0.4909915877326621,
+      "grad_norm": 0.26795658469200134,
+      "learning_rate": 0.00016727938502787958,
+      "loss": 1.6342,
+      "step": 59300
+    },
+    {
+      "epoch": 0.49181956680135125,
+      "grad_norm": 0.28816089034080505,
+      "learning_rate": 0.00016722418641872998,
+      "loss": 1.5939,
+      "step": 59400
+    },
+    {
+      "epoch": 0.4926475458700404,
+      "grad_norm": 0.3193409740924835,
+      "learning_rate": 0.0001671689878095804,
+      "loss": 1.6022,
+      "step": 59500
+    },
+    {
+      "epoch": 0.49347552493872954,
+      "grad_norm": 0.2833710014820099,
+      "learning_rate": 0.00016711378920043077,
+      "loss": 1.6164,
+      "step": 59600
+    },
+    {
+      "epoch": 0.4943035040074187,
+      "grad_norm": 0.2676531672477722,
+      "learning_rate": 0.00016705859059128116,
+      "loss": 1.6019,
+      "step": 59700
+    },
+    {
+      "epoch": 0.4951314830761078,
+      "grad_norm": 0.29554829001426697,
+      "learning_rate": 0.00016700339198213158,
+      "loss": 1.6028,
+      "step": 59800
+    },
+    {
+      "epoch": 0.49595946214479697,
+      "grad_norm": 0.24950549006462097,
+      "learning_rate": 0.00016694819337298195,
+      "loss": 1.5983,
+      "step": 59900
+    },
+    {
+      "epoch": 0.4967874412134861,
+      "grad_norm": 0.31418663263320923,
+      "learning_rate": 0.00016689299476383234,
+      "loss": 1.6168,
+      "step": 60000
+    },
+    {
+      "epoch": 0.49761542028217526,
+      "grad_norm": 0.2942350506782532,
+      "learning_rate": 0.00016683779615468276,
+      "loss": 1.6117,
+      "step": 60100
+    },
+    {
+      "epoch": 0.4984433993508644,
+      "grad_norm": 0.2196948081254959,
+      "learning_rate": 0.00016678259754553313,
+      "loss": 1.6221,
+      "step": 60200
+    },
+    {
+      "epoch": 0.49927137841955355,
+      "grad_norm": 0.25782310962677,
+      "learning_rate": 0.00016672739893638353,
+      "loss": 1.6207,
+      "step": 60300
+    },
+    {
+      "epoch": 0.5000993574882427,
+      "grad_norm": 0.24130311608314514,
+      "learning_rate": 0.00016667220032723395,
+      "loss": 1.6244,
+      "step": 60400
+    },
+    {
+      "epoch": 0.5009273365569319,
+      "grad_norm": 0.26923075318336487,
+      "learning_rate": 0.00016661700171808431,
+      "loss": 1.592,
+      "step": 60500
+    },
+    {
+      "epoch": 0.501755315625621,
+      "grad_norm": 0.23227745294570923,
+      "learning_rate": 0.00016656180310893474,
+      "loss": 1.5768,
+      "step": 60600
+    },
+    {
+      "epoch": 0.5025832946943102,
+      "grad_norm": 0.2719257175922394,
+      "learning_rate": 0.00016650660449978513,
+      "loss": 1.6187,
+      "step": 60700
+    },
+    {
+      "epoch": 0.5034112737629993,
+      "grad_norm": 0.26879358291625977,
+      "learning_rate": 0.0001664514058906355,
+      "loss": 1.6194,
+      "step": 60800
+    },
+    {
+      "epoch": 0.5042392528316885,
+      "grad_norm": 0.24769945442676544,
+      "learning_rate": 0.00016639620728148592,
+      "loss": 1.5787,
+      "step": 60900
+    },
+    {
+      "epoch": 0.5050672319003776,
+      "grad_norm": 0.26854538917541504,
+      "learning_rate": 0.0001663410086723363,
+      "loss": 1.6254,
+      "step": 61000
+    },
+    {
+      "epoch": 0.5058952109690668,
+      "grad_norm": 0.300074964761734,
+      "learning_rate": 0.00016628581006318668,
+      "loss": 1.6232,
+      "step": 61100
+    },
+    {
+      "epoch": 0.5067231900377559,
+      "grad_norm": 0.3866933584213257,
+      "learning_rate": 0.0001662306114540371,
+      "loss": 1.596,
+      "step": 61200
+    },
+    {
+      "epoch": 0.507551169106445,
+      "grad_norm": 0.23814737796783447,
+      "learning_rate": 0.0001661754128448875,
+      "loss": 1.6291,
+      "step": 61300
+    },
+    {
+      "epoch": 0.5083791481751342,
+      "grad_norm": 0.26245006918907166,
+      "learning_rate": 0.00016612021423573786,
+      "loss": 1.6209,
+      "step": 61400
+    },
+    {
+      "epoch": 0.5092071272438232,
+      "grad_norm": 0.24793724715709686,
+      "learning_rate": 0.00016606501562658828,
+      "loss": 1.5819,
+      "step": 61500
+    },
+    {
+      "epoch": 0.5100351063125124,
+      "grad_norm": 0.2816125452518463,
+      "learning_rate": 0.00016600981701743868,
+      "loss": 1.6488,
+      "step": 61600
+    },
+    {
+      "epoch": 0.5108630853812015,
+      "grad_norm": 0.2552174925804138,
+      "learning_rate": 0.00016595461840828904,
+      "loss": 1.6068,
+      "step": 61700
+    },
+    {
+      "epoch": 0.5116910644498907,
+      "grad_norm": 0.34865882992744446,
+      "learning_rate": 0.00016589941979913947,
+      "loss": 1.6184,
+      "step": 61800
+    },
+    {
+      "epoch": 0.5125190435185798,
+      "grad_norm": 0.238135427236557,
+      "learning_rate": 0.00016584422118998986,
+      "loss": 1.6167,
+      "step": 61900
+    },
+    {
+      "epoch": 0.5133470225872689,
+      "grad_norm": 0.26624831557273865,
+      "learning_rate": 0.00016578902258084023,
+      "loss": 1.6137,
+      "step": 62000
+    },
+    {
+      "epoch": 0.5141750016559581,
+      "grad_norm": 0.2653828561306,
+      "learning_rate": 0.00016573382397169065,
+      "loss": 1.63,
+      "step": 62100
+    },
+    {
+      "epoch": 0.5150029807246472,
+      "grad_norm": 0.25800028443336487,
+      "learning_rate": 0.00016567862536254104,
+      "loss": 1.6301,
+      "step": 62200
+    },
+    {
+      "epoch": 0.5158309597933364,
+      "grad_norm": 0.2781680226325989,
+      "learning_rate": 0.0001656234267533914,
+      "loss": 1.6505,
+      "step": 62300
+    },
+    {
+      "epoch": 0.5166589388620255,
+      "grad_norm": 0.2765764594078064,
+      "learning_rate": 0.00016556822814424183,
+      "loss": 1.6243,
+      "step": 62400
+    },
+    {
+      "epoch": 0.5174869179307147,
+      "grad_norm": 0.24544250965118408,
+      "learning_rate": 0.00016551302953509223,
+      "loss": 1.6221,
+      "step": 62500
+    },
+    {
+      "epoch": 0.5183148969994038,
+      "grad_norm": 0.3250795304775238,
+      "learning_rate": 0.0001654578309259426,
+      "loss": 1.6504,
+      "step": 62600
+    },
+    {
+      "epoch": 0.519142876068093,
+      "grad_norm": 0.23859694600105286,
+      "learning_rate": 0.00016540263231679301,
+      "loss": 1.6017,
+      "step": 62700
+    },
+    {
+      "epoch": 0.5199708551367821,
+      "grad_norm": 0.27489158511161804,
+      "learning_rate": 0.0001653474337076434,
+      "loss": 1.5901,
+      "step": 62800
+    },
+    {
+      "epoch": 0.5207988342054712,
+      "grad_norm": 0.25401830673217773,
+      "learning_rate": 0.00016529223509849378,
+      "loss": 1.6186,
+      "step": 62900
+    },
+    {
+      "epoch": 0.5216268132741604,
+      "grad_norm": 0.26960933208465576,
+      "learning_rate": 0.0001652370364893442,
+      "loss": 1.6156,
+      "step": 63000
+    },
+    {
+      "epoch": 0.5224547923428495,
+      "grad_norm": 0.256863534450531,
+      "learning_rate": 0.0001651818378801946,
+      "loss": 1.5954,
+      "step": 63100
+    },
+    {
+      "epoch": 0.5232827714115387,
+      "grad_norm": 0.2537617087364197,
+      "learning_rate": 0.00016512663927104496,
+      "loss": 1.5872,
+      "step": 63200
+    },
+    {
+      "epoch": 0.5241107504802278,
+      "grad_norm": 0.21423785388469696,
+      "learning_rate": 0.00016507144066189538,
+      "loss": 1.5982,
+      "step": 63300
+    },
+    {
+      "epoch": 0.524938729548917,
+      "grad_norm": 0.3197900354862213,
+      "learning_rate": 0.00016501624205274577,
+      "loss": 1.6362,
+      "step": 63400
+    },
+    {
+      "epoch": 0.5257667086176061,
+      "grad_norm": 0.2450074404478073,
+      "learning_rate": 0.00016496104344359614,
+      "loss": 1.6076,
+      "step": 63500
+    },
+    {
+      "epoch": 0.5265946876862952,
+      "grad_norm": 0.251626193523407,
+      "learning_rate": 0.00016490584483444656,
+      "loss": 1.6395,
+      "step": 63600
+    },
+    {
+      "epoch": 0.5274226667549844,
+      "grad_norm": 0.3137412965297699,
+      "learning_rate": 0.00016485064622529696,
+      "loss": 1.6047,
+      "step": 63700
+    },
+    {
+      "epoch": 0.5282506458236735,
+      "grad_norm": 0.28337448835372925,
+      "learning_rate": 0.00016479599960223884,
+      "loss": 1.621,
+      "step": 63800
+    },
+    {
+      "epoch": 0.5290786248923627,
+      "grad_norm": 0.22410038113594055,
+      "learning_rate": 0.00016474080099308924,
+      "loss": 1.5938,
+      "step": 63900
+    },
+    {
+      "epoch": 0.5299066039610518,
+      "grad_norm": 0.24572992324829102,
+      "learning_rate": 0.00016468560238393963,
+      "loss": 1.6213,
+      "step": 64000
+    },
+    {
+      "epoch": 0.530734583029741,
+      "grad_norm": 0.2508053481578827,
+      "learning_rate": 0.00016463040377479003,
+      "loss": 1.6016,
+      "step": 64100
+    },
+    {
+      "epoch": 0.5315625620984301,
+      "grad_norm": 0.26450827717781067,
+      "learning_rate": 0.00016457520516564042,
+      "loss": 1.6326,
+      "step": 64200
+    },
+    {
+      "epoch": 0.5323905411671193,
+      "grad_norm": 0.259705126285553,
+      "learning_rate": 0.00016452000655649081,
+      "loss": 1.622,
+      "step": 64300
+    },
+    {
+      "epoch": 0.5332185202358084,
+      "grad_norm": 0.3122607469558716,
+      "learning_rate": 0.0001644648079473412,
+      "loss": 1.6068,
+      "step": 64400
+    },
+    {
+      "epoch": 0.5340464993044975,
+      "grad_norm": 0.264130562543869,
+      "learning_rate": 0.0001644096093381916,
+      "loss": 1.6234,
+      "step": 64500
+    },
+    {
+      "epoch": 0.5348744783731867,
+      "grad_norm": 0.2097257822751999,
+      "learning_rate": 0.000164354410729042,
+      "loss": 1.6296,
+      "step": 64600
+    },
+    {
+      "epoch": 0.5357024574418758,
+      "grad_norm": 0.2408805638551712,
+      "learning_rate": 0.0001642992121198924,
+      "loss": 1.606,
+      "step": 64700
+    },
+    {
+      "epoch": 0.536530436510565,
+      "grad_norm": 0.25239992141723633,
+      "learning_rate": 0.00016424401351074279,
+      "loss": 1.5934,
+      "step": 64800
+    },
+    {
+      "epoch": 0.5373584155792541,
+      "grad_norm": 0.23828302323818207,
+      "learning_rate": 0.00016418881490159318,
+      "loss": 1.6142,
+      "step": 64900
+    },
+    {
+      "epoch": 0.5381863946479433,
+      "grad_norm": 0.24627448618412018,
+      "learning_rate": 0.00016413361629244357,
+      "loss": 1.6397,
+      "step": 65000
+    },
+    {
+      "epoch": 0.5390143737166324,
+      "grad_norm": 0.2534656226634979,
+      "learning_rate": 0.00016407841768329397,
+      "loss": 1.6293,
+      "step": 65100
+    },
+    {
+      "epoch": 0.5398423527853216,
+      "grad_norm": 0.31040382385253906,
+      "learning_rate": 0.00016402321907414436,
+      "loss": 1.61,
+      "step": 65200
+    },
+    {
+      "epoch": 0.5406703318540107,
+      "grad_norm": 0.23957033455371857,
+      "learning_rate": 0.00016396802046499476,
+      "loss": 1.6428,
+      "step": 65300
+    },
+    {
+      "epoch": 0.5414983109226998,
+      "grad_norm": 0.21376514434814453,
+      "learning_rate": 0.00016391282185584515,
+      "loss": 1.6473,
+      "step": 65400
+    },
+    {
+      "epoch": 0.542326289991389,
+      "grad_norm": 0.24178186058998108,
+      "learning_rate": 0.00016385762324669554,
+      "loss": 1.6311,
+      "step": 65500
+    },
+    {
+      "epoch": 0.5431542690600781,
+      "grad_norm": 0.28505873680114746,
+      "learning_rate": 0.00016380242463754594,
+      "loss": 1.6206,
+      "step": 65600
+    },
+    {
+      "epoch": 0.5439822481287673,
+      "grad_norm": 0.22037836909294128,
+      "learning_rate": 0.00016374722602839633,
+      "loss": 1.618,
+      "step": 65700
+    },
+    {
+      "epoch": 0.5448102271974564,
+      "grad_norm": 0.2399994432926178,
+      "learning_rate": 0.00016369202741924673,
+      "loss": 1.6268,
+      "step": 65800
+    },
+    {
+      "epoch": 0.5456382062661456,
+      "grad_norm": 0.24552536010742188,
+      "learning_rate": 0.00016363682881009712,
+      "loss": 1.6082,
+      "step": 65900
+    },
+    {
+      "epoch": 0.5464661853348347,
+      "grad_norm": 0.24302148818969727,
+      "learning_rate": 0.00016358163020094752,
+      "loss": 1.63,
+      "step": 66000
+    },
+    {
+      "epoch": 0.5472941644035239,
+      "grad_norm": 0.2581181228160858,
+      "learning_rate": 0.00016352643159179794,
+      "loss": 1.6249,
+      "step": 66100
+    },
+    {
+      "epoch": 0.548122143472213,
+      "grad_norm": 0.3073265552520752,
+      "learning_rate": 0.0001634712329826483,
+      "loss": 1.6163,
+      "step": 66200
+    },
+    {
+      "epoch": 0.5489501225409021,
+      "grad_norm": 0.24049556255340576,
+      "learning_rate": 0.0001634160343734987,
+      "loss": 1.63,
+      "step": 66300
+    },
+    {
+      "epoch": 0.5497781016095913,
+      "grad_norm": 0.2216363102197647,
+      "learning_rate": 0.00016336083576434912,
+      "loss": 1.6114,
+      "step": 66400
+    },
+    {
+      "epoch": 0.5506060806782804,
+      "grad_norm": 0.35404762625694275,
+      "learning_rate": 0.00016330618914129098,
+      "loss": 1.6514,
+      "step": 66500
+    },
+    {
+      "epoch": 0.5514340597469696,
+      "grad_norm": 0.2302398830652237,
+      "learning_rate": 0.0001632509905321414,
+      "loss": 1.6084,
+      "step": 66600
+    },
+    {
+      "epoch": 0.5522620388156587,
+      "grad_norm": 0.35909637808799744,
+      "learning_rate": 0.00016319579192299177,
+      "loss": 1.6509,
+      "step": 66700
+    },
+    {
+      "epoch": 0.5530900178843479,
+      "grad_norm": 0.2828253507614136,
+      "learning_rate": 0.00016314059331384216,
+      "loss": 1.612,
+      "step": 66800
+    },
+    {
+      "epoch": 0.553917996953037,
+      "grad_norm": 0.25532275438308716,
+      "learning_rate": 0.00016308539470469258,
+      "loss": 1.6508,
+      "step": 66900
+    },
+    {
+      "epoch": 0.5547459760217262,
+      "grad_norm": 0.2709483504295349,
+      "learning_rate": 0.00016303019609554295,
+      "loss": 1.6227,
+      "step": 67000
+    },
+    {
+      "epoch": 0.5555739550904153,
+      "grad_norm": 0.23612917959690094,
+      "learning_rate": 0.00016297499748639334,
+      "loss": 1.66,
+      "step": 67100
+    },
+    {
+      "epoch": 0.5564019341591044,
+      "grad_norm": 0.29938626289367676,
+      "learning_rate": 0.00016291979887724377,
+      "loss": 1.601,
+      "step": 67200
+    },
+    {
+      "epoch": 0.5572299132277936,
+      "grad_norm": 0.3122688829898834,
+      "learning_rate": 0.00016286460026809413,
+      "loss": 1.6137,
+      "step": 67300
+    },
+    {
+      "epoch": 0.5580578922964827,
+      "grad_norm": 0.2629024088382721,
+      "learning_rate": 0.00016280940165894453,
+      "loss": 1.6272,
+      "step": 67400
+    },
+    {
+      "epoch": 0.5588858713651719,
+      "grad_norm": 0.26424530148506165,
+      "learning_rate": 0.00016275420304979495,
+      "loss": 1.6248,
+      "step": 67500
+    },
+    {
+      "epoch": 0.559713850433861,
+      "grad_norm": 0.2512044608592987,
+      "learning_rate": 0.00016269900444064532,
+      "loss": 1.6147,
+      "step": 67600
+    },
+    {
+      "epoch": 0.5605418295025502,
+      "grad_norm": 0.22548918426036835,
+      "learning_rate": 0.0001626438058314957,
+      "loss": 1.5857,
+      "step": 67700
+    },
+    {
+      "epoch": 0.5613698085712393,
+      "grad_norm": 0.23559758067131042,
+      "learning_rate": 0.00016258860722234613,
+      "loss": 1.6165,
+      "step": 67800
+    },
+    {
+      "epoch": 0.5621977876399284,
+      "grad_norm": 0.28142249584198,
+      "learning_rate": 0.0001625334086131965,
+      "loss": 1.6195,
+      "step": 67900
+    },
+    {
+      "epoch": 0.5630257667086176,
+      "grad_norm": 0.2530786991119385,
+      "learning_rate": 0.0001624782100040469,
+      "loss": 1.6198,
+      "step": 68000
+    },
+    {
+      "epoch": 0.5638537457773067,
+      "grad_norm": 0.239883691072464,
+      "learning_rate": 0.00016242301139489731,
+      "loss": 1.6385,
+      "step": 68100
+    },
+    {
+      "epoch": 0.5646817248459959,
+      "grad_norm": 0.2762130796909332,
+      "learning_rate": 0.00016236781278574768,
+      "loss": 1.6237,
+      "step": 68200
+    },
+    {
+      "epoch": 0.565509703914685,
+      "grad_norm": 0.23887832462787628,
+      "learning_rate": 0.00016231261417659808,
+      "loss": 1.6299,
+      "step": 68300
+    },
+    {
+      "epoch": 0.5663376829833742,
+      "grad_norm": 0.23560328781604767,
+      "learning_rate": 0.0001622574155674485,
+      "loss": 1.6526,
+      "step": 68400
+    },
+    {
+      "epoch": 0.5671656620520633,
+      "grad_norm": 0.23129230737686157,
+      "learning_rate": 0.00016220221695829886,
+      "loss": 1.6188,
+      "step": 68500
+    },
+    {
+      "epoch": 0.5679936411207525,
+      "grad_norm": 0.22737343609333038,
+      "learning_rate": 0.00016214701834914926,
+      "loss": 1.6108,
+      "step": 68600
+    },
+    {
+      "epoch": 0.5688216201894416,
+      "grad_norm": 0.26043862104415894,
+      "learning_rate": 0.00016209181973999968,
+      "loss": 1.6087,
+      "step": 68700
+    },
+    {
+      "epoch": 0.5696495992581307,
+      "grad_norm": 0.30376845598220825,
+      "learning_rate": 0.00016203662113085005,
+      "loss": 1.6184,
+      "step": 68800
+    },
+    {
+      "epoch": 0.5704775783268199,
+      "grad_norm": 0.2718486487865448,
+      "learning_rate": 0.00016198142252170047,
+      "loss": 1.5929,
+      "step": 68900
+    },
+    {
+      "epoch": 0.571305557395509,
+      "grad_norm": 0.24742506444454193,
+      "learning_rate": 0.00016192622391255086,
+      "loss": 1.6409,
+      "step": 69000
+    },
+    {
+      "epoch": 0.5721335364641982,
+      "grad_norm": 0.2923363149166107,
+      "learning_rate": 0.00016187102530340123,
+      "loss": 1.6406,
+      "step": 69100
+    },
+    {
+      "epoch": 0.5729615155328873,
+      "grad_norm": 0.2717527151107788,
+      "learning_rate": 0.00016181582669425165,
+      "loss": 1.6028,
+      "step": 69200
+    },
+    {
+      "epoch": 0.5737894946015765,
+      "grad_norm": 0.24801529943943024,
+      "learning_rate": 0.00016176118007119354,
+      "loss": 1.6032,
+      "step": 69300
+    },
+    {
+      "epoch": 0.5746174736702656,
+      "grad_norm": 0.23781093955039978,
+      "learning_rate": 0.00016170598146204393,
+      "loss": 1.5806,
+      "step": 69400
+    },
+    {
+      "epoch": 0.5754454527389548,
+      "grad_norm": 0.2343488186597824,
+      "learning_rate": 0.00016165078285289433,
+      "loss": 1.6181,
+      "step": 69500
+    },
+    {
+      "epoch": 0.5762734318076439,
+      "grad_norm": 0.2726878225803375,
+      "learning_rate": 0.00016159558424374472,
+      "loss": 1.5645,
+      "step": 69600
+    },
+    {
+      "epoch": 0.577101410876333,
+      "grad_norm": 0.24465055763721466,
+      "learning_rate": 0.00016154038563459511,
+      "loss": 1.6262,
+      "step": 69700
+    },
+    {
+      "epoch": 0.5779293899450222,
+      "grad_norm": 0.23371031880378723,
+      "learning_rate": 0.0001614851870254455,
+      "loss": 1.6486,
+      "step": 69800
+    },
+    {
+      "epoch": 0.5787573690137113,
+      "grad_norm": 0.2924565076828003,
+      "learning_rate": 0.0001614299884162959,
+      "loss": 1.5835,
+      "step": 69900
+    },
+    {
+      "epoch": 0.5795853480824005,
+      "grad_norm": 0.25226888060569763,
+      "learning_rate": 0.0001613747898071463,
+      "loss": 1.595,
+      "step": 70000
+    },
+    {
+      "epoch": 0.5804133271510896,
+      "grad_norm": 0.24092262983322144,
+      "learning_rate": 0.0001613195911979967,
+      "loss": 1.6115,
+      "step": 70100
+    },
+    {
+      "epoch": 0.5812413062197788,
+      "grad_norm": 0.26227688789367676,
+      "learning_rate": 0.00016126439258884708,
+      "loss": 1.6405,
+      "step": 70200
+    },
+    {
+      "epoch": 0.5820692852884679,
+      "grad_norm": 0.24400807917118073,
+      "learning_rate": 0.00016120919397969748,
+      "loss": 1.6298,
+      "step": 70300
+    },
+    {
+      "epoch": 0.582897264357157,
+      "grad_norm": 0.24254167079925537,
+      "learning_rate": 0.00016115399537054787,
+      "loss": 1.5931,
+      "step": 70400
+    },
+    {
+      "epoch": 0.5837252434258462,
+      "grad_norm": 0.2771003246307373,
+      "learning_rate": 0.00016109879676139827,
+      "loss": 1.6287,
+      "step": 70500
+    },
+    {
+      "epoch": 0.5845532224945353,
+      "grad_norm": 0.2865670323371887,
+      "learning_rate": 0.00016104359815224866,
+      "loss": 1.6418,
+      "step": 70600
+    },
+    {
+      "epoch": 0.5853812015632245,
+      "grad_norm": 0.24516509473323822,
+      "learning_rate": 0.00016098839954309906,
+      "loss": 1.6244,
+      "step": 70700
+    },
+    {
+      "epoch": 0.5862091806319136,
+      "grad_norm": 0.28212985396385193,
+      "learning_rate": 0.00016093320093394945,
+      "loss": 1.6139,
+      "step": 70800
+    },
+    {
+      "epoch": 0.5870371597006028,
+      "grad_norm": 0.2579030990600586,
+      "learning_rate": 0.00016087800232479984,
+      "loss": 1.5961,
+      "step": 70900
+    },
+    {
+      "epoch": 0.5878651387692919,
+      "grad_norm": 0.22696354985237122,
+      "learning_rate": 0.00016082280371565024,
+      "loss": 1.6209,
+      "step": 71000
+    },
+    {
+      "epoch": 0.5886931178379811,
+      "grad_norm": 0.2646586298942566,
+      "learning_rate": 0.00016076760510650063,
+      "loss": 1.6111,
+      "step": 71100
+    },
+    {
+      "epoch": 0.5895210969066702,
+      "grad_norm": 0.25043609738349915,
+      "learning_rate": 0.00016071240649735103,
+      "loss": 1.5784,
+      "step": 71200
+    },
+    {
+      "epoch": 0.5903490759753593,
+      "grad_norm": 0.25381940603256226,
+      "learning_rate": 0.00016065720788820142,
+      "loss": 1.5925,
+      "step": 71300
+    },
+    {
+      "epoch": 0.5911770550440485,
+      "grad_norm": 0.2380959391593933,
+      "learning_rate": 0.00016060200927905182,
+      "loss": 1.6379,
+      "step": 71400
+    },
+    {
+      "epoch": 0.5920050341127376,
+      "grad_norm": 0.2419148087501526,
+      "learning_rate": 0.0001605479146420852,
+      "loss": 1.6047,
+      "step": 71500
+    },
+    {
+      "epoch": 0.5928330131814268,
+      "grad_norm": 0.2510753273963928,
+      "learning_rate": 0.0001604927160329356,
+      "loss": 1.6109,
+      "step": 71600
+    },
+    {
+      "epoch": 0.5936609922501159,
+      "grad_norm": 0.2659463584423065,
+      "learning_rate": 0.00016043751742378598,
+      "loss": 1.587,
+      "step": 71700
+    },
+    {
+      "epoch": 0.5944889713188051,
+      "grad_norm": 0.28224682807922363,
+      "learning_rate": 0.00016038231881463638,
+      "loss": 1.6516,
+      "step": 71800
+    },
+    {
+      "epoch": 0.5953169503874942,
+      "grad_norm": 0.24091459810733795,
+      "learning_rate": 0.00016032712020548677,
+      "loss": 1.6471,
+      "step": 71900
+    },
+    {
+      "epoch": 0.5961449294561834,
+      "grad_norm": 0.258425772190094,
+      "learning_rate": 0.00016027192159633717,
+      "loss": 1.5975,
+      "step": 72000
+    },
+    {
+      "epoch": 0.5969729085248725,
+      "grad_norm": 0.2684045135974884,
+      "learning_rate": 0.0001602167229871876,
+      "loss": 1.5932,
+      "step": 72100
+    },
+    {
+      "epoch": 0.5978008875935616,
+      "grad_norm": 0.19851814210414886,
+      "learning_rate": 0.00016016152437803795,
+      "loss": 1.6163,
+      "step": 72200
+    },
+    {
+      "epoch": 0.5986288666622508,
+      "grad_norm": 0.28404444456100464,
+      "learning_rate": 0.00016010632576888835,
+      "loss": 1.5884,
+      "step": 72300
+    },
+    {
+      "epoch": 0.5994568457309399,
+      "grad_norm": 0.24942117929458618,
+      "learning_rate": 0.00016005112715973877,
+      "loss": 1.6062,
+      "step": 72400
+    },
+    {
+      "epoch": 0.6002848247996291,
+      "grad_norm": 0.319736123085022,
+      "learning_rate": 0.00015999592855058914,
+      "loss": 1.6107,
+      "step": 72500
+    },
+    {
+      "epoch": 0.6011128038683182,
+      "grad_norm": 0.297432541847229,
+      "learning_rate": 0.00015994072994143953,
+      "loss": 1.6253,
+      "step": 72600
+    },
+    {
+      "epoch": 0.6019407829370074,
+      "grad_norm": 0.283583402633667,
+      "learning_rate": 0.00015988553133228995,
+      "loss": 1.6102,
+      "step": 72700
+    },
+    {
+      "epoch": 0.6027687620056965,
+      "grad_norm": 0.27938199043273926,
+      "learning_rate": 0.00015983033272314032,
+      "loss": 1.6206,
+      "step": 72800
+    },
+    {
+      "epoch": 0.6035967410743857,
+      "grad_norm": 0.2605910301208496,
+      "learning_rate": 0.0001597751341139907,
+      "loss": 1.6037,
+      "step": 72900
+    },
+    {
+      "epoch": 0.6044247201430748,
+      "grad_norm": 0.28445005416870117,
+      "learning_rate": 0.00015971993550484113,
+      "loss": 1.646,
+      "step": 73000
+    },
+    {
+      "epoch": 0.605252699211764,
+      "grad_norm": 0.26064708828926086,
+      "learning_rate": 0.0001596647368956915,
+      "loss": 1.614,
+      "step": 73100
+    },
+    {
+      "epoch": 0.6060806782804531,
+      "grad_norm": 0.2497030645608902,
+      "learning_rate": 0.0001596095382865419,
+      "loss": 1.6076,
+      "step": 73200
+    },
+    {
+      "epoch": 0.6069086573491422,
+      "grad_norm": 0.3160170614719391,
+      "learning_rate": 0.00015955433967739232,
+      "loss": 1.6347,
+      "step": 73300
+    },
+    {
+      "epoch": 0.6077366364178314,
+      "grad_norm": 0.2508627772331238,
+      "learning_rate": 0.00015949914106824268,
+      "loss": 1.6461,
+      "step": 73400
+    },
+    {
+      "epoch": 0.6085646154865205,
+      "grad_norm": 0.29909202456474304,
+      "learning_rate": 0.0001594444944451846,
+      "loss": 1.6179,
+      "step": 73500
+    },
+    {
+      "epoch": 0.6093925945552097,
+      "grad_norm": 0.22416061162948608,
+      "learning_rate": 0.000159389295836035,
+      "loss": 1.6013,
+      "step": 73600
+    },
+    {
+      "epoch": 0.6102205736238988,
+      "grad_norm": 0.3153701424598694,
+      "learning_rate": 0.00015933409722688536,
+      "loss": 1.6494,
+      "step": 73700
+    },
+    {
+      "epoch": 0.611048552692588,
+      "grad_norm": 0.2792322635650635,
+      "learning_rate": 0.00015927889861773578,
+      "loss": 1.6055,
+      "step": 73800
+    },
+    {
+      "epoch": 0.6118765317612771,
+      "grad_norm": 0.3239234685897827,
+      "learning_rate": 0.00015922370000858617,
+      "loss": 1.5841,
+      "step": 73900
+    },
+    {
+      "epoch": 0.6127045108299662,
+      "grad_norm": 0.4340994656085968,
+      "learning_rate": 0.00015916850139943654,
+      "loss": 1.6187,
+      "step": 74000
+    },
+    {
+      "epoch": 0.6135324898986554,
+      "grad_norm": 0.31595396995544434,
+      "learning_rate": 0.00015911330279028696,
+      "loss": 1.6438,
+      "step": 74100
+    },
+    {
+      "epoch": 0.6143604689673445,
+      "grad_norm": 0.22218069434165955,
+      "learning_rate": 0.00015905810418113736,
+      "loss": 1.6298,
+      "step": 74200
+    },
+    {
+      "epoch": 0.6151884480360337,
+      "grad_norm": 0.22682014107704163,
+      "learning_rate": 0.00015900290557198772,
+      "loss": 1.5793,
+      "step": 74300
+    },
+    {
+      "epoch": 0.6160164271047228,
+      "grad_norm": 0.23543739318847656,
+      "learning_rate": 0.00015894770696283815,
+      "loss": 1.6292,
+      "step": 74400
+    },
+    {
+      "epoch": 0.616844406173412,
+      "grad_norm": 0.2355438470840454,
+      "learning_rate": 0.00015889250835368854,
+      "loss": 1.636,
+      "step": 74500
+    },
+    {
+      "epoch": 0.6176723852421011,
+      "grad_norm": 0.2809857428073883,
+      "learning_rate": 0.0001588373097445389,
+      "loss": 1.6081,
+      "step": 74600
+    },
+    {
+      "epoch": 0.6185003643107903,
+      "grad_norm": 0.2578040361404419,
+      "learning_rate": 0.00015878211113538933,
+      "loss": 1.6238,
+      "step": 74700
+    },
+    {
+      "epoch": 0.6193283433794794,
+      "grad_norm": 0.23350849747657776,
+      "learning_rate": 0.00015872691252623972,
+      "loss": 1.627,
+      "step": 74800
+    },
+    {
+      "epoch": 0.6201563224481685,
+      "grad_norm": 0.28163543343544006,
+      "learning_rate": 0.00015867171391709012,
+      "loss": 1.6375,
+      "step": 74900
+    },
+    {
+      "epoch": 0.6209843015168577,
+      "grad_norm": 0.2725831866264343,
+      "learning_rate": 0.0001586165153079405,
+      "loss": 1.6054,
+      "step": 75000
+    },
+    {
+      "epoch": 0.6218122805855468,
+      "grad_norm": 0.294609397649765,
+      "learning_rate": 0.0001585613166987909,
+      "loss": 1.6532,
+      "step": 75100
+    },
+    {
+      "epoch": 0.622640259654236,
+      "grad_norm": 0.2077389657497406,
+      "learning_rate": 0.0001585061180896413,
+      "loss": 1.6033,
+      "step": 75200
+    },
+    {
+      "epoch": 0.6234682387229251,
+      "grad_norm": 0.25486066937446594,
+      "learning_rate": 0.0001584509194804917,
+      "loss": 1.6053,
+      "step": 75300
+    },
+    {
+      "epoch": 0.6242962177916143,
+      "grad_norm": 0.27974551916122437,
+      "learning_rate": 0.0001583957208713421,
+      "loss": 1.6003,
+      "step": 75400
+    },
+    {
+      "epoch": 0.6251241968603034,
+      "grad_norm": 0.22559039294719696,
+      "learning_rate": 0.00015834052226219248,
+      "loss": 1.6042,
+      "step": 75500
+    },
+    {
+      "epoch": 0.6259521759289925,
+      "grad_norm": 0.2382129728794098,
+      "learning_rate": 0.00015828532365304288,
+      "loss": 1.6016,
+      "step": 75600
+    },
+    {
+      "epoch": 0.6267801549976817,
+      "grad_norm": 0.22553443908691406,
+      "learning_rate": 0.00015823012504389327,
+      "loss": 1.5928,
+      "step": 75700
+    },
+    {
+      "epoch": 0.6276081340663708,
+      "grad_norm": 0.2299722135066986,
+      "learning_rate": 0.00015817492643474367,
+      "loss": 1.5972,
+      "step": 75800
+    },
+    {
+      "epoch": 0.62843611313506,
+      "grad_norm": 0.2587108910083771,
+      "learning_rate": 0.00015811972782559406,
+      "loss": 1.5785,
+      "step": 75900
+    },
+    {
+      "epoch": 0.6292640922037491,
+      "grad_norm": 0.23947428166866302,
+      "learning_rate": 0.00015806452921644445,
+      "loss": 1.6201,
+      "step": 76000
+    },
+    {
+      "epoch": 0.6300920712724383,
+      "grad_norm": 0.24543558061122894,
+      "learning_rate": 0.00015800933060729485,
+      "loss": 1.595,
+      "step": 76100
+    },
+    {
+      "epoch": 0.6309200503411274,
+      "grad_norm": 0.2749697268009186,
+      "learning_rate": 0.00015795413199814524,
+      "loss": 1.6041,
+      "step": 76200
+    },
+    {
+      "epoch": 0.6317480294098166,
+      "grad_norm": 0.2621932923793793,
+      "learning_rate": 0.00015789893338899564,
+      "loss": 1.6325,
+      "step": 76300
+    },
+    {
+      "epoch": 0.6325760084785057,
+      "grad_norm": 0.23789739608764648,
+      "learning_rate": 0.00015784373477984603,
+      "loss": 1.6411,
+      "step": 76400
+    },
+    {
+      "epoch": 0.6334039875471948,
+      "grad_norm": 0.3237096071243286,
+      "learning_rate": 0.00015778853617069642,
+      "loss": 1.6147,
+      "step": 76500
+    },
+    {
+      "epoch": 0.634231966615884,
+      "grad_norm": 0.28203284740448,
+      "learning_rate": 0.00015773333756154682,
+      "loss": 1.6395,
+      "step": 76600
+    },
+    {
+      "epoch": 0.6350599456845731,
+      "grad_norm": 0.25538840889930725,
+      "learning_rate": 0.0001576781389523972,
+      "loss": 1.5822,
+      "step": 76700
+    },
+    {
+      "epoch": 0.6358879247532623,
+      "grad_norm": 0.23973847925662994,
+      "learning_rate": 0.0001576229403432476,
+      "loss": 1.6125,
+      "step": 76800
+    },
+    {
+      "epoch": 0.6367159038219514,
+      "grad_norm": 0.22504395246505737,
+      "learning_rate": 0.000157567741734098,
+      "loss": 1.5907,
+      "step": 76900
+    },
+    {
+      "epoch": 0.6375438828906406,
+      "grad_norm": 0.21240180730819702,
+      "learning_rate": 0.0001575125431249484,
+      "loss": 1.5818,
+      "step": 77000
+    },
+    {
+      "epoch": 0.6383718619593297,
+      "grad_norm": 0.23655393719673157,
+      "learning_rate": 0.0001574573445157988,
+      "loss": 1.6278,
+      "step": 77100
+    },
+    {
+      "epoch": 0.6391998410280189,
+      "grad_norm": 0.24175895750522614,
+      "learning_rate": 0.00015740214590664918,
+      "loss": 1.6079,
+      "step": 77200
+    },
+    {
+      "epoch": 0.640027820096708,
+      "grad_norm": 0.2233435958623886,
+      "learning_rate": 0.00015734694729749958,
+      "loss": 1.6276,
+      "step": 77300
+    },
+    {
+      "epoch": 0.6408557991653971,
+      "grad_norm": 0.2756839990615845,
+      "learning_rate": 0.00015729174868834997,
+      "loss": 1.6015,
+      "step": 77400
+    },
+    {
+      "epoch": 0.6416837782340863,
+      "grad_norm": 0.24636799097061157,
+      "learning_rate": 0.00015723655007920037,
+      "loss": 1.5968,
+      "step": 77500
+    },
+    {
+      "epoch": 0.6425117573027754,
+      "grad_norm": 0.28196850419044495,
+      "learning_rate": 0.00015718135147005076,
+      "loss": 1.6366,
+      "step": 77600
+    },
+    {
+      "epoch": 0.6433397363714646,
+      "grad_norm": 0.2543933093547821,
+      "learning_rate": 0.00015712670484699265,
+      "loss": 1.6481,
+      "step": 77700
+    },
+    {
+      "epoch": 0.6441677154401537,
+      "grad_norm": 0.28510308265686035,
+      "learning_rate": 0.00015707150623784304,
+      "loss": 1.5877,
+      "step": 77800
+    },
+    {
+      "epoch": 0.6449956945088429,
+      "grad_norm": 0.22589468955993652,
+      "learning_rate": 0.00015701630762869344,
+      "loss": 1.5924,
+      "step": 77900
+    },
+    {
+      "epoch": 0.645823673577532,
+      "grad_norm": 0.33463847637176514,
+      "learning_rate": 0.00015696110901954383,
+      "loss": 1.6093,
+      "step": 78000
+    },
+    {
+      "epoch": 0.6466516526462212,
+      "grad_norm": 0.23685209453105927,
+      "learning_rate": 0.00015690591041039422,
+      "loss": 1.5936,
+      "step": 78100
+    },
+    {
+      "epoch": 0.6474796317149103,
+      "grad_norm": 0.2393782138824463,
+      "learning_rate": 0.00015685071180124462,
+      "loss": 1.5973,
+      "step": 78200
+    },
+    {
+      "epoch": 0.6483076107835994,
+      "grad_norm": 0.2803311049938202,
+      "learning_rate": 0.000156795513192095,
+      "loss": 1.6438,
+      "step": 78300
+    },
+    {
+      "epoch": 0.6491355898522886,
+      "grad_norm": 0.24263189733028412,
+      "learning_rate": 0.0001567403145829454,
+      "loss": 1.6008,
+      "step": 78400
+    }
+  ],
+  "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": 2.7106979758766653e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}