diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.6003632560955474,
+  "epoch": 0.650393527436843,
   "eval_steps": 500,
-  "global_step": 10908,
+  "global_step": 11817,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -76363,6 +76363,6369 @@
       "learning_rate": 7.961114987144781e-06,
       "loss": 0.6374,
       "step": 10908
+    },
+    {
+      "epoch": 0.600418294897903,
+      "grad_norm": 0.7149941921234131,
+      "learning_rate": 7.960765699838854e-06,
+      "loss": 0.8422,
+      "step": 10909
+    },
+    {
+      "epoch": 0.6004733337002587,
+      "grad_norm": 0.7040171027183533,
+      "learning_rate": 7.960416390280608e-06,
+      "loss": 0.8261,
+      "step": 10910
+    },
+    {
+      "epoch": 0.6005283725026144,
+      "grad_norm": 0.713591456413269,
+      "learning_rate": 7.960067058472663e-06,
+      "loss": 0.7908,
+      "step": 10911
+    },
+    {
+      "epoch": 0.60058341130497,
+      "grad_norm": 0.654086172580719,
+      "learning_rate": 7.959717704417645e-06,
+      "loss": 0.6971,
+      "step": 10912
+    },
+    {
+      "epoch": 0.6006384501073256,
+      "grad_norm": 0.7293223738670349,
+      "learning_rate": 7.959368328118183e-06,
+      "loss": 0.7032,
+      "step": 10913
+    },
+    {
+      "epoch": 0.6006934889096813,
+      "grad_norm": 0.705434262752533,
+      "learning_rate": 7.959018929576898e-06,
+      "loss": 0.7193,
+      "step": 10914
+    },
+    {
+      "epoch": 0.600748527712037,
+      "grad_norm": 0.7406907677650452,
+      "learning_rate": 7.958669508796422e-06,
+      "loss": 0.8464,
+      "step": 10915
+    },
+    {
+      "epoch": 0.6008035665143926,
+      "grad_norm": 0.6683858036994934,
+      "learning_rate": 7.958320065779377e-06,
+      "loss": 0.699,
+      "step": 10916
+    },
+    {
+      "epoch": 0.6008586053167483,
+      "grad_norm": 0.7380560636520386,
+      "learning_rate": 7.95797060052839e-06,
+      "loss": 0.7409,
+      "step": 10917
+    },
+    {
+      "epoch": 0.600913644119104,
+      "grad_norm": 0.7729377746582031,
+      "learning_rate": 7.957621113046088e-06,
+      "loss": 0.8838,
+      "step": 10918
+    },
+    {
+      "epoch": 0.6009686829214597,
+      "grad_norm": 0.6842743158340454,
+      "learning_rate": 7.957271603335097e-06,
+      "loss": 0.781,
+      "step": 10919
+    },
+    {
+      "epoch": 0.6010237217238152,
+      "grad_norm": 0.6864648461341858,
+      "learning_rate": 7.956922071398045e-06,
+      "loss": 0.6717,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6010787605261709,
+      "grad_norm": 0.7718262672424316,
+      "learning_rate": 7.956572517237557e-06,
+      "loss": 0.8023,
+      "step": 10921
+    },
+    {
+      "epoch": 0.6011337993285266,
+      "grad_norm": 0.686338484287262,
+      "learning_rate": 7.956222940856261e-06,
+      "loss": 0.7139,
+      "step": 10922
+    },
+    {
+      "epoch": 0.6011888381308823,
+      "grad_norm": 0.7064465284347534,
+      "learning_rate": 7.955873342256789e-06,
+      "loss": 0.845,
+      "step": 10923
+    },
+    {
+      "epoch": 0.6012438769332379,
+      "grad_norm": 0.6847875714302063,
+      "learning_rate": 7.955523721441761e-06,
+      "loss": 0.7078,
+      "step": 10924
+    },
+    {
+      "epoch": 0.6012989157355936,
+      "grad_norm": 0.6879494786262512,
+      "learning_rate": 7.955174078413806e-06,
+      "loss": 0.7532,
+      "step": 10925
+    },
+    {
+      "epoch": 0.6013539545379493,
+      "grad_norm": 0.6569855213165283,
+      "learning_rate": 7.954824413175554e-06,
+      "loss": 0.7529,
+      "step": 10926
+    },
+    {
+      "epoch": 0.601408993340305,
+      "grad_norm": 0.6225974559783936,
+      "learning_rate": 7.954474725729635e-06,
+      "loss": 0.6595,
+      "step": 10927
+    },
+    {
+      "epoch": 0.6014640321426605,
+      "grad_norm": 0.7067761421203613,
+      "learning_rate": 7.954125016078675e-06,
+      "loss": 0.7851,
+      "step": 10928
+    },
+    {
+      "epoch": 0.6015190709450162,
+      "grad_norm": 0.683030903339386,
+      "learning_rate": 7.9537752842253e-06,
+      "loss": 0.7461,
+      "step": 10929
+    },
+    {
+      "epoch": 0.6015741097473719,
+      "grad_norm": 0.6411080956459045,
+      "learning_rate": 7.953425530172143e-06,
+      "loss": 0.6945,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6016291485497276,
+      "grad_norm": 0.6254550814628601,
+      "learning_rate": 7.953075753921829e-06,
+      "loss": 0.7143,
+      "step": 10931
+    },
+    {
+      "epoch": 0.6016841873520832,
+      "grad_norm": 0.684100866317749,
+      "learning_rate": 7.952725955476987e-06,
+      "loss": 0.8137,
+      "step": 10932
+    },
+    {
+      "epoch": 0.6017392261544389,
+      "grad_norm": 0.6341036558151245,
+      "learning_rate": 7.95237613484025e-06,
+      "loss": 0.6692,
+      "step": 10933
+    },
+    {
+      "epoch": 0.6017942649567946,
+      "grad_norm": 0.7311153411865234,
+      "learning_rate": 7.952026292014242e-06,
+      "loss": 0.7091,
+      "step": 10934
+    },
+    {
+      "epoch": 0.6018493037591502,
+      "grad_norm": 0.7265943884849548,
+      "learning_rate": 7.951676427001596e-06,
+      "loss": 0.765,
+      "step": 10935
+    },
+    {
+      "epoch": 0.6019043425615058,
+      "grad_norm": 0.8777397274971008,
+      "learning_rate": 7.951326539804938e-06,
+      "loss": 0.7824,
+      "step": 10936
+    },
+    {
+      "epoch": 0.6019593813638615,
+      "grad_norm": 0.7241179347038269,
+      "learning_rate": 7.9509766304269e-06,
+      "loss": 0.7913,
+      "step": 10937
+    },
+    {
+      "epoch": 0.6020144201662172,
+      "grad_norm": 0.8090667128562927,
+      "learning_rate": 7.950626698870113e-06,
+      "loss": 0.8208,
+      "step": 10938
+    },
+    {
+      "epoch": 0.6020694589685729,
+      "grad_norm": 0.7376043796539307,
+      "learning_rate": 7.950276745137206e-06,
+      "loss": 0.7176,
+      "step": 10939
+    },
+    {
+      "epoch": 0.6021244977709285,
+      "grad_norm": 0.7149157524108887,
+      "learning_rate": 7.949926769230809e-06,
+      "loss": 0.7949,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6021795365732842,
+      "grad_norm": 0.8721579909324646,
+      "learning_rate": 7.949576771153549e-06,
+      "loss": 0.8433,
+      "step": 10941
+    },
+    {
+      "epoch": 0.6022345753756398,
+      "grad_norm": 0.7946182489395142,
+      "learning_rate": 7.949226750908062e-06,
+      "loss": 0.7412,
+      "step": 10942
+    },
+    {
+      "epoch": 0.6022896141779955,
+      "grad_norm": 0.6661237478256226,
+      "learning_rate": 7.948876708496975e-06,
+      "loss": 0.725,
+      "step": 10943
+    },
+    {
+      "epoch": 0.6023446529803511,
+      "grad_norm": 0.8346213698387146,
+      "learning_rate": 7.948526643922922e-06,
+      "loss": 0.6817,
+      "step": 10944
+    },
+    {
+      "epoch": 0.6023996917827068,
+      "grad_norm": 0.7911655306816101,
+      "learning_rate": 7.94817655718853e-06,
+      "loss": 0.7398,
+      "step": 10945
+    },
+    {
+      "epoch": 0.6024547305850625,
+      "grad_norm": 0.6480078101158142,
+      "learning_rate": 7.947826448296432e-06,
+      "loss": 0.6822,
+      "step": 10946
+    },
+    {
+      "epoch": 0.6025097693874182,
+      "grad_norm": 0.6950085759162903,
+      "learning_rate": 7.94747631724926e-06,
+      "loss": 0.8073,
+      "step": 10947
+    },
+    {
+      "epoch": 0.6025648081897738,
+      "grad_norm": 0.7142168879508972,
+      "learning_rate": 7.947126164049645e-06,
+      "loss": 0.6159,
+      "step": 10948
+    },
+    {
+      "epoch": 0.6026198469921294,
+      "grad_norm": 0.7459015846252441,
+      "learning_rate": 7.946775988700219e-06,
+      "loss": 0.8377,
+      "step": 10949
+    },
+    {
+      "epoch": 0.6026748857944851,
+      "grad_norm": 1.050179362297058,
+      "learning_rate": 7.946425791203614e-06,
+      "loss": 0.8098,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6027299245968408,
+      "grad_norm": 0.7473265528678894,
+      "learning_rate": 7.94607557156246e-06,
+      "loss": 0.6846,
+      "step": 10951
+    },
+    {
+      "epoch": 0.6027849633991964,
+      "grad_norm": 0.7990789413452148,
+      "learning_rate": 7.945725329779392e-06,
+      "loss": 0.8216,
+      "step": 10952
+    },
+    {
+      "epoch": 0.6028400022015521,
+      "grad_norm": 0.6461700201034546,
+      "learning_rate": 7.94537506585704e-06,
+      "loss": 0.7864,
+      "step": 10953
+    },
+    {
+      "epoch": 0.6028950410039078,
+      "grad_norm": 0.661123514175415,
+      "learning_rate": 7.945024779798038e-06,
+      "loss": 0.7466,
+      "step": 10954
+    },
+    {
+      "epoch": 0.6029500798062634,
+      "grad_norm": 0.6998088359832764,
+      "learning_rate": 7.944674471605018e-06,
+      "loss": 0.7846,
+      "step": 10955
+    },
+    {
+      "epoch": 0.603005118608619,
+      "grad_norm": 0.6917386651039124,
+      "learning_rate": 7.944324141280613e-06,
+      "loss": 0.7699,
+      "step": 10956
+    },
+    {
+      "epoch": 0.6030601574109747,
+      "grad_norm": 0.7304503321647644,
+      "learning_rate": 7.943973788827455e-06,
+      "loss": 0.8015,
+      "step": 10957
+    },
+    {
+      "epoch": 0.6031151962133304,
+      "grad_norm": 0.7996858358383179,
+      "learning_rate": 7.94362341424818e-06,
+      "loss": 0.7093,
+      "step": 10958
+    },
+    {
+      "epoch": 0.603170235015686,
+      "grad_norm": 0.7445322871208191,
+      "learning_rate": 7.943273017545419e-06,
+      "loss": 0.7388,
+      "step": 10959
+    },
+    {
+      "epoch": 0.6032252738180417,
+      "grad_norm": 0.6672174334526062,
+      "learning_rate": 7.942922598721805e-06,
+      "loss": 0.7703,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6032803126203974,
+      "grad_norm": 0.7313557267189026,
+      "learning_rate": 7.94257215777997e-06,
+      "loss": 0.6637,
+      "step": 10961
+    },
+    {
+      "epoch": 0.6033353514227531,
+      "grad_norm": 0.7248823642730713,
+      "learning_rate": 7.942221694722553e-06,
+      "loss": 0.836,
+      "step": 10962
+    },
+    {
+      "epoch": 0.6033903902251087,
+      "grad_norm": 0.6583372354507446,
+      "learning_rate": 7.941871209552187e-06,
+      "loss": 0.7582,
+      "step": 10963
+    },
+    {
+      "epoch": 0.6034454290274643,
+      "grad_norm": 0.7502591013908386,
+      "learning_rate": 7.941520702271503e-06,
+      "loss": 0.7455,
+      "step": 10964
+    },
+    {
+      "epoch": 0.60350046782982,
+      "grad_norm": 0.6899349689483643,
+      "learning_rate": 7.941170172883135e-06,
+      "loss": 0.7677,
+      "step": 10965
+    },
+    {
+      "epoch": 0.6035555066321757,
+      "grad_norm": 0.693321943283081,
+      "learning_rate": 7.940819621389722e-06,
+      "loss": 0.7754,
+      "step": 10966
+    },
+    {
+      "epoch": 0.6036105454345313,
+      "grad_norm": 0.7376342415809631,
+      "learning_rate": 7.940469047793893e-06,
+      "loss": 0.7761,
+      "step": 10967
+    },
+    {
+      "epoch": 0.603665584236887,
+      "grad_norm": 0.6377952694892883,
+      "learning_rate": 7.940118452098289e-06,
+      "loss": 0.6612,
+      "step": 10968
+    },
+    {
+      "epoch": 0.6037206230392427,
+      "grad_norm": 0.8041388988494873,
+      "learning_rate": 7.939767834305538e-06,
+      "loss": 0.8358,
+      "step": 10969
+    },
+    {
+      "epoch": 0.6037756618415984,
+      "grad_norm": 1.5993521213531494,
+      "learning_rate": 7.939417194418282e-06,
+      "loss": 0.8536,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6038307006439539,
+      "grad_norm": 0.6718295216560364,
+      "learning_rate": 7.939066532439153e-06,
+      "loss": 0.717,
+      "step": 10971
+    },
+    {
+      "epoch": 0.6038857394463096,
+      "grad_norm": 0.7951062917709351,
+      "learning_rate": 7.938715848370787e-06,
+      "loss": 0.6919,
+      "step": 10972
+    },
+    {
+      "epoch": 0.6039407782486653,
+      "grad_norm": 0.707804262638092,
+      "learning_rate": 7.938365142215816e-06,
+      "loss": 0.7346,
+      "step": 10973
+    },
+    {
+      "epoch": 0.603995817051021,
+      "grad_norm": 0.7244500517845154,
+      "learning_rate": 7.938014413976883e-06,
+      "loss": 0.708,
+      "step": 10974
+    },
+    {
+      "epoch": 0.6040508558533766,
+      "grad_norm": 0.7533566951751709,
+      "learning_rate": 7.937663663656617e-06,
+      "loss": 0.6761,
+      "step": 10975
+    },
+    {
+      "epoch": 0.6041058946557323,
+      "grad_norm": 0.8844665288925171,
+      "learning_rate": 7.93731289125766e-06,
+      "loss": 0.7833,
+      "step": 10976
+    },
+    {
+      "epoch": 0.604160933458088,
+      "grad_norm": 0.6413047313690186,
+      "learning_rate": 7.936962096782643e-06,
+      "loss": 0.7175,
+      "step": 10977
+    },
+    {
+      "epoch": 0.6042159722604437,
+      "grad_norm": 0.765943706035614,
+      "learning_rate": 7.936611280234206e-06,
+      "loss": 0.7654,
+      "step": 10978
+    },
+    {
+      "epoch": 0.6042710110627992,
+      "grad_norm": 0.6833398938179016,
+      "learning_rate": 7.936260441614985e-06,
+      "loss": 0.7459,
+      "step": 10979
+    },
+    {
+      "epoch": 0.6043260498651549,
+      "grad_norm": 0.6363481283187866,
+      "learning_rate": 7.935909580927617e-06,
+      "loss": 0.7173,
+      "step": 10980
+    },
+    {
+      "epoch": 0.6043810886675106,
+      "grad_norm": 0.7731046080589294,
+      "learning_rate": 7.935558698174738e-06,
+      "loss": 0.8428,
+      "step": 10981
+    },
+    {
+      "epoch": 0.6044361274698663,
+      "grad_norm": 0.7346602082252502,
+      "learning_rate": 7.935207793358986e-06,
+      "loss": 0.832,
+      "step": 10982
+    },
+    {
+      "epoch": 0.6044911662722219,
+      "grad_norm": 0.6711193919181824,
+      "learning_rate": 7.934856866482998e-06,
+      "loss": 0.742,
+      "step": 10983
+    },
+    {
+      "epoch": 0.6045462050745776,
+      "grad_norm": 0.6931266784667969,
+      "learning_rate": 7.934505917549411e-06,
+      "loss": 0.7779,
+      "step": 10984
+    },
+    {
+      "epoch": 0.6046012438769333,
+      "grad_norm": 0.7624725699424744,
+      "learning_rate": 7.934154946560862e-06,
+      "loss": 0.7229,
+      "step": 10985
+    },
+    {
+      "epoch": 0.604656282679289,
+      "grad_norm": 0.6594272255897522,
+      "learning_rate": 7.933803953519991e-06,
+      "loss": 0.7776,
+      "step": 10986
+    },
+    {
+      "epoch": 0.6047113214816445,
+      "grad_norm": 0.674521803855896,
+      "learning_rate": 7.933452938429435e-06,
+      "loss": 0.6904,
+      "step": 10987
+    },
+    {
+      "epoch": 0.6047663602840002,
+      "grad_norm": 0.7352569699287415,
+      "learning_rate": 7.933101901291831e-06,
+      "loss": 0.7655,
+      "step": 10988
+    },
+    {
+      "epoch": 0.6048213990863559,
+      "grad_norm": 0.8560347557067871,
+      "learning_rate": 7.932750842109817e-06,
+      "loss": 0.7894,
+      "step": 10989
+    },
+    {
+      "epoch": 0.6048764378887116,
+      "grad_norm": 0.769496500492096,
+      "learning_rate": 7.932399760886037e-06,
+      "loss": 0.8255,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6049314766910672,
+      "grad_norm": 0.9399588108062744,
+      "learning_rate": 7.932048657623122e-06,
+      "loss": 0.8554,
+      "step": 10991
+    },
+    {
+      "epoch": 0.6049865154934229,
+      "grad_norm": 0.6662001609802246,
+      "learning_rate": 7.931697532323716e-06,
+      "loss": 0.7788,
+      "step": 10992
+    },
+    {
+      "epoch": 0.6050415542957785,
+      "grad_norm": 0.758263111114502,
+      "learning_rate": 7.931346384990455e-06,
+      "loss": 0.7907,
+      "step": 10993
+    },
+    {
+      "epoch": 0.6050965930981342,
+      "grad_norm": 0.7283937335014343,
+      "learning_rate": 7.930995215625978e-06,
+      "loss": 0.8415,
+      "step": 10994
+    },
+    {
+      "epoch": 0.6051516319004898,
+      "grad_norm": 0.6611599922180176,
+      "learning_rate": 7.930644024232927e-06,
+      "loss": 0.7145,
+      "step": 10995
+    },
+    {
+      "epoch": 0.6052066707028455,
+      "grad_norm": 0.8450857400894165,
+      "learning_rate": 7.93029281081394e-06,
+      "loss": 0.7208,
+      "step": 10996
+    },
+    {
+      "epoch": 0.6052617095052012,
+      "grad_norm": 0.649010181427002,
+      "learning_rate": 7.929941575371655e-06,
+      "loss": 0.6928,
+      "step": 10997
+    },
+    {
+      "epoch": 0.6053167483075568,
+      "grad_norm": 0.7022100687026978,
+      "learning_rate": 7.929590317908718e-06,
+      "loss": 0.7329,
+      "step": 10998
+    },
+    {
+      "epoch": 0.6053717871099125,
+      "grad_norm": 0.768598198890686,
+      "learning_rate": 7.92923903842776e-06,
+      "loss": 0.7799,
+      "step": 10999
+    },
+    {
+      "epoch": 0.6054268259122682,
+      "grad_norm": 0.6648436784744263,
+      "learning_rate": 7.928887736931428e-06,
+      "loss": 0.7728,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6054818647146238,
+      "grad_norm": 0.6946157813072205,
+      "learning_rate": 7.928536413422357e-06,
+      "loss": 0.7609,
+      "step": 11001
+    },
+    {
+      "epoch": 0.6055369035169794,
+      "grad_norm": 0.7779337167739868,
+      "learning_rate": 7.928185067903191e-06,
+      "loss": 0.7679,
+      "step": 11002
+    },
+    {
+      "epoch": 0.6055919423193351,
+      "grad_norm": 0.6520814895629883,
+      "learning_rate": 7.927833700376573e-06,
+      "loss": 0.6734,
+      "step": 11003
+    },
+    {
+      "epoch": 0.6056469811216908,
+      "grad_norm": 0.7724258899688721,
+      "learning_rate": 7.927482310845138e-06,
+      "loss": 0.7564,
+      "step": 11004
+    },
+    {
+      "epoch": 0.6057020199240465,
+      "grad_norm": 0.6649174690246582,
+      "learning_rate": 7.927130899311529e-06,
+      "loss": 0.7217,
+      "step": 11005
+    },
+    {
+      "epoch": 0.6057570587264021,
+      "grad_norm": 0.6807287931442261,
+      "learning_rate": 7.926779465778389e-06,
+      "loss": 0.6966,
+      "step": 11006
+    },
+    {
+      "epoch": 0.6058120975287578,
+      "grad_norm": 0.6644826531410217,
+      "learning_rate": 7.926428010248357e-06,
+      "loss": 0.7238,
+      "step": 11007
+    },
+    {
+      "epoch": 0.6058671363311134,
+      "grad_norm": 0.7533535957336426,
+      "learning_rate": 7.926076532724077e-06,
+      "loss": 0.855,
+      "step": 11008
+    },
+    {
+      "epoch": 0.6059221751334691,
+      "grad_norm": 0.6457169055938721,
+      "learning_rate": 7.925725033208187e-06,
+      "loss": 0.6717,
+      "step": 11009
+    },
+    {
+      "epoch": 0.6059772139358247,
+      "grad_norm": 0.724719762802124,
+      "learning_rate": 7.925373511703332e-06,
+      "loss": 0.8701,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6060322527381804,
+      "grad_norm": 0.746755063533783,
+      "learning_rate": 7.925021968212153e-06,
+      "loss": 0.8509,
+      "step": 11011
+    },
+    {
+      "epoch": 0.6060872915405361,
+      "grad_norm": 0.7377174496650696,
+      "learning_rate": 7.924670402737292e-06,
+      "loss": 0.8053,
+      "step": 11012
+    },
+    {
+      "epoch": 0.6061423303428918,
+      "grad_norm": 0.9791839718818665,
+      "learning_rate": 7.92431881528139e-06,
+      "loss": 0.7893,
+      "step": 11013
+    },
+    {
+      "epoch": 0.6061973691452474,
+      "grad_norm": 0.7472195029258728,
+      "learning_rate": 7.923967205847089e-06,
+      "loss": 0.7195,
+      "step": 11014
+    },
+    {
+      "epoch": 0.606252407947603,
+      "grad_norm": 0.672851026058197,
+      "learning_rate": 7.923615574437037e-06,
+      "loss": 0.8234,
+      "step": 11015
+    },
+    {
+      "epoch": 0.6063074467499587,
+      "grad_norm": 0.739942729473114,
+      "learning_rate": 7.923263921053872e-06,
+      "loss": 0.8582,
+      "step": 11016
+    },
+    {
+      "epoch": 0.6063624855523144,
+      "grad_norm": 0.7337772846221924,
+      "learning_rate": 7.922912245700236e-06,
+      "loss": 0.8008,
+      "step": 11017
+    },
+    {
+      "epoch": 0.60641752435467,
+      "grad_norm": 0.6707174777984619,
+      "learning_rate": 7.922560548378774e-06,
+      "loss": 0.8531,
+      "step": 11018
+    },
+    {
+      "epoch": 0.6064725631570257,
+      "grad_norm": 0.6783839464187622,
+      "learning_rate": 7.922208829092133e-06,
+      "loss": 0.7963,
+      "step": 11019
+    },
+    {
+      "epoch": 0.6065276019593814,
+      "grad_norm": 0.6133253574371338,
+      "learning_rate": 7.92185708784295e-06,
+      "loss": 0.7375,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6065826407617371,
+      "grad_norm": 0.8300097584724426,
+      "learning_rate": 7.921505324633868e-06,
+      "loss": 0.7976,
+      "step": 11021
+    },
+    {
+      "epoch": 0.6066376795640926,
+      "grad_norm": 0.6800658702850342,
+      "learning_rate": 7.921153539467538e-06,
+      "loss": 0.7321,
+      "step": 11022
+    },
+    {
+      "epoch": 0.6066927183664483,
+      "grad_norm": 0.6849787831306458,
+      "learning_rate": 7.920801732346602e-06,
+      "loss": 0.7134,
+      "step": 11023
+    },
+    {
+      "epoch": 0.606747757168804,
+      "grad_norm": 0.7675080895423889,
+      "learning_rate": 7.920449903273697e-06,
+      "loss": 0.7402,
+      "step": 11024
+    },
+    {
+      "epoch": 0.6068027959711597,
+      "grad_norm": 0.7431055903434753,
+      "learning_rate": 7.920098052251476e-06,
+      "loss": 0.7872,
+      "step": 11025
+    },
+    {
+      "epoch": 0.6068578347735153,
+      "grad_norm": 0.6264036297798157,
+      "learning_rate": 7.919746179282577e-06,
+      "loss": 0.7496,
+      "step": 11026
+    },
+    {
+      "epoch": 0.606912873575871,
+      "grad_norm": 0.7800843715667725,
+      "learning_rate": 7.919394284369648e-06,
+      "loss": 0.7917,
+      "step": 11027
+    },
+    {
+      "epoch": 0.6069679123782267,
+      "grad_norm": 0.7665574550628662,
+      "learning_rate": 7.919042367515336e-06,
+      "loss": 0.7905,
+      "step": 11028
+    },
+    {
+      "epoch": 0.6070229511805824,
+      "grad_norm": 0.7473214864730835,
+      "learning_rate": 7.918690428722279e-06,
+      "loss": 0.7732,
+      "step": 11029
+    },
+    {
+      "epoch": 0.6070779899829379,
+      "grad_norm": 0.6717211008071899,
+      "learning_rate": 7.918338467993127e-06,
+      "loss": 0.8221,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6071330287852936,
+      "grad_norm": 0.6745431423187256,
+      "learning_rate": 7.917986485330525e-06,
+      "loss": 0.6899,
+      "step": 11031
+    },
+    {
+      "epoch": 0.6071880675876493,
+      "grad_norm": 0.6838263273239136,
+      "learning_rate": 7.917634480737117e-06,
+      "loss": 0.7133,
+      "step": 11032
+    },
+    {
+      "epoch": 0.607243106390005,
+      "grad_norm": 0.7975682020187378,
+      "learning_rate": 7.91728245421555e-06,
+      "loss": 0.8283,
+      "step": 11033
+    },
+    {
+      "epoch": 0.6072981451923606,
+      "grad_norm": 0.7112031579017639,
+      "learning_rate": 7.916930405768468e-06,
+      "loss": 0.7423,
+      "step": 11034
+    },
+    {
+      "epoch": 0.6073531839947163,
+      "grad_norm": 0.7006776928901672,
+      "learning_rate": 7.91657833539852e-06,
+      "loss": 0.716,
+      "step": 11035
+    },
+    {
+      "epoch": 0.607408222797072,
+      "grad_norm": 0.7523549795150757,
+      "learning_rate": 7.916226243108348e-06,
+      "loss": 0.8591,
+      "step": 11036
+    },
+    {
+      "epoch": 0.6074632615994277,
+      "grad_norm": 0.7257835268974304,
+      "learning_rate": 7.9158741289006e-06,
+      "loss": 0.7471,
+      "step": 11037
+    },
+    {
+      "epoch": 0.6075183004017832,
+      "grad_norm": 0.8100149631500244,
+      "learning_rate": 7.915521992777922e-06,
+      "loss": 0.8373,
+      "step": 11038
+    },
+    {
+      "epoch": 0.6075733392041389,
+      "grad_norm": 0.7781035304069519,
+      "learning_rate": 7.915169834742964e-06,
+      "loss": 0.8471,
+      "step": 11039
+    },
+    {
+      "epoch": 0.6076283780064946,
+      "grad_norm": 0.7426049709320068,
+      "learning_rate": 7.914817654798368e-06,
+      "loss": 0.753,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6076834168088502,
+      "grad_norm": 0.6990010738372803,
+      "learning_rate": 7.914465452946782e-06,
+      "loss": 0.7556,
+      "step": 11041
+    },
+    {
+      "epoch": 0.6077384556112059,
+      "grad_norm": 0.8038754463195801,
+      "learning_rate": 7.914113229190856e-06,
+      "loss": 0.7787,
+      "step": 11042
+    },
+    {
+      "epoch": 0.6077934944135616,
+      "grad_norm": 0.6434115767478943,
+      "learning_rate": 7.913760983533233e-06,
+      "loss": 0.7831,
+      "step": 11043
+    },
+    {
+      "epoch": 0.6078485332159173,
+      "grad_norm": 0.8119033575057983,
+      "learning_rate": 7.913408715976562e-06,
+      "loss": 0.7691,
+      "step": 11044
+    },
+    {
+      "epoch": 0.6079035720182728,
+      "grad_norm": 0.6710149049758911,
+      "learning_rate": 7.913056426523493e-06,
+      "loss": 0.7542,
+      "step": 11045
+    },
+    {
+      "epoch": 0.6079586108206285,
+      "grad_norm": 0.7458183765411377,
+      "learning_rate": 7.912704115176671e-06,
+      "loss": 0.7673,
+      "step": 11046
+    },
+    {
+      "epoch": 0.6080136496229842,
+      "grad_norm": 0.8061705827713013,
+      "learning_rate": 7.912351781938745e-06,
+      "loss": 0.9255,
+      "step": 11047
+    },
+    {
+      "epoch": 0.6080686884253399,
+      "grad_norm": 0.7193130850791931,
+      "learning_rate": 7.91199942681236e-06,
+      "loss": 0.8154,
+      "step": 11048
+    },
+    {
+      "epoch": 0.6081237272276955,
+      "grad_norm": 0.7785167098045349,
+      "learning_rate": 7.911647049800171e-06,
+      "loss": 0.7747,
+      "step": 11049
+    },
+    {
+      "epoch": 0.6081787660300512,
+      "grad_norm": 0.665765106678009,
+      "learning_rate": 7.911294650904818e-06,
+      "loss": 0.7573,
+      "step": 11050
+    },
+    {
+      "epoch": 0.6082338048324069,
+      "grad_norm": 0.7940623760223389,
+      "learning_rate": 7.910942230128956e-06,
+      "loss": 0.6628,
+      "step": 11051
+    },
+    {
+      "epoch": 0.6082888436347625,
+      "grad_norm": 0.8364549875259399,
+      "learning_rate": 7.910589787475232e-06,
+      "loss": 0.8103,
+      "step": 11052
+    },
+    {
+      "epoch": 0.6083438824371181,
+      "grad_norm": 0.6153101325035095,
+      "learning_rate": 7.910237322946292e-06,
+      "loss": 0.76,
+      "step": 11053
+    },
+    {
+      "epoch": 0.6083989212394738,
+      "grad_norm": 0.8381257653236389,
+      "learning_rate": 7.909884836544789e-06,
+      "loss": 0.8366,
+      "step": 11054
+    },
+    {
+      "epoch": 0.6084539600418295,
+      "grad_norm": 0.6602391600608826,
+      "learning_rate": 7.90953232827337e-06,
+      "loss": 0.7389,
+      "step": 11055
+    },
+    {
+      "epoch": 0.6085089988441852,
+      "grad_norm": 0.7329971194267273,
+      "learning_rate": 7.909179798134685e-06,
+      "loss": 0.8217,
+      "step": 11056
+    },
+    {
+      "epoch": 0.6085640376465408,
+      "grad_norm": 0.7319926023483276,
+      "learning_rate": 7.908827246131383e-06,
+      "loss": 0.78,
+      "step": 11057
+    },
+    {
+      "epoch": 0.6086190764488965,
+      "grad_norm": 0.6491387486457825,
+      "learning_rate": 7.908474672266114e-06,
+      "loss": 0.7496,
+      "step": 11058
+    },
+    {
+      "epoch": 0.6086741152512521,
+      "grad_norm": 0.656434953212738,
+      "learning_rate": 7.908122076541529e-06,
+      "loss": 0.7462,
+      "step": 11059
+    },
+    {
+      "epoch": 0.6087291540536078,
+      "grad_norm": 0.6908577680587769,
+      "learning_rate": 7.907769458960275e-06,
+      "loss": 0.7505,
+      "step": 11060
+    },
+    {
+      "epoch": 0.6087841928559634,
+      "grad_norm": 0.774424135684967,
+      "learning_rate": 7.907416819525007e-06,
+      "loss": 0.8275,
+      "step": 11061
+    },
+    {
+      "epoch": 0.6088392316583191,
+      "grad_norm": 0.6796718835830688,
+      "learning_rate": 7.90706415823837e-06,
+      "loss": 0.7606,
+      "step": 11062
+    },
+    {
+      "epoch": 0.6088942704606748,
+      "grad_norm": 0.9576514959335327,
+      "learning_rate": 7.906711475103016e-06,
+      "loss": 0.807,
+      "step": 11063
+    },
+    {
+      "epoch": 0.6089493092630305,
+      "grad_norm": 0.9848490953445435,
+      "learning_rate": 7.9063587701216e-06,
+      "loss": 0.7856,
+      "step": 11064
+    },
+    {
+      "epoch": 0.6090043480653861,
+      "grad_norm": 0.9490165710449219,
+      "learning_rate": 7.906006043296768e-06,
+      "loss": 0.8519,
+      "step": 11065
+    },
+    {
+      "epoch": 0.6090593868677417,
+      "grad_norm": 0.631382942199707,
+      "learning_rate": 7.905653294631172e-06,
+      "loss": 0.7041,
+      "step": 11066
+    },
+    {
+      "epoch": 0.6091144256700974,
+      "grad_norm": 0.6969574093818665,
+      "learning_rate": 7.905300524127464e-06,
+      "loss": 0.7556,
+      "step": 11067
+    },
+    {
+      "epoch": 0.6091694644724531,
+      "grad_norm": 0.6990532279014587,
+      "learning_rate": 7.904947731788295e-06,
+      "loss": 0.799,
+      "step": 11068
+    },
+    {
+      "epoch": 0.6092245032748087,
+      "grad_norm": 0.7216916084289551,
+      "learning_rate": 7.904594917616315e-06,
+      "loss": 0.7617,
+      "step": 11069
+    },
+    {
+      "epoch": 0.6092795420771644,
+      "grad_norm": 0.6874147653579712,
+      "learning_rate": 7.904242081614179e-06,
+      "loss": 0.7616,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6093345808795201,
+      "grad_norm": 0.6909550428390503,
+      "learning_rate": 7.903889223784535e-06,
+      "loss": 0.7649,
+      "step": 11071
+    },
+    {
+      "epoch": 0.6093896196818758,
+      "grad_norm": 0.7796370387077332,
+      "learning_rate": 7.90353634413004e-06,
+      "loss": 0.7557,
+      "step": 11072
+    },
+    {
+      "epoch": 0.6094446584842313,
+      "grad_norm": 0.807448148727417,
+      "learning_rate": 7.903183442653341e-06,
+      "loss": 0.7519,
+      "step": 11073
+    },
+    {
+      "epoch": 0.609499697286587,
+      "grad_norm": 0.846371054649353,
+      "learning_rate": 7.902830519357092e-06,
+      "loss": 0.9342,
+      "step": 11074
+    },
+    {
+      "epoch": 0.6095547360889427,
+      "grad_norm": 1.0386929512023926,
+      "learning_rate": 7.902477574243947e-06,
+      "loss": 0.6802,
+      "step": 11075
+    },
+    {
+      "epoch": 0.6096097748912984,
+      "grad_norm": 0.8011854887008667,
+      "learning_rate": 7.902124607316558e-06,
+      "loss": 0.7756,
+      "step": 11076
+    },
+    {
+      "epoch": 0.609664813693654,
+      "grad_norm": 0.6560170650482178,
+      "learning_rate": 7.901771618577574e-06,
+      "loss": 0.7831,
+      "step": 11077
+    },
+    {
+      "epoch": 0.6097198524960097,
+      "grad_norm": 0.656891942024231,
+      "learning_rate": 7.901418608029655e-06,
+      "loss": 0.7239,
+      "step": 11078
+    },
+    {
+      "epoch": 0.6097748912983654,
+      "grad_norm": 0.7451794743537903,
+      "learning_rate": 7.901065575675448e-06,
+      "loss": 0.7426,
+      "step": 11079
+    },
+    {
+      "epoch": 0.6098299301007211,
+      "grad_norm": 0.6805453300476074,
+      "learning_rate": 7.90071252151761e-06,
+      "loss": 0.7257,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6098849689030766,
+      "grad_norm": 0.7747140526771545,
+      "learning_rate": 7.900359445558791e-06,
+      "loss": 0.8554,
+      "step": 11081
+    },
+    {
+      "epoch": 0.6099400077054323,
+      "grad_norm": 0.7276260256767273,
+      "learning_rate": 7.900006347801649e-06,
+      "loss": 0.7608,
+      "step": 11082
+    },
+    {
+      "epoch": 0.609995046507788,
+      "grad_norm": 0.7496321201324463,
+      "learning_rate": 7.899653228248836e-06,
+      "loss": 0.7707,
+      "step": 11083
+    },
+    {
+      "epoch": 0.6100500853101436,
+      "grad_norm": 0.6810722947120667,
+      "learning_rate": 7.899300086903006e-06,
+      "loss": 0.7425,
+      "step": 11084
+    },
+    {
+      "epoch": 0.6101051241124993,
+      "grad_norm": 0.7245593070983887,
+      "learning_rate": 7.89894692376681e-06,
+      "loss": 0.8404,
+      "step": 11085
+    },
+    {
+      "epoch": 0.610160162914855,
+      "grad_norm": 0.7139402627944946,
+      "learning_rate": 7.898593738842906e-06,
+      "loss": 0.7219,
+      "step": 11086
+    },
+    {
+      "epoch": 0.6102152017172107,
+      "grad_norm": 0.6483772397041321,
+      "learning_rate": 7.898240532133947e-06,
+      "loss": 0.7571,
+      "step": 11087
+    },
+    {
+      "epoch": 0.6102702405195662,
+      "grad_norm": 0.7347467541694641,
+      "learning_rate": 7.89788730364259e-06,
+      "loss": 0.7666,
+      "step": 11088
+    },
+    {
+      "epoch": 0.6103252793219219,
+      "grad_norm": 0.8899261355400085,
+      "learning_rate": 7.897534053371485e-06,
+      "loss": 0.6886,
+      "step": 11089
+    },
+    {
+      "epoch": 0.6103803181242776,
+      "grad_norm": 0.7005650401115417,
+      "learning_rate": 7.89718078132329e-06,
+      "loss": 0.6771,
+      "step": 11090
+    },
+    {
+      "epoch": 0.6104353569266333,
+      "grad_norm": 0.776589035987854,
+      "learning_rate": 7.896827487500662e-06,
+      "loss": 0.7731,
+      "step": 11091
+    },
+    {
+      "epoch": 0.6104903957289889,
+      "grad_norm": 0.7039395570755005,
+      "learning_rate": 7.896474171906252e-06,
+      "loss": 0.7415,
+      "step": 11092
+    },
+    {
+      "epoch": 0.6105454345313446,
+      "grad_norm": 0.7453792095184326,
+      "learning_rate": 7.896120834542718e-06,
+      "loss": 0.8507,
+      "step": 11093
+    },
+    {
+      "epoch": 0.6106004733337003,
+      "grad_norm": 0.7516497373580933,
+      "learning_rate": 7.895767475412717e-06,
+      "loss": 0.8271,
+      "step": 11094
+    },
+    {
+      "epoch": 0.610655512136056,
+      "grad_norm": 0.6751283407211304,
+      "learning_rate": 7.895414094518901e-06,
+      "loss": 0.7788,
+      "step": 11095
+    },
+    {
+      "epoch": 0.6107105509384115,
+      "grad_norm": 0.7240836024284363,
+      "learning_rate": 7.895060691863927e-06,
+      "loss": 0.7507,
+      "step": 11096
+    },
+    {
+      "epoch": 0.6107655897407672,
+      "grad_norm": 0.8286149501800537,
+      "learning_rate": 7.894707267450451e-06,
+      "loss": 0.7033,
+      "step": 11097
+    },
+    {
+      "epoch": 0.6108206285431229,
+      "grad_norm": 0.8814655542373657,
+      "learning_rate": 7.894353821281131e-06,
+      "loss": 0.73,
+      "step": 11098
+    },
+    {
+      "epoch": 0.6108756673454786,
+      "grad_norm": 0.6792872548103333,
+      "learning_rate": 7.894000353358624e-06,
+      "loss": 0.7445,
+      "step": 11099
+    },
+    {
+      "epoch": 0.6109307061478342,
+      "grad_norm": 0.6442595720291138,
+      "learning_rate": 7.893646863685584e-06,
+      "loss": 0.7228,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6109857449501899,
+      "grad_norm": 0.6775944828987122,
+      "learning_rate": 7.89329335226467e-06,
+      "loss": 0.7937,
+      "step": 11101
+    },
+    {
+      "epoch": 0.6110407837525456,
+      "grad_norm": 0.6315211653709412,
+      "learning_rate": 7.892939819098534e-06,
+      "loss": 0.7328,
+      "step": 11102
+    },
+    {
+      "epoch": 0.6110958225549012,
+      "grad_norm": 0.7419382929801941,
+      "learning_rate": 7.89258626418984e-06,
+      "loss": 0.8088,
+      "step": 11103
+    },
+    {
+      "epoch": 0.6111508613572568,
+      "grad_norm": 0.6645117402076721,
+      "learning_rate": 7.89223268754124e-06,
+      "loss": 0.7844,
+      "step": 11104
+    },
+    {
+      "epoch": 0.6112059001596125,
+      "grad_norm": 0.6389926075935364,
+      "learning_rate": 7.891879089155397e-06,
+      "loss": 0.6353,
+      "step": 11105
+    },
+    {
+      "epoch": 0.6112609389619682,
+      "grad_norm": 0.8223785758018494,
+      "learning_rate": 7.891525469034963e-06,
+      "loss": 0.7377,
+      "step": 11106
+    },
+    {
+      "epoch": 0.6113159777643239,
+      "grad_norm": 0.7627747058868408,
+      "learning_rate": 7.891171827182595e-06,
+      "loss": 0.8317,
+      "step": 11107
+    },
+    {
+      "epoch": 0.6113710165666795,
+      "grad_norm": 0.8015971183776855,
+      "learning_rate": 7.890818163600956e-06,
+      "loss": 0.8324,
+      "step": 11108
+    },
+    {
+      "epoch": 0.6114260553690352,
+      "grad_norm": 0.7180280089378357,
+      "learning_rate": 7.8904644782927e-06,
+      "loss": 0.8211,
+      "step": 11109
+    },
+    {
+      "epoch": 0.6114810941713908,
+      "grad_norm": 0.7855646014213562,
+      "learning_rate": 7.890110771260487e-06,
+      "loss": 0.8629,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6115361329737465,
+      "grad_norm": 0.7389342784881592,
+      "learning_rate": 7.889757042506976e-06,
+      "loss": 0.6917,
+      "step": 11111
+    },
+    {
+      "epoch": 0.6115911717761021,
+      "grad_norm": 0.7996030449867249,
+      "learning_rate": 7.889403292034825e-06,
+      "loss": 0.7361,
+      "step": 11112
+    },
+    {
+      "epoch": 0.6116462105784578,
+      "grad_norm": 0.6658353805541992,
+      "learning_rate": 7.88904951984669e-06,
+      "loss": 0.7048,
+      "step": 11113
+    },
+    {
+      "epoch": 0.6117012493808135,
+      "grad_norm": 0.8128555417060852,
+      "learning_rate": 7.888695725945235e-06,
+      "loss": 0.7772,
+      "step": 11114
+    },
+    {
+      "epoch": 0.6117562881831692,
+      "grad_norm": 0.7597428560256958,
+      "learning_rate": 7.888341910333114e-06,
+      "loss": 0.7447,
+      "step": 11115
+    },
+    {
+      "epoch": 0.6118113269855248,
+      "grad_norm": 0.7330088019371033,
+      "learning_rate": 7.88798807301299e-06,
+      "loss": 0.849,
+      "step": 11116
+    },
+    {
+      "epoch": 0.6118663657878805,
+      "grad_norm": 0.8374074101448059,
+      "learning_rate": 7.88763421398752e-06,
+      "loss": 0.6149,
+      "step": 11117
+    },
+    {
+      "epoch": 0.6119214045902361,
+      "grad_norm": 0.7507160305976868,
+      "learning_rate": 7.887280333259364e-06,
+      "loss": 0.7737,
+      "step": 11118
+    },
+    {
+      "epoch": 0.6119764433925918,
+      "grad_norm": 0.7218281626701355,
+      "learning_rate": 7.886926430831181e-06,
+      "loss": 0.8151,
+      "step": 11119
+    },
+    {
+      "epoch": 0.6120314821949474,
+      "grad_norm": 0.6761744618415833,
+      "learning_rate": 7.886572506705634e-06,
+      "loss": 0.7429,
+      "step": 11120
+    },
+    {
+      "epoch": 0.6120865209973031,
+      "grad_norm": 0.8243520259857178,
+      "learning_rate": 7.886218560885379e-06,
+      "loss": 0.819,
+      "step": 11121
+    },
+    {
+      "epoch": 0.6121415597996588,
+      "grad_norm": 0.9675465822219849,
+      "learning_rate": 7.885864593373078e-06,
+      "loss": 0.7834,
+      "step": 11122
+    },
+    {
+      "epoch": 0.6121965986020145,
+      "grad_norm": 0.7220338582992554,
+      "learning_rate": 7.885510604171391e-06,
+      "loss": 0.8266,
+      "step": 11123
+    },
+    {
+      "epoch": 0.61225163740437,
+      "grad_norm": 0.7185316681861877,
+      "learning_rate": 7.88515659328298e-06,
+      "loss": 0.7949,
+      "step": 11124
+    },
+    {
+      "epoch": 0.6123066762067257,
+      "grad_norm": 0.67637038230896,
+      "learning_rate": 7.884802560710503e-06,
+      "loss": 0.7456,
+      "step": 11125
+    },
+    {
+      "epoch": 0.6123617150090814,
+      "grad_norm": 0.7886855602264404,
+      "learning_rate": 7.884448506456622e-06,
+      "loss": 0.7181,
+      "step": 11126
+    },
+    {
+      "epoch": 0.612416753811437,
+      "grad_norm": 0.7250227928161621,
+      "learning_rate": 7.884094430523999e-06,
+      "loss": 0.7537,
+      "step": 11127
+    },
+    {
+      "epoch": 0.6124717926137927,
+      "grad_norm": 0.6771906614303589,
+      "learning_rate": 7.883740332915295e-06,
+      "loss": 0.7642,
+      "step": 11128
+    },
+    {
+      "epoch": 0.6125268314161484,
+      "grad_norm": 0.8375886082649231,
+      "learning_rate": 7.88338621363317e-06,
+      "loss": 0.7231,
+      "step": 11129
+    },
+    {
+      "epoch": 0.6125818702185041,
+      "grad_norm": 0.6782773733139038,
+      "learning_rate": 7.883032072680285e-06,
+      "loss": 0.8391,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6126369090208597,
+      "grad_norm": 0.7103945016860962,
+      "learning_rate": 7.882677910059304e-06,
+      "loss": 0.7838,
+      "step": 11131
+    },
+    {
+      "epoch": 0.6126919478232153,
+      "grad_norm": 0.7037224769592285,
+      "learning_rate": 7.882323725772887e-06,
+      "loss": 0.7906,
+      "step": 11132
+    },
+    {
+      "epoch": 0.612746986625571,
+      "grad_norm": 0.6872009634971619,
+      "learning_rate": 7.881969519823695e-06,
+      "loss": 0.7764,
+      "step": 11133
+    },
+    {
+      "epoch": 0.6128020254279267,
+      "grad_norm": 0.7377448678016663,
+      "learning_rate": 7.881615292214393e-06,
+      "loss": 0.8231,
+      "step": 11134
+    },
+    {
+      "epoch": 0.6128570642302823,
+      "grad_norm": 0.62479168176651,
+      "learning_rate": 7.881261042947642e-06,
+      "loss": 0.6522,
+      "step": 11135
+    },
+    {
+      "epoch": 0.612912103032638,
+      "grad_norm": 0.7989023923873901,
+      "learning_rate": 7.880906772026105e-06,
+      "loss": 0.7326,
+      "step": 11136
+    },
+    {
+      "epoch": 0.6129671418349937,
+      "grad_norm": 0.6322734951972961,
+      "learning_rate": 7.880552479452441e-06,
+      "loss": 0.6775,
+      "step": 11137
+    },
+    {
+      "epoch": 0.6130221806373494,
+      "grad_norm": 0.8628767132759094,
+      "learning_rate": 7.880198165229318e-06,
+      "loss": 0.7705,
+      "step": 11138
+    },
+    {
+      "epoch": 0.613077219439705,
+      "grad_norm": 0.7386173605918884,
+      "learning_rate": 7.879843829359396e-06,
+      "loss": 0.7297,
+      "step": 11139
+    },
+    {
+      "epoch": 0.6131322582420606,
+      "grad_norm": 0.6882045269012451,
+      "learning_rate": 7.879489471845339e-06,
+      "loss": 0.6875,
+      "step": 11140
+    },
+    {
+      "epoch": 0.6131872970444163,
+      "grad_norm": 0.5986032485961914,
+      "learning_rate": 7.879135092689809e-06,
+      "loss": 0.6329,
+      "step": 11141
+    },
+    {
+      "epoch": 0.613242335846772,
+      "grad_norm": 0.7973099946975708,
+      "learning_rate": 7.878780691895472e-06,
+      "loss": 0.809,
+      "step": 11142
+    },
+    {
+      "epoch": 0.6132973746491276,
+      "grad_norm": 0.6828579902648926,
+      "learning_rate": 7.878426269464989e-06,
+      "loss": 0.7777,
+      "step": 11143
+    },
+    {
+      "epoch": 0.6133524134514833,
+      "grad_norm": 0.8179183006286621,
+      "learning_rate": 7.878071825401024e-06,
+      "loss": 0.7275,
+      "step": 11144
+    },
+    {
+      "epoch": 0.613407452253839,
+      "grad_norm": 0.7290762066841125,
+      "learning_rate": 7.877717359706242e-06,
+      "loss": 0.7424,
+      "step": 11145
+    },
+    {
+      "epoch": 0.6134624910561947,
+      "grad_norm": 0.732510507106781,
+      "learning_rate": 7.877362872383305e-06,
+      "loss": 0.6157,
+      "step": 11146
+    },
+    {
+      "epoch": 0.6135175298585502,
+      "grad_norm": 0.9205982685089111,
+      "learning_rate": 7.877008363434881e-06,
+      "loss": 0.7723,
+      "step": 11147
+    },
+    {
+      "epoch": 0.6135725686609059,
+      "grad_norm": 0.7138587832450867,
+      "learning_rate": 7.876653832863633e-06,
+      "loss": 0.7773,
+      "step": 11148
+    },
+    {
+      "epoch": 0.6136276074632616,
+      "grad_norm": 0.7323171496391296,
+      "learning_rate": 7.876299280672224e-06,
+      "loss": 0.8265,
+      "step": 11149
+    },
+    {
+      "epoch": 0.6136826462656173,
+      "grad_norm": 0.6717494130134583,
+      "learning_rate": 7.875944706863318e-06,
+      "loss": 0.788,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6137376850679729,
+      "grad_norm": 0.7779331207275391,
+      "learning_rate": 7.875590111439582e-06,
+      "loss": 0.7864,
+      "step": 11151
+    },
+    {
+      "epoch": 0.6137927238703286,
+      "grad_norm": 0.6706684827804565,
+      "learning_rate": 7.875235494403683e-06,
+      "loss": 0.6673,
+      "step": 11152
+    },
+    {
+      "epoch": 0.6138477626726843,
+      "grad_norm": 0.7142137885093689,
+      "learning_rate": 7.874880855758281e-06,
+      "loss": 0.8031,
+      "step": 11153
+    },
+    {
+      "epoch": 0.61390280147504,
+      "grad_norm": 0.6962595582008362,
+      "learning_rate": 7.874526195506045e-06,
+      "loss": 0.692,
+      "step": 11154
+    },
+    {
+      "epoch": 0.6139578402773955,
+      "grad_norm": 0.7237100601196289,
+      "learning_rate": 7.874171513649638e-06,
+      "loss": 0.7504,
+      "step": 11155
+    },
+    {
+      "epoch": 0.6140128790797512,
+      "grad_norm": 0.8235127925872803,
+      "learning_rate": 7.87381681019173e-06,
+      "loss": 0.8132,
+      "step": 11156
+    },
+    {
+      "epoch": 0.6140679178821069,
+      "grad_norm": 0.7483351826667786,
+      "learning_rate": 7.873462085134981e-06,
+      "loss": 0.7589,
+      "step": 11157
+    },
+    {
+      "epoch": 0.6141229566844626,
+      "grad_norm": 0.7309976816177368,
+      "learning_rate": 7.873107338482062e-06,
+      "loss": 0.7722,
+      "step": 11158
+    },
+    {
+      "epoch": 0.6141779954868182,
+      "grad_norm": 0.8871245384216309,
+      "learning_rate": 7.872752570235639e-06,
+      "loss": 0.882,
+      "step": 11159
+    },
+    {
+      "epoch": 0.6142330342891739,
+      "grad_norm": 0.5987886190414429,
+      "learning_rate": 7.872397780398374e-06,
+      "loss": 0.6312,
+      "step": 11160
+    },
+    {
+      "epoch": 0.6142880730915296,
+      "grad_norm": 0.7320038080215454,
+      "learning_rate": 7.872042968972937e-06,
+      "loss": 0.7444,
+      "step": 11161
+    },
+    {
+      "epoch": 0.6143431118938852,
+      "grad_norm": 0.8111129999160767,
+      "learning_rate": 7.871688135961995e-06,
+      "loss": 0.7413,
+      "step": 11162
+    },
+    {
+      "epoch": 0.6143981506962408,
+      "grad_norm": 0.7497085332870483,
+      "learning_rate": 7.871333281368211e-06,
+      "loss": 0.8413,
+      "step": 11163
+    },
+    {
+      "epoch": 0.6144531894985965,
+      "grad_norm": 0.8341198563575745,
+      "learning_rate": 7.870978405194256e-06,
+      "loss": 0.7959,
+      "step": 11164
+    },
+    {
+      "epoch": 0.6145082283009522,
+      "grad_norm": 0.6293482780456543,
+      "learning_rate": 7.870623507442797e-06,
+      "loss": 0.6429,
+      "step": 11165
+    },
+    {
+      "epoch": 0.6145632671033079,
+      "grad_norm": 1.2423945665359497,
+      "learning_rate": 7.870268588116499e-06,
+      "loss": 0.6309,
+      "step": 11166
+    },
+    {
+      "epoch": 0.6146183059056635,
+      "grad_norm": 0.7811731100082397,
+      "learning_rate": 7.86991364721803e-06,
+      "loss": 0.738,
+      "step": 11167
+    },
+    {
+      "epoch": 0.6146733447080192,
+      "grad_norm": 0.6904361248016357,
+      "learning_rate": 7.869558684750061e-06,
+      "loss": 0.7995,
+      "step": 11168
+    },
+    {
+      "epoch": 0.6147283835103748,
+      "grad_norm": 0.7267210483551025,
+      "learning_rate": 7.869203700715254e-06,
+      "loss": 0.6989,
+      "step": 11169
+    },
+    {
+      "epoch": 0.6147834223127304,
+      "grad_norm": 0.7183068990707397,
+      "learning_rate": 7.868848695116282e-06,
+      "loss": 0.7872,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6148384611150861,
+      "grad_norm": 0.6774286031723022,
+      "learning_rate": 7.868493667955808e-06,
+      "loss": 0.7502,
+      "step": 11171
+    },
+    {
+      "epoch": 0.6148934999174418,
+      "grad_norm": 0.7587934732437134,
+      "learning_rate": 7.868138619236507e-06,
+      "loss": 0.8037,
+      "step": 11172
+    },
+    {
+      "epoch": 0.6149485387197975,
+      "grad_norm": 0.6825854182243347,
+      "learning_rate": 7.867783548961043e-06,
+      "loss": 0.7924,
+      "step": 11173
+    },
+    {
+      "epoch": 0.6150035775221531,
+      "grad_norm": 0.6243380904197693,
+      "learning_rate": 7.867428457132084e-06,
+      "loss": 0.5953,
+      "step": 11174
+    },
+    {
+      "epoch": 0.6150586163245088,
+      "grad_norm": 0.6630006432533264,
+      "learning_rate": 7.8670733437523e-06,
+      "loss": 0.7102,
+      "step": 11175
+    },
+    {
+      "epoch": 0.6151136551268644,
+      "grad_norm": 0.7059652805328369,
+      "learning_rate": 7.866718208824362e-06,
+      "loss": 0.6847,
+      "step": 11176
+    },
+    {
+      "epoch": 0.6151686939292201,
+      "grad_norm": 0.6768305897712708,
+      "learning_rate": 7.866363052350938e-06,
+      "loss": 0.7152,
+      "step": 11177
+    },
+    {
+      "epoch": 0.6152237327315757,
+      "grad_norm": 0.6850628852844238,
+      "learning_rate": 7.866007874334696e-06,
+      "loss": 0.767,
+      "step": 11178
+    },
+    {
+      "epoch": 0.6152787715339314,
+      "grad_norm": 0.6767143607139587,
+      "learning_rate": 7.865652674778305e-06,
+      "loss": 0.6826,
+      "step": 11179
+    },
+    {
+      "epoch": 0.6153338103362871,
+      "grad_norm": 0.8240014314651489,
+      "learning_rate": 7.865297453684436e-06,
+      "loss": 0.8493,
+      "step": 11180
+    },
+    {
+      "epoch": 0.6153888491386428,
+      "grad_norm": 0.7725485563278198,
+      "learning_rate": 7.864942211055758e-06,
+      "loss": 0.8704,
+      "step": 11181
+    },
+    {
+      "epoch": 0.6154438879409984,
+      "grad_norm": 0.9260931015014648,
+      "learning_rate": 7.864586946894941e-06,
+      "loss": 0.7926,
+      "step": 11182
+    },
+    {
+      "epoch": 0.615498926743354,
+      "grad_norm": 0.7558152079582214,
+      "learning_rate": 7.864231661204655e-06,
+      "loss": 0.8436,
+      "step": 11183
+    },
+    {
+      "epoch": 0.6155539655457097,
+      "grad_norm": 0.7899817824363708,
+      "learning_rate": 7.863876353987571e-06,
+      "loss": 0.7579,
+      "step": 11184
+    },
+    {
+      "epoch": 0.6156090043480654,
+      "grad_norm": 0.7757478952407837,
+      "learning_rate": 7.863521025246362e-06,
+      "loss": 0.7534,
+      "step": 11185
+    },
+    {
+      "epoch": 0.615664043150421,
+      "grad_norm": 0.6563131809234619,
+      "learning_rate": 7.863165674983693e-06,
+      "loss": 0.728,
+      "step": 11186
+    },
+    {
+      "epoch": 0.6157190819527767,
+      "grad_norm": 0.6516488790512085,
+      "learning_rate": 7.862810303202234e-06,
+      "loss": 0.736,
+      "step": 11187
+    },
+    {
+      "epoch": 0.6157741207551324,
+      "grad_norm": 0.6867820620536804,
+      "learning_rate": 7.862454909904665e-06,
+      "loss": 0.8032,
+      "step": 11188
+    },
+    {
+      "epoch": 0.6158291595574881,
+      "grad_norm": 0.7399753928184509,
+      "learning_rate": 7.862099495093647e-06,
+      "loss": 0.8681,
+      "step": 11189
+    },
+    {
+      "epoch": 0.6158841983598436,
+      "grad_norm": 0.7249311804771423,
+      "learning_rate": 7.861744058771857e-06,
+      "loss": 0.7868,
+      "step": 11190
+    },
+    {
+      "epoch": 0.6159392371621993,
+      "grad_norm": 0.8579045534133911,
+      "learning_rate": 7.861388600941964e-06,
+      "loss": 0.7915,
+      "step": 11191
+    },
+    {
+      "epoch": 0.615994275964555,
+      "grad_norm": 0.6855454444885254,
+      "learning_rate": 7.86103312160664e-06,
+      "loss": 0.8442,
+      "step": 11192
+    },
+    {
+      "epoch": 0.6160493147669107,
+      "grad_norm": 0.7412910461425781,
+      "learning_rate": 7.860677620768558e-06,
+      "loss": 0.7684,
+      "step": 11193
+    },
+    {
+      "epoch": 0.6161043535692663,
+      "grad_norm": 0.8567430377006531,
+      "learning_rate": 7.860322098430389e-06,
+      "loss": 0.8801,
+      "step": 11194
+    },
+    {
+      "epoch": 0.616159392371622,
+      "grad_norm": 0.7504804134368896,
+      "learning_rate": 7.859966554594802e-06,
+      "loss": 0.7359,
+      "step": 11195
+    },
+    {
+      "epoch": 0.6162144311739777,
+      "grad_norm": 0.7086803317070007,
+      "learning_rate": 7.859610989264474e-06,
+      "loss": 0.8498,
+      "step": 11196
+    },
+    {
+      "epoch": 0.6162694699763334,
+      "grad_norm": 0.7201757431030273,
+      "learning_rate": 7.859255402442075e-06,
+      "loss": 0.608,
+      "step": 11197
+    },
+    {
+      "epoch": 0.6163245087786889,
+      "grad_norm": 0.8968291282653809,
+      "learning_rate": 7.858899794130279e-06,
+      "loss": 0.8067,
+      "step": 11198
+    },
+    {
+      "epoch": 0.6163795475810446,
+      "grad_norm": 0.7474254965782166,
+      "learning_rate": 7.858544164331756e-06,
+      "loss": 0.8355,
+      "step": 11199
+    },
+    {
+      "epoch": 0.6164345863834003,
+      "grad_norm": 0.6907560229301453,
+      "learning_rate": 7.85818851304918e-06,
+      "loss": 0.788,
+      "step": 11200
+    },
+    {
+      "epoch": 0.616489625185756,
+      "grad_norm": 0.725330650806427,
+      "learning_rate": 7.857832840285224e-06,
+      "loss": 0.8157,
+      "step": 11201
+    },
+    {
+      "epoch": 0.6165446639881116,
+      "grad_norm": 0.682722270488739,
+      "learning_rate": 7.857477146042562e-06,
+      "loss": 0.7939,
+      "step": 11202
+    },
+    {
+      "epoch": 0.6165997027904673,
+      "grad_norm": 0.661533534526825,
+      "learning_rate": 7.857121430323866e-06,
+      "loss": 0.7173,
+      "step": 11203
+    },
+    {
+      "epoch": 0.616654741592823,
+      "grad_norm": 0.6922706961631775,
+      "learning_rate": 7.856765693131811e-06,
+      "loss": 0.7719,
+      "step": 11204
+    },
+    {
+      "epoch": 0.6167097803951787,
+      "grad_norm": 0.72809898853302,
+      "learning_rate": 7.856409934469071e-06,
+      "loss": 0.7362,
+      "step": 11205
+    },
+    {
+      "epoch": 0.6167648191975342,
+      "grad_norm": 0.7540956735610962,
+      "learning_rate": 7.856054154338317e-06,
+      "loss": 0.7883,
+      "step": 11206
+    },
+    {
+      "epoch": 0.6168198579998899,
+      "grad_norm": 0.6777094006538391,
+      "learning_rate": 7.855698352742224e-06,
+      "loss": 0.6938,
+      "step": 11207
+    },
+    {
+      "epoch": 0.6168748968022456,
+      "grad_norm": 0.6771852970123291,
+      "learning_rate": 7.855342529683467e-06,
+      "loss": 0.697,
+      "step": 11208
+    },
+    {
+      "epoch": 0.6169299356046013,
+      "grad_norm": 0.7810118198394775,
+      "learning_rate": 7.854986685164721e-06,
+      "loss": 0.6875,
+      "step": 11209
+    },
+    {
+      "epoch": 0.6169849744069569,
+      "grad_norm": 0.6992766261100769,
+      "learning_rate": 7.854630819188658e-06,
+      "loss": 0.6553,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6170400132093126,
+      "grad_norm": 0.7409703135490417,
+      "learning_rate": 7.854274931757954e-06,
+      "loss": 0.7685,
+      "step": 11211
+    },
+    {
+      "epoch": 0.6170950520116683,
+      "grad_norm": 0.7263410687446594,
+      "learning_rate": 7.853919022875285e-06,
+      "loss": 0.7939,
+      "step": 11212
+    },
+    {
+      "epoch": 0.6171500908140238,
+      "grad_norm": 0.8451918959617615,
+      "learning_rate": 7.853563092543323e-06,
+      "loss": 0.7522,
+      "step": 11213
+    },
+    {
+      "epoch": 0.6172051296163795,
+      "grad_norm": 0.672926664352417,
+      "learning_rate": 7.853207140764745e-06,
+      "loss": 0.732,
+      "step": 11214
+    },
+    {
+      "epoch": 0.6172601684187352,
+      "grad_norm": 0.6607885956764221,
+      "learning_rate": 7.852851167542226e-06,
+      "loss": 0.7441,
+      "step": 11215
+    },
+    {
+      "epoch": 0.6173152072210909,
+      "grad_norm": 0.730385422706604,
+      "learning_rate": 7.85249517287844e-06,
+      "loss": 0.7925,
+      "step": 11216
+    },
+    {
+      "epoch": 0.6173702460234465,
+      "grad_norm": 0.7338821887969971,
+      "learning_rate": 7.852139156776067e-06,
+      "loss": 0.8106,
+      "step": 11217
+    },
+    {
+      "epoch": 0.6174252848258022,
+      "grad_norm": 0.7662163376808167,
+      "learning_rate": 7.851783119237777e-06,
+      "loss": 0.8166,
+      "step": 11218
+    },
+    {
+      "epoch": 0.6174803236281579,
+      "grad_norm": 0.7738409042358398,
+      "learning_rate": 7.85142706026625e-06,
+      "loss": 0.7898,
+      "step": 11219
+    },
+    {
+      "epoch": 0.6175353624305135,
+      "grad_norm": 0.8129978775978088,
+      "learning_rate": 7.851070979864159e-06,
+      "loss": 0.7618,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6175904012328691,
+      "grad_norm": 0.7923482060432434,
+      "learning_rate": 7.850714878034183e-06,
+      "loss": 0.7341,
+      "step": 11221
+    },
+    {
+      "epoch": 0.6176454400352248,
+      "grad_norm": 0.7189306020736694,
+      "learning_rate": 7.850358754778996e-06,
+      "loss": 0.7775,
+      "step": 11222
+    },
+    {
+      "epoch": 0.6177004788375805,
+      "grad_norm": 0.9873724579811096,
+      "learning_rate": 7.850002610101276e-06,
+      "loss": 0.8521,
+      "step": 11223
+    },
+    {
+      "epoch": 0.6177555176399362,
+      "grad_norm": 0.6350038051605225,
+      "learning_rate": 7.8496464440037e-06,
+      "loss": 0.6356,
+      "step": 11224
+    },
+    {
+      "epoch": 0.6178105564422918,
+      "grad_norm": 0.8059771060943604,
+      "learning_rate": 7.849290256488941e-06,
+      "loss": 0.821,
+      "step": 11225
+    },
+    {
+      "epoch": 0.6178655952446475,
+      "grad_norm": 0.7469610571861267,
+      "learning_rate": 7.848934047559684e-06,
+      "loss": 0.7782,
+      "step": 11226
+    },
+    {
+      "epoch": 0.6179206340470031,
+      "grad_norm": 0.6423176527023315,
+      "learning_rate": 7.848577817218597e-06,
+      "loss": 0.6693,
+      "step": 11227
+    },
+    {
+      "epoch": 0.6179756728493588,
+      "grad_norm": 0.7298387885093689,
+      "learning_rate": 7.848221565468363e-06,
+      "loss": 0.775,
+      "step": 11228
+    },
+    {
+      "epoch": 0.6180307116517144,
+      "grad_norm": 0.7125145196914673,
+      "learning_rate": 7.84786529231166e-06,
+      "loss": 0.7507,
+      "step": 11229
+    },
+    {
+      "epoch": 0.6180857504540701,
+      "grad_norm": 0.6658627390861511,
+      "learning_rate": 7.847508997751163e-06,
+      "loss": 0.7506,
+      "step": 11230
+    },
+    {
+      "epoch": 0.6181407892564258,
+      "grad_norm": 0.6425275206565857,
+      "learning_rate": 7.847152681789549e-06,
+      "loss": 0.657,
+      "step": 11231
+    },
+    {
+      "epoch": 0.6181958280587815,
+      "grad_norm": 0.8075960278511047,
+      "learning_rate": 7.846796344429498e-06,
+      "loss": 0.5434,
+      "step": 11232
+    },
+    {
+      "epoch": 0.6182508668611371,
+      "grad_norm": 0.8481889367103577,
+      "learning_rate": 7.846439985673689e-06,
+      "loss": 0.8303,
+      "step": 11233
+    },
+    {
+      "epoch": 0.6183059056634927,
+      "grad_norm": 0.7216358184814453,
+      "learning_rate": 7.846083605524799e-06,
+      "loss": 0.7589,
+      "step": 11234
+    },
+    {
+      "epoch": 0.6183609444658484,
+      "grad_norm": 0.8399745225906372,
+      "learning_rate": 7.845727203985504e-06,
+      "loss": 0.8096,
+      "step": 11235
+    },
+    {
+      "epoch": 0.6184159832682041,
+      "grad_norm": 0.6708692908287048,
+      "learning_rate": 7.845370781058489e-06,
+      "loss": 0.6858,
+      "step": 11236
+    },
+    {
+      "epoch": 0.6184710220705597,
+      "grad_norm": 0.6309100389480591,
+      "learning_rate": 7.845014336746426e-06,
+      "loss": 0.6093,
+      "step": 11237
+    },
+    {
+      "epoch": 0.6185260608729154,
+      "grad_norm": 0.8138728141784668,
+      "learning_rate": 7.844657871051997e-06,
+      "loss": 0.8259,
+      "step": 11238
+    },
+    {
+      "epoch": 0.6185810996752711,
+      "grad_norm": 0.6763564348220825,
+      "learning_rate": 7.844301383977882e-06,
+      "loss": 0.7056,
+      "step": 11239
+    },
+    {
+      "epoch": 0.6186361384776268,
+      "grad_norm": 0.792085587978363,
+      "learning_rate": 7.843944875526758e-06,
+      "loss": 0.7364,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6186911772799824,
+      "grad_norm": 0.8738027811050415,
+      "learning_rate": 7.843588345701306e-06,
+      "loss": 0.7092,
+      "step": 11241
+    },
+    {
+      "epoch": 0.618746216082338,
+      "grad_norm": 0.7694413065910339,
+      "learning_rate": 7.843231794504205e-06,
+      "loss": 0.852,
+      "step": 11242
+    },
+    {
+      "epoch": 0.6188012548846937,
+      "grad_norm": 0.8211640119552612,
+      "learning_rate": 7.842875221938135e-06,
+      "loss": 0.8218,
+      "step": 11243
+    },
+    {
+      "epoch": 0.6188562936870494,
+      "grad_norm": 0.620566189289093,
+      "learning_rate": 7.842518628005776e-06,
+      "loss": 0.7176,
+      "step": 11244
+    },
+    {
+      "epoch": 0.618911332489405,
+      "grad_norm": 0.7044099569320679,
+      "learning_rate": 7.84216201270981e-06,
+      "loss": 0.8068,
+      "step": 11245
+    },
+    {
+      "epoch": 0.6189663712917607,
+      "grad_norm": 0.765209436416626,
+      "learning_rate": 7.841805376052912e-06,
+      "loss": 0.8002,
+      "step": 11246
+    },
+    {
+      "epoch": 0.6190214100941164,
+      "grad_norm": 0.7565444707870483,
+      "learning_rate": 7.841448718037765e-06,
+      "loss": 0.7997,
+      "step": 11247
+    },
+    {
+      "epoch": 0.6190764488964721,
+      "grad_norm": 0.9544101357460022,
+      "learning_rate": 7.841092038667052e-06,
+      "loss": 0.647,
+      "step": 11248
+    },
+    {
+      "epoch": 0.6191314876988276,
+      "grad_norm": 0.7319634556770325,
+      "learning_rate": 7.840735337943452e-06,
+      "loss": 0.7982,
+      "step": 11249
+    },
+    {
+      "epoch": 0.6191865265011833,
+      "grad_norm": 0.6017479300498962,
+      "learning_rate": 7.840378615869645e-06,
+      "loss": 0.6817,
+      "step": 11250
+    },
+    {
+      "epoch": 0.619241565303539,
+      "grad_norm": 0.6936477422714233,
+      "learning_rate": 7.840021872448312e-06,
+      "loss": 0.7227,
+      "step": 11251
+    },
+    {
+      "epoch": 0.6192966041058947,
+      "grad_norm": 0.6962631940841675,
+      "learning_rate": 7.839665107682135e-06,
+      "loss": 0.779,
+      "step": 11252
+    },
+    {
+      "epoch": 0.6193516429082503,
+      "grad_norm": 0.9580947160720825,
+      "learning_rate": 7.839308321573797e-06,
+      "loss": 0.8821,
+      "step": 11253
+    },
+    {
+      "epoch": 0.619406681710606,
+      "grad_norm": 0.7721261978149414,
+      "learning_rate": 7.838951514125977e-06,
+      "loss": 0.7146,
+      "step": 11254
+    },
+    {
+      "epoch": 0.6194617205129617,
+      "grad_norm": 0.7349434494972229,
+      "learning_rate": 7.838594685341354e-06,
+      "loss": 0.7601,
+      "step": 11255
+    },
+    {
+      "epoch": 0.6195167593153172,
+      "grad_norm": 0.6787356734275818,
+      "learning_rate": 7.838237835222618e-06,
+      "loss": 0.706,
+      "step": 11256
+    },
+    {
+      "epoch": 0.6195717981176729,
+      "grad_norm": 0.7658288478851318,
+      "learning_rate": 7.837880963772445e-06,
+      "loss": 0.7102,
+      "step": 11257
+    },
+    {
+      "epoch": 0.6196268369200286,
+      "grad_norm": 0.8083927035331726,
+      "learning_rate": 7.837524070993516e-06,
+      "loss": 0.8501,
+      "step": 11258
+    },
+    {
+      "epoch": 0.6196818757223843,
+      "grad_norm": 0.7656283974647522,
+      "learning_rate": 7.837167156888516e-06,
+      "loss": 0.7558,
+      "step": 11259
+    },
+    {
+      "epoch": 0.6197369145247399,
+      "grad_norm": 0.7897886037826538,
+      "learning_rate": 7.836810221460128e-06,
+      "loss": 0.7567,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6197919533270956,
+      "grad_norm": 0.6858190298080444,
+      "learning_rate": 7.836453264711035e-06,
+      "loss": 0.717,
+      "step": 11261
+    },
+    {
+      "epoch": 0.6198469921294513,
+      "grad_norm": 0.7423431873321533,
+      "learning_rate": 7.836096286643917e-06,
+      "loss": 0.7047,
+      "step": 11262
+    },
+    {
+      "epoch": 0.619902030931807,
+      "grad_norm": 0.8277921676635742,
+      "learning_rate": 7.835739287261458e-06,
+      "loss": 0.7418,
+      "step": 11263
+    },
+    {
+      "epoch": 0.6199570697341625,
+      "grad_norm": 0.7102510929107666,
+      "learning_rate": 7.835382266566343e-06,
+      "loss": 0.8202,
+      "step": 11264
+    },
+    {
+      "epoch": 0.6200121085365182,
+      "grad_norm": 0.6705429553985596,
+      "learning_rate": 7.835025224561252e-06,
+      "loss": 0.7332,
+      "step": 11265
+    },
+    {
+      "epoch": 0.6200671473388739,
+      "grad_norm": 0.6529950499534607,
+      "learning_rate": 7.834668161248873e-06,
+      "loss": 0.7579,
+      "step": 11266
+    },
+    {
+      "epoch": 0.6201221861412296,
+      "grad_norm": 0.7189938426017761,
+      "learning_rate": 7.834311076631885e-06,
+      "loss": 0.7323,
+      "step": 11267
+    },
+    {
+      "epoch": 0.6201772249435852,
+      "grad_norm": 0.6559470891952515,
+      "learning_rate": 7.833953970712973e-06,
+      "loss": 0.5973,
+      "step": 11268
+    },
+    {
+      "epoch": 0.6202322637459409,
+      "grad_norm": 0.7971723675727844,
+      "learning_rate": 7.833596843494824e-06,
+      "loss": 0.804,
+      "step": 11269
+    },
+    {
+      "epoch": 0.6202873025482966,
+      "grad_norm": 0.7800958752632141,
+      "learning_rate": 7.833239694980118e-06,
+      "loss": 0.772,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6203423413506522,
+      "grad_norm": 0.6831466555595398,
+      "learning_rate": 7.83288252517154e-06,
+      "loss": 0.7341,
+      "step": 11271
+    },
+    {
+      "epoch": 0.6203973801530078,
+      "grad_norm": 0.6504807472229004,
+      "learning_rate": 7.832525334071776e-06,
+      "loss": 0.6462,
+      "step": 11272
+    },
+    {
+      "epoch": 0.6204524189553635,
+      "grad_norm": 0.6973552703857422,
+      "learning_rate": 7.832168121683512e-06,
+      "loss": 0.7504,
+      "step": 11273
+    },
+    {
+      "epoch": 0.6205074577577192,
+      "grad_norm": 0.6772480607032776,
+      "learning_rate": 7.831810888009427e-06,
+      "loss": 0.7273,
+      "step": 11274
+    },
+    {
+      "epoch": 0.6205624965600749,
+      "grad_norm": 0.7077416777610779,
+      "learning_rate": 7.831453633052212e-06,
+      "loss": 0.7365,
+      "step": 11275
+    },
+    {
+      "epoch": 0.6206175353624305,
+      "grad_norm": 0.7338337898254395,
+      "learning_rate": 7.831096356814548e-06,
+      "loss": 0.7959,
+      "step": 11276
+    },
+    {
+      "epoch": 0.6206725741647862,
+      "grad_norm": 0.6313255429267883,
+      "learning_rate": 7.830739059299123e-06,
+      "loss": 0.7027,
+      "step": 11277
+    },
+    {
+      "epoch": 0.6207276129671419,
+      "grad_norm": 0.7377570867538452,
+      "learning_rate": 7.830381740508619e-06,
+      "loss": 0.6903,
+      "step": 11278
+    },
+    {
+      "epoch": 0.6207826517694975,
+      "grad_norm": 0.6868650317192078,
+      "learning_rate": 7.830024400445724e-06,
+      "loss": 0.6882,
+      "step": 11279
+    },
+    {
+      "epoch": 0.6208376905718531,
+      "grad_norm": 0.7632661461830139,
+      "learning_rate": 7.829667039113124e-06,
+      "loss": 0.8437,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6208927293742088,
+      "grad_norm": 0.9241608381271362,
+      "learning_rate": 7.829309656513504e-06,
+      "loss": 0.779,
+      "step": 11281
+    },
+    {
+      "epoch": 0.6209477681765645,
+      "grad_norm": 0.6857842206954956,
+      "learning_rate": 7.828952252649551e-06,
+      "loss": 0.7882,
+      "step": 11282
+    },
+    {
+      "epoch": 0.6210028069789202,
+      "grad_norm": 0.695659875869751,
+      "learning_rate": 7.828594827523947e-06,
+      "loss": 0.7471,
+      "step": 11283
+    },
+    {
+      "epoch": 0.6210578457812758,
+      "grad_norm": 0.6398521661758423,
+      "learning_rate": 7.828237381139383e-06,
+      "loss": 0.7328,
+      "step": 11284
+    },
+    {
+      "epoch": 0.6211128845836315,
+      "grad_norm": 0.7386063933372498,
+      "learning_rate": 7.827879913498544e-06,
+      "loss": 0.748,
+      "step": 11285
+    },
+    {
+      "epoch": 0.6211679233859871,
+      "grad_norm": 0.6740923523902893,
+      "learning_rate": 7.827522424604117e-06,
+      "loss": 0.6866,
+      "step": 11286
+    },
+    {
+      "epoch": 0.6212229621883428,
+      "grad_norm": 0.6794413924217224,
+      "learning_rate": 7.82716491445879e-06,
+      "loss": 0.7299,
+      "step": 11287
+    },
+    {
+      "epoch": 0.6212780009906984,
+      "grad_norm": 0.6471715569496155,
+      "learning_rate": 7.826807383065245e-06,
+      "loss": 0.7071,
+      "step": 11288
+    },
+    {
+      "epoch": 0.6213330397930541,
+      "grad_norm": 0.9716162085533142,
+      "learning_rate": 7.826449830426174e-06,
+      "loss": 0.7417,
+      "step": 11289
+    },
+    {
+      "epoch": 0.6213880785954098,
+      "grad_norm": 0.6928716897964478,
+      "learning_rate": 7.826092256544263e-06,
+      "loss": 0.7757,
+      "step": 11290
+    },
+    {
+      "epoch": 0.6214431173977655,
+      "grad_norm": 0.6739227175712585,
+      "learning_rate": 7.825734661422197e-06,
+      "loss": 0.7576,
+      "step": 11291
+    },
+    {
+      "epoch": 0.621498156200121,
+      "grad_norm": 1.2619935274124146,
+      "learning_rate": 7.825377045062668e-06,
+      "loss": 0.7454,
+      "step": 11292
+    },
+    {
+      "epoch": 0.6215531950024767,
+      "grad_norm": 0.6713572144508362,
+      "learning_rate": 7.825019407468361e-06,
+      "loss": 0.7916,
+      "step": 11293
+    },
+    {
+      "epoch": 0.6216082338048324,
+      "grad_norm": 0.6143541932106018,
+      "learning_rate": 7.824661748641964e-06,
+      "loss": 0.6765,
+      "step": 11294
+    },
+    {
+      "epoch": 0.6216632726071881,
+      "grad_norm": 0.7141658067703247,
+      "learning_rate": 7.824304068586163e-06,
+      "loss": 0.7773,
+      "step": 11295
+    },
+    {
+      "epoch": 0.6217183114095437,
+      "grad_norm": 0.7320290803909302,
+      "learning_rate": 7.823946367303653e-06,
+      "loss": 0.8062,
+      "step": 11296
+    },
+    {
+      "epoch": 0.6217733502118994,
+      "grad_norm": 0.7523403167724609,
+      "learning_rate": 7.823588644797115e-06,
+      "loss": 0.7126,
+      "step": 11297
+    },
+    {
+      "epoch": 0.6218283890142551,
+      "grad_norm": 0.6512221097946167,
+      "learning_rate": 7.823230901069242e-06,
+      "loss": 0.7563,
+      "step": 11298
+    },
+    {
+      "epoch": 0.6218834278166107,
+      "grad_norm": 0.6512733697891235,
+      "learning_rate": 7.82287313612272e-06,
+      "loss": 0.7603,
+      "step": 11299
+    },
+    {
+      "epoch": 0.6219384666189663,
+      "grad_norm": 1.0590927600860596,
+      "learning_rate": 7.82251534996024e-06,
+      "loss": 0.8325,
+      "step": 11300
+    },
+    {
+      "epoch": 0.621993505421322,
+      "grad_norm": 0.6763397455215454,
+      "learning_rate": 7.82215754258449e-06,
+      "loss": 0.7915,
+      "step": 11301
+    },
+    {
+      "epoch": 0.6220485442236777,
+      "grad_norm": 0.6640639901161194,
+      "learning_rate": 7.82179971399816e-06,
+      "loss": 0.6953,
+      "step": 11302
+    },
+    {
+      "epoch": 0.6221035830260333,
+      "grad_norm": 0.6611515283584595,
+      "learning_rate": 7.821441864203938e-06,
+      "loss": 0.8331,
+      "step": 11303
+    },
+    {
+      "epoch": 0.622158621828389,
+      "grad_norm": 0.8226057887077332,
+      "learning_rate": 7.821083993204514e-06,
+      "loss": 0.7448,
+      "step": 11304
+    },
+    {
+      "epoch": 0.6222136606307447,
+      "grad_norm": 0.6798059940338135,
+      "learning_rate": 7.820726101002578e-06,
+      "loss": 0.717,
+      "step": 11305
+    },
+    {
+      "epoch": 0.6222686994331004,
+      "grad_norm": 0.7623499631881714,
+      "learning_rate": 7.820368187600821e-06,
+      "loss": 0.7343,
+      "step": 11306
+    },
+    {
+      "epoch": 0.622323738235456,
+      "grad_norm": 0.703886866569519,
+      "learning_rate": 7.82001025300193e-06,
+      "loss": 0.8008,
+      "step": 11307
+    },
+    {
+      "epoch": 0.6223787770378116,
+      "grad_norm": 0.6817659735679626,
+      "learning_rate": 7.819652297208597e-06,
+      "loss": 0.7534,
+      "step": 11308
+    },
+    {
+      "epoch": 0.6224338158401673,
+      "grad_norm": 0.8991402983665466,
+      "learning_rate": 7.819294320223513e-06,
+      "loss": 0.6236,
+      "step": 11309
+    },
+    {
+      "epoch": 0.622488854642523,
+      "grad_norm": 0.791199803352356,
+      "learning_rate": 7.818936322049366e-06,
+      "loss": 0.772,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6225438934448786,
+      "grad_norm": 0.6401470303535461,
+      "learning_rate": 7.81857830268885e-06,
+      "loss": 0.7749,
+      "step": 11311
+    },
+    {
+      "epoch": 0.6225989322472343,
+      "grad_norm": 0.6731516122817993,
+      "learning_rate": 7.818220262144653e-06,
+      "loss": 0.7506,
+      "step": 11312
+    },
+    {
+      "epoch": 0.62265397104959,
+      "grad_norm": 0.7391661405563354,
+      "learning_rate": 7.817862200419467e-06,
+      "loss": 0.7288,
+      "step": 11313
+    },
+    {
+      "epoch": 0.6227090098519457,
+      "grad_norm": 0.7363784909248352,
+      "learning_rate": 7.817504117515984e-06,
+      "loss": 0.7087,
+      "step": 11314
+    },
+    {
+      "epoch": 0.6227640486543012,
+      "grad_norm": 0.7609296441078186,
+      "learning_rate": 7.817146013436893e-06,
+      "loss": 0.7553,
+      "step": 11315
+    },
+    {
+      "epoch": 0.6228190874566569,
+      "grad_norm": 0.6818829774856567,
+      "learning_rate": 7.816787888184886e-06,
+      "loss": 0.7534,
+      "step": 11316
+    },
+    {
+      "epoch": 0.6228741262590126,
+      "grad_norm": 0.7434844374656677,
+      "learning_rate": 7.816429741762657e-06,
+      "loss": 0.8008,
+      "step": 11317
+    },
+    {
+      "epoch": 0.6229291650613683,
+      "grad_norm": 0.6881742477416992,
+      "learning_rate": 7.816071574172895e-06,
+      "loss": 0.7324,
+      "step": 11318
+    },
+    {
+      "epoch": 0.6229842038637239,
+      "grad_norm": 0.7109540104866028,
+      "learning_rate": 7.815713385418293e-06,
+      "loss": 0.7954,
+      "step": 11319
+    },
+    {
+      "epoch": 0.6230392426660796,
+      "grad_norm": 0.6868860721588135,
+      "learning_rate": 7.815355175501542e-06,
+      "loss": 0.6703,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6230942814684353,
+      "grad_norm": 0.7851449847221375,
+      "learning_rate": 7.814996944425337e-06,
+      "loss": 0.8321,
+      "step": 11321
+    },
+    {
+      "epoch": 0.623149320270791,
+      "grad_norm": 0.7966809272766113,
+      "learning_rate": 7.814638692192367e-06,
+      "loss": 0.7603,
+      "step": 11322
+    },
+    {
+      "epoch": 0.6232043590731465,
+      "grad_norm": 0.6612964272499084,
+      "learning_rate": 7.814280418805327e-06,
+      "loss": 0.8096,
+      "step": 11323
+    },
+    {
+      "epoch": 0.6232593978755022,
+      "grad_norm": 0.6398881077766418,
+      "learning_rate": 7.813922124266908e-06,
+      "loss": 0.7559,
+      "step": 11324
+    },
+    {
+      "epoch": 0.6233144366778579,
+      "grad_norm": 0.8062521815299988,
+      "learning_rate": 7.813563808579804e-06,
+      "loss": 0.7863,
+      "step": 11325
+    },
+    {
+      "epoch": 0.6233694754802136,
+      "grad_norm": 0.7083317041397095,
+      "learning_rate": 7.813205471746708e-06,
+      "loss": 0.7358,
+      "step": 11326
+    },
+    {
+      "epoch": 0.6234245142825692,
+      "grad_norm": 0.6190419793128967,
+      "learning_rate": 7.812847113770312e-06,
+      "loss": 0.637,
+      "step": 11327
+    },
+    {
+      "epoch": 0.6234795530849249,
+      "grad_norm": 0.7036548256874084,
+      "learning_rate": 7.812488734653309e-06,
+      "loss": 0.8049,
+      "step": 11328
+    },
+    {
+      "epoch": 0.6235345918872806,
+      "grad_norm": 0.7952288389205933,
+      "learning_rate": 7.812130334398395e-06,
+      "loss": 0.781,
+      "step": 11329
+    },
+    {
+      "epoch": 0.6235896306896362,
+      "grad_norm": 0.7925593852996826,
+      "learning_rate": 7.811771913008262e-06,
+      "loss": 0.7913,
+      "step": 11330
+    },
+    {
+      "epoch": 0.6236446694919918,
+      "grad_norm": 0.7190900444984436,
+      "learning_rate": 7.811413470485604e-06,
+      "loss": 0.7464,
+      "step": 11331
+    },
+    {
+      "epoch": 0.6236997082943475,
+      "grad_norm": 0.6476338505744934,
+      "learning_rate": 7.811055006833114e-06,
+      "loss": 0.699,
+      "step": 11332
+    },
+    {
+      "epoch": 0.6237547470967032,
+      "grad_norm": 0.7412729263305664,
+      "learning_rate": 7.810696522053487e-06,
+      "loss": 0.7958,
+      "step": 11333
+    },
+    {
+      "epoch": 0.6238097858990589,
+      "grad_norm": 0.6646767854690552,
+      "learning_rate": 7.81033801614942e-06,
+      "loss": 0.6276,
+      "step": 11334
+    },
+    {
+      "epoch": 0.6238648247014145,
+      "grad_norm": 0.6912583112716675,
+      "learning_rate": 7.809979489123601e-06,
+      "loss": 0.7611,
+      "step": 11335
+    },
+    {
+      "epoch": 0.6239198635037702,
+      "grad_norm": 0.7324331998825073,
+      "learning_rate": 7.80962094097873e-06,
+      "loss": 0.7436,
+      "step": 11336
+    },
+    {
+      "epoch": 0.6239749023061258,
+      "grad_norm": 0.7046643495559692,
+      "learning_rate": 7.809262371717501e-06,
+      "loss": 0.7287,
+      "step": 11337
+    },
+    {
+      "epoch": 0.6240299411084815,
+      "grad_norm": 0.6013771891593933,
+      "learning_rate": 7.808903781342607e-06,
+      "loss": 0.6822,
+      "step": 11338
+    },
+    {
+      "epoch": 0.6240849799108371,
+      "grad_norm": 0.633074164390564,
+      "learning_rate": 7.808545169856745e-06,
+      "loss": 0.7758,
+      "step": 11339
+    },
+    {
+      "epoch": 0.6241400187131928,
+      "grad_norm": 0.6603411436080933,
+      "learning_rate": 7.808186537262608e-06,
+      "loss": 0.6797,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6241950575155485,
+      "grad_norm": 0.8316327929496765,
+      "learning_rate": 7.807827883562894e-06,
+      "loss": 0.777,
+      "step": 11341
+    },
+    {
+      "epoch": 0.6242500963179041,
+      "grad_norm": 0.7954252362251282,
+      "learning_rate": 7.807469208760295e-06,
+      "loss": 0.6581,
+      "step": 11342
+    },
+    {
+      "epoch": 0.6243051351202598,
+      "grad_norm": 0.6108134984970093,
+      "learning_rate": 7.80711051285751e-06,
+      "loss": 0.7126,
+      "step": 11343
+    },
+    {
+      "epoch": 0.6243601739226154,
+      "grad_norm": 0.7224909067153931,
+      "learning_rate": 7.806751795857235e-06,
+      "loss": 0.8677,
+      "step": 11344
+    },
+    {
+      "epoch": 0.6244152127249711,
+      "grad_norm": 0.720923125743866,
+      "learning_rate": 7.806393057762165e-06,
+      "loss": 0.7174,
+      "step": 11345
+    },
+    {
+      "epoch": 0.6244702515273267,
+      "grad_norm": 0.6837444305419922,
+      "learning_rate": 7.806034298574993e-06,
+      "loss": 0.7431,
+      "step": 11346
+    },
+    {
+      "epoch": 0.6245252903296824,
+      "grad_norm": 0.8486534953117371,
+      "learning_rate": 7.80567551829842e-06,
+      "loss": 0.7955,
+      "step": 11347
+    },
+    {
+      "epoch": 0.6245803291320381,
+      "grad_norm": 0.6459395885467529,
+      "learning_rate": 7.805316716935143e-06,
+      "loss": 0.7681,
+      "step": 11348
+    },
+    {
+      "epoch": 0.6246353679343938,
+      "grad_norm": 0.8414636850357056,
+      "learning_rate": 7.804957894487854e-06,
+      "loss": 0.8985,
+      "step": 11349
+    },
+    {
+      "epoch": 0.6246904067367494,
+      "grad_norm": 0.7930828928947449,
+      "learning_rate": 7.804599050959254e-06,
+      "loss": 0.7389,
+      "step": 11350
+    },
+    {
+      "epoch": 0.624745445539105,
+      "grad_norm": 0.7102516889572144,
+      "learning_rate": 7.804240186352038e-06,
+      "loss": 0.8072,
+      "step": 11351
+    },
+    {
+      "epoch": 0.6248004843414607,
+      "grad_norm": 0.773341178894043,
+      "learning_rate": 7.803881300668901e-06,
+      "loss": 0.7531,
+      "step": 11352
+    },
+    {
+      "epoch": 0.6248555231438164,
+      "grad_norm": 0.6354981064796448,
+      "learning_rate": 7.803522393912544e-06,
+      "loss": 0.6761,
+      "step": 11353
+    },
+    {
+      "epoch": 0.624910561946172,
+      "grad_norm": 0.7833859324455261,
+      "learning_rate": 7.803163466085663e-06,
+      "loss": 0.7768,
+      "step": 11354
+    },
+    {
+      "epoch": 0.6249656007485277,
+      "grad_norm": 0.6982376575469971,
+      "learning_rate": 7.802804517190957e-06,
+      "loss": 0.7472,
+      "step": 11355
+    },
+    {
+      "epoch": 0.6250206395508834,
+      "grad_norm": 0.7214694023132324,
+      "learning_rate": 7.80244554723112e-06,
+      "loss": 0.7919,
+      "step": 11356
+    },
+    {
+      "epoch": 0.6250756783532391,
+      "grad_norm": 0.8002933859825134,
+      "learning_rate": 7.802086556208855e-06,
+      "loss": 0.8278,
+      "step": 11357
+    },
+    {
+      "epoch": 0.6251307171555947,
+      "grad_norm": 0.7619680762290955,
+      "learning_rate": 7.801727544126858e-06,
+      "loss": 0.7775,
+      "step": 11358
+    },
+    {
+      "epoch": 0.6251857559579503,
+      "grad_norm": 0.6340392827987671,
+      "learning_rate": 7.801368510987825e-06,
+      "loss": 0.7324,
+      "step": 11359
+    },
+    {
+      "epoch": 0.625240794760306,
+      "grad_norm": 0.6754844784736633,
+      "learning_rate": 7.801009456794457e-06,
+      "loss": 0.7296,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6252958335626617,
+      "grad_norm": 0.6871771216392517,
+      "learning_rate": 7.80065038154945e-06,
+      "loss": 0.7398,
+      "step": 11361
+    },
+    {
+      "epoch": 0.6253508723650173,
+      "grad_norm": 0.6610772013664246,
+      "learning_rate": 7.800291285255505e-06,
+      "loss": 0.738,
+      "step": 11362
+    },
+    {
+      "epoch": 0.625405911167373,
+      "grad_norm": 0.6858081221580505,
+      "learning_rate": 7.799932167915322e-06,
+      "loss": 0.7353,
+      "step": 11363
+    },
+    {
+      "epoch": 0.6254609499697287,
+      "grad_norm": 0.6698840856552124,
+      "learning_rate": 7.799573029531597e-06,
+      "loss": 0.7505,
+      "step": 11364
+    },
+    {
+      "epoch": 0.6255159887720844,
+      "grad_norm": 0.7374000549316406,
+      "learning_rate": 7.799213870107031e-06,
+      "loss": 0.7974,
+      "step": 11365
+    },
+    {
+      "epoch": 0.6255710275744399,
+      "grad_norm": 0.6962621808052063,
+      "learning_rate": 7.798854689644324e-06,
+      "loss": 0.8183,
+      "step": 11366
+    },
+    {
+      "epoch": 0.6256260663767956,
+      "grad_norm": 0.8477681279182434,
+      "learning_rate": 7.798495488146173e-06,
+      "loss": 0.7533,
+      "step": 11367
+    },
+    {
+      "epoch": 0.6256811051791513,
+      "grad_norm": 0.6963459253311157,
+      "learning_rate": 7.798136265615278e-06,
+      "loss": 0.6362,
+      "step": 11368
+    },
+    {
+      "epoch": 0.625736143981507,
+      "grad_norm": 0.7125601172447205,
+      "learning_rate": 7.79777702205434e-06,
+      "loss": 0.7296,
+      "step": 11369
+    },
+    {
+      "epoch": 0.6257911827838626,
+      "grad_norm": 0.6650554537773132,
+      "learning_rate": 7.79741775746606e-06,
+      "loss": 0.8231,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6258462215862183,
+      "grad_norm": 0.6556620597839355,
+      "learning_rate": 7.797058471853138e-06,
+      "loss": 0.6952,
+      "step": 11371
+    },
+    {
+      "epoch": 0.625901260388574,
+      "grad_norm": 0.6350956559181213,
+      "learning_rate": 7.79669916521827e-06,
+      "loss": 0.686,
+      "step": 11372
+    },
+    {
+      "epoch": 0.6259562991909297,
+      "grad_norm": 0.6346702575683594,
+      "learning_rate": 7.796339837564163e-06,
+      "loss": 0.7234,
+      "step": 11373
+    },
+    {
+      "epoch": 0.6260113379932852,
+      "grad_norm": 0.741437554359436,
+      "learning_rate": 7.795980488893514e-06,
+      "loss": 0.8096,
+      "step": 11374
+    },
+    {
+      "epoch": 0.6260663767956409,
+      "grad_norm": 0.7057582139968872,
+      "learning_rate": 7.795621119209021e-06,
+      "loss": 0.8022,
+      "step": 11375
+    },
+    {
+      "epoch": 0.6261214155979966,
+      "grad_norm": 0.658107578754425,
+      "learning_rate": 7.79526172851339e-06,
+      "loss": 0.7564,
+      "step": 11376
+    },
+    {
+      "epoch": 0.6261764544003523,
+      "grad_norm": 0.7974086403846741,
+      "learning_rate": 7.79490231680932e-06,
+      "loss": 0.7721,
+      "step": 11377
+    },
+    {
+      "epoch": 0.6262314932027079,
+      "grad_norm": 0.6669130921363831,
+      "learning_rate": 7.794542884099513e-06,
+      "loss": 0.7652,
+      "step": 11378
+    },
+    {
+      "epoch": 0.6262865320050636,
+      "grad_norm": 0.7364919185638428,
+      "learning_rate": 7.794183430386669e-06,
+      "loss": 0.8679,
+      "step": 11379
+    },
+    {
+      "epoch": 0.6263415708074193,
+      "grad_norm": 0.7383667230606079,
+      "learning_rate": 7.793823955673489e-06,
+      "loss": 0.7715,
+      "step": 11380
+    },
+    {
+      "epoch": 0.626396609609775,
+      "grad_norm": 0.6688774228096008,
+      "learning_rate": 7.793464459962679e-06,
+      "loss": 0.7503,
+      "step": 11381
+    },
+    {
+      "epoch": 0.6264516484121305,
+      "grad_norm": 0.6771709322929382,
+      "learning_rate": 7.793104943256935e-06,
+      "loss": 0.7479,
+      "step": 11382
+    },
+    {
+      "epoch": 0.6265066872144862,
+      "grad_norm": 0.7121349573135376,
+      "learning_rate": 7.792745405558964e-06,
+      "loss": 0.7655,
+      "step": 11383
+    },
+    {
+      "epoch": 0.6265617260168419,
+      "grad_norm": 0.6315643787384033,
+      "learning_rate": 7.792385846871465e-06,
+      "loss": 0.7418,
+      "step": 11384
+    },
+    {
+      "epoch": 0.6266167648191975,
+      "grad_norm": 0.6701569557189941,
+      "learning_rate": 7.792026267197142e-06,
+      "loss": 0.7669,
+      "step": 11385
+    },
+    {
+      "epoch": 0.6266718036215532,
+      "grad_norm": 0.6890652179718018,
+      "learning_rate": 7.791666666538697e-06,
+      "loss": 0.7659,
+      "step": 11386
+    },
+    {
+      "epoch": 0.6267268424239089,
+      "grad_norm": 0.7636297345161438,
+      "learning_rate": 7.791307044898833e-06,
+      "loss": 0.7272,
+      "step": 11387
+    },
+    {
+      "epoch": 0.6267818812262645,
+      "grad_norm": 0.6563602089881897,
+      "learning_rate": 7.790947402280252e-06,
+      "loss": 0.7603,
+      "step": 11388
+    },
+    {
+      "epoch": 0.6268369200286201,
+      "grad_norm": 0.7252678275108337,
+      "learning_rate": 7.790587738685655e-06,
+      "loss": 0.7789,
+      "step": 11389
+    },
+    {
+      "epoch": 0.6268919588309758,
+      "grad_norm": 0.6703618764877319,
+      "learning_rate": 7.79022805411775e-06,
+      "loss": 0.6883,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6269469976333315,
+      "grad_norm": 0.7165848612785339,
+      "learning_rate": 7.789868348579239e-06,
+      "loss": 0.7944,
+      "step": 11391
+    },
+    {
+      "epoch": 0.6270020364356872,
+      "grad_norm": 0.9325329065322876,
+      "learning_rate": 7.789508622072822e-06,
+      "loss": 0.9059,
+      "step": 11392
+    },
+    {
+      "epoch": 0.6270570752380428,
+      "grad_norm": 0.6875555515289307,
+      "learning_rate": 7.789148874601204e-06,
+      "loss": 0.7115,
+      "step": 11393
+    },
+    {
+      "epoch": 0.6271121140403985,
+      "grad_norm": 0.6470181941986084,
+      "learning_rate": 7.788789106167093e-06,
+      "loss": 0.7603,
+      "step": 11394
+    },
+    {
+      "epoch": 0.6271671528427541,
+      "grad_norm": 0.688685417175293,
+      "learning_rate": 7.788429316773188e-06,
+      "loss": 0.8397,
+      "step": 11395
+    },
+    {
+      "epoch": 0.6272221916451098,
+      "grad_norm": 0.6299887895584106,
+      "learning_rate": 7.788069506422193e-06,
+      "loss": 0.7026,
+      "step": 11396
+    },
+    {
+      "epoch": 0.6272772304474654,
+      "grad_norm": 0.8046191930770874,
+      "learning_rate": 7.787709675116817e-06,
+      "loss": 0.8573,
+      "step": 11397
+    },
+    {
+      "epoch": 0.6273322692498211,
+      "grad_norm": 0.6700685620307922,
+      "learning_rate": 7.78734982285976e-06,
+      "loss": 0.7225,
+      "step": 11398
+    },
+    {
+      "epoch": 0.6273873080521768,
+      "grad_norm": 0.6968538761138916,
+      "learning_rate": 7.786989949653726e-06,
+      "loss": 0.6571,
+      "step": 11399
+    },
+    {
+      "epoch": 0.6274423468545325,
+      "grad_norm": 0.6857314705848694,
+      "learning_rate": 7.786630055501425e-06,
+      "loss": 0.8131,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6274973856568881,
+      "grad_norm": 0.702316403388977,
+      "learning_rate": 7.786270140405557e-06,
+      "loss": 0.7222,
+      "step": 11401
+    },
+    {
+      "epoch": 0.6275524244592438,
+      "grad_norm": 0.6987283825874329,
+      "learning_rate": 7.785910204368827e-06,
+      "loss": 0.7171,
+      "step": 11402
+    },
+    {
+      "epoch": 0.6276074632615994,
+      "grad_norm": 0.6835529208183289,
+      "learning_rate": 7.785550247393943e-06,
+      "loss": 0.8077,
+      "step": 11403
+    },
+    {
+      "epoch": 0.6276625020639551,
+      "grad_norm": 0.6423392295837402,
+      "learning_rate": 7.785190269483609e-06,
+      "loss": 0.6689,
+      "step": 11404
+    },
+    {
+      "epoch": 0.6277175408663107,
+      "grad_norm": 0.6995517611503601,
+      "learning_rate": 7.78483027064053e-06,
+      "loss": 0.7417,
+      "step": 11405
+    },
+    {
+      "epoch": 0.6277725796686664,
+      "grad_norm": 0.6639729142189026,
+      "learning_rate": 7.784470250867413e-06,
+      "loss": 0.6521,
+      "step": 11406
+    },
+    {
+      "epoch": 0.6278276184710221,
+      "grad_norm": 0.7280262112617493,
+      "learning_rate": 7.784110210166961e-06,
+      "loss": 0.7686,
+      "step": 11407
+    },
+    {
+      "epoch": 0.6278826572733778,
+      "grad_norm": 0.6741863489151001,
+      "learning_rate": 7.783750148541884e-06,
+      "loss": 0.7794,
+      "step": 11408
+    },
+    {
+      "epoch": 0.6279376960757334,
+      "grad_norm": 0.8160151243209839,
+      "learning_rate": 7.783390065994885e-06,
+      "loss": 0.7065,
+      "step": 11409
+    },
+    {
+      "epoch": 0.627992734878089,
+      "grad_norm": 0.7288973927497864,
+      "learning_rate": 7.783029962528672e-06,
+      "loss": 0.8337,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6280477736804447,
+      "grad_norm": 0.7764643430709839,
+      "learning_rate": 7.782669838145952e-06,
+      "loss": 0.8812,
+      "step": 11411
+    },
+    {
+      "epoch": 0.6281028124828004,
+      "grad_norm": 0.8145303130149841,
+      "learning_rate": 7.782309692849425e-06,
+      "loss": 0.9206,
+      "step": 11412
+    },
+    {
+      "epoch": 0.628157851285156,
+      "grad_norm": 0.6883288621902466,
+      "learning_rate": 7.781949526641808e-06,
+      "loss": 0.7779,
+      "step": 11413
+    },
+    {
+      "epoch": 0.6282128900875117,
+      "grad_norm": 0.7281043529510498,
+      "learning_rate": 7.781589339525803e-06,
+      "loss": 0.7933,
+      "step": 11414
+    },
+    {
+      "epoch": 0.6282679288898674,
+      "grad_norm": 0.7998347878456116,
+      "learning_rate": 7.781229131504115e-06,
+      "loss": 0.8772,
+      "step": 11415
+    },
+    {
+      "epoch": 0.6283229676922231,
+      "grad_norm": 0.7591177225112915,
+      "learning_rate": 7.780868902579455e-06,
+      "loss": 0.9054,
+      "step": 11416
+    },
+    {
+      "epoch": 0.6283780064945786,
+      "grad_norm": 0.7209650278091431,
+      "learning_rate": 7.780508652754528e-06,
+      "loss": 0.7781,
+      "step": 11417
+    },
+    {
+      "epoch": 0.6284330452969343,
+      "grad_norm": 1.2373511791229248,
+      "learning_rate": 7.780148382032042e-06,
+      "loss": 0.7501,
+      "step": 11418
+    },
+    {
+      "epoch": 0.62848808409929,
+      "grad_norm": 0.6281551122665405,
+      "learning_rate": 7.779788090414704e-06,
+      "loss": 0.8122,
+      "step": 11419
+    },
+    {
+      "epoch": 0.6285431229016457,
+      "grad_norm": 0.6954115629196167,
+      "learning_rate": 7.779427777905224e-06,
+      "loss": 0.7815,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6285981617040013,
+      "grad_norm": 0.727043628692627,
+      "learning_rate": 7.77906744450631e-06,
+      "loss": 0.7116,
+      "step": 11421
+    },
+    {
+      "epoch": 0.628653200506357,
+      "grad_norm": 0.6979809403419495,
+      "learning_rate": 7.778707090220667e-06,
+      "loss": 0.7707,
+      "step": 11422
+    },
+    {
+      "epoch": 0.6287082393087127,
+      "grad_norm": 0.6851169466972351,
+      "learning_rate": 7.778346715051006e-06,
+      "loss": 0.811,
+      "step": 11423
+    },
+    {
+      "epoch": 0.6287632781110684,
+      "grad_norm": 0.70259028673172,
+      "learning_rate": 7.777986319000036e-06,
+      "loss": 0.7766,
+      "step": 11424
+    },
+    {
+      "epoch": 0.6288183169134239,
+      "grad_norm": 0.7436364889144897,
+      "learning_rate": 7.777625902070463e-06,
+      "loss": 0.8449,
+      "step": 11425
+    },
+    {
+      "epoch": 0.6288733557157796,
+      "grad_norm": 0.6452080607414246,
+      "learning_rate": 7.777265464264998e-06,
+      "loss": 0.7138,
+      "step": 11426
+    },
+    {
+      "epoch": 0.6289283945181353,
+      "grad_norm": 0.6329460144042969,
+      "learning_rate": 7.776905005586349e-06,
+      "loss": 0.6482,
+      "step": 11427
+    },
+    {
+      "epoch": 0.6289834333204909,
+      "grad_norm": 0.7521186470985413,
+      "learning_rate": 7.776544526037225e-06,
+      "loss": 0.751,
+      "step": 11428
+    },
+    {
+      "epoch": 0.6290384721228466,
+      "grad_norm": 0.7105319499969482,
+      "learning_rate": 7.776184025620334e-06,
+      "loss": 0.843,
+      "step": 11429
+    },
+    {
+      "epoch": 0.6290935109252023,
+      "grad_norm": 0.7329964637756348,
+      "learning_rate": 7.77582350433839e-06,
+      "loss": 0.6992,
+      "step": 11430
+    },
+    {
+      "epoch": 0.629148549727558,
+      "grad_norm": 0.7492092847824097,
+      "learning_rate": 7.775462962194098e-06,
+      "loss": 0.7579,
+      "step": 11431
+    },
+    {
+      "epoch": 0.6292035885299135,
+      "grad_norm": 0.7332866191864014,
+      "learning_rate": 7.77510239919017e-06,
+      "loss": 0.7758,
+      "step": 11432
+    },
+    {
+      "epoch": 0.6292586273322692,
+      "grad_norm": 0.7532867193222046,
+      "learning_rate": 7.774741815329315e-06,
+      "loss": 0.8157,
+      "step": 11433
+    },
+    {
+      "epoch": 0.6293136661346249,
+      "grad_norm": 0.7498316168785095,
+      "learning_rate": 7.774381210614244e-06,
+      "loss": 0.7671,
+      "step": 11434
+    },
+    {
+      "epoch": 0.6293687049369806,
+      "grad_norm": 0.8017444610595703,
+      "learning_rate": 7.774020585047666e-06,
+      "loss": 0.6989,
+      "step": 11435
+    },
+    {
+      "epoch": 0.6294237437393362,
+      "grad_norm": 0.7827737927436829,
+      "learning_rate": 7.77365993863229e-06,
+      "loss": 0.852,
+      "step": 11436
+    },
+    {
+      "epoch": 0.6294787825416919,
+      "grad_norm": 1.1411668062210083,
+      "learning_rate": 7.77329927137083e-06,
+      "loss": 0.9303,
+      "step": 11437
+    },
+    {
+      "epoch": 0.6295338213440476,
+      "grad_norm": 1.2931067943572998,
+      "learning_rate": 7.772938583265995e-06,
+      "loss": 0.8913,
+      "step": 11438
+    },
+    {
+      "epoch": 0.6295888601464033,
+      "grad_norm": 0.7407616376876831,
+      "learning_rate": 7.772577874320494e-06,
+      "loss": 0.9247,
+      "step": 11439
+    },
+    {
+      "epoch": 0.6296438989487588,
+      "grad_norm": 0.6544716954231262,
+      "learning_rate": 7.772217144537043e-06,
+      "loss": 0.7879,
+      "step": 11440
+    },
+    {
+      "epoch": 0.6296989377511145,
+      "grad_norm": 0.7467932105064392,
+      "learning_rate": 7.77185639391835e-06,
+      "loss": 0.7624,
+      "step": 11441
+    },
+    {
+      "epoch": 0.6297539765534702,
+      "grad_norm": 0.6845136880874634,
+      "learning_rate": 7.771495622467123e-06,
+      "loss": 0.691,
+      "step": 11442
+    },
+    {
+      "epoch": 0.6298090153558259,
+      "grad_norm": 0.7881575226783752,
+      "learning_rate": 7.771134830186079e-06,
+      "loss": 0.7567,
+      "step": 11443
+    },
+    {
+      "epoch": 0.6298640541581815,
+      "grad_norm": 0.6910528540611267,
+      "learning_rate": 7.770774017077928e-06,
+      "loss": 0.7527,
+      "step": 11444
+    },
+    {
+      "epoch": 0.6299190929605372,
+      "grad_norm": 0.7395550608634949,
+      "learning_rate": 7.770413183145379e-06,
+      "loss": 0.8288,
+      "step": 11445
+    },
+    {
+      "epoch": 0.6299741317628929,
+      "grad_norm": 0.6876364350318909,
+      "learning_rate": 7.770052328391147e-06,
+      "loss": 0.7759,
+      "step": 11446
+    },
+    {
+      "epoch": 0.6300291705652485,
+      "grad_norm": 0.7936999201774597,
+      "learning_rate": 7.769691452817945e-06,
+      "loss": 0.6885,
+      "step": 11447
+    },
+    {
+      "epoch": 0.6300842093676041,
+      "grad_norm": 0.721479058265686,
+      "learning_rate": 7.769330556428482e-06,
+      "loss": 0.7215,
+      "step": 11448
+    },
+    {
+      "epoch": 0.6301392481699598,
+      "grad_norm": 0.6549312472343445,
+      "learning_rate": 7.76896963922547e-06,
+      "loss": 0.7523,
+      "step": 11449
+    },
+    {
+      "epoch": 0.6301942869723155,
+      "grad_norm": 0.6684648394584656,
+      "learning_rate": 7.768608701211627e-06,
+      "loss": 0.768,
+      "step": 11450
+    },
+    {
+      "epoch": 0.6302493257746712,
+      "grad_norm": 0.7014286518096924,
+      "learning_rate": 7.76824774238966e-06,
+      "loss": 0.7534,
+      "step": 11451
+    },
+    {
+      "epoch": 0.6303043645770268,
+      "grad_norm": 0.9186445474624634,
+      "learning_rate": 7.767886762762284e-06,
+      "loss": 0.8398,
+      "step": 11452
+    },
+    {
+      "epoch": 0.6303594033793825,
+      "grad_norm": 0.787187933921814,
+      "learning_rate": 7.76752576233221e-06,
+      "loss": 0.8035,
+      "step": 11453
+    },
+    {
+      "epoch": 0.6304144421817381,
+      "grad_norm": 0.7471121549606323,
+      "learning_rate": 7.767164741102157e-06,
+      "loss": 0.7983,
+      "step": 11454
+    },
+    {
+      "epoch": 0.6304694809840938,
+      "grad_norm": 0.6810591816902161,
+      "learning_rate": 7.766803699074834e-06,
+      "loss": 0.7132,
+      "step": 11455
+    },
+    {
+      "epoch": 0.6305245197864494,
+      "grad_norm": 0.7154163122177124,
+      "learning_rate": 7.766442636252953e-06,
+      "loss": 0.7942,
+      "step": 11456
+    },
+    {
+      "epoch": 0.6305795585888051,
+      "grad_norm": 0.6990880966186523,
+      "learning_rate": 7.766081552639231e-06,
+      "loss": 0.7296,
+      "step": 11457
+    },
+    {
+      "epoch": 0.6306345973911608,
+      "grad_norm": 0.8848066926002502,
+      "learning_rate": 7.76572044823638e-06,
+      "loss": 0.621,
+      "step": 11458
+    },
+    {
+      "epoch": 0.6306896361935165,
+      "grad_norm": 0.6929910182952881,
+      "learning_rate": 7.765359323047116e-06,
+      "loss": 0.5917,
+      "step": 11459
+    },
+    {
+      "epoch": 0.6307446749958721,
+      "grad_norm": 0.6874505281448364,
+      "learning_rate": 7.764998177074149e-06,
+      "loss": 0.7244,
+      "step": 11460
+    },
+    {
+      "epoch": 0.6307997137982277,
+      "grad_norm": 0.6823066473007202,
+      "learning_rate": 7.764637010320197e-06,
+      "loss": 0.7299,
+      "step": 11461
+    },
+    {
+      "epoch": 0.6308547526005834,
+      "grad_norm": 0.7315061688423157,
+      "learning_rate": 7.764275822787972e-06,
+      "loss": 0.7759,
+      "step": 11462
+    },
+    {
+      "epoch": 0.6309097914029391,
+      "grad_norm": 0.6186662316322327,
+      "learning_rate": 7.763914614480192e-06,
+      "loss": 0.6746,
+      "step": 11463
+    },
+    {
+      "epoch": 0.6309648302052947,
+      "grad_norm": 0.6751530170440674,
+      "learning_rate": 7.763553385399569e-06,
+      "loss": 0.8371,
+      "step": 11464
+    },
+    {
+      "epoch": 0.6310198690076504,
+      "grad_norm": 1.0283396244049072,
+      "learning_rate": 7.763192135548818e-06,
+      "loss": 0.7743,
+      "step": 11465
+    },
+    {
+      "epoch": 0.6310749078100061,
+      "grad_norm": 0.7695029973983765,
+      "learning_rate": 7.762830864930655e-06,
+      "loss": 0.7387,
+      "step": 11466
+    },
+    {
+      "epoch": 0.6311299466123618,
+      "grad_norm": 0.8087024688720703,
+      "learning_rate": 7.762469573547795e-06,
+      "loss": 0.8357,
+      "step": 11467
+    },
+    {
+      "epoch": 0.6311849854147173,
+      "grad_norm": 0.9203382134437561,
+      "learning_rate": 7.762108261402951e-06,
+      "loss": 0.8191,
+      "step": 11468
+    },
+    {
+      "epoch": 0.631240024217073,
+      "grad_norm": 0.6569168567657471,
+      "learning_rate": 7.761746928498843e-06,
+      "loss": 0.7035,
+      "step": 11469
+    },
+    {
+      "epoch": 0.6312950630194287,
+      "grad_norm": 0.7903677225112915,
+      "learning_rate": 7.761385574838183e-06,
+      "loss": 0.8295,
+      "step": 11470
+    },
+    {
+      "epoch": 0.6313501018217843,
+      "grad_norm": 0.6780279278755188,
+      "learning_rate": 7.76102420042369e-06,
+      "loss": 0.6497,
+      "step": 11471
+    },
+    {
+      "epoch": 0.63140514062414,
+      "grad_norm": 0.7150516510009766,
+      "learning_rate": 7.760662805258076e-06,
+      "loss": 0.7979,
+      "step": 11472
+    },
+    {
+      "epoch": 0.6314601794264957,
+      "grad_norm": 0.7278215885162354,
+      "learning_rate": 7.760301389344061e-06,
+      "loss": 0.8503,
+      "step": 11473
+    },
+    {
+      "epoch": 0.6315152182288514,
+      "grad_norm": 0.8695063591003418,
+      "learning_rate": 7.75993995268436e-06,
+      "loss": 0.7796,
+      "step": 11474
+    },
+    {
+      "epoch": 0.631570257031207,
+      "grad_norm": 0.7154332399368286,
+      "learning_rate": 7.759578495281688e-06,
+      "loss": 0.725,
+      "step": 11475
+    },
+    {
+      "epoch": 0.6316252958335626,
+      "grad_norm": 0.7151778936386108,
+      "learning_rate": 7.759217017138763e-06,
+      "loss": 0.6932,
+      "step": 11476
+    },
+    {
+      "epoch": 0.6316803346359183,
+      "grad_norm": 0.6328319311141968,
+      "learning_rate": 7.758855518258301e-06,
+      "loss": 0.7382,
+      "step": 11477
+    },
+    {
+      "epoch": 0.631735373438274,
+      "grad_norm": 0.8377438187599182,
+      "learning_rate": 7.75849399864302e-06,
+      "loss": 0.7782,
+      "step": 11478
+    },
+    {
+      "epoch": 0.6317904122406296,
+      "grad_norm": 0.6654751896858215,
+      "learning_rate": 7.758132458295637e-06,
+      "loss": 0.8076,
+      "step": 11479
+    },
+    {
+      "epoch": 0.6318454510429853,
+      "grad_norm": 0.6841873526573181,
+      "learning_rate": 7.757770897218869e-06,
+      "loss": 0.7195,
+      "step": 11480
+    },
+    {
+      "epoch": 0.631900489845341,
+      "grad_norm": 0.7791223526000977,
+      "learning_rate": 7.757409315415431e-06,
+      "loss": 0.7858,
+      "step": 11481
+    },
+    {
+      "epoch": 0.6319555286476967,
+      "grad_norm": 0.6412019729614258,
+      "learning_rate": 7.757047712888044e-06,
+      "loss": 0.6853,
+      "step": 11482
+    },
+    {
+      "epoch": 0.6320105674500522,
+      "grad_norm": 0.7058777213096619,
+      "learning_rate": 7.756686089639425e-06,
+      "loss": 0.8955,
+      "step": 11483
+    },
+    {
+      "epoch": 0.6320656062524079,
+      "grad_norm": 0.6950271725654602,
+      "learning_rate": 7.75632444567229e-06,
+      "loss": 0.7213,
+      "step": 11484
+    },
+    {
+      "epoch": 0.6321206450547636,
+      "grad_norm": 0.6938642859458923,
+      "learning_rate": 7.755962780989359e-06,
+      "loss": 0.749,
+      "step": 11485
+    },
+    {
+      "epoch": 0.6321756838571193,
+      "grad_norm": 4.447030544281006,
+      "learning_rate": 7.755601095593348e-06,
+      "loss": 0.7603,
+      "step": 11486
+    },
+    {
+      "epoch": 0.6322307226594749,
+      "grad_norm": 0.6693708896636963,
+      "learning_rate": 7.755239389486979e-06,
+      "loss": 0.769,
+      "step": 11487
+    },
+    {
+      "epoch": 0.6322857614618306,
+      "grad_norm": 0.830352246761322,
+      "learning_rate": 7.754877662672968e-06,
+      "loss": 0.8069,
+      "step": 11488
+    },
+    {
+      "epoch": 0.6323408002641863,
+      "grad_norm": 0.7211840748786926,
+      "learning_rate": 7.754515915154033e-06,
+      "loss": 0.7972,
+      "step": 11489
+    },
+    {
+      "epoch": 0.632395839066542,
+      "grad_norm": 0.723101019859314,
+      "learning_rate": 7.754154146932893e-06,
+      "loss": 0.7385,
+      "step": 11490
+    },
+    {
+      "epoch": 0.6324508778688975,
+      "grad_norm": 0.6515377759933472,
+      "learning_rate": 7.75379235801227e-06,
+      "loss": 0.7527,
+      "step": 11491
+    },
+    {
+      "epoch": 0.6325059166712532,
+      "grad_norm": 0.6296554803848267,
+      "learning_rate": 7.75343054839488e-06,
+      "loss": 0.7135,
+      "step": 11492
+    },
+    {
+      "epoch": 0.6325609554736089,
+      "grad_norm": 0.8153911232948303,
+      "learning_rate": 7.753068718083441e-06,
+      "loss": 0.7298,
+      "step": 11493
+    },
+    {
+      "epoch": 0.6326159942759646,
+      "grad_norm": 0.6735014915466309,
+      "learning_rate": 7.752706867080676e-06,
+      "loss": 0.6851,
+      "step": 11494
+    },
+    {
+      "epoch": 0.6326710330783202,
+      "grad_norm": 0.7077293992042542,
+      "learning_rate": 7.752344995389303e-06,
+      "loss": 0.7806,
+      "step": 11495
+    },
+    {
+      "epoch": 0.6327260718806759,
+      "grad_norm": 0.6928272843360901,
+      "learning_rate": 7.751983103012042e-06,
+      "loss": 0.7538,
+      "step": 11496
+    },
+    {
+      "epoch": 0.6327811106830316,
+      "grad_norm": 0.7058837413787842,
+      "learning_rate": 7.751621189951612e-06,
+      "loss": 0.7065,
+      "step": 11497
+    },
+    {
+      "epoch": 0.6328361494853872,
+      "grad_norm": 0.7272600531578064,
+      "learning_rate": 7.751259256210735e-06,
+      "loss": 0.7468,
+      "step": 11498
+    },
+    {
+      "epoch": 0.6328911882877428,
+      "grad_norm": 0.6175968050956726,
+      "learning_rate": 7.75089730179213e-06,
+      "loss": 0.7195,
+      "step": 11499
+    },
+    {
+      "epoch": 0.6329462270900985,
+      "grad_norm": 0.6567386984825134,
+      "learning_rate": 7.750535326698514e-06,
+      "loss": 0.8147,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6330012658924542,
+      "grad_norm": 0.6325315237045288,
+      "learning_rate": 7.750173330932613e-06,
+      "loss": 0.7087,
+      "step": 11501
+    },
+    {
+      "epoch": 0.6330563046948099,
+      "grad_norm": 0.8607509732246399,
+      "learning_rate": 7.749811314497147e-06,
+      "loss": 0.8009,
+      "step": 11502
+    },
+    {
+      "epoch": 0.6331113434971655,
+      "grad_norm": 0.7452824711799622,
+      "learning_rate": 7.749449277394833e-06,
+      "loss": 0.7497,
+      "step": 11503
+    },
+    {
+      "epoch": 0.6331663822995212,
+      "grad_norm": 0.7371357679367065,
+      "learning_rate": 7.749087219628395e-06,
+      "loss": 0.8936,
+      "step": 11504
+    },
+    {
+      "epoch": 0.6332214211018768,
+      "grad_norm": 0.7177306413650513,
+      "learning_rate": 7.748725141200552e-06,
+      "loss": 0.8327,
+      "step": 11505
+    },
+    {
+      "epoch": 0.6332764599042325,
+      "grad_norm": 0.5938527584075928,
+      "learning_rate": 7.748363042114028e-06,
+      "loss": 0.6471,
+      "step": 11506
+    },
+    {
+      "epoch": 0.6333314987065881,
+      "grad_norm": 0.8827341198921204,
+      "learning_rate": 7.748000922371543e-06,
+      "loss": 0.7247,
+      "step": 11507
+    },
+    {
+      "epoch": 0.6333865375089438,
+      "grad_norm": 0.7008641958236694,
+      "learning_rate": 7.747638781975818e-06,
+      "loss": 0.684,
+      "step": 11508
+    },
+    {
+      "epoch": 0.6334415763112995,
+      "grad_norm": 0.7752355337142944,
+      "learning_rate": 7.747276620929576e-06,
+      "loss": 0.7993,
+      "step": 11509
+    },
+    {
+      "epoch": 0.6334966151136552,
+      "grad_norm": 0.6928088068962097,
+      "learning_rate": 7.74691443923554e-06,
+      "loss": 0.7213,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6335516539160108,
+      "grad_norm": 0.8197296261787415,
+      "learning_rate": 7.746552236896428e-06,
+      "loss": 0.847,
+      "step": 11511
+    },
+    {
+      "epoch": 0.6336066927183664,
+      "grad_norm": 0.7912493348121643,
+      "learning_rate": 7.746190013914966e-06,
+      "loss": 0.8217,
+      "step": 11512
+    },
+    {
+      "epoch": 0.6336617315207221,
+      "grad_norm": 0.7726556062698364,
+      "learning_rate": 7.745827770293871e-06,
+      "loss": 0.7626,
+      "step": 11513
+    },
+    {
+      "epoch": 0.6337167703230777,
+      "grad_norm": 0.668569028377533,
+      "learning_rate": 7.745465506035873e-06,
+      "loss": 0.7141,
+      "step": 11514
+    },
+    {
+      "epoch": 0.6337718091254334,
+      "grad_norm": 0.7226139903068542,
+      "learning_rate": 7.745103221143694e-06,
+      "loss": 0.7262,
+      "step": 11515
+    },
+    {
+      "epoch": 0.6338268479277891,
+      "grad_norm": 0.7315354943275452,
+      "learning_rate": 7.744740915620051e-06,
+      "loss": 0.7955,
+      "step": 11516
+    },
+    {
+      "epoch": 0.6338818867301448,
+      "grad_norm": 0.6815279126167297,
+      "learning_rate": 7.744378589467668e-06,
+      "loss": 0.7347,
+      "step": 11517
+    },
+    {
+      "epoch": 0.6339369255325004,
+      "grad_norm": 0.6931445598602295,
+      "learning_rate": 7.744016242689272e-06,
+      "loss": 0.7959,
+      "step": 11518
+    },
+    {
+      "epoch": 0.633991964334856,
+      "grad_norm": 0.7156991362571716,
+      "learning_rate": 7.743653875287584e-06,
+      "loss": 0.7793,
+      "step": 11519
+    },
+    {
+      "epoch": 0.6340470031372117,
+      "grad_norm": 0.8503926396369934,
+      "learning_rate": 7.74329148726533e-06,
+      "loss": 0.823,
+      "step": 11520
+    },
+    {
+      "epoch": 0.6341020419395674,
+      "grad_norm": 0.6280057430267334,
+      "learning_rate": 7.742929078625228e-06,
+      "loss": 0.6729,
+      "step": 11521
+    },
+    {
+      "epoch": 0.634157080741923,
+      "grad_norm": 0.7004517316818237,
+      "learning_rate": 7.742566649370008e-06,
+      "loss": 0.7578,
+      "step": 11522
+    },
+    {
+      "epoch": 0.6342121195442787,
+      "grad_norm": 0.7147908210754395,
+      "learning_rate": 7.74220419950239e-06,
+      "loss": 0.7705,
+      "step": 11523
+    },
+    {
+      "epoch": 0.6342671583466344,
+      "grad_norm": 0.7191137671470642,
+      "learning_rate": 7.7418417290251e-06,
+      "loss": 0.789,
+      "step": 11524
+    },
+    {
+      "epoch": 0.6343221971489901,
+      "grad_norm": 0.7288943529129028,
+      "learning_rate": 7.741479237940862e-06,
+      "loss": 0.8204,
+      "step": 11525
+    },
+    {
+      "epoch": 0.6343772359513457,
+      "grad_norm": 0.714821994304657,
+      "learning_rate": 7.741116726252398e-06,
+      "loss": 0.8252,
+      "step": 11526
+    },
+    {
+      "epoch": 0.6344322747537013,
+      "grad_norm": 0.6869103312492371,
+      "learning_rate": 7.740754193962435e-06,
+      "loss": 0.8136,
+      "step": 11527
+    },
+    {
+      "epoch": 0.634487313556057,
+      "grad_norm": 0.6629248857498169,
+      "learning_rate": 7.740391641073698e-06,
+      "loss": 0.7049,
+      "step": 11528
+    },
+    {
+      "epoch": 0.6345423523584127,
+      "grad_norm": 0.7078685164451599,
+      "learning_rate": 7.74002906758891e-06,
+      "loss": 0.7345,
+      "step": 11529
+    },
+    {
+      "epoch": 0.6345973911607683,
+      "grad_norm": 0.7748367190361023,
+      "learning_rate": 7.739666473510798e-06,
+      "loss": 0.7085,
+      "step": 11530
+    },
+    {
+      "epoch": 0.634652429963124,
+      "grad_norm": 0.6661930084228516,
+      "learning_rate": 7.739303858842086e-06,
+      "loss": 0.7795,
+      "step": 11531
+    },
+    {
+      "epoch": 0.6347074687654797,
+      "grad_norm": 0.6847965121269226,
+      "learning_rate": 7.738941223585499e-06,
+      "loss": 0.797,
+      "step": 11532
+    },
+    {
+      "epoch": 0.6347625075678354,
+      "grad_norm": 0.695184051990509,
+      "learning_rate": 7.738578567743762e-06,
+      "loss": 0.8184,
+      "step": 11533
+    },
+    {
+      "epoch": 0.6348175463701909,
+      "grad_norm": 0.6620088815689087,
+      "learning_rate": 7.738215891319603e-06,
+      "loss": 0.721,
+      "step": 11534
+    },
+    {
+      "epoch": 0.6348725851725466,
+      "grad_norm": 0.6802023649215698,
+      "learning_rate": 7.737853194315745e-06,
+      "loss": 0.9207,
+      "step": 11535
+    },
+    {
+      "epoch": 0.6349276239749023,
+      "grad_norm": 1.0193618535995483,
+      "learning_rate": 7.737490476734916e-06,
+      "loss": 0.8495,
+      "step": 11536
+    },
+    {
+      "epoch": 0.634982662777258,
+      "grad_norm": 0.6578189730644226,
+      "learning_rate": 7.737127738579841e-06,
+      "loss": 0.7455,
+      "step": 11537
+    },
+    {
+      "epoch": 0.6350377015796136,
+      "grad_norm": 0.70018470287323,
+      "learning_rate": 7.736764979853248e-06,
+      "loss": 0.7414,
+      "step": 11538
+    },
+    {
+      "epoch": 0.6350927403819693,
+      "grad_norm": 0.8136304616928101,
+      "learning_rate": 7.736402200557862e-06,
+      "loss": 0.7327,
+      "step": 11539
+    },
+    {
+      "epoch": 0.635147779184325,
+      "grad_norm": 0.7805309295654297,
+      "learning_rate": 7.736039400696408e-06,
+      "loss": 0.7659,
+      "step": 11540
+    },
+    {
+      "epoch": 0.6352028179866807,
+      "grad_norm": 0.675215482711792,
+      "learning_rate": 7.735676580271615e-06,
+      "loss": 0.7532,
+      "step": 11541
+    },
+    {
+      "epoch": 0.6352578567890362,
+      "grad_norm": 0.6873239874839783,
+      "learning_rate": 7.735313739286208e-06,
+      "loss": 0.8123,
+      "step": 11542
+    },
+    {
+      "epoch": 0.6353128955913919,
+      "grad_norm": 0.6624773144721985,
+      "learning_rate": 7.734950877742917e-06,
+      "loss": 0.7642,
+      "step": 11543
+    },
+    {
+      "epoch": 0.6353679343937476,
+      "grad_norm": 0.8047438859939575,
+      "learning_rate": 7.734587995644468e-06,
+      "loss": 0.7452,
+      "step": 11544
+    },
+    {
+      "epoch": 0.6354229731961033,
+      "grad_norm": 0.7449815273284912,
+      "learning_rate": 7.734225092993585e-06,
+      "loss": 0.7756,
+      "step": 11545
+    },
+    {
+      "epoch": 0.6354780119984589,
+      "grad_norm": 0.693081259727478,
+      "learning_rate": 7.733862169792999e-06,
+      "loss": 0.7029,
+      "step": 11546
+    },
+    {
+      "epoch": 0.6355330508008146,
+      "grad_norm": 0.6593700051307678,
+      "learning_rate": 7.733499226045437e-06,
+      "loss": 0.6009,
+      "step": 11547
+    },
+    {
+      "epoch": 0.6355880896031703,
+      "grad_norm": 0.7402041554450989,
+      "learning_rate": 7.733136261753627e-06,
+      "loss": 0.6921,
+      "step": 11548
+    },
+    {
+      "epoch": 0.635643128405526,
+      "grad_norm": 0.7686228156089783,
+      "learning_rate": 7.732773276920294e-06,
+      "loss": 0.855,
+      "step": 11549
+    },
+    {
+      "epoch": 0.6356981672078815,
+      "grad_norm": 0.6776669025421143,
+      "learning_rate": 7.732410271548171e-06,
+      "loss": 0.7146,
+      "step": 11550
+    },
+    {
+      "epoch": 0.6357532060102372,
+      "grad_norm": 0.6055952906608582,
+      "learning_rate": 7.732047245639983e-06,
+      "loss": 0.6926,
+      "step": 11551
+    },
+    {
+      "epoch": 0.6358082448125929,
+      "grad_norm": 0.7452635765075684,
+      "learning_rate": 7.731684199198461e-06,
+      "loss": 0.7766,
+      "step": 11552
+    },
+    {
+      "epoch": 0.6358632836149486,
+      "grad_norm": 0.7482720017433167,
+      "learning_rate": 7.73132113222633e-06,
+      "loss": 0.7725,
+      "step": 11553
+    },
+    {
+      "epoch": 0.6359183224173042,
+      "grad_norm": 0.6534025073051453,
+      "learning_rate": 7.73095804472632e-06,
+      "loss": 0.7902,
+      "step": 11554
+    },
+    {
+      "epoch": 0.6359733612196599,
+      "grad_norm": 0.7364560961723328,
+      "learning_rate": 7.730594936701162e-06,
+      "loss": 0.7998,
+      "step": 11555
+    },
+    {
+      "epoch": 0.6360284000220155,
+      "grad_norm": 0.6881458163261414,
+      "learning_rate": 7.730231808153582e-06,
+      "loss": 0.7586,
+      "step": 11556
+    },
+    {
+      "epoch": 0.6360834388243711,
+      "grad_norm": 0.6574262976646423,
+      "learning_rate": 7.72986865908631e-06,
+      "loss": 0.6999,
+      "step": 11557
+    },
+    {
+      "epoch": 0.6361384776267268,
+      "grad_norm": 0.6976385712623596,
+      "learning_rate": 7.729505489502078e-06,
+      "loss": 0.7387,
+      "step": 11558
+    },
+    {
+      "epoch": 0.6361935164290825,
+      "grad_norm": 0.6482532620429993,
+      "learning_rate": 7.729142299403613e-06,
+      "loss": 0.7715,
+      "step": 11559
+    },
+    {
+      "epoch": 0.6362485552314382,
+      "grad_norm": 0.7140287160873413,
+      "learning_rate": 7.728779088793643e-06,
+      "loss": 0.8562,
+      "step": 11560
+    },
+    {
+      "epoch": 0.6363035940337938,
+      "grad_norm": 0.6579470634460449,
+      "learning_rate": 7.728415857674901e-06,
+      "loss": 0.727,
+      "step": 11561
+    },
+    {
+      "epoch": 0.6363586328361495,
+      "grad_norm": 0.8670933246612549,
+      "learning_rate": 7.728052606050116e-06,
+      "loss": 0.7459,
+      "step": 11562
+    },
+    {
+      "epoch": 0.6364136716385052,
+      "grad_norm": 0.7995489835739136,
+      "learning_rate": 7.72768933392202e-06,
+      "loss": 0.8228,
+      "step": 11563
+    },
+    {
+      "epoch": 0.6364687104408608,
+      "grad_norm": 0.6467362642288208,
+      "learning_rate": 7.727326041293336e-06,
+      "loss": 0.7545,
+      "step": 11564
+    },
+    {
+      "epoch": 0.6365237492432164,
+      "grad_norm": 0.6646577715873718,
+      "learning_rate": 7.726962728166803e-06,
+      "loss": 0.7824,
+      "step": 11565
+    },
+    {
+      "epoch": 0.6365787880455721,
+      "grad_norm": 0.6576912999153137,
+      "learning_rate": 7.726599394545149e-06,
+      "loss": 0.7324,
+      "step": 11566
+    },
+    {
+      "epoch": 0.6366338268479278,
+      "grad_norm": 0.7514963150024414,
+      "learning_rate": 7.726236040431101e-06,
+      "loss": 0.7712,
+      "step": 11567
+    },
+    {
+      "epoch": 0.6366888656502835,
+      "grad_norm": 0.7313328981399536,
+      "learning_rate": 7.725872665827394e-06,
+      "loss": 0.7361,
+      "step": 11568
+    },
+    {
+      "epoch": 0.6367439044526391,
+      "grad_norm": 0.7109994292259216,
+      "learning_rate": 7.725509270736759e-06,
+      "loss": 0.812,
+      "step": 11569
+    },
+    {
+      "epoch": 0.6367989432549948,
+      "grad_norm": 1.128675103187561,
+      "learning_rate": 7.725145855161924e-06,
+      "loss": 0.726,
+      "step": 11570
+    },
+    {
+      "epoch": 0.6368539820573504,
+      "grad_norm": 0.7357437014579773,
+      "learning_rate": 7.724782419105622e-06,
+      "loss": 0.7958,
+      "step": 11571
+    },
+    {
+      "epoch": 0.6369090208597061,
+      "grad_norm": 0.6874725222587585,
+      "learning_rate": 7.724418962570587e-06,
+      "loss": 0.751,
+      "step": 11572
+    },
+    {
+      "epoch": 0.6369640596620617,
+      "grad_norm": 0.7175989747047424,
+      "learning_rate": 7.724055485559545e-06,
+      "loss": 0.7191,
+      "step": 11573
+    },
+    {
+      "epoch": 0.6370190984644174,
+      "grad_norm": 0.6424688100814819,
+      "learning_rate": 7.723691988075235e-06,
+      "loss": 0.608,
+      "step": 11574
+    },
+    {
+      "epoch": 0.6370741372667731,
+      "grad_norm": 0.6845381855964661,
+      "learning_rate": 7.723328470120383e-06,
+      "loss": 0.7465,
+      "step": 11575
+    },
+    {
+      "epoch": 0.6371291760691288,
+      "grad_norm": 0.7955030202865601,
+      "learning_rate": 7.722964931697723e-06,
+      "loss": 0.745,
+      "step": 11576
+    },
+    {
+      "epoch": 0.6371842148714844,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 7.722601372809989e-06,
+      "loss": 0.7764,
+      "step": 11577
+    },
+    {
+      "epoch": 0.63723925367384,
+      "grad_norm": 0.7505692839622498,
+      "learning_rate": 7.722237793459909e-06,
+      "loss": 0.8324,
+      "step": 11578
+    },
+    {
+      "epoch": 0.6372942924761957,
+      "grad_norm": 0.6852842569351196,
+      "learning_rate": 7.721874193650221e-06,
+      "loss": 0.7599,
+      "step": 11579
+    },
+    {
+      "epoch": 0.6373493312785514,
+      "grad_norm": 0.698210597038269,
+      "learning_rate": 7.721510573383654e-06,
+      "loss": 0.843,
+      "step": 11580
+    },
+    {
+      "epoch": 0.637404370080907,
+      "grad_norm": 0.8344444632530212,
+      "learning_rate": 7.721146932662942e-06,
+      "loss": 0.8602,
+      "step": 11581
+    },
+    {
+      "epoch": 0.6374594088832627,
+      "grad_norm": 0.6385721564292908,
+      "learning_rate": 7.72078327149082e-06,
+      "loss": 0.7449,
+      "step": 11582
+    },
+    {
+      "epoch": 0.6375144476856184,
+      "grad_norm": 0.6474401354789734,
+      "learning_rate": 7.720419589870016e-06,
+      "loss": 0.6328,
+      "step": 11583
+    },
+    {
+      "epoch": 0.6375694864879741,
+      "grad_norm": 0.6554263234138489,
+      "learning_rate": 7.720055887803268e-06,
+      "loss": 0.6672,
+      "step": 11584
+    },
+    {
+      "epoch": 0.6376245252903296,
+      "grad_norm": 0.6551910638809204,
+      "learning_rate": 7.719692165293309e-06,
+      "loss": 0.8024,
+      "step": 11585
+    },
+    {
+      "epoch": 0.6376795640926853,
+      "grad_norm": 0.693418025970459,
+      "learning_rate": 7.719328422342871e-06,
+      "loss": 0.726,
+      "step": 11586
+    },
+    {
+      "epoch": 0.637734602895041,
+      "grad_norm": 0.8642090559005737,
+      "learning_rate": 7.718964658954689e-06,
+      "loss": 0.8274,
+      "step": 11587
+    },
+    {
+      "epoch": 0.6377896416973967,
+      "grad_norm": 0.8255778551101685,
+      "learning_rate": 7.718600875131494e-06,
+      "loss": 0.7259,
+      "step": 11588
+    },
+    {
+      "epoch": 0.6378446804997523,
+      "grad_norm": 0.7492913007736206,
+      "learning_rate": 7.718237070876025e-06,
+      "loss": 0.7093,
+      "step": 11589
+    },
+    {
+      "epoch": 0.637899719302108,
+      "grad_norm": 0.7154868245124817,
+      "learning_rate": 7.717873246191013e-06,
+      "loss": 0.7909,
+      "step": 11590
+    },
+    {
+      "epoch": 0.6379547581044637,
+      "grad_norm": 0.7751424312591553,
+      "learning_rate": 7.717509401079194e-06,
+      "loss": 0.8528,
+      "step": 11591
+    },
+    {
+      "epoch": 0.6380097969068194,
+      "grad_norm": 0.68199223279953,
+      "learning_rate": 7.7171455355433e-06,
+      "loss": 0.7077,
+      "step": 11592
+    },
+    {
+      "epoch": 0.6380648357091749,
+      "grad_norm": 0.7340414524078369,
+      "learning_rate": 7.716781649586069e-06,
+      "loss": 0.693,
+      "step": 11593
+    },
+    {
+      "epoch": 0.6381198745115306,
+      "grad_norm": 0.6278988122940063,
+      "learning_rate": 7.716417743210234e-06,
+      "loss": 0.7049,
+      "step": 11594
+    },
+    {
+      "epoch": 0.6381749133138863,
+      "grad_norm": 0.9113193154335022,
+      "learning_rate": 7.716053816418532e-06,
+      "loss": 0.7757,
+      "step": 11595
+    },
+    {
+      "epoch": 0.638229952116242,
+      "grad_norm": 0.7059371471405029,
+      "learning_rate": 7.715689869213694e-06,
+      "loss": 0.7805,
+      "step": 11596
+    },
+    {
+      "epoch": 0.6382849909185976,
+      "grad_norm": 0.7508488297462463,
+      "learning_rate": 7.71532590159846e-06,
+      "loss": 0.7394,
+      "step": 11597
+    },
+    {
+      "epoch": 0.6383400297209533,
+      "grad_norm": 0.8222774863243103,
+      "learning_rate": 7.71496191357556e-06,
+      "loss": 0.7675,
+      "step": 11598
+    },
+    {
+      "epoch": 0.638395068523309,
+      "grad_norm": 0.7295246124267578,
+      "learning_rate": 7.714597905147736e-06,
+      "loss": 0.7766,
+      "step": 11599
+    },
+    {
+      "epoch": 0.6384501073256645,
+      "grad_norm": 0.7482065558433533,
+      "learning_rate": 7.71423387631772e-06,
+      "loss": 0.7334,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6385051461280202,
+      "grad_norm": 0.7654659748077393,
+      "learning_rate": 7.71386982708825e-06,
+      "loss": 0.8097,
+      "step": 11601
+    },
+    {
+      "epoch": 0.6385601849303759,
+      "grad_norm": 0.9125531911849976,
+      "learning_rate": 7.71350575746206e-06,
+      "loss": 0.7776,
+      "step": 11602
+    },
+    {
+      "epoch": 0.6386152237327316,
+      "grad_norm": 0.8063878417015076,
+      "learning_rate": 7.713141667441886e-06,
+      "loss": 0.7899,
+      "step": 11603
+    },
+    {
+      "epoch": 0.6386702625350872,
+      "grad_norm": 0.7315171360969543,
+      "learning_rate": 7.712777557030466e-06,
+      "loss": 0.7884,
+      "step": 11604
+    },
+    {
+      "epoch": 0.6387253013374429,
+      "grad_norm": 0.7306345105171204,
+      "learning_rate": 7.712413426230536e-06,
+      "loss": 0.8646,
+      "step": 11605
+    },
+    {
+      "epoch": 0.6387803401397986,
+      "grad_norm": 0.8300313353538513,
+      "learning_rate": 7.712049275044833e-06,
+      "loss": 0.8131,
+      "step": 11606
+    },
+    {
+      "epoch": 0.6388353789421543,
+      "grad_norm": 0.7513623237609863,
+      "learning_rate": 7.711685103476093e-06,
+      "loss": 0.8115,
+      "step": 11607
+    },
+    {
+      "epoch": 0.6388904177445098,
+      "grad_norm": 0.7126060128211975,
+      "learning_rate": 7.711320911527054e-06,
+      "loss": 0.8198,
+      "step": 11608
+    },
+    {
+      "epoch": 0.6389454565468655,
+      "grad_norm": 0.7017398476600647,
+      "learning_rate": 7.710956699200454e-06,
+      "loss": 0.8088,
+      "step": 11609
+    },
+    {
+      "epoch": 0.6390004953492212,
+      "grad_norm": 0.7345026135444641,
+      "learning_rate": 7.710592466499027e-06,
+      "loss": 0.8228,
+      "step": 11610
+    },
+    {
+      "epoch": 0.6390555341515769,
+      "grad_norm": 0.6903058886528015,
+      "learning_rate": 7.710228213425514e-06,
+      "loss": 0.7058,
+      "step": 11611
+    },
+    {
+      "epoch": 0.6391105729539325,
+      "grad_norm": 0.6838604211807251,
+      "learning_rate": 7.70986393998265e-06,
+      "loss": 0.7091,
+      "step": 11612
+    },
+    {
+      "epoch": 0.6391656117562882,
+      "grad_norm": 0.7067943811416626,
+      "learning_rate": 7.709499646173177e-06,
+      "loss": 0.7631,
+      "step": 11613
+    },
+    {
+      "epoch": 0.6392206505586439,
+      "grad_norm": 0.7577057480812073,
+      "learning_rate": 7.709135331999827e-06,
+      "loss": 0.7545,
+      "step": 11614
+    },
+    {
+      "epoch": 0.6392756893609995,
+      "grad_norm": 0.6425572633743286,
+      "learning_rate": 7.70877099746534e-06,
+      "loss": 0.7188,
+      "step": 11615
+    },
+    {
+      "epoch": 0.6393307281633551,
+      "grad_norm": 0.7257497310638428,
+      "learning_rate": 7.708406642572459e-06,
+      "loss": 0.7514,
+      "step": 11616
+    },
+    {
+      "epoch": 0.6393857669657108,
+      "grad_norm": 0.8214251399040222,
+      "learning_rate": 7.708042267323916e-06,
+      "loss": 0.7824,
+      "step": 11617
+    },
+    {
+      "epoch": 0.6394408057680665,
+      "grad_norm": 0.7879108786582947,
+      "learning_rate": 7.707677871722453e-06,
+      "loss": 0.6122,
+      "step": 11618
+    },
+    {
+      "epoch": 0.6394958445704222,
+      "grad_norm": 0.6656795740127563,
+      "learning_rate": 7.707313455770808e-06,
+      "loss": 0.754,
+      "step": 11619
+    },
+    {
+      "epoch": 0.6395508833727778,
+      "grad_norm": 0.7196451425552368,
+      "learning_rate": 7.70694901947172e-06,
+      "loss": 0.7662,
+      "step": 11620
+    },
+    {
+      "epoch": 0.6396059221751335,
+      "grad_norm": 0.8213779926300049,
+      "learning_rate": 7.706584562827928e-06,
+      "loss": 0.8732,
+      "step": 11621
+    },
+    {
+      "epoch": 0.6396609609774891,
+      "grad_norm": 0.7114893794059753,
+      "learning_rate": 7.70622008584217e-06,
+      "loss": 0.8493,
+      "step": 11622
+    },
+    {
+      "epoch": 0.6397159997798448,
+      "grad_norm": 0.7009783983230591,
+      "learning_rate": 7.705855588517188e-06,
+      "loss": 0.738,
+      "step": 11623
+    },
+    {
+      "epoch": 0.6397710385822004,
+      "grad_norm": 0.7576995491981506,
+      "learning_rate": 7.705491070855717e-06,
+      "loss": 0.8839,
+      "step": 11624
+    },
+    {
+      "epoch": 0.6398260773845561,
+      "grad_norm": 0.705784022808075,
+      "learning_rate": 7.7051265328605e-06,
+      "loss": 0.7246,
+      "step": 11625
+    },
+    {
+      "epoch": 0.6398811161869118,
+      "grad_norm": 0.6696903109550476,
+      "learning_rate": 7.704761974534277e-06,
+      "loss": 0.7418,
+      "step": 11626
+    },
+    {
+      "epoch": 0.6399361549892675,
+      "grad_norm": 0.8617024421691895,
+      "learning_rate": 7.704397395879786e-06,
+      "loss": 0.8109,
+      "step": 11627
+    },
+    {
+      "epoch": 0.6399911937916231,
+      "grad_norm": 0.6819054484367371,
+      "learning_rate": 7.70403279689977e-06,
+      "loss": 0.6438,
+      "step": 11628
+    },
+    {
+      "epoch": 0.6400462325939787,
+      "grad_norm": 0.6145044565200806,
+      "learning_rate": 7.703668177596966e-06,
+      "loss": 0.6712,
+      "step": 11629
+    },
+    {
+      "epoch": 0.6401012713963344,
+      "grad_norm": 0.6946390271186829,
+      "learning_rate": 7.703303537974116e-06,
+      "loss": 0.8099,
+      "step": 11630
+    },
+    {
+      "epoch": 0.6401563101986901,
+      "grad_norm": 0.6791605949401855,
+      "learning_rate": 7.702938878033961e-06,
+      "loss": 0.7494,
+      "step": 11631
+    },
+    {
+      "epoch": 0.6402113490010457,
+      "grad_norm": 0.6718626618385315,
+      "learning_rate": 7.70257419777924e-06,
+      "loss": 0.7471,
+      "step": 11632
+    },
+    {
+      "epoch": 0.6402663878034014,
+      "grad_norm": 0.8051798343658447,
+      "learning_rate": 7.702209497212694e-06,
+      "loss": 0.8569,
+      "step": 11633
+    },
+    {
+      "epoch": 0.6403214266057571,
+      "grad_norm": 0.6602774858474731,
+      "learning_rate": 7.701844776337067e-06,
+      "loss": 0.7396,
+      "step": 11634
+    },
+    {
+      "epoch": 0.6403764654081128,
+      "grad_norm": 0.672363817691803,
+      "learning_rate": 7.701480035155096e-06,
+      "loss": 0.7584,
+      "step": 11635
+    },
+    {
+      "epoch": 0.6404315042104683,
+      "grad_norm": 0.7363641262054443,
+      "learning_rate": 7.701115273669524e-06,
+      "loss": 0.8149,
+      "step": 11636
+    },
+    {
+      "epoch": 0.640486543012824,
+      "grad_norm": 0.7238422632217407,
+      "learning_rate": 7.700750491883094e-06,
+      "loss": 0.7598,
+      "step": 11637
+    },
+    {
+      "epoch": 0.6405415818151797,
+      "grad_norm": 1.3627614974975586,
+      "learning_rate": 7.700385689798544e-06,
+      "loss": 0.8303,
+      "step": 11638
+    },
+    {
+      "epoch": 0.6405966206175354,
+      "grad_norm": 0.6339633464813232,
+      "learning_rate": 7.70002086741862e-06,
+      "loss": 0.7308,
+      "step": 11639
+    },
+    {
+      "epoch": 0.640651659419891,
+      "grad_norm": 0.6821589469909668,
+      "learning_rate": 7.699656024746062e-06,
+      "loss": 0.6728,
+      "step": 11640
+    },
+    {
+      "epoch": 0.6407066982222467,
+      "grad_norm": 0.8514766097068787,
+      "learning_rate": 7.699291161783611e-06,
+      "loss": 0.8693,
+      "step": 11641
+    },
+    {
+      "epoch": 0.6407617370246024,
+      "grad_norm": 0.649075984954834,
+      "learning_rate": 7.698926278534011e-06,
+      "loss": 0.7482,
+      "step": 11642
+    },
+    {
+      "epoch": 0.640816775826958,
+      "grad_norm": 0.6507017016410828,
+      "learning_rate": 7.698561375000001e-06,
+      "loss": 0.7841,
+      "step": 11643
+    },
+    {
+      "epoch": 0.6408718146293136,
+      "grad_norm": 0.6736069321632385,
+      "learning_rate": 7.69819645118433e-06,
+      "loss": 0.74,
+      "step": 11644
+    },
+    {
+      "epoch": 0.6409268534316693,
+      "grad_norm": 0.6727941632270813,
+      "learning_rate": 7.697831507089734e-06,
+      "loss": 0.806,
+      "step": 11645
+    },
+    {
+      "epoch": 0.640981892234025,
+      "grad_norm": 0.7089083194732666,
+      "learning_rate": 7.697466542718959e-06,
+      "loss": 0.8091,
+      "step": 11646
+    },
+    {
+      "epoch": 0.6410369310363806,
+      "grad_norm": 0.6355387568473816,
+      "learning_rate": 7.69710155807475e-06,
+      "loss": 0.7033,
+      "step": 11647
+    },
+    {
+      "epoch": 0.6410919698387363,
+      "grad_norm": 0.6327098608016968,
+      "learning_rate": 7.696736553159846e-06,
+      "loss": 0.7664,
+      "step": 11648
+    },
+    {
+      "epoch": 0.641147008641092,
+      "grad_norm": 0.6971945762634277,
+      "learning_rate": 7.69637152797699e-06,
+      "loss": 0.7441,
+      "step": 11649
+    },
+    {
+      "epoch": 0.6412020474434477,
+      "grad_norm": 0.7420539855957031,
+      "learning_rate": 7.696006482528929e-06,
+      "loss": 0.7909,
+      "step": 11650
+    },
+    {
+      "epoch": 0.6412570862458032,
+      "grad_norm": 0.6877853274345398,
+      "learning_rate": 7.695641416818405e-06,
+      "loss": 0.7624,
+      "step": 11651
+    },
+    {
+      "epoch": 0.6413121250481589,
+      "grad_norm": 0.7337075471878052,
+      "learning_rate": 7.695276330848162e-06,
+      "loss": 0.7829,
+      "step": 11652
+    },
+    {
+      "epoch": 0.6413671638505146,
+      "grad_norm": 0.6423582434654236,
+      "learning_rate": 7.694911224620944e-06,
+      "loss": 0.6686,
+      "step": 11653
+    },
+    {
+      "epoch": 0.6414222026528703,
+      "grad_norm": 0.7826602458953857,
+      "learning_rate": 7.694546098139492e-06,
+      "loss": 0.774,
+      "step": 11654
+    },
+    {
+      "epoch": 0.6414772414552259,
+      "grad_norm": 0.7678147554397583,
+      "learning_rate": 7.694180951406556e-06,
+      "loss": 0.8067,
+      "step": 11655
+    },
+    {
+      "epoch": 0.6415322802575816,
+      "grad_norm": 0.6400566101074219,
+      "learning_rate": 7.693815784424875e-06,
+      "loss": 0.7796,
+      "step": 11656
+    },
+    {
+      "epoch": 0.6415873190599373,
+      "grad_norm": 0.6606197357177734,
+      "learning_rate": 7.693450597197196e-06,
+      "loss": 0.7381,
+      "step": 11657
+    },
+    {
+      "epoch": 0.641642357862293,
+      "grad_norm": 0.7953683137893677,
+      "learning_rate": 7.693085389726262e-06,
+      "loss": 0.8867,
+      "step": 11658
+    },
+    {
+      "epoch": 0.6416973966646485,
+      "grad_norm": 0.6763843894004822,
+      "learning_rate": 7.692720162014822e-06,
+      "loss": 0.7579,
+      "step": 11659
+    },
+    {
+      "epoch": 0.6417524354670042,
+      "grad_norm": 0.6456292867660522,
+      "learning_rate": 7.692354914065617e-06,
+      "loss": 0.7814,
+      "step": 11660
+    },
+    {
+      "epoch": 0.6418074742693599,
+      "grad_norm": 0.702803373336792,
+      "learning_rate": 7.691989645881393e-06,
+      "loss": 0.7393,
+      "step": 11661
+    },
+    {
+      "epoch": 0.6418625130717156,
+      "grad_norm": 0.8328298926353455,
+      "learning_rate": 7.691624357464895e-06,
+      "loss": 0.6587,
+      "step": 11662
+    },
+    {
+      "epoch": 0.6419175518740712,
+      "grad_norm": 0.8409613966941833,
+      "learning_rate": 7.691259048818871e-06,
+      "loss": 0.8075,
+      "step": 11663
+    },
+    {
+      "epoch": 0.6419725906764269,
+      "grad_norm": 0.6969256401062012,
+      "learning_rate": 7.690893719946062e-06,
+      "loss": 0.8061,
+      "step": 11664
+    },
+    {
+      "epoch": 0.6420276294787826,
+      "grad_norm": 0.7689732313156128,
+      "learning_rate": 7.690528370849217e-06,
+      "loss": 0.7709,
+      "step": 11665
+    },
+    {
+      "epoch": 0.6420826682811382,
+      "grad_norm": 0.8239523768424988,
+      "learning_rate": 7.69016300153108e-06,
+      "loss": 0.7421,
+      "step": 11666
+    },
+    {
+      "epoch": 0.6421377070834938,
+      "grad_norm": 0.7199227809906006,
+      "learning_rate": 7.689797611994398e-06,
+      "loss": 0.7877,
+      "step": 11667
+    },
+    {
+      "epoch": 0.6421927458858495,
+      "grad_norm": 0.8315985798835754,
+      "learning_rate": 7.689432202241919e-06,
+      "loss": 0.8458,
+      "step": 11668
+    },
+    {
+      "epoch": 0.6422477846882052,
+      "grad_norm": 0.7213512063026428,
+      "learning_rate": 7.689066772276385e-06,
+      "loss": 0.7199,
+      "step": 11669
+    },
+    {
+      "epoch": 0.6423028234905609,
+      "grad_norm": 0.6023604273796082,
+      "learning_rate": 7.688701322100547e-06,
+      "loss": 0.6485,
+      "step": 11670
+    },
+    {
+      "epoch": 0.6423578622929165,
+      "grad_norm": 0.8171319365501404,
+      "learning_rate": 7.688335851717148e-06,
+      "loss": 0.7561,
+      "step": 11671
+    },
+    {
+      "epoch": 0.6424129010952722,
+      "grad_norm": 0.6545816659927368,
+      "learning_rate": 7.687970361128937e-06,
+      "loss": 0.6796,
+      "step": 11672
+    },
+    {
+      "epoch": 0.6424679398976278,
+      "grad_norm": 0.8093686103820801,
+      "learning_rate": 7.687604850338661e-06,
+      "loss": 0.8538,
+      "step": 11673
+    },
+    {
+      "epoch": 0.6425229786999835,
+      "grad_norm": 0.6438135504722595,
+      "learning_rate": 7.687239319349066e-06,
+      "loss": 0.7046,
+      "step": 11674
+    },
+    {
+      "epoch": 0.6425780175023391,
+      "grad_norm": 0.685100257396698,
+      "learning_rate": 7.6868737681629e-06,
+      "loss": 0.7568,
+      "step": 11675
+    },
+    {
+      "epoch": 0.6426330563046948,
+      "grad_norm": 0.6850112676620483,
+      "learning_rate": 7.68650819678291e-06,
+      "loss": 0.7082,
+      "step": 11676
+    },
+    {
+      "epoch": 0.6426880951070505,
+      "grad_norm": 0.7524490356445312,
+      "learning_rate": 7.686142605211843e-06,
+      "loss": 0.7285,
+      "step": 11677
+    },
+    {
+      "epoch": 0.6427431339094062,
+      "grad_norm": 0.7706617116928101,
+      "learning_rate": 7.685776993452446e-06,
+      "loss": 0.7934,
+      "step": 11678
+    },
+    {
+      "epoch": 0.6427981727117618,
+      "grad_norm": 0.6612235307693481,
+      "learning_rate": 7.68541136150747e-06,
+      "loss": 0.6538,
+      "step": 11679
+    },
+    {
+      "epoch": 0.6428532115141175,
+      "grad_norm": 0.6380587816238403,
+      "learning_rate": 7.68504570937966e-06,
+      "loss": 0.7,
+      "step": 11680
+    },
+    {
+      "epoch": 0.6429082503164731,
+      "grad_norm": 0.6563882231712341,
+      "learning_rate": 7.684680037071765e-06,
+      "loss": 0.6912,
+      "step": 11681
+    },
+    {
+      "epoch": 0.6429632891188288,
+      "grad_norm": 0.6579793095588684,
+      "learning_rate": 7.684314344586534e-06,
+      "loss": 0.7263,
+      "step": 11682
+    },
+    {
+      "epoch": 0.6430183279211844,
+      "grad_norm": 0.7029374837875366,
+      "learning_rate": 7.683948631926713e-06,
+      "loss": 0.7151,
+      "step": 11683
+    },
+    {
+      "epoch": 0.6430733667235401,
+      "grad_norm": 0.6683217883110046,
+      "learning_rate": 7.683582899095056e-06,
+      "loss": 0.7643,
+      "step": 11684
+    },
+    {
+      "epoch": 0.6431284055258958,
+      "grad_norm": 1.0482646226882935,
+      "learning_rate": 7.683217146094308e-06,
+      "loss": 0.8889,
+      "step": 11685
+    },
+    {
+      "epoch": 0.6431834443282514,
+      "grad_norm": 0.7101102471351624,
+      "learning_rate": 7.682851372927216e-06,
+      "loss": 0.7762,
+      "step": 11686
+    },
+    {
+      "epoch": 0.643238483130607,
+      "grad_norm": 0.674961268901825,
+      "learning_rate": 7.682485579596533e-06,
+      "loss": 0.736,
+      "step": 11687
+    },
+    {
+      "epoch": 0.6432935219329627,
+      "grad_norm": 0.7071837782859802,
+      "learning_rate": 7.682119766105005e-06,
+      "loss": 0.7231,
+      "step": 11688
+    },
+    {
+      "epoch": 0.6433485607353184,
+      "grad_norm": 0.6982744932174683,
+      "learning_rate": 7.681753932455383e-06,
+      "loss": 0.7498,
+      "step": 11689
+    },
+    {
+      "epoch": 0.643403599537674,
+      "grad_norm": 0.6927201747894287,
+      "learning_rate": 7.681388078650415e-06,
+      "loss": 0.803,
+      "step": 11690
+    },
+    {
+      "epoch": 0.6434586383400297,
+      "grad_norm": 0.7299236059188843,
+      "learning_rate": 7.681022204692854e-06,
+      "loss": 0.7386,
+      "step": 11691
+    },
+    {
+      "epoch": 0.6435136771423854,
+      "grad_norm": 0.8809047937393188,
+      "learning_rate": 7.680656310585449e-06,
+      "loss": 0.741,
+      "step": 11692
+    },
+    {
+      "epoch": 0.6435687159447411,
+      "grad_norm": 0.862843930721283,
+      "learning_rate": 7.680290396330947e-06,
+      "loss": 0.8357,
+      "step": 11693
+    },
+    {
+      "epoch": 0.6436237547470967,
+      "grad_norm": 0.7436664700508118,
+      "learning_rate": 7.679924461932098e-06,
+      "loss": 0.8352,
+      "step": 11694
+    },
+    {
+      "epoch": 0.6436787935494523,
+      "grad_norm": 0.6582232713699341,
+      "learning_rate": 7.679558507391657e-06,
+      "loss": 0.7107,
+      "step": 11695
+    },
+    {
+      "epoch": 0.643733832351808,
+      "grad_norm": 0.6798850297927856,
+      "learning_rate": 7.67919253271237e-06,
+      "loss": 0.6968,
+      "step": 11696
+    },
+    {
+      "epoch": 0.6437888711541637,
+      "grad_norm": 0.7747187614440918,
+      "learning_rate": 7.67882653789699e-06,
+      "loss": 0.7611,
+      "step": 11697
+    },
+    {
+      "epoch": 0.6438439099565193,
+      "grad_norm": 0.7097567915916443,
+      "learning_rate": 7.678460522948267e-06,
+      "loss": 0.7275,
+      "step": 11698
+    },
+    {
+      "epoch": 0.643898948758875,
+      "grad_norm": 0.6958394050598145,
+      "learning_rate": 7.678094487868952e-06,
+      "loss": 0.7441,
+      "step": 11699
+    },
+    {
+      "epoch": 0.6439539875612307,
+      "grad_norm": 0.9129040837287903,
+      "learning_rate": 7.677728432661794e-06,
+      "loss": 0.7693,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6440090263635864,
+      "grad_norm": 1.1396137475967407,
+      "learning_rate": 7.677362357329548e-06,
+      "loss": 0.7479,
+      "step": 11701
+    },
+    {
+      "epoch": 0.644064065165942,
+      "grad_norm": 0.8163042664527893,
+      "learning_rate": 7.67699626187496e-06,
+      "loss": 0.835,
+      "step": 11702
+    },
+    {
+      "epoch": 0.6441191039682976,
+      "grad_norm": 0.9869117736816406,
+      "learning_rate": 7.676630146300787e-06,
+      "loss": 0.769,
+      "step": 11703
+    },
+    {
+      "epoch": 0.6441741427706533,
+      "grad_norm": 0.7439526915550232,
+      "learning_rate": 7.676264010609777e-06,
+      "loss": 0.8239,
+      "step": 11704
+    },
+    {
+      "epoch": 0.644229181573009,
+      "grad_norm": 0.6943735480308533,
+      "learning_rate": 7.675897854804685e-06,
+      "loss": 0.7702,
+      "step": 11705
+    },
+    {
+      "epoch": 0.6442842203753646,
+      "grad_norm": 0.7384238243103027,
+      "learning_rate": 7.67553167888826e-06,
+      "loss": 0.6911,
+      "step": 11706
+    },
+    {
+      "epoch": 0.6443392591777203,
+      "grad_norm": 0.660022497177124,
+      "learning_rate": 7.675165482863254e-06,
+      "loss": 0.7359,
+      "step": 11707
+    },
+    {
+      "epoch": 0.644394297980076,
+      "grad_norm": 0.6956108808517456,
+      "learning_rate": 7.674799266732422e-06,
+      "loss": 0.7845,
+      "step": 11708
+    },
+    {
+      "epoch": 0.6444493367824317,
+      "grad_norm": 0.7361618280410767,
+      "learning_rate": 7.674433030498513e-06,
+      "loss": 0.7391,
+      "step": 11709
+    },
+    {
+      "epoch": 0.6445043755847872,
+      "grad_norm": 0.7655043005943298,
+      "learning_rate": 7.674066774164284e-06,
+      "loss": 0.8305,
+      "step": 11710
+    },
+    {
+      "epoch": 0.6445594143871429,
+      "grad_norm": 0.7160911560058594,
+      "learning_rate": 7.673700497732483e-06,
+      "loss": 0.7654,
+      "step": 11711
+    },
+    {
+      "epoch": 0.6446144531894986,
+      "grad_norm": 0.7812016010284424,
+      "learning_rate": 7.673334201205866e-06,
+      "loss": 0.8212,
+      "step": 11712
+    },
+    {
+      "epoch": 0.6446694919918543,
+      "grad_norm": 0.7457767128944397,
+      "learning_rate": 7.672967884587184e-06,
+      "loss": 0.8084,
+      "step": 11713
+    },
+    {
+      "epoch": 0.6447245307942099,
+      "grad_norm": 0.7524051070213318,
+      "learning_rate": 7.672601547879189e-06,
+      "loss": 0.7525,
+      "step": 11714
+    },
+    {
+      "epoch": 0.6447795695965656,
+      "grad_norm": 0.7271043062210083,
+      "learning_rate": 7.672235191084638e-06,
+      "loss": 0.7627,
+      "step": 11715
+    },
+    {
+      "epoch": 0.6448346083989213,
+      "grad_norm": 0.6893014907836914,
+      "learning_rate": 7.671868814206283e-06,
+      "loss": 0.7969,
+      "step": 11716
+    },
+    {
+      "epoch": 0.644889647201277,
+      "grad_norm": 0.7057414054870605,
+      "learning_rate": 7.671502417246876e-06,
+      "loss": 0.7448,
+      "step": 11717
+    },
+    {
+      "epoch": 0.6449446860036325,
+      "grad_norm": 0.7490910887718201,
+      "learning_rate": 7.671136000209172e-06,
+      "loss": 0.8046,
+      "step": 11718
+    },
+    {
+      "epoch": 0.6449997248059882,
+      "grad_norm": 0.7338950634002686,
+      "learning_rate": 7.670769563095926e-06,
+      "loss": 0.8521,
+      "step": 11719
+    },
+    {
+      "epoch": 0.6450547636083439,
+      "grad_norm": 0.8669398427009583,
+      "learning_rate": 7.670403105909891e-06,
+      "loss": 0.7803,
+      "step": 11720
+    },
+    {
+      "epoch": 0.6451098024106996,
+      "grad_norm": 0.7012562155723572,
+      "learning_rate": 7.67003662865382e-06,
+      "loss": 0.8047,
+      "step": 11721
+    },
+    {
+      "epoch": 0.6451648412130552,
+      "grad_norm": 0.9933050274848938,
+      "learning_rate": 7.66967013133047e-06,
+      "loss": 0.7081,
+      "step": 11722
+    },
+    {
+      "epoch": 0.6452198800154109,
+      "grad_norm": 1.12044358253479,
+      "learning_rate": 7.669303613942592e-06,
+      "loss": 0.7315,
+      "step": 11723
+    },
+    {
+      "epoch": 0.6452749188177666,
+      "grad_norm": 0.8654733300209045,
+      "learning_rate": 7.668937076492943e-06,
+      "loss": 0.6849,
+      "step": 11724
+    },
+    {
+      "epoch": 0.6453299576201222,
+      "grad_norm": 0.7081291675567627,
+      "learning_rate": 7.668570518984277e-06,
+      "loss": 0.7584,
+      "step": 11725
+    },
+    {
+      "epoch": 0.6453849964224778,
+      "grad_norm": 0.7473898530006409,
+      "learning_rate": 7.66820394141935e-06,
+      "loss": 0.8364,
+      "step": 11726
+    },
+    {
+      "epoch": 0.6454400352248335,
+      "grad_norm": 0.7863657474517822,
+      "learning_rate": 7.667837343800916e-06,
+      "loss": 0.7235,
+      "step": 11727
+    },
+    {
+      "epoch": 0.6454950740271892,
+      "grad_norm": 0.6664546728134155,
+      "learning_rate": 7.667470726131732e-06,
+      "loss": 0.7203,
+      "step": 11728
+    },
+    {
+      "epoch": 0.6455501128295448,
+      "grad_norm": 0.7182374596595764,
+      "learning_rate": 7.667104088414552e-06,
+      "loss": 0.7376,
+      "step": 11729
+    },
+    {
+      "epoch": 0.6456051516319005,
+      "grad_norm": 0.6518070697784424,
+      "learning_rate": 7.666737430652128e-06,
+      "loss": 0.6804,
+      "step": 11730
+    },
+    {
+      "epoch": 0.6456601904342562,
+      "grad_norm": 0.7354047894477844,
+      "learning_rate": 7.666370752847223e-06,
+      "loss": 0.7648,
+      "step": 11731
+    },
+    {
+      "epoch": 0.6457152292366118,
+      "grad_norm": 0.7440805435180664,
+      "learning_rate": 7.666004055002588e-06,
+      "loss": 0.7674,
+      "step": 11732
+    },
+    {
+      "epoch": 0.6457702680389674,
+      "grad_norm": 1.6423569917678833,
+      "learning_rate": 7.665637337120981e-06,
+      "loss": 0.8957,
+      "step": 11733
+    },
+    {
+      "epoch": 0.6458253068413231,
+      "grad_norm": 0.6960558295249939,
+      "learning_rate": 7.665270599205156e-06,
+      "loss": 0.7278,
+      "step": 11734
+    },
+    {
+      "epoch": 0.6458803456436788,
+      "grad_norm": 0.6983850002288818,
+      "learning_rate": 7.664903841257871e-06,
+      "loss": 0.7351,
+      "step": 11735
+    },
+    {
+      "epoch": 0.6459353844460345,
+      "grad_norm": 0.6905686855316162,
+      "learning_rate": 7.664537063281883e-06,
+      "loss": 0.7558,
+      "step": 11736
+    },
+    {
+      "epoch": 0.6459904232483901,
+      "grad_norm": 0.7483980655670166,
+      "learning_rate": 7.664170265279946e-06,
+      "loss": 0.813,
+      "step": 11737
+    },
+    {
+      "epoch": 0.6460454620507458,
+      "grad_norm": 0.767756998538971,
+      "learning_rate": 7.66380344725482e-06,
+      "loss": 0.8397,
+      "step": 11738
+    },
+    {
+      "epoch": 0.6461005008531014,
+      "grad_norm": 0.7813250422477722,
+      "learning_rate": 7.66343660920926e-06,
+      "loss": 0.8034,
+      "step": 11739
+    },
+    {
+      "epoch": 0.6461555396554571,
+      "grad_norm": 0.7357046604156494,
+      "learning_rate": 7.663069751146022e-06,
+      "loss": 0.7604,
+      "step": 11740
+    },
+    {
+      "epoch": 0.6462105784578127,
+      "grad_norm": 0.620285153388977,
+      "learning_rate": 7.662702873067866e-06,
+      "loss": 0.6191,
+      "step": 11741
+    },
+    {
+      "epoch": 0.6462656172601684,
+      "grad_norm": 0.6711301803588867,
+      "learning_rate": 7.662335974977549e-06,
+      "loss": 0.7674,
+      "step": 11742
+    },
+    {
+      "epoch": 0.6463206560625241,
+      "grad_norm": 0.756258487701416,
+      "learning_rate": 7.661969056877824e-06,
+      "loss": 0.7074,
+      "step": 11743
+    },
+    {
+      "epoch": 0.6463756948648798,
+      "grad_norm": 0.8121050596237183,
+      "learning_rate": 7.661602118771456e-06,
+      "loss": 0.8028,
+      "step": 11744
+    },
+    {
+      "epoch": 0.6464307336672354,
+      "grad_norm": 0.735906720161438,
+      "learning_rate": 7.661235160661197e-06,
+      "loss": 0.7197,
+      "step": 11745
+    },
+    {
+      "epoch": 0.646485772469591,
+      "grad_norm": 0.644490122795105,
+      "learning_rate": 7.660868182549807e-06,
+      "loss": 0.6172,
+      "step": 11746
+    },
+    {
+      "epoch": 0.6465408112719467,
+      "grad_norm": 0.7228739261627197,
+      "learning_rate": 7.660501184440045e-06,
+      "loss": 0.8302,
+      "step": 11747
+    },
+    {
+      "epoch": 0.6465958500743024,
+      "grad_norm": 0.8292868137359619,
+      "learning_rate": 7.660134166334668e-06,
+      "loss": 0.7506,
+      "step": 11748
+    },
+    {
+      "epoch": 0.646650888876658,
+      "grad_norm": 0.7224695086479187,
+      "learning_rate": 7.659767128236433e-06,
+      "loss": 0.8043,
+      "step": 11749
+    },
+    {
+      "epoch": 0.6467059276790137,
+      "grad_norm": 0.7092188000679016,
+      "learning_rate": 7.659400070148102e-06,
+      "loss": 0.7838,
+      "step": 11750
+    },
+    {
+      "epoch": 0.6467609664813694,
+      "grad_norm": 0.6975178122520447,
+      "learning_rate": 7.65903299207243e-06,
+      "loss": 0.7576,
+      "step": 11751
+    },
+    {
+      "epoch": 0.6468160052837251,
+      "grad_norm": 0.6524471044540405,
+      "learning_rate": 7.658665894012179e-06,
+      "loss": 0.7822,
+      "step": 11752
+    },
+    {
+      "epoch": 0.6468710440860806,
+      "grad_norm": 0.8134269118309021,
+      "learning_rate": 7.658298775970107e-06,
+      "loss": 0.8116,
+      "step": 11753
+    },
+    {
+      "epoch": 0.6469260828884363,
+      "grad_norm": 0.7166362404823303,
+      "learning_rate": 7.657931637948974e-06,
+      "loss": 0.768,
+      "step": 11754
+    },
+    {
+      "epoch": 0.646981121690792,
+      "grad_norm": 0.6418643593788147,
+      "learning_rate": 7.657564479951535e-06,
+      "loss": 0.7488,
+      "step": 11755
+    },
+    {
+      "epoch": 0.6470361604931477,
+      "grad_norm": 0.7104085087776184,
+      "learning_rate": 7.657197301980556e-06,
+      "loss": 0.7518,
+      "step": 11756
+    },
+    {
+      "epoch": 0.6470911992955033,
+      "grad_norm": 0.7297894358634949,
+      "learning_rate": 7.656830104038793e-06,
+      "loss": 0.7877,
+      "step": 11757
+    },
+    {
+      "epoch": 0.647146238097859,
+      "grad_norm": 0.8037092089653015,
+      "learning_rate": 7.656462886129006e-06,
+      "loss": 0.7375,
+      "step": 11758
+    },
+    {
+      "epoch": 0.6472012769002147,
+      "grad_norm": 0.7498913407325745,
+      "learning_rate": 7.656095648253955e-06,
+      "loss": 0.7899,
+      "step": 11759
+    },
+    {
+      "epoch": 0.6472563157025704,
+      "grad_norm": 0.7383849620819092,
+      "learning_rate": 7.655728390416398e-06,
+      "loss": 0.8276,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6473113545049259,
+      "grad_norm": 0.750481367111206,
+      "learning_rate": 7.6553611126191e-06,
+      "loss": 0.7649,
+      "step": 11761
+    },
+    {
+      "epoch": 0.6473663933072816,
+      "grad_norm": 0.8483286499977112,
+      "learning_rate": 7.654993814864817e-06,
+      "loss": 0.877,
+      "step": 11762
+    },
+    {
+      "epoch": 0.6474214321096373,
+      "grad_norm": 0.7938307523727417,
+      "learning_rate": 7.654626497156311e-06,
+      "loss": 0.8159,
+      "step": 11763
+    },
+    {
+      "epoch": 0.647476470911993,
+      "grad_norm": 0.6576653122901917,
+      "learning_rate": 7.654259159496343e-06,
+      "loss": 0.797,
+      "step": 11764
+    },
+    {
+      "epoch": 0.6475315097143486,
+      "grad_norm": 0.6495664715766907,
+      "learning_rate": 7.653891801887675e-06,
+      "loss": 0.6641,
+      "step": 11765
+    },
+    {
+      "epoch": 0.6475865485167043,
+      "grad_norm": 0.7447353601455688,
+      "learning_rate": 7.653524424333065e-06,
+      "loss": 0.667,
+      "step": 11766
+    },
+    {
+      "epoch": 0.64764158731906,
+      "grad_norm": 0.6565769910812378,
+      "learning_rate": 7.653157026835277e-06,
+      "loss": 0.7123,
+      "step": 11767
+    },
+    {
+      "epoch": 0.6476966261214157,
+      "grad_norm": 0.8406145572662354,
+      "learning_rate": 7.652789609397072e-06,
+      "loss": 0.7582,
+      "step": 11768
+    },
+    {
+      "epoch": 0.6477516649237712,
+      "grad_norm": 0.8478217720985413,
+      "learning_rate": 7.652422172021207e-06,
+      "loss": 0.6758,
+      "step": 11769
+    },
+    {
+      "epoch": 0.6478067037261269,
+      "grad_norm": 0.7230110168457031,
+      "learning_rate": 7.652054714710448e-06,
+      "loss": 0.8216,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6478617425284826,
+      "grad_norm": 0.6718668341636658,
+      "learning_rate": 7.651687237467558e-06,
+      "loss": 0.7204,
+      "step": 11771
+    },
+    {
+      "epoch": 0.6479167813308382,
+      "grad_norm": 1.062383770942688,
+      "learning_rate": 7.651319740295296e-06,
+      "loss": 0.6853,
+      "step": 11772
+    },
+    {
+      "epoch": 0.6479718201331939,
+      "grad_norm": 0.7157385945320129,
+      "learning_rate": 7.650952223196423e-06,
+      "loss": 0.6826,
+      "step": 11773
+    },
+    {
+      "epoch": 0.6480268589355496,
+      "grad_norm": 0.6762190461158752,
+      "learning_rate": 7.650584686173703e-06,
+      "loss": 0.7673,
+      "step": 11774
+    },
+    {
+      "epoch": 0.6480818977379053,
+      "grad_norm": 0.7540121674537659,
+      "learning_rate": 7.650217129229897e-06,
+      "loss": 0.7361,
+      "step": 11775
+    },
+    {
+      "epoch": 0.6481369365402608,
+      "grad_norm": 1.0383096933364868,
+      "learning_rate": 7.649849552367771e-06,
+      "loss": 0.7936,
+      "step": 11776
+    },
+    {
+      "epoch": 0.6481919753426165,
+      "grad_norm": 0.6430917382240295,
+      "learning_rate": 7.649481955590084e-06,
+      "loss": 0.7738,
+      "step": 11777
+    },
+    {
+      "epoch": 0.6482470141449722,
+      "grad_norm": 0.7846735715866089,
+      "learning_rate": 7.6491143388996e-06,
+      "loss": 0.6892,
+      "step": 11778
+    },
+    {
+      "epoch": 0.6483020529473279,
+      "grad_norm": 0.7154437899589539,
+      "learning_rate": 7.64874670229908e-06,
+      "loss": 0.6889,
+      "step": 11779
+    },
+    {
+      "epoch": 0.6483570917496835,
+      "grad_norm": 0.731270432472229,
+      "learning_rate": 7.648379045791291e-06,
+      "loss": 0.6405,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6484121305520392,
+      "grad_norm": 0.6782581210136414,
+      "learning_rate": 7.648011369378993e-06,
+      "loss": 0.7822,
+      "step": 11781
+    },
+    {
+      "epoch": 0.6484671693543949,
+      "grad_norm": 0.7025747299194336,
+      "learning_rate": 7.64764367306495e-06,
+      "loss": 0.6929,
+      "step": 11782
+    },
+    {
+      "epoch": 0.6485222081567505,
+      "grad_norm": 0.6791071891784668,
+      "learning_rate": 7.647275956851928e-06,
+      "loss": 0.7507,
+      "step": 11783
+    },
+    {
+      "epoch": 0.6485772469591061,
+      "grad_norm": 0.7598931193351746,
+      "learning_rate": 7.646908220742686e-06,
+      "loss": 0.776,
+      "step": 11784
+    },
+    {
+      "epoch": 0.6486322857614618,
+      "grad_norm": 0.6930273771286011,
+      "learning_rate": 7.646540464739993e-06,
+      "loss": 0.7653,
+      "step": 11785
+    },
+    {
+      "epoch": 0.6486873245638175,
+      "grad_norm": 0.7276393175125122,
+      "learning_rate": 7.646172688846608e-06,
+      "loss": 0.8102,
+      "step": 11786
+    },
+    {
+      "epoch": 0.6487423633661732,
+      "grad_norm": 0.6826562285423279,
+      "learning_rate": 7.645804893065298e-06,
+      "loss": 0.6182,
+      "step": 11787
+    },
+    {
+      "epoch": 0.6487974021685288,
+      "grad_norm": 0.7837507128715515,
+      "learning_rate": 7.645437077398827e-06,
+      "loss": 0.8124,
+      "step": 11788
+    },
+    {
+      "epoch": 0.6488524409708845,
+      "grad_norm": 0.6937540769577026,
+      "learning_rate": 7.645069241849959e-06,
+      "loss": 0.7831,
+      "step": 11789
+    },
+    {
+      "epoch": 0.6489074797732401,
+      "grad_norm": 0.6531546115875244,
+      "learning_rate": 7.644701386421458e-06,
+      "loss": 0.755,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6489625185755958,
+      "grad_norm": 0.8563246726989746,
+      "learning_rate": 7.644333511116088e-06,
+      "loss": 0.7715,
+      "step": 11791
+    },
+    {
+      "epoch": 0.6490175573779514,
+      "grad_norm": 0.8330580592155457,
+      "learning_rate": 7.643965615936619e-06,
+      "loss": 0.6651,
+      "step": 11792
+    },
+    {
+      "epoch": 0.6490725961803071,
+      "grad_norm": 0.6478384137153625,
+      "learning_rate": 7.643597700885809e-06,
+      "loss": 0.7063,
+      "step": 11793
+    },
+    {
+      "epoch": 0.6491276349826628,
+      "grad_norm": 0.7169124484062195,
+      "learning_rate": 7.643229765966428e-06,
+      "loss": 0.7578,
+      "step": 11794
+    },
+    {
+      "epoch": 0.6491826737850185,
+      "grad_norm": 0.726198136806488,
+      "learning_rate": 7.642861811181239e-06,
+      "loss": 0.783,
+      "step": 11795
+    },
+    {
+      "epoch": 0.6492377125873741,
+      "grad_norm": 0.7167587280273438,
+      "learning_rate": 7.642493836533008e-06,
+      "loss": 0.81,
+      "step": 11796
+    },
+    {
+      "epoch": 0.6492927513897297,
+      "grad_norm": 0.7215337157249451,
+      "learning_rate": 7.642125842024502e-06,
+      "loss": 0.8176,
+      "step": 11797
+    },
+    {
+      "epoch": 0.6493477901920854,
+      "grad_norm": 0.7041502594947815,
+      "learning_rate": 7.641757827658484e-06,
+      "loss": 0.8117,
+      "step": 11798
+    },
+    {
+      "epoch": 0.6494028289944411,
+      "grad_norm": 1.0303698778152466,
+      "learning_rate": 7.64138979343772e-06,
+      "loss": 0.781,
+      "step": 11799
+    },
+    {
+      "epoch": 0.6494578677967967,
+      "grad_norm": 0.626518189907074,
+      "learning_rate": 7.64102173936498e-06,
+      "loss": 0.6668,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6495129065991524,
+      "grad_norm": 0.8889065980911255,
+      "learning_rate": 7.640653665443025e-06,
+      "loss": 0.8076,
+      "step": 11801
+    },
+    {
+      "epoch": 0.6495679454015081,
+      "grad_norm": 0.8333556652069092,
+      "learning_rate": 7.640285571674626e-06,
+      "loss": 0.8111,
+      "step": 11802
+    },
+    {
+      "epoch": 0.6496229842038638,
+      "grad_norm": 0.7248615622520447,
+      "learning_rate": 7.639917458062547e-06,
+      "loss": 0.7876,
+      "step": 11803
+    },
+    {
+      "epoch": 0.6496780230062194,
+      "grad_norm": 0.8870820999145508,
+      "learning_rate": 7.639549324609554e-06,
+      "loss": 0.8586,
+      "step": 11804
+    },
+    {
+      "epoch": 0.649733061808575,
+      "grad_norm": 0.7777245044708252,
+      "learning_rate": 7.639181171318417e-06,
+      "loss": 0.7793,
+      "step": 11805
+    },
+    {
+      "epoch": 0.6497881006109307,
+      "grad_norm": 0.7858467102050781,
+      "learning_rate": 7.638812998191897e-06,
+      "loss": 0.7842,
+      "step": 11806
+    },
+    {
+      "epoch": 0.6498431394132864,
+      "grad_norm": 0.6278610825538635,
+      "learning_rate": 7.638444805232769e-06,
+      "loss": 0.6659,
+      "step": 11807
+    },
+    {
+      "epoch": 0.649898178215642,
+      "grad_norm": 0.6758826971054077,
+      "learning_rate": 7.638076592443795e-06,
+      "loss": 0.7047,
+      "step": 11808
+    },
+    {
+      "epoch": 0.6499532170179977,
+      "grad_norm": 0.745007336139679,
+      "learning_rate": 7.637708359827743e-06,
+      "loss": 0.8557,
+      "step": 11809
+    },
+    {
+      "epoch": 0.6500082558203534,
+      "grad_norm": 0.8092321157455444,
+      "learning_rate": 7.63734010738738e-06,
+      "loss": 0.7895,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6500632946227091,
+      "grad_norm": 0.7055220603942871,
+      "learning_rate": 7.636971835125476e-06,
+      "loss": 0.7678,
+      "step": 11811
+    },
+    {
+      "epoch": 0.6501183334250646,
+      "grad_norm": 0.7130264043807983,
+      "learning_rate": 7.636603543044797e-06,
+      "loss": 0.7648,
+      "step": 11812
+    },
+    {
+      "epoch": 0.6501733722274203,
+      "grad_norm": 0.7494268417358398,
+      "learning_rate": 7.636235231148112e-06,
+      "loss": 0.7883,
+      "step": 11813
+    },
+    {
+      "epoch": 0.650228411029776,
+      "grad_norm": 0.7998068332672119,
+      "learning_rate": 7.635866899438189e-06,
+      "loss": 0.7849,
+      "step": 11814
+    },
+    {
+      "epoch": 0.6502834498321316,
+      "grad_norm": 0.6749094128608704,
+      "learning_rate": 7.635498547917795e-06,
+      "loss": 0.8488,
+      "step": 11815
+    },
+    {
+      "epoch": 0.6503384886344873,
+      "grad_norm": 0.743679940700531,
+      "learning_rate": 7.635130176589698e-06,
+      "loss": 0.7562,
+      "step": 11816
+    },
+    {
+      "epoch": 0.650393527436843,
+      "grad_norm": 0.8368289470672607,
+      "learning_rate": 7.634761785456671e-06,
+      "loss": 0.7012,
+      "step": 11817
     }
   ],
   "logging_steps": 1,
@@ -76382,7 +82745,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.219017656946393e+19,
+  "total_flos": 3.4872691283585925e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null