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": 0.7504540701194342,
+  "epoch": 0.8004843414607298,
   "eval_steps": 500,
-  "global_step": 13635,
+  "global_step": 14544,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -95452,6 +95452,6369 @@
       "learning_rate": 6.935121468327907e-06,
       "loss": 0.7858,
       "step": 13635
+    },
+    {
+      "epoch": 0.7505091089217899,
+      "grad_norm": 0.7102827429771423,
+      "learning_rate": 6.934721774701824e-06,
+      "loss": 0.7485,
+      "step": 13636
+    },
+    {
+      "epoch": 0.7505641477241455,
+      "grad_norm": 0.7031061053276062,
+      "learning_rate": 6.934322066534891e-06,
+      "loss": 0.7154,
+      "step": 13637
+    },
+    {
+      "epoch": 0.7506191865265012,
+      "grad_norm": 0.6468148231506348,
+      "learning_rate": 6.933922343830112e-06,
+      "loss": 0.729,
+      "step": 13638
+    },
+    {
+      "epoch": 0.7506742253288569,
+      "grad_norm": 0.8570408225059509,
+      "learning_rate": 6.933522606590489e-06,
+      "loss": 0.6922,
+      "step": 13639
+    },
+    {
+      "epoch": 0.7507292641312125,
+      "grad_norm": 0.6836286783218384,
+      "learning_rate": 6.933122854819027e-06,
+      "loss": 0.7982,
+      "step": 13640
+    },
+    {
+      "epoch": 0.7507843029335681,
+      "grad_norm": 1.052017092704773,
+      "learning_rate": 6.9327230885187344e-06,
+      "loss": 0.7522,
+      "step": 13641
+    },
+    {
+      "epoch": 0.7508393417359238,
+      "grad_norm": 0.6352099180221558,
+      "learning_rate": 6.932323307692611e-06,
+      "loss": 0.6724,
+      "step": 13642
+    },
+    {
+      "epoch": 0.7508943805382795,
+      "grad_norm": 0.7046655416488647,
+      "learning_rate": 6.931923512343663e-06,
+      "loss": 0.7732,
+      "step": 13643
+    },
+    {
+      "epoch": 0.7509494193406352,
+      "grad_norm": 0.7600587010383606,
+      "learning_rate": 6.931523702474893e-06,
+      "loss": 0.7013,
+      "step": 13644
+    },
+    {
+      "epoch": 0.7510044581429908,
+      "grad_norm": 0.674828052520752,
+      "learning_rate": 6.9311238780893095e-06,
+      "loss": 0.7022,
+      "step": 13645
+    },
+    {
+      "epoch": 0.7510594969453465,
+      "grad_norm": 0.7517798542976379,
+      "learning_rate": 6.930724039189916e-06,
+      "loss": 0.7248,
+      "step": 13646
+    },
+    {
+      "epoch": 0.7511145357477022,
+      "grad_norm": 0.7851112484931946,
+      "learning_rate": 6.930324185779716e-06,
+      "loss": 0.8025,
+      "step": 13647
+    },
+    {
+      "epoch": 0.7511695745500578,
+      "grad_norm": 0.6545413732528687,
+      "learning_rate": 6.929924317861717e-06,
+      "loss": 0.781,
+      "step": 13648
+    },
+    {
+      "epoch": 0.7512246133524134,
+      "grad_norm": 0.7079984545707703,
+      "learning_rate": 6.929524435438923e-06,
+      "loss": 0.8033,
+      "step": 13649
+    },
+    {
+      "epoch": 0.7512796521547691,
+      "grad_norm": 0.6501914262771606,
+      "learning_rate": 6.929124538514341e-06,
+      "loss": 0.7525,
+      "step": 13650
+    },
+    {
+      "epoch": 0.7513346909571248,
+      "grad_norm": 0.7697597742080688,
+      "learning_rate": 6.928724627090975e-06,
+      "loss": 0.7358,
+      "step": 13651
+    },
+    {
+      "epoch": 0.7513897297594805,
+      "grad_norm": 0.8155171275138855,
+      "learning_rate": 6.928324701171832e-06,
+      "loss": 0.7389,
+      "step": 13652
+    },
+    {
+      "epoch": 0.7514447685618361,
+      "grad_norm": 0.6969262361526489,
+      "learning_rate": 6.927924760759914e-06,
+      "loss": 0.8349,
+      "step": 13653
+    },
+    {
+      "epoch": 0.7514998073641918,
+      "grad_norm": 0.6736776828765869,
+      "learning_rate": 6.927524805858233e-06,
+      "loss": 0.7379,
+      "step": 13654
+    },
+    {
+      "epoch": 0.7515548461665474,
+      "grad_norm": 0.6362389922142029,
+      "learning_rate": 6.927124836469788e-06,
+      "loss": 0.7479,
+      "step": 13655
+    },
+    {
+      "epoch": 0.7516098849689031,
+      "grad_norm": 0.688922643661499,
+      "learning_rate": 6.92672485259759e-06,
+      "loss": 0.7828,
+      "step": 13656
+    },
+    {
+      "epoch": 0.7516649237712587,
+      "grad_norm": 0.7098214030265808,
+      "learning_rate": 6.926324854244644e-06,
+      "loss": 0.6084,
+      "step": 13657
+    },
+    {
+      "epoch": 0.7517199625736144,
+      "grad_norm": 0.6436209678649902,
+      "learning_rate": 6.925924841413956e-06,
+      "loss": 0.687,
+      "step": 13658
+    },
+    {
+      "epoch": 0.7517750013759701,
+      "grad_norm": 0.6051730513572693,
+      "learning_rate": 6.925524814108533e-06,
+      "loss": 0.6884,
+      "step": 13659
+    },
+    {
+      "epoch": 0.7518300401783257,
+      "grad_norm": 0.6347759962081909,
+      "learning_rate": 6.92512477233138e-06,
+      "loss": 0.7057,
+      "step": 13660
+    },
+    {
+      "epoch": 0.7518850789806814,
+      "grad_norm": 0.6917054653167725,
+      "learning_rate": 6.924724716085505e-06,
+      "loss": 0.8374,
+      "step": 13661
+    },
+    {
+      "epoch": 0.751940117783037,
+      "grad_norm": 0.7676698565483093,
+      "learning_rate": 6.924324645373914e-06,
+      "loss": 0.7435,
+      "step": 13662
+    },
+    {
+      "epoch": 0.7519951565853927,
+      "grad_norm": 0.6601388454437256,
+      "learning_rate": 6.923924560199613e-06,
+      "loss": 0.7168,
+      "step": 13663
+    },
+    {
+      "epoch": 0.7520501953877483,
+      "grad_norm": 0.6342683434486389,
+      "learning_rate": 6.923524460565611e-06,
+      "loss": 0.7382,
+      "step": 13664
+    },
+    {
+      "epoch": 0.752105234190104,
+      "grad_norm": 0.6703974604606628,
+      "learning_rate": 6.923124346474915e-06,
+      "loss": 0.7687,
+      "step": 13665
+    },
+    {
+      "epoch": 0.7521602729924597,
+      "grad_norm": 0.6937074661254883,
+      "learning_rate": 6.922724217930531e-06,
+      "loss": 0.7687,
+      "step": 13666
+    },
+    {
+      "epoch": 0.7522153117948154,
+      "grad_norm": 0.7919568419456482,
+      "learning_rate": 6.922324074935466e-06,
+      "loss": 0.7328,
+      "step": 13667
+    },
+    {
+      "epoch": 0.752270350597171,
+      "grad_norm": 0.668331503868103,
+      "learning_rate": 6.9219239174927275e-06,
+      "loss": 0.7654,
+      "step": 13668
+    },
+    {
+      "epoch": 0.7523253893995266,
+      "grad_norm": 0.6298941969871521,
+      "learning_rate": 6.921523745605323e-06,
+      "loss": 0.719,
+      "step": 13669
+    },
+    {
+      "epoch": 0.7523804282018823,
+      "grad_norm": 0.6539381146430969,
+      "learning_rate": 6.921123559276262e-06,
+      "loss": 0.6681,
+      "step": 13670
+    },
+    {
+      "epoch": 0.752435467004238,
+      "grad_norm": 1.0692330598831177,
+      "learning_rate": 6.920723358508548e-06,
+      "loss": 0.7914,
+      "step": 13671
+    },
+    {
+      "epoch": 0.7524905058065936,
+      "grad_norm": 0.7410482168197632,
+      "learning_rate": 6.920323143305193e-06,
+      "loss": 0.8331,
+      "step": 13672
+    },
+    {
+      "epoch": 0.7525455446089493,
+      "grad_norm": 0.6976327300071716,
+      "learning_rate": 6.919922913669203e-06,
+      "loss": 0.8131,
+      "step": 13673
+    },
+    {
+      "epoch": 0.752600583411305,
+      "grad_norm": 0.646442174911499,
+      "learning_rate": 6.919522669603587e-06,
+      "loss": 0.7658,
+      "step": 13674
+    },
+    {
+      "epoch": 0.7526556222136607,
+      "grad_norm": 0.6257727146148682,
+      "learning_rate": 6.919122411111352e-06,
+      "loss": 0.666,
+      "step": 13675
+    },
+    {
+      "epoch": 0.7527106610160162,
+      "grad_norm": 0.6913230419158936,
+      "learning_rate": 6.918722138195506e-06,
+      "loss": 0.6935,
+      "step": 13676
+    },
+    {
+      "epoch": 0.7527656998183719,
+      "grad_norm": 0.6282557249069214,
+      "learning_rate": 6.918321850859059e-06,
+      "loss": 0.7042,
+      "step": 13677
+    },
+    {
+      "epoch": 0.7528207386207276,
+      "grad_norm": 0.6980175971984863,
+      "learning_rate": 6.917921549105018e-06,
+      "loss": 0.6757,
+      "step": 13678
+    },
+    {
+      "epoch": 0.7528757774230833,
+      "grad_norm": 0.6954337954521179,
+      "learning_rate": 6.917521232936393e-06,
+      "loss": 0.729,
+      "step": 13679
+    },
+    {
+      "epoch": 0.7529308162254389,
+      "grad_norm": 0.6813758015632629,
+      "learning_rate": 6.91712090235619e-06,
+      "loss": 0.6964,
+      "step": 13680
+    },
+    {
+      "epoch": 0.7529858550277946,
+      "grad_norm": 1.0940780639648438,
+      "learning_rate": 6.916720557367419e-06,
+      "loss": 0.7853,
+      "step": 13681
+    },
+    {
+      "epoch": 0.7530408938301503,
+      "grad_norm": 0.6899382472038269,
+      "learning_rate": 6.9163201979730906e-06,
+      "loss": 0.7639,
+      "step": 13682
+    },
+    {
+      "epoch": 0.753095932632506,
+      "grad_norm": 0.660252034664154,
+      "learning_rate": 6.915919824176213e-06,
+      "loss": 0.7068,
+      "step": 13683
+    },
+    {
+      "epoch": 0.7531509714348615,
+      "grad_norm": 0.6454583406448364,
+      "learning_rate": 6.915519435979795e-06,
+      "loss": 0.7268,
+      "step": 13684
+    },
+    {
+      "epoch": 0.7532060102372172,
+      "grad_norm": 0.7292754650115967,
+      "learning_rate": 6.915119033386843e-06,
+      "loss": 0.8131,
+      "step": 13685
+    },
+    {
+      "epoch": 0.7532610490395729,
+      "grad_norm": 0.6312932372093201,
+      "learning_rate": 6.914718616400372e-06,
+      "loss": 0.6977,
+      "step": 13686
+    },
+    {
+      "epoch": 0.7533160878419286,
+      "grad_norm": 0.8528029322624207,
+      "learning_rate": 6.914318185023388e-06,
+      "loss": 0.8403,
+      "step": 13687
+    },
+    {
+      "epoch": 0.7533711266442842,
+      "grad_norm": 0.758721649646759,
+      "learning_rate": 6.9139177392589e-06,
+      "loss": 0.7,
+      "step": 13688
+    },
+    {
+      "epoch": 0.7534261654466399,
+      "grad_norm": 0.6678142547607422,
+      "learning_rate": 6.913517279109919e-06,
+      "loss": 0.6251,
+      "step": 13689
+    },
+    {
+      "epoch": 0.7534812042489956,
+      "grad_norm": 0.6136146783828735,
+      "learning_rate": 6.913116804579455e-06,
+      "loss": 0.653,
+      "step": 13690
+    },
+    {
+      "epoch": 0.7535362430513513,
+      "grad_norm": 0.7546648383140564,
+      "learning_rate": 6.912716315670517e-06,
+      "loss": 0.8202,
+      "step": 13691
+    },
+    {
+      "epoch": 0.7535912818537068,
+      "grad_norm": 0.7232012152671814,
+      "learning_rate": 6.912315812386114e-06,
+      "loss": 0.7993,
+      "step": 13692
+    },
+    {
+      "epoch": 0.7536463206560625,
+      "grad_norm": 0.7288710474967957,
+      "learning_rate": 6.911915294729258e-06,
+      "loss": 0.7702,
+      "step": 13693
+    },
+    {
+      "epoch": 0.7537013594584182,
+      "grad_norm": 0.6847403049468994,
+      "learning_rate": 6.9115147627029575e-06,
+      "loss": 0.8141,
+      "step": 13694
+    },
+    {
+      "epoch": 0.7537563982607739,
+      "grad_norm": 0.62345951795578,
+      "learning_rate": 6.9111142163102255e-06,
+      "loss": 0.6832,
+      "step": 13695
+    },
+    {
+      "epoch": 0.7538114370631295,
+      "grad_norm": 0.7275232672691345,
+      "learning_rate": 6.9107136555540695e-06,
+      "loss": 0.7548,
+      "step": 13696
+    },
+    {
+      "epoch": 0.7538664758654852,
+      "grad_norm": 0.6724695563316345,
+      "learning_rate": 6.910313080437501e-06,
+      "loss": 0.7755,
+      "step": 13697
+    },
+    {
+      "epoch": 0.7539215146678409,
+      "grad_norm": 0.8446974754333496,
+      "learning_rate": 6.90991249096353e-06,
+      "loss": 0.827,
+      "step": 13698
+    },
+    {
+      "epoch": 0.7539765534701965,
+      "grad_norm": 0.7124913930892944,
+      "learning_rate": 6.9095118871351705e-06,
+      "loss": 0.7463,
+      "step": 13699
+    },
+    {
+      "epoch": 0.7540315922725521,
+      "grad_norm": 0.6916043162345886,
+      "learning_rate": 6.90911126895543e-06,
+      "loss": 0.714,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7540866310749078,
+      "grad_norm": 0.7585330009460449,
+      "learning_rate": 6.908710636427319e-06,
+      "loss": 0.6731,
+      "step": 13701
+    },
+    {
+      "epoch": 0.7541416698772635,
+      "grad_norm": 0.6905520558357239,
+      "learning_rate": 6.90830998955385e-06,
+      "loss": 0.726,
+      "step": 13702
+    },
+    {
+      "epoch": 0.7541967086796191,
+      "grad_norm": 0.7482494115829468,
+      "learning_rate": 6.907909328338035e-06,
+      "loss": 0.7269,
+      "step": 13703
+    },
+    {
+      "epoch": 0.7542517474819748,
+      "grad_norm": 0.7565957307815552,
+      "learning_rate": 6.907508652782884e-06,
+      "loss": 0.6959,
+      "step": 13704
+    },
+    {
+      "epoch": 0.7543067862843305,
+      "grad_norm": 0.7458370923995972,
+      "learning_rate": 6.9071079628914075e-06,
+      "loss": 0.7448,
+      "step": 13705
+    },
+    {
+      "epoch": 0.7543618250866861,
+      "grad_norm": 1.3538293838500977,
+      "learning_rate": 6.9067072586666185e-06,
+      "loss": 0.8164,
+      "step": 13706
+    },
+    {
+      "epoch": 0.7544168638890417,
+      "grad_norm": 0.6217493414878845,
+      "learning_rate": 6.906306540111528e-06,
+      "loss": 0.7001,
+      "step": 13707
+    },
+    {
+      "epoch": 0.7544719026913974,
+      "grad_norm": 0.6862730383872986,
+      "learning_rate": 6.9059058072291485e-06,
+      "loss": 0.7921,
+      "step": 13708
+    },
+    {
+      "epoch": 0.7545269414937531,
+      "grad_norm": 0.6684688925743103,
+      "learning_rate": 6.905505060022491e-06,
+      "loss": 0.6736,
+      "step": 13709
+    },
+    {
+      "epoch": 0.7545819802961088,
+      "grad_norm": 0.6581160426139832,
+      "learning_rate": 6.905104298494567e-06,
+      "loss": 0.7581,
+      "step": 13710
+    },
+    {
+      "epoch": 0.7546370190984644,
+      "grad_norm": 0.7772610783576965,
+      "learning_rate": 6.9047035226483885e-06,
+      "loss": 0.7984,
+      "step": 13711
+    },
+    {
+      "epoch": 0.7546920579008201,
+      "grad_norm": 0.6856822371482849,
+      "learning_rate": 6.90430273248697e-06,
+      "loss": 0.8232,
+      "step": 13712
+    },
+    {
+      "epoch": 0.7547470967031757,
+      "grad_norm": 0.7250725626945496,
+      "learning_rate": 6.903901928013322e-06,
+      "loss": 0.7844,
+      "step": 13713
+    },
+    {
+      "epoch": 0.7548021355055314,
+      "grad_norm": 0.7034164667129517,
+      "learning_rate": 6.9035011092304545e-06,
+      "loss": 0.8293,
+      "step": 13714
+    },
+    {
+      "epoch": 0.754857174307887,
+      "grad_norm": 0.6783095002174377,
+      "learning_rate": 6.903100276141383e-06,
+      "loss": 0.6841,
+      "step": 13715
+    },
+    {
+      "epoch": 0.7549122131102427,
+      "grad_norm": 0.6180121302604675,
+      "learning_rate": 6.90269942874912e-06,
+      "loss": 0.7111,
+      "step": 13716
+    },
+    {
+      "epoch": 0.7549672519125984,
+      "grad_norm": 0.70428466796875,
+      "learning_rate": 6.902298567056677e-06,
+      "loss": 0.8758,
+      "step": 13717
+    },
+    {
+      "epoch": 0.7550222907149541,
+      "grad_norm": 0.8130238652229309,
+      "learning_rate": 6.9018976910670665e-06,
+      "loss": 0.6443,
+      "step": 13718
+    },
+    {
+      "epoch": 0.7550773295173097,
+      "grad_norm": 0.6910800933837891,
+      "learning_rate": 6.901496800783302e-06,
+      "loss": 0.7231,
+      "step": 13719
+    },
+    {
+      "epoch": 0.7551323683196653,
+      "grad_norm": 0.700933575630188,
+      "learning_rate": 6.901095896208398e-06,
+      "loss": 0.6785,
+      "step": 13720
+    },
+    {
+      "epoch": 0.755187407122021,
+      "grad_norm": 0.7407829761505127,
+      "learning_rate": 6.9006949773453656e-06,
+      "loss": 0.694,
+      "step": 13721
+    },
+    {
+      "epoch": 0.7552424459243767,
+      "grad_norm": 0.7907935380935669,
+      "learning_rate": 6.900294044197218e-06,
+      "loss": 0.7674,
+      "step": 13722
+    },
+    {
+      "epoch": 0.7552974847267323,
+      "grad_norm": 0.6585111021995544,
+      "learning_rate": 6.89989309676697e-06,
+      "loss": 0.6785,
+      "step": 13723
+    },
+    {
+      "epoch": 0.755352523529088,
+      "grad_norm": 0.7611724138259888,
+      "learning_rate": 6.899492135057633e-06,
+      "loss": 0.8028,
+      "step": 13724
+    },
+    {
+      "epoch": 0.7554075623314437,
+      "grad_norm": 0.6412070989608765,
+      "learning_rate": 6.899091159072222e-06,
+      "loss": 0.7634,
+      "step": 13725
+    },
+    {
+      "epoch": 0.7554626011337994,
+      "grad_norm": 0.7712366580963135,
+      "learning_rate": 6.898690168813751e-06,
+      "loss": 0.8275,
+      "step": 13726
+    },
+    {
+      "epoch": 0.755517639936155,
+      "grad_norm": 0.6826579570770264,
+      "learning_rate": 6.898289164285232e-06,
+      "loss": 0.7949,
+      "step": 13727
+    },
+    {
+      "epoch": 0.7555726787385106,
+      "grad_norm": 0.7501955628395081,
+      "learning_rate": 6.897888145489681e-06,
+      "loss": 0.7846,
+      "step": 13728
+    },
+    {
+      "epoch": 0.7556277175408663,
+      "grad_norm": 0.6493077874183655,
+      "learning_rate": 6.8974871124301075e-06,
+      "loss": 0.7294,
+      "step": 13729
+    },
+    {
+      "epoch": 0.755682756343222,
+      "grad_norm": 0.6854347586631775,
+      "learning_rate": 6.897086065109532e-06,
+      "loss": 0.7121,
+      "step": 13730
+    },
+    {
+      "epoch": 0.7557377951455776,
+      "grad_norm": 0.7376317977905273,
+      "learning_rate": 6.896685003530964e-06,
+      "loss": 0.7719,
+      "step": 13731
+    },
+    {
+      "epoch": 0.7557928339479333,
+      "grad_norm": 0.8477175235748291,
+      "learning_rate": 6.89628392769742e-06,
+      "loss": 0.7981,
+      "step": 13732
+    },
+    {
+      "epoch": 0.755847872750289,
+      "grad_norm": 0.6611722111701965,
+      "learning_rate": 6.8958828376119125e-06,
+      "loss": 0.7628,
+      "step": 13733
+    },
+    {
+      "epoch": 0.7559029115526447,
+      "grad_norm": 0.6898290514945984,
+      "learning_rate": 6.895481733277458e-06,
+      "loss": 0.7578,
+      "step": 13734
+    },
+    {
+      "epoch": 0.7559579503550002,
+      "grad_norm": 0.6566810607910156,
+      "learning_rate": 6.89508061469707e-06,
+      "loss": 0.6919,
+      "step": 13735
+    },
+    {
+      "epoch": 0.7560129891573559,
+      "grad_norm": 0.6395933032035828,
+      "learning_rate": 6.894679481873763e-06,
+      "loss": 0.7334,
+      "step": 13736
+    },
+    {
+      "epoch": 0.7560680279597116,
+      "grad_norm": 0.7060876488685608,
+      "learning_rate": 6.8942783348105535e-06,
+      "loss": 0.7405,
+      "step": 13737
+    },
+    {
+      "epoch": 0.7561230667620673,
+      "grad_norm": 0.7303228974342346,
+      "learning_rate": 6.893877173510454e-06,
+      "loss": 0.8563,
+      "step": 13738
+    },
+    {
+      "epoch": 0.7561781055644229,
+      "grad_norm": 0.663474977016449,
+      "learning_rate": 6.893475997976481e-06,
+      "loss": 0.703,
+      "step": 13739
+    },
+    {
+      "epoch": 0.7562331443667786,
+      "grad_norm": 0.8005428910255432,
+      "learning_rate": 6.893074808211649e-06,
+      "loss": 0.7219,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7562881831691343,
+      "grad_norm": 1.3285688161849976,
+      "learning_rate": 6.892673604218972e-06,
+      "loss": 0.672,
+      "step": 13741
+    },
+    {
+      "epoch": 0.75634322197149,
+      "grad_norm": 0.6958948373794556,
+      "learning_rate": 6.892272386001469e-06,
+      "loss": 0.7728,
+      "step": 13742
+    },
+    {
+      "epoch": 0.7563982607738455,
+      "grad_norm": 0.6840598583221436,
+      "learning_rate": 6.891871153562153e-06,
+      "loss": 0.7881,
+      "step": 13743
+    },
+    {
+      "epoch": 0.7564532995762012,
+      "grad_norm": 0.7184257507324219,
+      "learning_rate": 6.891469906904039e-06,
+      "loss": 0.736,
+      "step": 13744
+    },
+    {
+      "epoch": 0.7565083383785569,
+      "grad_norm": 0.6611571311950684,
+      "learning_rate": 6.891068646030143e-06,
+      "loss": 0.7171,
+      "step": 13745
+    },
+    {
+      "epoch": 0.7565633771809125,
+      "grad_norm": 0.8237559795379639,
+      "learning_rate": 6.890667370943482e-06,
+      "loss": 0.8669,
+      "step": 13746
+    },
+    {
+      "epoch": 0.7566184159832682,
+      "grad_norm": 0.6898388266563416,
+      "learning_rate": 6.890266081647072e-06,
+      "loss": 0.6654,
+      "step": 13747
+    },
+    {
+      "epoch": 0.7566734547856239,
+      "grad_norm": 0.6541711688041687,
+      "learning_rate": 6.889864778143928e-06,
+      "loss": 0.7455,
+      "step": 13748
+    },
+    {
+      "epoch": 0.7567284935879796,
+      "grad_norm": 0.6518157124519348,
+      "learning_rate": 6.8894634604370655e-06,
+      "loss": 0.7174,
+      "step": 13749
+    },
+    {
+      "epoch": 0.7567835323903351,
+      "grad_norm": 0.7992080450057983,
+      "learning_rate": 6.889062128529502e-06,
+      "loss": 0.7349,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7568385711926908,
+      "grad_norm": 0.5748338103294373,
+      "learning_rate": 6.888660782424253e-06,
+      "loss": 0.5398,
+      "step": 13751
+    },
+    {
+      "epoch": 0.7568936099950465,
+      "grad_norm": 0.6507781744003296,
+      "learning_rate": 6.8882594221243344e-06,
+      "loss": 0.6762,
+      "step": 13752
+    },
+    {
+      "epoch": 0.7569486487974022,
+      "grad_norm": 0.6908432841300964,
+      "learning_rate": 6.887858047632764e-06,
+      "loss": 0.8034,
+      "step": 13753
+    },
+    {
+      "epoch": 0.7570036875997578,
+      "grad_norm": 0.6497751474380493,
+      "learning_rate": 6.887456658952557e-06,
+      "loss": 0.6351,
+      "step": 13754
+    },
+    {
+      "epoch": 0.7570587264021135,
+      "grad_norm": 0.7233273386955261,
+      "learning_rate": 6.887055256086732e-06,
+      "loss": 0.7096,
+      "step": 13755
+    },
+    {
+      "epoch": 0.7571137652044692,
+      "grad_norm": 0.6587454676628113,
+      "learning_rate": 6.886653839038305e-06,
+      "loss": 0.7354,
+      "step": 13756
+    },
+    {
+      "epoch": 0.7571688040068248,
+      "grad_norm": 0.6654310822486877,
+      "learning_rate": 6.886252407810292e-06,
+      "loss": 0.7776,
+      "step": 13757
+    },
+    {
+      "epoch": 0.7572238428091804,
+      "grad_norm": 0.796604573726654,
+      "learning_rate": 6.885850962405711e-06,
+      "loss": 0.7925,
+      "step": 13758
+    },
+    {
+      "epoch": 0.7572788816115361,
+      "grad_norm": 0.7053457498550415,
+      "learning_rate": 6.8854495028275795e-06,
+      "loss": 0.7893,
+      "step": 13759
+    },
+    {
+      "epoch": 0.7573339204138918,
+      "grad_norm": 0.7201200127601624,
+      "learning_rate": 6.885048029078914e-06,
+      "loss": 0.8346,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7573889592162475,
+      "grad_norm": 0.8437653183937073,
+      "learning_rate": 6.884646541162731e-06,
+      "loss": 0.7468,
+      "step": 13761
+    },
+    {
+      "epoch": 0.7574439980186031,
+      "grad_norm": 0.6910028457641602,
+      "learning_rate": 6.884245039082052e-06,
+      "loss": 0.7362,
+      "step": 13762
+    },
+    {
+      "epoch": 0.7574990368209588,
+      "grad_norm": 0.6896274089813232,
+      "learning_rate": 6.883843522839889e-06,
+      "loss": 0.6515,
+      "step": 13763
+    },
+    {
+      "epoch": 0.7575540756233144,
+      "grad_norm": 0.9833560585975647,
+      "learning_rate": 6.8834419924392636e-06,
+      "loss": 0.8764,
+      "step": 13764
+    },
+    {
+      "epoch": 0.7576091144256701,
+      "grad_norm": 0.7130032181739807,
+      "learning_rate": 6.88304044788319e-06,
+      "loss": 0.7631,
+      "step": 13765
+    },
+    {
+      "epoch": 0.7576641532280257,
+      "grad_norm": 0.7059195041656494,
+      "learning_rate": 6.882638889174691e-06,
+      "loss": 0.8147,
+      "step": 13766
+    },
+    {
+      "epoch": 0.7577191920303814,
+      "grad_norm": 0.6451989412307739,
+      "learning_rate": 6.882237316316781e-06,
+      "loss": 0.6638,
+      "step": 13767
+    },
+    {
+      "epoch": 0.7577742308327371,
+      "grad_norm": 0.7541074752807617,
+      "learning_rate": 6.881835729312481e-06,
+      "loss": 0.6918,
+      "step": 13768
+    },
+    {
+      "epoch": 0.7578292696350928,
+      "grad_norm": 0.7227535843849182,
+      "learning_rate": 6.881434128164805e-06,
+      "loss": 0.7759,
+      "step": 13769
+    },
+    {
+      "epoch": 0.7578843084374484,
+      "grad_norm": 0.673112154006958,
+      "learning_rate": 6.881032512876774e-06,
+      "loss": 0.7328,
+      "step": 13770
+    },
+    {
+      "epoch": 0.757939347239804,
+      "grad_norm": 0.6536681056022644,
+      "learning_rate": 6.880630883451407e-06,
+      "loss": 0.7677,
+      "step": 13771
+    },
+    {
+      "epoch": 0.7579943860421597,
+      "grad_norm": 0.8517894148826599,
+      "learning_rate": 6.880229239891721e-06,
+      "loss": 0.8566,
+      "step": 13772
+    },
+    {
+      "epoch": 0.7580494248445154,
+      "grad_norm": 0.8260573148727417,
+      "learning_rate": 6.879827582200737e-06,
+      "loss": 0.8228,
+      "step": 13773
+    },
+    {
+      "epoch": 0.758104463646871,
+      "grad_norm": 0.7460072040557861,
+      "learning_rate": 6.87942591038147e-06,
+      "loss": 0.8047,
+      "step": 13774
+    },
+    {
+      "epoch": 0.7581595024492267,
+      "grad_norm": 0.7648436427116394,
+      "learning_rate": 6.879024224436942e-06,
+      "loss": 0.852,
+      "step": 13775
+    },
+    {
+      "epoch": 0.7582145412515824,
+      "grad_norm": 0.7161253094673157,
+      "learning_rate": 6.878622524370171e-06,
+      "loss": 0.7638,
+      "step": 13776
+    },
+    {
+      "epoch": 0.7582695800539381,
+      "grad_norm": 0.6559579372406006,
+      "learning_rate": 6.878220810184175e-06,
+      "loss": 0.6932,
+      "step": 13777
+    },
+    {
+      "epoch": 0.7583246188562937,
+      "grad_norm": 0.6846898198127747,
+      "learning_rate": 6.877819081881975e-06,
+      "loss": 0.7098,
+      "step": 13778
+    },
+    {
+      "epoch": 0.7583796576586493,
+      "grad_norm": 0.7569675445556641,
+      "learning_rate": 6.87741733946659e-06,
+      "loss": 0.687,
+      "step": 13779
+    },
+    {
+      "epoch": 0.758434696461005,
+      "grad_norm": 0.7513766288757324,
+      "learning_rate": 6.877015582941038e-06,
+      "loss": 0.8673,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7584897352633607,
+      "grad_norm": 0.7158082127571106,
+      "learning_rate": 6.876613812308338e-06,
+      "loss": 0.7563,
+      "step": 13781
+    },
+    {
+      "epoch": 0.7585447740657163,
+      "grad_norm": 0.6307277083396912,
+      "learning_rate": 6.876212027571513e-06,
+      "loss": 0.6725,
+      "step": 13782
+    },
+    {
+      "epoch": 0.758599812868072,
+      "grad_norm": 0.735090434551239,
+      "learning_rate": 6.87581022873358e-06,
+      "loss": 0.763,
+      "step": 13783
+    },
+    {
+      "epoch": 0.7586548516704277,
+      "grad_norm": 0.6412403583526611,
+      "learning_rate": 6.8754084157975594e-06,
+      "loss": 0.5992,
+      "step": 13784
+    },
+    {
+      "epoch": 0.7587098904727834,
+      "grad_norm": 0.639854907989502,
+      "learning_rate": 6.875006588766472e-06,
+      "loss": 0.7372,
+      "step": 13785
+    },
+    {
+      "epoch": 0.7587649292751389,
+      "grad_norm": 0.6855082511901855,
+      "learning_rate": 6.8746047476433365e-06,
+      "loss": 0.7709,
+      "step": 13786
+    },
+    {
+      "epoch": 0.7588199680774946,
+      "grad_norm": 0.6838769912719727,
+      "learning_rate": 6.874202892431173e-06,
+      "loss": 0.7545,
+      "step": 13787
+    },
+    {
+      "epoch": 0.7588750068798503,
+      "grad_norm": 1.1560181379318237,
+      "learning_rate": 6.873801023133002e-06,
+      "loss": 0.7291,
+      "step": 13788
+    },
+    {
+      "epoch": 0.7589300456822059,
+      "grad_norm": 0.7140469551086426,
+      "learning_rate": 6.873399139751844e-06,
+      "loss": 0.7214,
+      "step": 13789
+    },
+    {
+      "epoch": 0.7589850844845616,
+      "grad_norm": 0.6856355667114258,
+      "learning_rate": 6.8729972422907195e-06,
+      "loss": 0.7417,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7590401232869173,
+      "grad_norm": 0.7856155633926392,
+      "learning_rate": 6.8725953307526505e-06,
+      "loss": 0.7484,
+      "step": 13791
+    },
+    {
+      "epoch": 0.759095162089273,
+      "grad_norm": 0.8107255697250366,
+      "learning_rate": 6.8721934051406555e-06,
+      "loss": 0.7568,
+      "step": 13792
+    },
+    {
+      "epoch": 0.7591502008916285,
+      "grad_norm": 0.6590837240219116,
+      "learning_rate": 6.871791465457757e-06,
+      "loss": 0.7495,
+      "step": 13793
+    },
+    {
+      "epoch": 0.7592052396939842,
+      "grad_norm": 0.7531588077545166,
+      "learning_rate": 6.8713895117069715e-06,
+      "loss": 0.7434,
+      "step": 13794
+    },
+    {
+      "epoch": 0.7592602784963399,
+      "grad_norm": 0.6818329095840454,
+      "learning_rate": 6.870987543891326e-06,
+      "loss": 0.7128,
+      "step": 13795
+    },
+    {
+      "epoch": 0.7593153172986956,
+      "grad_norm": 0.6082884669303894,
+      "learning_rate": 6.8705855620138395e-06,
+      "loss": 0.7437,
+      "step": 13796
+    },
+    {
+      "epoch": 0.7593703561010512,
+      "grad_norm": 0.9583787322044373,
+      "learning_rate": 6.870183566077532e-06,
+      "loss": 0.7779,
+      "step": 13797
+    },
+    {
+      "epoch": 0.7594253949034069,
+      "grad_norm": 0.6684621572494507,
+      "learning_rate": 6.869781556085425e-06,
+      "loss": 0.5856,
+      "step": 13798
+    },
+    {
+      "epoch": 0.7594804337057626,
+      "grad_norm": 0.6225603222846985,
+      "learning_rate": 6.869379532040541e-06,
+      "loss": 0.7407,
+      "step": 13799
+    },
+    {
+      "epoch": 0.7595354725081183,
+      "grad_norm": 0.6973103284835815,
+      "learning_rate": 6.8689774939459005e-06,
+      "loss": 0.7789,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7595905113104738,
+      "grad_norm": 0.6655399203300476,
+      "learning_rate": 6.868575441804526e-06,
+      "loss": 0.7489,
+      "step": 13801
+    },
+    {
+      "epoch": 0.7596455501128295,
+      "grad_norm": 0.7066664695739746,
+      "learning_rate": 6.868173375619437e-06,
+      "loss": 0.7035,
+      "step": 13802
+    },
+    {
+      "epoch": 0.7597005889151852,
+      "grad_norm": 1.0646852254867554,
+      "learning_rate": 6.867771295393658e-06,
+      "loss": 0.8488,
+      "step": 13803
+    },
+    {
+      "epoch": 0.7597556277175409,
+      "grad_norm": 0.6551353335380554,
+      "learning_rate": 6.867369201130209e-06,
+      "loss": 0.7147,
+      "step": 13804
+    },
+    {
+      "epoch": 0.7598106665198965,
+      "grad_norm": 0.6749850511550903,
+      "learning_rate": 6.866967092832115e-06,
+      "loss": 0.7963,
+      "step": 13805
+    },
+    {
+      "epoch": 0.7598657053222522,
+      "grad_norm": 0.6704042553901672,
+      "learning_rate": 6.866564970502394e-06,
+      "loss": 0.7992,
+      "step": 13806
+    },
+    {
+      "epoch": 0.7599207441246079,
+      "grad_norm": 0.7027791142463684,
+      "learning_rate": 6.866162834144071e-06,
+      "loss": 0.7931,
+      "step": 13807
+    },
+    {
+      "epoch": 0.7599757829269636,
+      "grad_norm": 0.7925322651863098,
+      "learning_rate": 6.865760683760169e-06,
+      "loss": 0.7826,
+      "step": 13808
+    },
+    {
+      "epoch": 0.7600308217293191,
+      "grad_norm": 0.7152161002159119,
+      "learning_rate": 6.865358519353708e-06,
+      "loss": 0.7481,
+      "step": 13809
+    },
+    {
+      "epoch": 0.7600858605316748,
+      "grad_norm": 0.6572757959365845,
+      "learning_rate": 6.864956340927711e-06,
+      "loss": 0.785,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7601408993340305,
+      "grad_norm": 0.6848406791687012,
+      "learning_rate": 6.864554148485203e-06,
+      "loss": 0.6423,
+      "step": 13811
+    },
+    {
+      "epoch": 0.7601959381363862,
+      "grad_norm": 0.747597873210907,
+      "learning_rate": 6.864151942029205e-06,
+      "loss": 0.7901,
+      "step": 13812
+    },
+    {
+      "epoch": 0.7602509769387418,
+      "grad_norm": 0.7106720805168152,
+      "learning_rate": 6.863749721562738e-06,
+      "loss": 0.7488,
+      "step": 13813
+    },
+    {
+      "epoch": 0.7603060157410975,
+      "grad_norm": 0.6864057779312134,
+      "learning_rate": 6.8633474870888275e-06,
+      "loss": 0.7066,
+      "step": 13814
+    },
+    {
+      "epoch": 0.7603610545434532,
+      "grad_norm": 0.7022056579589844,
+      "learning_rate": 6.862945238610496e-06,
+      "loss": 0.6851,
+      "step": 13815
+    },
+    {
+      "epoch": 0.7604160933458088,
+      "grad_norm": 0.7361913919448853,
+      "learning_rate": 6.862542976130769e-06,
+      "loss": 0.7425,
+      "step": 13816
+    },
+    {
+      "epoch": 0.7604711321481644,
+      "grad_norm": 0.6723676323890686,
+      "learning_rate": 6.862140699652666e-06,
+      "loss": 0.7937,
+      "step": 13817
+    },
+    {
+      "epoch": 0.7605261709505201,
+      "grad_norm": 0.7491924166679382,
+      "learning_rate": 6.861738409179212e-06,
+      "loss": 0.7585,
+      "step": 13818
+    },
+    {
+      "epoch": 0.7605812097528758,
+      "grad_norm": 0.6772211790084839,
+      "learning_rate": 6.86133610471343e-06,
+      "loss": 0.7617,
+      "step": 13819
+    },
+    {
+      "epoch": 0.7606362485552315,
+      "grad_norm": 0.7819864153862,
+      "learning_rate": 6.860933786258344e-06,
+      "loss": 0.7924,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7606912873575871,
+      "grad_norm": 0.6992526650428772,
+      "learning_rate": 6.86053145381698e-06,
+      "loss": 0.7054,
+      "step": 13821
+    },
+    {
+      "epoch": 0.7607463261599428,
+      "grad_norm": 0.7189231514930725,
+      "learning_rate": 6.860129107392357e-06,
+      "loss": 0.7603,
+      "step": 13822
+    },
+    {
+      "epoch": 0.7608013649622984,
+      "grad_norm": 0.7165294885635376,
+      "learning_rate": 6.859726746987503e-06,
+      "loss": 0.8118,
+      "step": 13823
+    },
+    {
+      "epoch": 0.7608564037646541,
+      "grad_norm": 0.6510334014892578,
+      "learning_rate": 6.85932437260544e-06,
+      "loss": 0.7584,
+      "step": 13824
+    },
+    {
+      "epoch": 0.7609114425670097,
+      "grad_norm": 0.7113379836082458,
+      "learning_rate": 6.8589219842491935e-06,
+      "loss": 0.7799,
+      "step": 13825
+    },
+    {
+      "epoch": 0.7609664813693654,
+      "grad_norm": 0.7441100478172302,
+      "learning_rate": 6.8585195819217856e-06,
+      "loss": 0.6468,
+      "step": 13826
+    },
+    {
+      "epoch": 0.7610215201717211,
+      "grad_norm": 1.0703508853912354,
+      "learning_rate": 6.858117165626244e-06,
+      "loss": 0.7922,
+      "step": 13827
+    },
+    {
+      "epoch": 0.7610765589740768,
+      "grad_norm": 0.7097275853157043,
+      "learning_rate": 6.857714735365589e-06,
+      "loss": 0.7594,
+      "step": 13828
+    },
+    {
+      "epoch": 0.7611315977764324,
+      "grad_norm": 0.7001124620437622,
+      "learning_rate": 6.857312291142848e-06,
+      "loss": 0.7679,
+      "step": 13829
+    },
+    {
+      "epoch": 0.761186636578788,
+      "grad_norm": 0.6898123621940613,
+      "learning_rate": 6.856909832961045e-06,
+      "loss": 0.7684,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7612416753811437,
+      "grad_norm": 0.6535243391990662,
+      "learning_rate": 6.856507360823206e-06,
+      "loss": 0.6143,
+      "step": 13831
+    },
+    {
+      "epoch": 0.7612967141834993,
+      "grad_norm": 0.6726056933403015,
+      "learning_rate": 6.856104874732353e-06,
+      "loss": 0.7566,
+      "step": 13832
+    },
+    {
+      "epoch": 0.761351752985855,
+      "grad_norm": 0.8741437196731567,
+      "learning_rate": 6.855702374691513e-06,
+      "loss": 0.723,
+      "step": 13833
+    },
+    {
+      "epoch": 0.7614067917882107,
+      "grad_norm": 0.7025718092918396,
+      "learning_rate": 6.855299860703712e-06,
+      "loss": 0.8035,
+      "step": 13834
+    },
+    {
+      "epoch": 0.7614618305905664,
+      "grad_norm": 1.08286452293396,
+      "learning_rate": 6.8548973327719726e-06,
+      "loss": 0.7347,
+      "step": 13835
+    },
+    {
+      "epoch": 0.761516869392922,
+      "grad_norm": 0.6483243107795715,
+      "learning_rate": 6.854494790899322e-06,
+      "loss": 0.7326,
+      "step": 13836
+    },
+    {
+      "epoch": 0.7615719081952776,
+      "grad_norm": 0.6611089110374451,
+      "learning_rate": 6.854092235088784e-06,
+      "loss": 0.7619,
+      "step": 13837
+    },
+    {
+      "epoch": 0.7616269469976333,
+      "grad_norm": 0.8394322991371155,
+      "learning_rate": 6.853689665343385e-06,
+      "loss": 0.7017,
+      "step": 13838
+    },
+    {
+      "epoch": 0.761681985799989,
+      "grad_norm": 0.7131583094596863,
+      "learning_rate": 6.853287081666151e-06,
+      "loss": 0.7367,
+      "step": 13839
+    },
+    {
+      "epoch": 0.7617370246023446,
+      "grad_norm": 0.7316367626190186,
+      "learning_rate": 6.852884484060108e-06,
+      "loss": 0.7323,
+      "step": 13840
+    },
+    {
+      "epoch": 0.7617920634047003,
+      "grad_norm": 0.7639010548591614,
+      "learning_rate": 6.852481872528281e-06,
+      "loss": 0.819,
+      "step": 13841
+    },
+    {
+      "epoch": 0.761847102207056,
+      "grad_norm": 0.7118390202522278,
+      "learning_rate": 6.852079247073695e-06,
+      "loss": 0.7645,
+      "step": 13842
+    },
+    {
+      "epoch": 0.7619021410094117,
+      "grad_norm": 0.6885393857955933,
+      "learning_rate": 6.851676607699379e-06,
+      "loss": 0.8052,
+      "step": 13843
+    },
+    {
+      "epoch": 0.7619571798117672,
+      "grad_norm": 0.7034374475479126,
+      "learning_rate": 6.851273954408356e-06,
+      "loss": 0.8464,
+      "step": 13844
+    },
+    {
+      "epoch": 0.7620122186141229,
+      "grad_norm": 0.6531803607940674,
+      "learning_rate": 6.850871287203654e-06,
+      "loss": 0.7871,
+      "step": 13845
+    },
+    {
+      "epoch": 0.7620672574164786,
+      "grad_norm": 0.6637283563613892,
+      "learning_rate": 6.8504686060882995e-06,
+      "loss": 0.7326,
+      "step": 13846
+    },
+    {
+      "epoch": 0.7621222962188343,
+      "grad_norm": 0.6467694640159607,
+      "learning_rate": 6.850065911065318e-06,
+      "loss": 0.7936,
+      "step": 13847
+    },
+    {
+      "epoch": 0.7621773350211899,
+      "grad_norm": 0.6829109191894531,
+      "learning_rate": 6.849663202137735e-06,
+      "loss": 0.7003,
+      "step": 13848
+    },
+    {
+      "epoch": 0.7622323738235456,
+      "grad_norm": 0.7321386933326721,
+      "learning_rate": 6.84926047930858e-06,
+      "loss": 0.6921,
+      "step": 13849
+    },
+    {
+      "epoch": 0.7622874126259013,
+      "grad_norm": 0.6900202631950378,
+      "learning_rate": 6.8488577425808766e-06,
+      "loss": 0.7496,
+      "step": 13850
+    },
+    {
+      "epoch": 0.762342451428257,
+      "grad_norm": 0.6304247975349426,
+      "learning_rate": 6.848454991957655e-06,
+      "loss": 0.7135,
+      "step": 13851
+    },
+    {
+      "epoch": 0.7623974902306125,
+      "grad_norm": 0.7087798118591309,
+      "learning_rate": 6.8480522274419404e-06,
+      "loss": 0.7032,
+      "step": 13852
+    },
+    {
+      "epoch": 0.7624525290329682,
+      "grad_norm": 0.7777289152145386,
+      "learning_rate": 6.84764944903676e-06,
+      "loss": 0.7345,
+      "step": 13853
+    },
+    {
+      "epoch": 0.7625075678353239,
+      "grad_norm": 0.7282242774963379,
+      "learning_rate": 6.847246656745139e-06,
+      "loss": 0.6408,
+      "step": 13854
+    },
+    {
+      "epoch": 0.7625626066376796,
+      "grad_norm": 0.7798221707344055,
+      "learning_rate": 6.846843850570107e-06,
+      "loss": 0.9058,
+      "step": 13855
+    },
+    {
+      "epoch": 0.7626176454400352,
+      "grad_norm": 0.6145210266113281,
+      "learning_rate": 6.846441030514692e-06,
+      "loss": 0.6331,
+      "step": 13856
+    },
+    {
+      "epoch": 0.7626726842423909,
+      "grad_norm": 0.7079364061355591,
+      "learning_rate": 6.846038196581921e-06,
+      "loss": 0.7511,
+      "step": 13857
+    },
+    {
+      "epoch": 0.7627277230447466,
+      "grad_norm": 0.733635425567627,
+      "learning_rate": 6.845635348774821e-06,
+      "loss": 0.6957,
+      "step": 13858
+    },
+    {
+      "epoch": 0.7627827618471023,
+      "grad_norm": 0.8099489808082581,
+      "learning_rate": 6.845232487096419e-06,
+      "loss": 0.8068,
+      "step": 13859
+    },
+    {
+      "epoch": 0.7628378006494578,
+      "grad_norm": 0.6241937875747681,
+      "learning_rate": 6.844829611549744e-06,
+      "loss": 0.7102,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7628928394518135,
+      "grad_norm": 0.8009611368179321,
+      "learning_rate": 6.8444267221378235e-06,
+      "loss": 0.8369,
+      "step": 13861
+    },
+    {
+      "epoch": 0.7629478782541692,
+      "grad_norm": 0.6700903177261353,
+      "learning_rate": 6.844023818863685e-06,
+      "loss": 0.8075,
+      "step": 13862
+    },
+    {
+      "epoch": 0.7630029170565249,
+      "grad_norm": 0.9378371834754944,
+      "learning_rate": 6.843620901730357e-06,
+      "loss": 0.7539,
+      "step": 13863
+    },
+    {
+      "epoch": 0.7630579558588805,
+      "grad_norm": 0.6704423427581787,
+      "learning_rate": 6.843217970740867e-06,
+      "loss": 0.7285,
+      "step": 13864
+    },
+    {
+      "epoch": 0.7631129946612362,
+      "grad_norm": 0.7236818075180054,
+      "learning_rate": 6.842815025898246e-06,
+      "loss": 0.7223,
+      "step": 13865
+    },
+    {
+      "epoch": 0.7631680334635919,
+      "grad_norm": 0.676184356212616,
+      "learning_rate": 6.84241206720552e-06,
+      "loss": 0.7286,
+      "step": 13866
+    },
+    {
+      "epoch": 0.7632230722659475,
+      "grad_norm": 0.6443304419517517,
+      "learning_rate": 6.842009094665717e-06,
+      "loss": 0.6806,
+      "step": 13867
+    },
+    {
+      "epoch": 0.7632781110683031,
+      "grad_norm": 0.7931790947914124,
+      "learning_rate": 6.841606108281868e-06,
+      "loss": 0.7801,
+      "step": 13868
+    },
+    {
+      "epoch": 0.7633331498706588,
+      "grad_norm": 0.7440798878669739,
+      "learning_rate": 6.841203108057e-06,
+      "loss": 0.8044,
+      "step": 13869
+    },
+    {
+      "epoch": 0.7633881886730145,
+      "grad_norm": 0.7226675748825073,
+      "learning_rate": 6.840800093994142e-06,
+      "loss": 0.718,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7634432274753702,
+      "grad_norm": 0.7351265549659729,
+      "learning_rate": 6.8403970660963245e-06,
+      "loss": 0.8389,
+      "step": 13871
+    },
+    {
+      "epoch": 0.7634982662777258,
+      "grad_norm": 0.8326215744018555,
+      "learning_rate": 6.839994024366574e-06,
+      "loss": 0.8583,
+      "step": 13872
+    },
+    {
+      "epoch": 0.7635533050800815,
+      "grad_norm": 0.6841259002685547,
+      "learning_rate": 6.839590968807922e-06,
+      "loss": 0.7553,
+      "step": 13873
+    },
+    {
+      "epoch": 0.7636083438824371,
+      "grad_norm": 0.7305078506469727,
+      "learning_rate": 6.839187899423395e-06,
+      "loss": 0.7825,
+      "step": 13874
+    },
+    {
+      "epoch": 0.7636633826847927,
+      "grad_norm": 0.7235193252563477,
+      "learning_rate": 6.838784816216025e-06,
+      "loss": 0.7653,
+      "step": 13875
+    },
+    {
+      "epoch": 0.7637184214871484,
+      "grad_norm": 0.6468761563301086,
+      "learning_rate": 6.838381719188842e-06,
+      "loss": 0.6901,
+      "step": 13876
+    },
+    {
+      "epoch": 0.7637734602895041,
+      "grad_norm": 0.6806310415267944,
+      "learning_rate": 6.837978608344872e-06,
+      "loss": 0.6876,
+      "step": 13877
+    },
+    {
+      "epoch": 0.7638284990918598,
+      "grad_norm": 0.692081093788147,
+      "learning_rate": 6.837575483687147e-06,
+      "loss": 0.7506,
+      "step": 13878
+    },
+    {
+      "epoch": 0.7638835378942154,
+      "grad_norm": 0.6447135806083679,
+      "learning_rate": 6.837172345218697e-06,
+      "loss": 0.6841,
+      "step": 13879
+    },
+    {
+      "epoch": 0.7639385766965711,
+      "grad_norm": 0.7352014183998108,
+      "learning_rate": 6.8367691929425516e-06,
+      "loss": 0.8066,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7639936154989267,
+      "grad_norm": 0.7305072546005249,
+      "learning_rate": 6.8363660268617405e-06,
+      "loss": 0.717,
+      "step": 13881
+    },
+    {
+      "epoch": 0.7640486543012824,
+      "grad_norm": 0.6580411195755005,
+      "learning_rate": 6.835962846979294e-06,
+      "loss": 0.7585,
+      "step": 13882
+    },
+    {
+      "epoch": 0.764103693103638,
+      "grad_norm": 0.7568425536155701,
+      "learning_rate": 6.835559653298242e-06,
+      "loss": 0.8273,
+      "step": 13883
+    },
+    {
+      "epoch": 0.7641587319059937,
+      "grad_norm": 0.8121107816696167,
+      "learning_rate": 6.835156445821616e-06,
+      "loss": 0.9064,
+      "step": 13884
+    },
+    {
+      "epoch": 0.7642137707083494,
+      "grad_norm": 0.6522091031074524,
+      "learning_rate": 6.834753224552444e-06,
+      "loss": 0.767,
+      "step": 13885
+    },
+    {
+      "epoch": 0.7642688095107051,
+      "grad_norm": 1.0779389142990112,
+      "learning_rate": 6.8343499894937574e-06,
+      "loss": 0.7702,
+      "step": 13886
+    },
+    {
+      "epoch": 0.7643238483130607,
+      "grad_norm": 0.6902838349342346,
+      "learning_rate": 6.833946740648588e-06,
+      "loss": 0.6529,
+      "step": 13887
+    },
+    {
+      "epoch": 0.7643788871154164,
+      "grad_norm": 0.692480742931366,
+      "learning_rate": 6.833543478019966e-06,
+      "loss": 0.7404,
+      "step": 13888
+    },
+    {
+      "epoch": 0.764433925917772,
+      "grad_norm": 0.633627712726593,
+      "learning_rate": 6.833140201610923e-06,
+      "loss": 0.711,
+      "step": 13889
+    },
+    {
+      "epoch": 0.7644889647201277,
+      "grad_norm": 0.8653294444084167,
+      "learning_rate": 6.832736911424487e-06,
+      "loss": 0.8102,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7645440035224833,
+      "grad_norm": 0.7864197492599487,
+      "learning_rate": 6.832333607463692e-06,
+      "loss": 0.7064,
+      "step": 13891
+    },
+    {
+      "epoch": 0.764599042324839,
+      "grad_norm": 0.6703711748123169,
+      "learning_rate": 6.831930289731569e-06,
+      "loss": 0.7653,
+      "step": 13892
+    },
+    {
+      "epoch": 0.7646540811271947,
+      "grad_norm": 0.7420178651809692,
+      "learning_rate": 6.831526958231147e-06,
+      "loss": 0.8137,
+      "step": 13893
+    },
+    {
+      "epoch": 0.7647091199295504,
+      "grad_norm": 0.7372543215751648,
+      "learning_rate": 6.831123612965459e-06,
+      "loss": 0.6871,
+      "step": 13894
+    },
+    {
+      "epoch": 0.764764158731906,
+      "grad_norm": 0.77486652135849,
+      "learning_rate": 6.830720253937536e-06,
+      "loss": 0.727,
+      "step": 13895
+    },
+    {
+      "epoch": 0.7648191975342616,
+      "grad_norm": 0.7087406516075134,
+      "learning_rate": 6.83031688115041e-06,
+      "loss": 0.7743,
+      "step": 13896
+    },
+    {
+      "epoch": 0.7648742363366173,
+      "grad_norm": 0.8415336608886719,
+      "learning_rate": 6.829913494607112e-06,
+      "loss": 0.774,
+      "step": 13897
+    },
+    {
+      "epoch": 0.764929275138973,
+      "grad_norm": 0.7736749053001404,
+      "learning_rate": 6.829510094310674e-06,
+      "loss": 0.7541,
+      "step": 13898
+    },
+    {
+      "epoch": 0.7649843139413286,
+      "grad_norm": 0.6749987602233887,
+      "learning_rate": 6.829106680264128e-06,
+      "loss": 0.7139,
+      "step": 13899
+    },
+    {
+      "epoch": 0.7650393527436843,
+      "grad_norm": 0.7079635262489319,
+      "learning_rate": 6.8287032524705055e-06,
+      "loss": 0.75,
+      "step": 13900
+    },
+    {
+      "epoch": 0.76509439154604,
+      "grad_norm": 0.6906388401985168,
+      "learning_rate": 6.828299810932839e-06,
+      "loss": 0.6895,
+      "step": 13901
+    },
+    {
+      "epoch": 0.7651494303483957,
+      "grad_norm": 0.7045881152153015,
+      "learning_rate": 6.82789635565416e-06,
+      "loss": 0.8728,
+      "step": 13902
+    },
+    {
+      "epoch": 0.7652044691507512,
+      "grad_norm": 0.6836426258087158,
+      "learning_rate": 6.827492886637501e-06,
+      "loss": 0.7315,
+      "step": 13903
+    },
+    {
+      "epoch": 0.7652595079531069,
+      "grad_norm": 0.6467520594596863,
+      "learning_rate": 6.827089403885896e-06,
+      "loss": 0.7556,
+      "step": 13904
+    },
+    {
+      "epoch": 0.7653145467554626,
+      "grad_norm": 0.7118285894393921,
+      "learning_rate": 6.826685907402376e-06,
+      "loss": 0.8686,
+      "step": 13905
+    },
+    {
+      "epoch": 0.7653695855578183,
+      "grad_norm": 0.6093236207962036,
+      "learning_rate": 6.826282397189974e-06,
+      "loss": 0.7066,
+      "step": 13906
+    },
+    {
+      "epoch": 0.7654246243601739,
+      "grad_norm": 0.6839649677276611,
+      "learning_rate": 6.825878873251721e-06,
+      "loss": 0.7025,
+      "step": 13907
+    },
+    {
+      "epoch": 0.7654796631625296,
+      "grad_norm": 0.7582715153694153,
+      "learning_rate": 6.825475335590652e-06,
+      "loss": 0.7301,
+      "step": 13908
+    },
+    {
+      "epoch": 0.7655347019648853,
+      "grad_norm": 0.6580978631973267,
+      "learning_rate": 6.8250717842098e-06,
+      "loss": 0.6771,
+      "step": 13909
+    },
+    {
+      "epoch": 0.765589740767241,
+      "grad_norm": 0.6754937171936035,
+      "learning_rate": 6.824668219112195e-06,
+      "loss": 0.7446,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7656447795695965,
+      "grad_norm": 0.7541018724441528,
+      "learning_rate": 6.8242646403008725e-06,
+      "loss": 0.802,
+      "step": 13911
+    },
+    {
+      "epoch": 0.7656998183719522,
+      "grad_norm": 0.6714808344841003,
+      "learning_rate": 6.823861047778866e-06,
+      "loss": 0.7334,
+      "step": 13912
+    },
+    {
+      "epoch": 0.7657548571743079,
+      "grad_norm": 0.6972425580024719,
+      "learning_rate": 6.823457441549209e-06,
+      "loss": 0.7859,
+      "step": 13913
+    },
+    {
+      "epoch": 0.7658098959766636,
+      "grad_norm": 0.6660878658294678,
+      "learning_rate": 6.823053821614931e-06,
+      "loss": 0.6594,
+      "step": 13914
+    },
+    {
+      "epoch": 0.7658649347790192,
+      "grad_norm": 0.7392181158065796,
+      "learning_rate": 6.82265018797907e-06,
+      "loss": 0.6667,
+      "step": 13915
+    },
+    {
+      "epoch": 0.7659199735813749,
+      "grad_norm": 0.7601449489593506,
+      "learning_rate": 6.822246540644659e-06,
+      "loss": 0.7349,
+      "step": 13916
+    },
+    {
+      "epoch": 0.7659750123837306,
+      "grad_norm": 0.6648421287536621,
+      "learning_rate": 6.821842879614731e-06,
+      "loss": 0.7597,
+      "step": 13917
+    },
+    {
+      "epoch": 0.7660300511860861,
+      "grad_norm": 0.6369950175285339,
+      "learning_rate": 6.821439204892317e-06,
+      "loss": 0.7452,
+      "step": 13918
+    },
+    {
+      "epoch": 0.7660850899884418,
+      "grad_norm": 0.747653603553772,
+      "learning_rate": 6.821035516480457e-06,
+      "loss": 0.693,
+      "step": 13919
+    },
+    {
+      "epoch": 0.7661401287907975,
+      "grad_norm": 0.6450137495994568,
+      "learning_rate": 6.8206318143821795e-06,
+      "loss": 0.6492,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7661951675931532,
+      "grad_norm": 0.707801878452301,
+      "learning_rate": 6.8202280986005205e-06,
+      "loss": 0.7284,
+      "step": 13921
+    },
+    {
+      "epoch": 0.7662502063955088,
+      "grad_norm": 0.7191962003707886,
+      "learning_rate": 6.8198243691385146e-06,
+      "loss": 0.7714,
+      "step": 13922
+    },
+    {
+      "epoch": 0.7663052451978645,
+      "grad_norm": 0.7477172613143921,
+      "learning_rate": 6.819420625999196e-06,
+      "loss": 0.7076,
+      "step": 13923
+    },
+    {
+      "epoch": 0.7663602840002202,
+      "grad_norm": 0.6221175193786621,
+      "learning_rate": 6.819016869185599e-06,
+      "loss": 0.6848,
+      "step": 13924
+    },
+    {
+      "epoch": 0.7664153228025758,
+      "grad_norm": 0.7840436697006226,
+      "learning_rate": 6.818613098700758e-06,
+      "loss": 0.7028,
+      "step": 13925
+    },
+    {
+      "epoch": 0.7664703616049314,
+      "grad_norm": 0.7147907018661499,
+      "learning_rate": 6.818209314547707e-06,
+      "loss": 0.7242,
+      "step": 13926
+    },
+    {
+      "epoch": 0.7665254004072871,
+      "grad_norm": 0.6627985835075378,
+      "learning_rate": 6.817805516729482e-06,
+      "loss": 0.7177,
+      "step": 13927
+    },
+    {
+      "epoch": 0.7665804392096428,
+      "grad_norm": 0.8019070625305176,
+      "learning_rate": 6.817401705249118e-06,
+      "loss": 0.6594,
+      "step": 13928
+    },
+    {
+      "epoch": 0.7666354780119985,
+      "grad_norm": 0.7127207517623901,
+      "learning_rate": 6.816997880109649e-06,
+      "loss": 0.8282,
+      "step": 13929
+    },
+    {
+      "epoch": 0.7666905168143541,
+      "grad_norm": 0.7335825562477112,
+      "learning_rate": 6.816594041314111e-06,
+      "loss": 0.7593,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7667455556167098,
+      "grad_norm": 0.6878668069839478,
+      "learning_rate": 6.816190188865538e-06,
+      "loss": 0.7898,
+      "step": 13931
+    },
+    {
+      "epoch": 0.7668005944190655,
+      "grad_norm": 0.6441968679428101,
+      "learning_rate": 6.815786322766965e-06,
+      "loss": 0.6795,
+      "step": 13932
+    },
+    {
+      "epoch": 0.7668556332214211,
+      "grad_norm": 0.6503410339355469,
+      "learning_rate": 6.815382443021429e-06,
+      "loss": 0.753,
+      "step": 13933
+    },
+    {
+      "epoch": 0.7669106720237767,
+      "grad_norm": 0.6734908223152161,
+      "learning_rate": 6.8149785496319645e-06,
+      "loss": 0.7145,
+      "step": 13934
+    },
+    {
+      "epoch": 0.7669657108261324,
+      "grad_norm": 0.8363823890686035,
+      "learning_rate": 6.814574642601606e-06,
+      "loss": 0.8499,
+      "step": 13935
+    },
+    {
+      "epoch": 0.7670207496284881,
+      "grad_norm": 0.6986021995544434,
+      "learning_rate": 6.81417072193339e-06,
+      "loss": 0.7101,
+      "step": 13936
+    },
+    {
+      "epoch": 0.7670757884308438,
+      "grad_norm": 0.9656592011451721,
+      "learning_rate": 6.813766787630354e-06,
+      "loss": 0.7841,
+      "step": 13937
+    },
+    {
+      "epoch": 0.7671308272331994,
+      "grad_norm": 0.6830777525901794,
+      "learning_rate": 6.813362839695532e-06,
+      "loss": 0.7443,
+      "step": 13938
+    },
+    {
+      "epoch": 0.767185866035555,
+      "grad_norm": 0.6358513236045837,
+      "learning_rate": 6.812958878131959e-06,
+      "loss": 0.7017,
+      "step": 13939
+    },
+    {
+      "epoch": 0.7672409048379107,
+      "grad_norm": 0.9075862169265747,
+      "learning_rate": 6.812554902942673e-06,
+      "loss": 0.6991,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7672959436402664,
+      "grad_norm": 0.7004347443580627,
+      "learning_rate": 6.812150914130709e-06,
+      "loss": 0.6519,
+      "step": 13941
+    },
+    {
+      "epoch": 0.767350982442622,
+      "grad_norm": 0.6648300886154175,
+      "learning_rate": 6.811746911699105e-06,
+      "loss": 0.7044,
+      "step": 13942
+    },
+    {
+      "epoch": 0.7674060212449777,
+      "grad_norm": 0.7050208449363708,
+      "learning_rate": 6.811342895650896e-06,
+      "loss": 0.78,
+      "step": 13943
+    },
+    {
+      "epoch": 0.7674610600473334,
+      "grad_norm": 0.6387132406234741,
+      "learning_rate": 6.810938865989119e-06,
+      "loss": 0.6062,
+      "step": 13944
+    },
+    {
+      "epoch": 0.7675160988496891,
+      "grad_norm": 0.6441114544868469,
+      "learning_rate": 6.81053482271681e-06,
+      "loss": 0.7252,
+      "step": 13945
+    },
+    {
+      "epoch": 0.7675711376520447,
+      "grad_norm": 0.7309751510620117,
+      "learning_rate": 6.810130765837006e-06,
+      "loss": 0.6407,
+      "step": 13946
+    },
+    {
+      "epoch": 0.7676261764544003,
+      "grad_norm": 0.7132161259651184,
+      "learning_rate": 6.809726695352742e-06,
+      "loss": 0.8341,
+      "step": 13947
+    },
+    {
+      "epoch": 0.767681215256756,
+      "grad_norm": 0.7214738726615906,
+      "learning_rate": 6.809322611267058e-06,
+      "loss": 0.8357,
+      "step": 13948
+    },
+    {
+      "epoch": 0.7677362540591117,
+      "grad_norm": 0.6410175561904907,
+      "learning_rate": 6.80891851358299e-06,
+      "loss": 0.6718,
+      "step": 13949
+    },
+    {
+      "epoch": 0.7677912928614673,
+      "grad_norm": 0.8888845443725586,
+      "learning_rate": 6.8085144023035745e-06,
+      "loss": 0.7823,
+      "step": 13950
+    },
+    {
+      "epoch": 0.767846331663823,
+      "grad_norm": 0.7327878475189209,
+      "learning_rate": 6.808110277431848e-06,
+      "loss": 0.7083,
+      "step": 13951
+    },
+    {
+      "epoch": 0.7679013704661787,
+      "grad_norm": 0.6871985793113708,
+      "learning_rate": 6.807706138970849e-06,
+      "loss": 0.7808,
+      "step": 13952
+    },
+    {
+      "epoch": 0.7679564092685344,
+      "grad_norm": 0.6939501762390137,
+      "learning_rate": 6.8073019869236134e-06,
+      "loss": 0.693,
+      "step": 13953
+    },
+    {
+      "epoch": 0.76801144807089,
+      "grad_norm": 0.7377064824104309,
+      "learning_rate": 6.8068978212931814e-06,
+      "loss": 0.9322,
+      "step": 13954
+    },
+    {
+      "epoch": 0.7680664868732456,
+      "grad_norm": 0.8165044188499451,
+      "learning_rate": 6.80649364208259e-06,
+      "loss": 0.6846,
+      "step": 13955
+    },
+    {
+      "epoch": 0.7681215256756013,
+      "grad_norm": 0.6774152517318726,
+      "learning_rate": 6.806089449294875e-06,
+      "loss": 0.8503,
+      "step": 13956
+    },
+    {
+      "epoch": 0.768176564477957,
+      "grad_norm": 0.7773441076278687,
+      "learning_rate": 6.805685242933074e-06,
+      "loss": 0.8775,
+      "step": 13957
+    },
+    {
+      "epoch": 0.7682316032803126,
+      "grad_norm": 0.6710473895072937,
+      "learning_rate": 6.805281023000227e-06,
+      "loss": 0.7831,
+      "step": 13958
+    },
+    {
+      "epoch": 0.7682866420826683,
+      "grad_norm": 0.6163424849510193,
+      "learning_rate": 6.80487678949937e-06,
+      "loss": 0.7309,
+      "step": 13959
+    },
+    {
+      "epoch": 0.768341680885024,
+      "grad_norm": 0.6851963400840759,
+      "learning_rate": 6.804472542433543e-06,
+      "loss": 0.6556,
+      "step": 13960
+    },
+    {
+      "epoch": 0.7683967196873795,
+      "grad_norm": 0.6881004571914673,
+      "learning_rate": 6.804068281805784e-06,
+      "loss": 0.7115,
+      "step": 13961
+    },
+    {
+      "epoch": 0.7684517584897352,
+      "grad_norm": 0.7372351884841919,
+      "learning_rate": 6.8036640076191304e-06,
+      "loss": 0.7869,
+      "step": 13962
+    },
+    {
+      "epoch": 0.7685067972920909,
+      "grad_norm": 0.7900989055633545,
+      "learning_rate": 6.8032597198766205e-06,
+      "loss": 0.7419,
+      "step": 13963
+    },
+    {
+      "epoch": 0.7685618360944466,
+      "grad_norm": 0.7245132327079773,
+      "learning_rate": 6.802855418581294e-06,
+      "loss": 0.8175,
+      "step": 13964
+    },
+    {
+      "epoch": 0.7686168748968022,
+      "grad_norm": 0.6681550741195679,
+      "learning_rate": 6.802451103736188e-06,
+      "loss": 0.773,
+      "step": 13965
+    },
+    {
+      "epoch": 0.7686719136991579,
+      "grad_norm": 0.6316970586776733,
+      "learning_rate": 6.802046775344343e-06,
+      "loss": 0.6597,
+      "step": 13966
+    },
+    {
+      "epoch": 0.7687269525015136,
+      "grad_norm": 0.7201604843139648,
+      "learning_rate": 6.801642433408796e-06,
+      "loss": 0.7205,
+      "step": 13967
+    },
+    {
+      "epoch": 0.7687819913038693,
+      "grad_norm": 0.6226171851158142,
+      "learning_rate": 6.801238077932587e-06,
+      "loss": 0.7271,
+      "step": 13968
+    },
+    {
+      "epoch": 0.7688370301062248,
+      "grad_norm": 0.833369255065918,
+      "learning_rate": 6.800833708918755e-06,
+      "loss": 0.7731,
+      "step": 13969
+    },
+    {
+      "epoch": 0.7688920689085805,
+      "grad_norm": 0.7280329465866089,
+      "learning_rate": 6.800429326370339e-06,
+      "loss": 0.7833,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7689471077109362,
+      "grad_norm": 0.7581672072410583,
+      "learning_rate": 6.800024930290376e-06,
+      "loss": 0.8008,
+      "step": 13971
+    },
+    {
+      "epoch": 0.7690021465132919,
+      "grad_norm": 0.7931516170501709,
+      "learning_rate": 6.79962052068191e-06,
+      "loss": 0.8884,
+      "step": 13972
+    },
+    {
+      "epoch": 0.7690571853156475,
+      "grad_norm": 0.8455879092216492,
+      "learning_rate": 6.799216097547977e-06,
+      "loss": 0.8109,
+      "step": 13973
+    },
+    {
+      "epoch": 0.7691122241180032,
+      "grad_norm": 0.687336266040802,
+      "learning_rate": 6.798811660891618e-06,
+      "loss": 0.783,
+      "step": 13974
+    },
+    {
+      "epoch": 0.7691672629203589,
+      "grad_norm": 0.7661089897155762,
+      "learning_rate": 6.7984072107158696e-06,
+      "loss": 0.8448,
+      "step": 13975
+    },
+    {
+      "epoch": 0.7692223017227146,
+      "grad_norm": 0.6965043544769287,
+      "learning_rate": 6.798002747023776e-06,
+      "loss": 0.7421,
+      "step": 13976
+    },
+    {
+      "epoch": 0.7692773405250701,
+      "grad_norm": 0.7373656630516052,
+      "learning_rate": 6.797598269818375e-06,
+      "loss": 0.7093,
+      "step": 13977
+    },
+    {
+      "epoch": 0.7693323793274258,
+      "grad_norm": 0.6387331485748291,
+      "learning_rate": 6.7971937791027064e-06,
+      "loss": 0.7811,
+      "step": 13978
+    },
+    {
+      "epoch": 0.7693874181297815,
+      "grad_norm": 0.7566075325012207,
+      "learning_rate": 6.796789274879811e-06,
+      "loss": 0.8245,
+      "step": 13979
+    },
+    {
+      "epoch": 0.7694424569321372,
+      "grad_norm": 0.7035738229751587,
+      "learning_rate": 6.796384757152729e-06,
+      "loss": 0.7674,
+      "step": 13980
+    },
+    {
+      "epoch": 0.7694974957344928,
+      "grad_norm": 0.8265605568885803,
+      "learning_rate": 6.795980225924499e-06,
+      "loss": 0.7755,
+      "step": 13981
+    },
+    {
+      "epoch": 0.7695525345368485,
+      "grad_norm": 0.709454357624054,
+      "learning_rate": 6.7955756811981625e-06,
+      "loss": 0.8651,
+      "step": 13982
+    },
+    {
+      "epoch": 0.7696075733392042,
+      "grad_norm": 0.7075764536857605,
+      "learning_rate": 6.795171122976758e-06,
+      "loss": 0.7371,
+      "step": 13983
+    },
+    {
+      "epoch": 0.7696626121415598,
+      "grad_norm": 0.7027561664581299,
+      "learning_rate": 6.79476655126333e-06,
+      "loss": 0.7763,
+      "step": 13984
+    },
+    {
+      "epoch": 0.7697176509439154,
+      "grad_norm": 0.7922375202178955,
+      "learning_rate": 6.794361966060916e-06,
+      "loss": 0.7677,
+      "step": 13985
+    },
+    {
+      "epoch": 0.7697726897462711,
+      "grad_norm": 0.7185537219047546,
+      "learning_rate": 6.793957367372559e-06,
+      "loss": 0.7229,
+      "step": 13986
+    },
+    {
+      "epoch": 0.7698277285486268,
+      "grad_norm": 0.7173545956611633,
+      "learning_rate": 6.793552755201297e-06,
+      "loss": 0.7508,
+      "step": 13987
+    },
+    {
+      "epoch": 0.7698827673509825,
+      "grad_norm": 0.7743139863014221,
+      "learning_rate": 6.793148129550175e-06,
+      "loss": 0.7305,
+      "step": 13988
+    },
+    {
+      "epoch": 0.7699378061533381,
+      "grad_norm": 0.7992164492607117,
+      "learning_rate": 6.792743490422229e-06,
+      "loss": 0.7212,
+      "step": 13989
+    },
+    {
+      "epoch": 0.7699928449556938,
+      "grad_norm": 0.7437503337860107,
+      "learning_rate": 6.792338837820504e-06,
+      "loss": 0.6396,
+      "step": 13990
+    },
+    {
+      "epoch": 0.7700478837580494,
+      "grad_norm": 0.6908634305000305,
+      "learning_rate": 6.79193417174804e-06,
+      "loss": 0.7279,
+      "step": 13991
+    },
+    {
+      "epoch": 0.7701029225604051,
+      "grad_norm": 0.6894391775131226,
+      "learning_rate": 6.7915294922078805e-06,
+      "loss": 0.7615,
+      "step": 13992
+    },
+    {
+      "epoch": 0.7701579613627607,
+      "grad_norm": 0.7162172794342041,
+      "learning_rate": 6.791124799203062e-06,
+      "loss": 0.7404,
+      "step": 13993
+    },
+    {
+      "epoch": 0.7702130001651164,
+      "grad_norm": 0.6469258069992065,
+      "learning_rate": 6.79072009273663e-06,
+      "loss": 0.7035,
+      "step": 13994
+    },
+    {
+      "epoch": 0.7702680389674721,
+      "grad_norm": 0.6456457376480103,
+      "learning_rate": 6.790315372811625e-06,
+      "loss": 0.708,
+      "step": 13995
+    },
+    {
+      "epoch": 0.7703230777698278,
+      "grad_norm": 0.7880644798278809,
+      "learning_rate": 6.789910639431089e-06,
+      "loss": 0.7723,
+      "step": 13996
+    },
+    {
+      "epoch": 0.7703781165721834,
+      "grad_norm": 0.7847834229469299,
+      "learning_rate": 6.789505892598063e-06,
+      "loss": 0.8585,
+      "step": 13997
+    },
+    {
+      "epoch": 0.770433155374539,
+      "grad_norm": 0.6909215450286865,
+      "learning_rate": 6.789101132315591e-06,
+      "loss": 0.7107,
+      "step": 13998
+    },
+    {
+      "epoch": 0.7704881941768947,
+      "grad_norm": 0.7883939146995544,
+      "learning_rate": 6.788696358586713e-06,
+      "loss": 0.7575,
+      "step": 13999
+    },
+    {
+      "epoch": 0.7705432329792504,
+      "grad_norm": 0.6629998087882996,
+      "learning_rate": 6.788291571414472e-06,
+      "loss": 0.7273,
+      "step": 14000
+    },
+    {
+      "epoch": 0.770598271781606,
+      "grad_norm": 0.7548647522926331,
+      "learning_rate": 6.7878867708019106e-06,
+      "loss": 0.8214,
+      "step": 14001
+    },
+    {
+      "epoch": 0.7706533105839617,
+      "grad_norm": 0.6721330881118774,
+      "learning_rate": 6.78748195675207e-06,
+      "loss": 0.7153,
+      "step": 14002
+    },
+    {
+      "epoch": 0.7707083493863174,
+      "grad_norm": 0.6921262145042419,
+      "learning_rate": 6.787077129267994e-06,
+      "loss": 0.7099,
+      "step": 14003
+    },
+    {
+      "epoch": 0.770763388188673,
+      "grad_norm": 0.956937849521637,
+      "learning_rate": 6.786672288352725e-06,
+      "loss": 0.6765,
+      "step": 14004
+    },
+    {
+      "epoch": 0.7708184269910286,
+      "grad_norm": 0.7265778183937073,
+      "learning_rate": 6.786267434009306e-06,
+      "loss": 0.7653,
+      "step": 14005
+    },
+    {
+      "epoch": 0.7708734657933843,
+      "grad_norm": 0.7429845929145813,
+      "learning_rate": 6.785862566240778e-06,
+      "loss": 0.8064,
+      "step": 14006
+    },
+    {
+      "epoch": 0.77092850459574,
+      "grad_norm": 0.7437632083892822,
+      "learning_rate": 6.785457685050184e-06,
+      "loss": 0.7138,
+      "step": 14007
+    },
+    {
+      "epoch": 0.7709835433980956,
+      "grad_norm": 0.7218232750892639,
+      "learning_rate": 6.7850527904405695e-06,
+      "loss": 0.7785,
+      "step": 14008
+    },
+    {
+      "epoch": 0.7710385822004513,
+      "grad_norm": 0.7131973505020142,
+      "learning_rate": 6.784647882414977e-06,
+      "loss": 0.7651,
+      "step": 14009
+    },
+    {
+      "epoch": 0.771093621002807,
+      "grad_norm": 0.739919126033783,
+      "learning_rate": 6.784242960976447e-06,
+      "loss": 0.7993,
+      "step": 14010
+    },
+    {
+      "epoch": 0.7711486598051627,
+      "grad_norm": 0.6655608415603638,
+      "learning_rate": 6.783838026128025e-06,
+      "loss": 0.7394,
+      "step": 14011
+    },
+    {
+      "epoch": 0.7712036986075183,
+      "grad_norm": 0.9327310919761658,
+      "learning_rate": 6.783433077872753e-06,
+      "loss": 0.8737,
+      "step": 14012
+    },
+    {
+      "epoch": 0.7712587374098739,
+      "grad_norm": 0.5928294062614441,
+      "learning_rate": 6.783028116213677e-06,
+      "loss": 0.5819,
+      "step": 14013
+    },
+    {
+      "epoch": 0.7713137762122296,
+      "grad_norm": 0.6752136945724487,
+      "learning_rate": 6.782623141153838e-06,
+      "loss": 0.8021,
+      "step": 14014
+    },
+    {
+      "epoch": 0.7713688150145853,
+      "grad_norm": 0.6452222466468811,
+      "learning_rate": 6.78221815269628e-06,
+      "loss": 0.7806,
+      "step": 14015
+    },
+    {
+      "epoch": 0.7714238538169409,
+      "grad_norm": 0.7725237607955933,
+      "learning_rate": 6.78181315084405e-06,
+      "loss": 0.7679,
+      "step": 14016
+    },
+    {
+      "epoch": 0.7714788926192966,
+      "grad_norm": 0.6594743728637695,
+      "learning_rate": 6.781408135600187e-06,
+      "loss": 0.7254,
+      "step": 14017
+    },
+    {
+      "epoch": 0.7715339314216523,
+      "grad_norm": 0.7008917927742004,
+      "learning_rate": 6.7810031069677385e-06,
+      "loss": 0.705,
+      "step": 14018
+    },
+    {
+      "epoch": 0.771588970224008,
+      "grad_norm": 0.9435684084892273,
+      "learning_rate": 6.780598064949746e-06,
+      "loss": 0.7787,
+      "step": 14019
+    },
+    {
+      "epoch": 0.7716440090263635,
+      "grad_norm": 0.6615981459617615,
+      "learning_rate": 6.780193009549256e-06,
+      "loss": 0.7592,
+      "step": 14020
+    },
+    {
+      "epoch": 0.7716990478287192,
+      "grad_norm": 0.7042600512504578,
+      "learning_rate": 6.7797879407693115e-06,
+      "loss": 0.719,
+      "step": 14021
+    },
+    {
+      "epoch": 0.7717540866310749,
+      "grad_norm": 0.7135425209999084,
+      "learning_rate": 6.779382858612957e-06,
+      "loss": 0.739,
+      "step": 14022
+    },
+    {
+      "epoch": 0.7718091254334306,
+      "grad_norm": 0.6546016931533813,
+      "learning_rate": 6.778977763083238e-06,
+      "loss": 0.7039,
+      "step": 14023
+    },
+    {
+      "epoch": 0.7718641642357862,
+      "grad_norm": 0.8549250960350037,
+      "learning_rate": 6.778572654183198e-06,
+      "loss": 0.8384,
+      "step": 14024
+    },
+    {
+      "epoch": 0.7719192030381419,
+      "grad_norm": 0.7008731365203857,
+      "learning_rate": 6.778167531915882e-06,
+      "loss": 0.776,
+      "step": 14025
+    },
+    {
+      "epoch": 0.7719742418404976,
+      "grad_norm": 0.7047393321990967,
+      "learning_rate": 6.7777623962843355e-06,
+      "loss": 0.819,
+      "step": 14026
+    },
+    {
+      "epoch": 0.7720292806428533,
+      "grad_norm": 0.7015580534934998,
+      "learning_rate": 6.777357247291601e-06,
+      "loss": 0.8339,
+      "step": 14027
+    },
+    {
+      "epoch": 0.7720843194452088,
+      "grad_norm": 0.7008551955223083,
+      "learning_rate": 6.776952084940727e-06,
+      "loss": 0.783,
+      "step": 14028
+    },
+    {
+      "epoch": 0.7721393582475645,
+      "grad_norm": 1.0310637950897217,
+      "learning_rate": 6.776546909234757e-06,
+      "loss": 0.7447,
+      "step": 14029
+    },
+    {
+      "epoch": 0.7721943970499202,
+      "grad_norm": 0.6264338493347168,
+      "learning_rate": 6.776141720176734e-06,
+      "loss": 0.5542,
+      "step": 14030
+    },
+    {
+      "epoch": 0.7722494358522759,
+      "grad_norm": 0.6249508261680603,
+      "learning_rate": 6.775736517769707e-06,
+      "loss": 0.6514,
+      "step": 14031
+    },
+    {
+      "epoch": 0.7723044746546315,
+      "grad_norm": 0.6741732954978943,
+      "learning_rate": 6.775331302016719e-06,
+      "loss": 0.6967,
+      "step": 14032
+    },
+    {
+      "epoch": 0.7723595134569872,
+      "grad_norm": 0.7342913746833801,
+      "learning_rate": 6.774926072920815e-06,
+      "loss": 0.8279,
+      "step": 14033
+    },
+    {
+      "epoch": 0.7724145522593429,
+      "grad_norm": 0.7702916264533997,
+      "learning_rate": 6.774520830485044e-06,
+      "loss": 0.8539,
+      "step": 14034
+    },
+    {
+      "epoch": 0.7724695910616985,
+      "grad_norm": 0.7873550057411194,
+      "learning_rate": 6.774115574712448e-06,
+      "loss": 0.6999,
+      "step": 14035
+    },
+    {
+      "epoch": 0.7725246298640541,
+      "grad_norm": 0.6832353472709656,
+      "learning_rate": 6.773710305606074e-06,
+      "loss": 0.7246,
+      "step": 14036
+    },
+    {
+      "epoch": 0.7725796686664098,
+      "grad_norm": 0.7547367215156555,
+      "learning_rate": 6.773305023168969e-06,
+      "loss": 0.7357,
+      "step": 14037
+    },
+    {
+      "epoch": 0.7726347074687655,
+      "grad_norm": 0.7146826386451721,
+      "learning_rate": 6.772899727404178e-06,
+      "loss": 0.6742,
+      "step": 14038
+    },
+    {
+      "epoch": 0.7726897462711212,
+      "grad_norm": 0.7623558640480042,
+      "learning_rate": 6.772494418314748e-06,
+      "loss": 0.7729,
+      "step": 14039
+    },
+    {
+      "epoch": 0.7727447850734768,
+      "grad_norm": 0.637706458568573,
+      "learning_rate": 6.772089095903723e-06,
+      "loss": 0.6662,
+      "step": 14040
+    },
+    {
+      "epoch": 0.7727998238758325,
+      "grad_norm": 0.7293589115142822,
+      "learning_rate": 6.771683760174151e-06,
+      "loss": 0.7899,
+      "step": 14041
+    },
+    {
+      "epoch": 0.7728548626781881,
+      "grad_norm": 0.7191390991210938,
+      "learning_rate": 6.771278411129079e-06,
+      "loss": 0.6912,
+      "step": 14042
+    },
+    {
+      "epoch": 0.7729099014805438,
+      "grad_norm": 0.8264575004577637,
+      "learning_rate": 6.770873048771552e-06,
+      "loss": 0.7027,
+      "step": 14043
+    },
+    {
+      "epoch": 0.7729649402828994,
+      "grad_norm": 0.7490931749343872,
+      "learning_rate": 6.770467673104617e-06,
+      "loss": 0.6917,
+      "step": 14044
+    },
+    {
+      "epoch": 0.7730199790852551,
+      "grad_norm": 0.6901552081108093,
+      "learning_rate": 6.77006228413132e-06,
+      "loss": 0.8097,
+      "step": 14045
+    },
+    {
+      "epoch": 0.7730750178876108,
+      "grad_norm": 0.6340280175209045,
+      "learning_rate": 6.76965688185471e-06,
+      "loss": 0.6309,
+      "step": 14046
+    },
+    {
+      "epoch": 0.7731300566899664,
+      "grad_norm": 0.6807279586791992,
+      "learning_rate": 6.7692514662778315e-06,
+      "loss": 0.7744,
+      "step": 14047
+    },
+    {
+      "epoch": 0.7731850954923221,
+      "grad_norm": 1.2796865701675415,
+      "learning_rate": 6.7688460374037335e-06,
+      "loss": 0.7499,
+      "step": 14048
+    },
+    {
+      "epoch": 0.7732401342946778,
+      "grad_norm": 0.7059674263000488,
+      "learning_rate": 6.768440595235463e-06,
+      "loss": 0.8705,
+      "step": 14049
+    },
+    {
+      "epoch": 0.7732951730970334,
+      "grad_norm": 0.7626641392707825,
+      "learning_rate": 6.768035139776066e-06,
+      "loss": 0.8448,
+      "step": 14050
+    },
+    {
+      "epoch": 0.773350211899389,
+      "grad_norm": 0.6590229868888855,
+      "learning_rate": 6.767629671028588e-06,
+      "loss": 0.6796,
+      "step": 14051
+    },
+    {
+      "epoch": 0.7734052507017447,
+      "grad_norm": 0.6702030301094055,
+      "learning_rate": 6.767224188996081e-06,
+      "loss": 0.7087,
+      "step": 14052
+    },
+    {
+      "epoch": 0.7734602895041004,
+      "grad_norm": 0.670612096786499,
+      "learning_rate": 6.76681869368159e-06,
+      "loss": 0.7203,
+      "step": 14053
+    },
+    {
+      "epoch": 0.7735153283064561,
+      "grad_norm": 0.6892215013504028,
+      "learning_rate": 6.766413185088161e-06,
+      "loss": 0.6891,
+      "step": 14054
+    },
+    {
+      "epoch": 0.7735703671088117,
+      "grad_norm": 0.8354474902153015,
+      "learning_rate": 6.766007663218843e-06,
+      "loss": 0.7378,
+      "step": 14055
+    },
+    {
+      "epoch": 0.7736254059111674,
+      "grad_norm": 0.7633876204490662,
+      "learning_rate": 6.765602128076686e-06,
+      "loss": 0.6916,
+      "step": 14056
+    },
+    {
+      "epoch": 0.773680444713523,
+      "grad_norm": 0.7249060869216919,
+      "learning_rate": 6.765196579664736e-06,
+      "loss": 0.791,
+      "step": 14057
+    },
+    {
+      "epoch": 0.7737354835158787,
+      "grad_norm": 0.7033042311668396,
+      "learning_rate": 6.7647910179860395e-06,
+      "loss": 0.6799,
+      "step": 14058
+    },
+    {
+      "epoch": 0.7737905223182343,
+      "grad_norm": 0.7087684273719788,
+      "learning_rate": 6.7643854430436466e-06,
+      "loss": 0.6389,
+      "step": 14059
+    },
+    {
+      "epoch": 0.77384556112059,
+      "grad_norm": 0.6433978080749512,
+      "learning_rate": 6.763979854840606e-06,
+      "loss": 0.7214,
+      "step": 14060
+    },
+    {
+      "epoch": 0.7739005999229457,
+      "grad_norm": 0.7777101993560791,
+      "learning_rate": 6.763574253379964e-06,
+      "loss": 0.7458,
+      "step": 14061
+    },
+    {
+      "epoch": 0.7739556387253014,
+      "grad_norm": 0.7065346240997314,
+      "learning_rate": 6.763168638664771e-06,
+      "loss": 0.7663,
+      "step": 14062
+    },
+    {
+      "epoch": 0.774010677527657,
+      "grad_norm": 0.7136278748512268,
+      "learning_rate": 6.762763010698074e-06,
+      "loss": 0.667,
+      "step": 14063
+    },
+    {
+      "epoch": 0.7740657163300126,
+      "grad_norm": 0.6670508980751038,
+      "learning_rate": 6.762357369482921e-06,
+      "loss": 0.7462,
+      "step": 14064
+    },
+    {
+      "epoch": 0.7741207551323683,
+      "grad_norm": 0.6366799473762512,
+      "learning_rate": 6.7619517150223635e-06,
+      "loss": 0.7147,
+      "step": 14065
+    },
+    {
+      "epoch": 0.774175793934724,
+      "grad_norm": 0.5999431610107422,
+      "learning_rate": 6.761546047319447e-06,
+      "loss": 0.667,
+      "step": 14066
+    },
+    {
+      "epoch": 0.7742308327370796,
+      "grad_norm": 0.6751196980476379,
+      "learning_rate": 6.761140366377222e-06,
+      "loss": 0.7255,
+      "step": 14067
+    },
+    {
+      "epoch": 0.7742858715394353,
+      "grad_norm": 0.6786272525787354,
+      "learning_rate": 6.760734672198738e-06,
+      "loss": 0.7694,
+      "step": 14068
+    },
+    {
+      "epoch": 0.774340910341791,
+      "grad_norm": 0.6915947794914246,
+      "learning_rate": 6.760328964787044e-06,
+      "loss": 0.7955,
+      "step": 14069
+    },
+    {
+      "epoch": 0.7743959491441467,
+      "grad_norm": 0.7041972279548645,
+      "learning_rate": 6.759923244145188e-06,
+      "loss": 0.6542,
+      "step": 14070
+    },
+    {
+      "epoch": 0.7744509879465022,
+      "grad_norm": 0.6384761333465576,
+      "learning_rate": 6.759517510276221e-06,
+      "loss": 0.7384,
+      "step": 14071
+    },
+    {
+      "epoch": 0.7745060267488579,
+      "grad_norm": 0.7430800199508667,
+      "learning_rate": 6.759111763183189e-06,
+      "loss": 0.7587,
+      "step": 14072
+    },
+    {
+      "epoch": 0.7745610655512136,
+      "grad_norm": 0.6568213701248169,
+      "learning_rate": 6.758706002869146e-06,
+      "loss": 0.7118,
+      "step": 14073
+    },
+    {
+      "epoch": 0.7746161043535693,
+      "grad_norm": 0.8791618943214417,
+      "learning_rate": 6.75830022933714e-06,
+      "loss": 0.8049,
+      "step": 14074
+    },
+    {
+      "epoch": 0.7746711431559249,
+      "grad_norm": 0.6377304792404175,
+      "learning_rate": 6.75789444259022e-06,
+      "loss": 0.737,
+      "step": 14075
+    },
+    {
+      "epoch": 0.7747261819582806,
+      "grad_norm": 0.7253721356391907,
+      "learning_rate": 6.757488642631434e-06,
+      "loss": 0.8432,
+      "step": 14076
+    },
+    {
+      "epoch": 0.7747812207606363,
+      "grad_norm": 0.684626042842865,
+      "learning_rate": 6.757082829463835e-06,
+      "loss": 0.7845,
+      "step": 14077
+    },
+    {
+      "epoch": 0.774836259562992,
+      "grad_norm": 0.7737520337104797,
+      "learning_rate": 6.756677003090471e-06,
+      "loss": 0.8055,
+      "step": 14078
+    },
+    {
+      "epoch": 0.7748912983653475,
+      "grad_norm": 0.7294824719429016,
+      "learning_rate": 6.756271163514394e-06,
+      "loss": 0.7666,
+      "step": 14079
+    },
+    {
+      "epoch": 0.7749463371677032,
+      "grad_norm": 0.7728607654571533,
+      "learning_rate": 6.755865310738651e-06,
+      "loss": 0.7748,
+      "step": 14080
+    },
+    {
+      "epoch": 0.7750013759700589,
+      "grad_norm": 0.6738442778587341,
+      "learning_rate": 6.755459444766297e-06,
+      "loss": 0.6711,
+      "step": 14081
+    },
+    {
+      "epoch": 0.7750564147724146,
+      "grad_norm": 0.7041414976119995,
+      "learning_rate": 6.7550535656003794e-06,
+      "loss": 0.7126,
+      "step": 14082
+    },
+    {
+      "epoch": 0.7751114535747702,
+      "grad_norm": 1.0205422639846802,
+      "learning_rate": 6.754647673243948e-06,
+      "loss": 0.7394,
+      "step": 14083
+    },
+    {
+      "epoch": 0.7751664923771259,
+      "grad_norm": 0.6594380736351013,
+      "learning_rate": 6.754241767700054e-06,
+      "loss": 0.7599,
+      "step": 14084
+    },
+    {
+      "epoch": 0.7752215311794816,
+      "grad_norm": 0.6800520420074463,
+      "learning_rate": 6.753835848971749e-06,
+      "loss": 0.7579,
+      "step": 14085
+    },
+    {
+      "epoch": 0.7752765699818372,
+      "grad_norm": 0.7658087611198425,
+      "learning_rate": 6.7534299170620846e-06,
+      "loss": 0.7705,
+      "step": 14086
+    },
+    {
+      "epoch": 0.7753316087841928,
+      "grad_norm": 0.7242750525474548,
+      "learning_rate": 6.7530239719741084e-06,
+      "loss": 0.7683,
+      "step": 14087
+    },
+    {
+      "epoch": 0.7753866475865485,
+      "grad_norm": 0.6997398138046265,
+      "learning_rate": 6.752618013710874e-06,
+      "loss": 0.8023,
+      "step": 14088
+    },
+    {
+      "epoch": 0.7754416863889042,
+      "grad_norm": 0.7041590809822083,
+      "learning_rate": 6.752212042275431e-06,
+      "loss": 0.7013,
+      "step": 14089
+    },
+    {
+      "epoch": 0.7754967251912598,
+      "grad_norm": 0.7027721405029297,
+      "learning_rate": 6.751806057670832e-06,
+      "loss": 0.7678,
+      "step": 14090
+    },
+    {
+      "epoch": 0.7755517639936155,
+      "grad_norm": 0.714290201663971,
+      "learning_rate": 6.751400059900128e-06,
+      "loss": 0.6769,
+      "step": 14091
+    },
+    {
+      "epoch": 0.7756068027959712,
+      "grad_norm": 0.7385110855102539,
+      "learning_rate": 6.750994048966369e-06,
+      "loss": 0.6576,
+      "step": 14092
+    },
+    {
+      "epoch": 0.7756618415983269,
+      "grad_norm": 0.7665147185325623,
+      "learning_rate": 6.750588024872607e-06,
+      "loss": 0.8127,
+      "step": 14093
+    },
+    {
+      "epoch": 0.7757168804006824,
+      "grad_norm": 0.6774508953094482,
+      "learning_rate": 6.750181987621895e-06,
+      "loss": 0.8112,
+      "step": 14094
+    },
+    {
+      "epoch": 0.7757719192030381,
+      "grad_norm": 0.666394054889679,
+      "learning_rate": 6.749775937217285e-06,
+      "loss": 0.6444,
+      "step": 14095
+    },
+    {
+      "epoch": 0.7758269580053938,
+      "grad_norm": 0.6557022929191589,
+      "learning_rate": 6.749369873661825e-06,
+      "loss": 0.7613,
+      "step": 14096
+    },
+    {
+      "epoch": 0.7758819968077495,
+      "grad_norm": 0.7090621590614319,
+      "learning_rate": 6.74896379695857e-06,
+      "loss": 0.7229,
+      "step": 14097
+    },
+    {
+      "epoch": 0.7759370356101051,
+      "grad_norm": 0.8117626309394836,
+      "learning_rate": 6.7485577071105734e-06,
+      "loss": 0.8002,
+      "step": 14098
+    },
+    {
+      "epoch": 0.7759920744124608,
+      "grad_norm": 0.6743370294570923,
+      "learning_rate": 6.748151604120883e-06,
+      "loss": 0.7457,
+      "step": 14099
+    },
+    {
+      "epoch": 0.7760471132148165,
+      "grad_norm": 0.7637452483177185,
+      "learning_rate": 6.747745487992553e-06,
+      "loss": 0.7471,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7761021520171721,
+      "grad_norm": 0.6732922196388245,
+      "learning_rate": 6.747339358728636e-06,
+      "loss": 0.7471,
+      "step": 14101
+    },
+    {
+      "epoch": 0.7761571908195277,
+      "grad_norm": 0.7510336637496948,
+      "learning_rate": 6.746933216332184e-06,
+      "loss": 0.7252,
+      "step": 14102
+    },
+    {
+      "epoch": 0.7762122296218834,
+      "grad_norm": 0.731719434261322,
+      "learning_rate": 6.746527060806251e-06,
+      "loss": 0.8706,
+      "step": 14103
+    },
+    {
+      "epoch": 0.7762672684242391,
+      "grad_norm": 0.7625692486763,
+      "learning_rate": 6.746120892153886e-06,
+      "loss": 0.7518,
+      "step": 14104
+    },
+    {
+      "epoch": 0.7763223072265948,
+      "grad_norm": 0.6809547543525696,
+      "learning_rate": 6.745714710378145e-06,
+      "loss": 0.7172,
+      "step": 14105
+    },
+    {
+      "epoch": 0.7763773460289504,
+      "grad_norm": 0.709996223449707,
+      "learning_rate": 6.745308515482079e-06,
+      "loss": 0.7925,
+      "step": 14106
+    },
+    {
+      "epoch": 0.776432384831306,
+      "grad_norm": 0.6675372123718262,
+      "learning_rate": 6.744902307468742e-06,
+      "loss": 0.8175,
+      "step": 14107
+    },
+    {
+      "epoch": 0.7764874236336617,
+      "grad_norm": 0.6978115439414978,
+      "learning_rate": 6.744496086341186e-06,
+      "loss": 0.7895,
+      "step": 14108
+    },
+    {
+      "epoch": 0.7765424624360174,
+      "grad_norm": 0.6593814492225647,
+      "learning_rate": 6.7440898521024634e-06,
+      "loss": 0.7791,
+      "step": 14109
+    },
+    {
+      "epoch": 0.776597501238373,
+      "grad_norm": 0.7169299721717834,
+      "learning_rate": 6.743683604755631e-06,
+      "loss": 0.7944,
+      "step": 14110
+    },
+    {
+      "epoch": 0.7766525400407287,
+      "grad_norm": 0.6805511713027954,
+      "learning_rate": 6.743277344303738e-06,
+      "loss": 0.7671,
+      "step": 14111
+    },
+    {
+      "epoch": 0.7767075788430844,
+      "grad_norm": 0.7300780415534973,
+      "learning_rate": 6.742871070749838e-06,
+      "loss": 0.7789,
+      "step": 14112
+    },
+    {
+      "epoch": 0.7767626176454401,
+      "grad_norm": 0.6475857496261597,
+      "learning_rate": 6.742464784096987e-06,
+      "loss": 0.6652,
+      "step": 14113
+    },
+    {
+      "epoch": 0.7768176564477957,
+      "grad_norm": 0.6941269040107727,
+      "learning_rate": 6.742058484348236e-06,
+      "loss": 0.8138,
+      "step": 14114
+    },
+    {
+      "epoch": 0.7768726952501513,
+      "grad_norm": 0.6175981760025024,
+      "learning_rate": 6.7416521715066405e-06,
+      "loss": 0.7667,
+      "step": 14115
+    },
+    {
+      "epoch": 0.776927734052507,
+      "grad_norm": 0.6499401330947876,
+      "learning_rate": 6.741245845575252e-06,
+      "loss": 0.7415,
+      "step": 14116
+    },
+    {
+      "epoch": 0.7769827728548627,
+      "grad_norm": 0.6601547598838806,
+      "learning_rate": 6.740839506557127e-06,
+      "loss": 0.732,
+      "step": 14117
+    },
+    {
+      "epoch": 0.7770378116572183,
+      "grad_norm": 0.7939042448997498,
+      "learning_rate": 6.740433154455319e-06,
+      "loss": 0.7043,
+      "step": 14118
+    },
+    {
+      "epoch": 0.777092850459574,
+      "grad_norm": 0.7381628751754761,
+      "learning_rate": 6.740026789272881e-06,
+      "loss": 0.8256,
+      "step": 14119
+    },
+    {
+      "epoch": 0.7771478892619297,
+      "grad_norm": 0.6131769418716431,
+      "learning_rate": 6.739620411012866e-06,
+      "loss": 0.726,
+      "step": 14120
+    },
+    {
+      "epoch": 0.7772029280642854,
+      "grad_norm": 1.201745867729187,
+      "learning_rate": 6.739214019678332e-06,
+      "loss": 0.7097,
+      "step": 14121
+    },
+    {
+      "epoch": 0.777257966866641,
+      "grad_norm": 0.6618456244468689,
+      "learning_rate": 6.7388076152723295e-06,
+      "loss": 0.6396,
+      "step": 14122
+    },
+    {
+      "epoch": 0.7773130056689966,
+      "grad_norm": 0.7490836977958679,
+      "learning_rate": 6.738401197797915e-06,
+      "loss": 0.6475,
+      "step": 14123
+    },
+    {
+      "epoch": 0.7773680444713523,
+      "grad_norm": 0.8125407099723816,
+      "learning_rate": 6.737994767258142e-06,
+      "loss": 0.7693,
+      "step": 14124
+    },
+    {
+      "epoch": 0.777423083273708,
+      "grad_norm": 0.7501794099807739,
+      "learning_rate": 6.737588323656065e-06,
+      "loss": 0.7333,
+      "step": 14125
+    },
+    {
+      "epoch": 0.7774781220760636,
+      "grad_norm": 1.3062889575958252,
+      "learning_rate": 6.73718186699474e-06,
+      "loss": 0.6909,
+      "step": 14126
+    },
+    {
+      "epoch": 0.7775331608784193,
+      "grad_norm": 0.6784525513648987,
+      "learning_rate": 6.736775397277221e-06,
+      "loss": 0.7256,
+      "step": 14127
+    },
+    {
+      "epoch": 0.777588199680775,
+      "grad_norm": 0.7018646597862244,
+      "learning_rate": 6.736368914506562e-06,
+      "loss": 0.7632,
+      "step": 14128
+    },
+    {
+      "epoch": 0.7776432384831307,
+      "grad_norm": 0.7596307992935181,
+      "learning_rate": 6.735962418685821e-06,
+      "loss": 0.7117,
+      "step": 14129
+    },
+    {
+      "epoch": 0.7776982772854862,
+      "grad_norm": 0.7582107186317444,
+      "learning_rate": 6.7355559098180504e-06,
+      "loss": 0.7808,
+      "step": 14130
+    },
+    {
+      "epoch": 0.7777533160878419,
+      "grad_norm": 0.6460647583007812,
+      "learning_rate": 6.7351493879063056e-06,
+      "loss": 0.675,
+      "step": 14131
+    },
+    {
+      "epoch": 0.7778083548901976,
+      "grad_norm": 0.6801304221153259,
+      "learning_rate": 6.7347428529536415e-06,
+      "loss": 0.6504,
+      "step": 14132
+    },
+    {
+      "epoch": 0.7778633936925532,
+      "grad_norm": 0.8122933506965637,
+      "learning_rate": 6.7343363049631176e-06,
+      "loss": 0.7949,
+      "step": 14133
+    },
+    {
+      "epoch": 0.7779184324949089,
+      "grad_norm": 0.6750267744064331,
+      "learning_rate": 6.733929743937784e-06,
+      "loss": 0.7689,
+      "step": 14134
+    },
+    {
+      "epoch": 0.7779734712972646,
+      "grad_norm": 0.7141891121864319,
+      "learning_rate": 6.7335231698807005e-06,
+      "loss": 0.7099,
+      "step": 14135
+    },
+    {
+      "epoch": 0.7780285100996203,
+      "grad_norm": 0.7904065251350403,
+      "learning_rate": 6.733116582794918e-06,
+      "loss": 0.8458,
+      "step": 14136
+    },
+    {
+      "epoch": 0.7780835489019758,
+      "grad_norm": 0.6905248165130615,
+      "learning_rate": 6.732709982683496e-06,
+      "loss": 0.7848,
+      "step": 14137
+    },
+    {
+      "epoch": 0.7781385877043315,
+      "grad_norm": 0.6707245707511902,
+      "learning_rate": 6.732303369549491e-06,
+      "loss": 0.8319,
+      "step": 14138
+    },
+    {
+      "epoch": 0.7781936265066872,
+      "grad_norm": 0.6611519455909729,
+      "learning_rate": 6.731896743395957e-06,
+      "loss": 0.7025,
+      "step": 14139
+    },
+    {
+      "epoch": 0.7782486653090429,
+      "grad_norm": 0.7113156914710999,
+      "learning_rate": 6.73149010422595e-06,
+      "loss": 0.8297,
+      "step": 14140
+    },
+    {
+      "epoch": 0.7783037041113985,
+      "grad_norm": 0.7279486060142517,
+      "learning_rate": 6.7310834520425265e-06,
+      "loss": 0.8134,
+      "step": 14141
+    },
+    {
+      "epoch": 0.7783587429137542,
+      "grad_norm": 0.7561796307563782,
+      "learning_rate": 6.730676786848744e-06,
+      "loss": 0.806,
+      "step": 14142
+    },
+    {
+      "epoch": 0.7784137817161099,
+      "grad_norm": 0.6724728345870972,
+      "learning_rate": 6.7302701086476585e-06,
+      "loss": 0.7782,
+      "step": 14143
+    },
+    {
+      "epoch": 0.7784688205184656,
+      "grad_norm": 0.6363211274147034,
+      "learning_rate": 6.729863417442325e-06,
+      "loss": 0.6298,
+      "step": 14144
+    },
+    {
+      "epoch": 0.7785238593208211,
+      "grad_norm": 0.6920950412750244,
+      "learning_rate": 6.729456713235803e-06,
+      "loss": 0.5804,
+      "step": 14145
+    },
+    {
+      "epoch": 0.7785788981231768,
+      "grad_norm": 0.7388806343078613,
+      "learning_rate": 6.729049996031145e-06,
+      "loss": 0.6594,
+      "step": 14146
+    },
+    {
+      "epoch": 0.7786339369255325,
+      "grad_norm": 0.7736972570419312,
+      "learning_rate": 6.728643265831412e-06,
+      "loss": 0.8244,
+      "step": 14147
+    },
+    {
+      "epoch": 0.7786889757278882,
+      "grad_norm": 0.6928302049636841,
+      "learning_rate": 6.728236522639658e-06,
+      "loss": 0.6713,
+      "step": 14148
+    },
+    {
+      "epoch": 0.7787440145302438,
+      "grad_norm": 0.8058464527130127,
+      "learning_rate": 6.72782976645894e-06,
+      "loss": 0.7647,
+      "step": 14149
+    },
+    {
+      "epoch": 0.7787990533325995,
+      "grad_norm": 0.7111127376556396,
+      "learning_rate": 6.727422997292317e-06,
+      "loss": 0.7629,
+      "step": 14150
+    },
+    {
+      "epoch": 0.7788540921349552,
+      "grad_norm": 0.9375373721122742,
+      "learning_rate": 6.7270162151428455e-06,
+      "loss": 0.8306,
+      "step": 14151
+    },
+    {
+      "epoch": 0.7789091309373108,
+      "grad_norm": 0.6894392371177673,
+      "learning_rate": 6.726609420013581e-06,
+      "loss": 0.6995,
+      "step": 14152
+    },
+    {
+      "epoch": 0.7789641697396664,
+      "grad_norm": 0.7058690786361694,
+      "learning_rate": 6.726202611907583e-06,
+      "loss": 0.844,
+      "step": 14153
+    },
+    {
+      "epoch": 0.7790192085420221,
+      "grad_norm": 0.7672932744026184,
+      "learning_rate": 6.725795790827909e-06,
+      "loss": 0.6613,
+      "step": 14154
+    },
+    {
+      "epoch": 0.7790742473443778,
+      "grad_norm": 0.8575173020362854,
+      "learning_rate": 6.7253889567776146e-06,
+      "loss": 0.6946,
+      "step": 14155
+    },
+    {
+      "epoch": 0.7791292861467335,
+      "grad_norm": 0.6832261085510254,
+      "learning_rate": 6.724982109759759e-06,
+      "loss": 0.7121,
+      "step": 14156
+    },
+    {
+      "epoch": 0.7791843249490891,
+      "grad_norm": 0.8188209533691406,
+      "learning_rate": 6.724575249777401e-06,
+      "loss": 0.6479,
+      "step": 14157
+    },
+    {
+      "epoch": 0.7792393637514448,
+      "grad_norm": 0.6514336466789246,
+      "learning_rate": 6.724168376833595e-06,
+      "loss": 0.6117,
+      "step": 14158
+    },
+    {
+      "epoch": 0.7792944025538004,
+      "grad_norm": 0.7283767461776733,
+      "learning_rate": 6.723761490931403e-06,
+      "loss": 0.6882,
+      "step": 14159
+    },
+    {
+      "epoch": 0.7793494413561561,
+      "grad_norm": 0.7681146860122681,
+      "learning_rate": 6.7233545920738785e-06,
+      "loss": 0.8028,
+      "step": 14160
+    },
+    {
+      "epoch": 0.7794044801585117,
+      "grad_norm": 0.6202995181083679,
+      "learning_rate": 6.722947680264084e-06,
+      "loss": 0.713,
+      "step": 14161
+    },
+    {
+      "epoch": 0.7794595189608674,
+      "grad_norm": 0.7137139439582825,
+      "learning_rate": 6.722540755505076e-06,
+      "loss": 0.7842,
+      "step": 14162
+    },
+    {
+      "epoch": 0.7795145577632231,
+      "grad_norm": 0.6852554678916931,
+      "learning_rate": 6.722133817799913e-06,
+      "loss": 0.7329,
+      "step": 14163
+    },
+    {
+      "epoch": 0.7795695965655788,
+      "grad_norm": 0.7520774602890015,
+      "learning_rate": 6.7217268671516525e-06,
+      "loss": 0.7498,
+      "step": 14164
+    },
+    {
+      "epoch": 0.7796246353679344,
+      "grad_norm": 0.708577573299408,
+      "learning_rate": 6.7213199035633525e-06,
+      "loss": 0.675,
+      "step": 14165
+    },
+    {
+      "epoch": 0.77967967417029,
+      "grad_norm": 0.8061410188674927,
+      "learning_rate": 6.7209129270380744e-06,
+      "loss": 0.7176,
+      "step": 14166
+    },
+    {
+      "epoch": 0.7797347129726457,
+      "grad_norm": 0.8070787787437439,
+      "learning_rate": 6.720505937578876e-06,
+      "loss": 0.8138,
+      "step": 14167
+    },
+    {
+      "epoch": 0.7797897517750014,
+      "grad_norm": 0.7127004265785217,
+      "learning_rate": 6.720098935188815e-06,
+      "loss": 0.7004,
+      "step": 14168
+    },
+    {
+      "epoch": 0.779844790577357,
+      "grad_norm": 0.7188708782196045,
+      "learning_rate": 6.719691919870951e-06,
+      "loss": 0.6996,
+      "step": 14169
+    },
+    {
+      "epoch": 0.7798998293797127,
+      "grad_norm": 0.6346360445022583,
+      "learning_rate": 6.719284891628342e-06,
+      "loss": 0.7349,
+      "step": 14170
+    },
+    {
+      "epoch": 0.7799548681820684,
+      "grad_norm": 0.6262187361717224,
+      "learning_rate": 6.71887785046405e-06,
+      "loss": 0.7279,
+      "step": 14171
+    },
+    {
+      "epoch": 0.7800099069844241,
+      "grad_norm": 0.7538053393363953,
+      "learning_rate": 6.718470796381129e-06,
+      "loss": 0.754,
+      "step": 14172
+    },
+    {
+      "epoch": 0.7800649457867797,
+      "grad_norm": 0.6569569706916809,
+      "learning_rate": 6.718063729382643e-06,
+      "loss": 0.6787,
+      "step": 14173
+    },
+    {
+      "epoch": 0.7801199845891353,
+      "grad_norm": 0.6446678042411804,
+      "learning_rate": 6.71765664947165e-06,
+      "loss": 0.6338,
+      "step": 14174
+    },
+    {
+      "epoch": 0.780175023391491,
+      "grad_norm": 0.7559269666671753,
+      "learning_rate": 6.7172495566512095e-06,
+      "loss": 0.7472,
+      "step": 14175
+    },
+    {
+      "epoch": 0.7802300621938466,
+      "grad_norm": 0.6920101642608643,
+      "learning_rate": 6.71684245092438e-06,
+      "loss": 0.7189,
+      "step": 14176
+    },
+    {
+      "epoch": 0.7802851009962023,
+      "grad_norm": 0.6513105034828186,
+      "learning_rate": 6.716435332294223e-06,
+      "loss": 0.6104,
+      "step": 14177
+    },
+    {
+      "epoch": 0.780340139798558,
+      "grad_norm": 0.7076418399810791,
+      "learning_rate": 6.716028200763798e-06,
+      "loss": 0.7974,
+      "step": 14178
+    },
+    {
+      "epoch": 0.7803951786009137,
+      "grad_norm": 0.7291662693023682,
+      "learning_rate": 6.715621056336164e-06,
+      "loss": 0.7661,
+      "step": 14179
+    },
+    {
+      "epoch": 0.7804502174032693,
+      "grad_norm": 0.682321310043335,
+      "learning_rate": 6.715213899014381e-06,
+      "loss": 0.7345,
+      "step": 14180
+    },
+    {
+      "epoch": 0.7805052562056249,
+      "grad_norm": 0.7170400619506836,
+      "learning_rate": 6.71480672880151e-06,
+      "loss": 0.6968,
+      "step": 14181
+    },
+    {
+      "epoch": 0.7805602950079806,
+      "grad_norm": 0.7504192590713501,
+      "learning_rate": 6.714399545700611e-06,
+      "loss": 0.7868,
+      "step": 14182
+    },
+    {
+      "epoch": 0.7806153338103363,
+      "grad_norm": 0.7334801554679871,
+      "learning_rate": 6.713992349714744e-06,
+      "loss": 0.8806,
+      "step": 14183
+    },
+    {
+      "epoch": 0.7806703726126919,
+      "grad_norm": 0.6495537161827087,
+      "learning_rate": 6.713585140846969e-06,
+      "loss": 0.7272,
+      "step": 14184
+    },
+    {
+      "epoch": 0.7807254114150476,
+      "grad_norm": 0.7101101279258728,
+      "learning_rate": 6.713177919100347e-06,
+      "loss": 0.8038,
+      "step": 14185
+    },
+    {
+      "epoch": 0.7807804502174033,
+      "grad_norm": 0.7013083100318909,
+      "learning_rate": 6.712770684477937e-06,
+      "loss": 0.7576,
+      "step": 14186
+    },
+    {
+      "epoch": 0.780835489019759,
+      "grad_norm": 0.7535369992256165,
+      "learning_rate": 6.712363436982802e-06,
+      "loss": 0.6537,
+      "step": 14187
+    },
+    {
+      "epoch": 0.7808905278221145,
+      "grad_norm": 0.7432667016983032,
+      "learning_rate": 6.711956176618001e-06,
+      "loss": 0.7734,
+      "step": 14188
+    },
+    {
+      "epoch": 0.7809455666244702,
+      "grad_norm": 0.718006432056427,
+      "learning_rate": 6.711548903386597e-06,
+      "loss": 0.7291,
+      "step": 14189
+    },
+    {
+      "epoch": 0.7810006054268259,
+      "grad_norm": 0.7983072400093079,
+      "learning_rate": 6.711141617291649e-06,
+      "loss": 0.8403,
+      "step": 14190
+    },
+    {
+      "epoch": 0.7810556442291816,
+      "grad_norm": 0.7017259001731873,
+      "learning_rate": 6.710734318336218e-06,
+      "loss": 0.7293,
+      "step": 14191
+    },
+    {
+      "epoch": 0.7811106830315372,
+      "grad_norm": 0.6061737537384033,
+      "learning_rate": 6.710327006523366e-06,
+      "loss": 0.6624,
+      "step": 14192
+    },
+    {
+      "epoch": 0.7811657218338929,
+      "grad_norm": 0.6876726746559143,
+      "learning_rate": 6.709919681856155e-06,
+      "loss": 0.723,
+      "step": 14193
+    },
+    {
+      "epoch": 0.7812207606362486,
+      "grad_norm": 0.6926757097244263,
+      "learning_rate": 6.709512344337646e-06,
+      "loss": 0.7392,
+      "step": 14194
+    },
+    {
+      "epoch": 0.7812757994386043,
+      "grad_norm": 0.6464381217956543,
+      "learning_rate": 6.7091049939708985e-06,
+      "loss": 0.7301,
+      "step": 14195
+    },
+    {
+      "epoch": 0.7813308382409598,
+      "grad_norm": 0.7292629480361938,
+      "learning_rate": 6.708697630758974e-06,
+      "loss": 0.7511,
+      "step": 14196
+    },
+    {
+      "epoch": 0.7813858770433155,
+      "grad_norm": 0.7483099102973938,
+      "learning_rate": 6.708290254704937e-06,
+      "loss": 0.7981,
+      "step": 14197
+    },
+    {
+      "epoch": 0.7814409158456712,
+      "grad_norm": 0.6766877770423889,
+      "learning_rate": 6.707882865811848e-06,
+      "loss": 0.7987,
+      "step": 14198
+    },
+    {
+      "epoch": 0.7814959546480269,
+      "grad_norm": 0.7340181469917297,
+      "learning_rate": 6.707475464082769e-06,
+      "loss": 0.799,
+      "step": 14199
+    },
+    {
+      "epoch": 0.7815509934503825,
+      "grad_norm": 0.6247759461402893,
+      "learning_rate": 6.707068049520759e-06,
+      "loss": 0.7299,
+      "step": 14200
+    },
+    {
+      "epoch": 0.7816060322527382,
+      "grad_norm": 0.6783067584037781,
+      "learning_rate": 6.706660622128885e-06,
+      "loss": 0.6987,
+      "step": 14201
+    },
+    {
+      "epoch": 0.7816610710550939,
+      "grad_norm": 0.7613719701766968,
+      "learning_rate": 6.706253181910205e-06,
+      "loss": 0.7894,
+      "step": 14202
+    },
+    {
+      "epoch": 0.7817161098574495,
+      "grad_norm": 0.6673761606216431,
+      "learning_rate": 6.705845728867784e-06,
+      "loss": 0.8015,
+      "step": 14203
+    },
+    {
+      "epoch": 0.7817711486598051,
+      "grad_norm": 0.6551307439804077,
+      "learning_rate": 6.705438263004683e-06,
+      "loss": 0.7057,
+      "step": 14204
+    },
+    {
+      "epoch": 0.7818261874621608,
+      "grad_norm": 0.6815405488014221,
+      "learning_rate": 6.705030784323965e-06,
+      "loss": 0.7466,
+      "step": 14205
+    },
+    {
+      "epoch": 0.7818812262645165,
+      "grad_norm": 0.6838087439537048,
+      "learning_rate": 6.704623292828692e-06,
+      "loss": 0.8226,
+      "step": 14206
+    },
+    {
+      "epoch": 0.7819362650668722,
+      "grad_norm": 0.6704637408256531,
+      "learning_rate": 6.704215788521925e-06,
+      "loss": 0.8101,
+      "step": 14207
+    },
+    {
+      "epoch": 0.7819913038692278,
+      "grad_norm": 0.6606172919273376,
+      "learning_rate": 6.70380827140673e-06,
+      "loss": 0.7824,
+      "step": 14208
+    },
+    {
+      "epoch": 0.7820463426715835,
+      "grad_norm": 0.6641090512275696,
+      "learning_rate": 6.703400741486166e-06,
+      "loss": 0.7507,
+      "step": 14209
+    },
+    {
+      "epoch": 0.7821013814739392,
+      "grad_norm": 1.6413429975509644,
+      "learning_rate": 6.702993198763299e-06,
+      "loss": 0.7793,
+      "step": 14210
+    },
+    {
+      "epoch": 0.7821564202762948,
+      "grad_norm": 0.6664854884147644,
+      "learning_rate": 6.7025856432411915e-06,
+      "loss": 0.7304,
+      "step": 14211
+    },
+    {
+      "epoch": 0.7822114590786504,
+      "grad_norm": 0.6968172192573547,
+      "learning_rate": 6.7021780749229075e-06,
+      "loss": 0.7506,
+      "step": 14212
+    },
+    {
+      "epoch": 0.7822664978810061,
+      "grad_norm": 0.6443943381309509,
+      "learning_rate": 6.701770493811506e-06,
+      "loss": 0.7511,
+      "step": 14213
+    },
+    {
+      "epoch": 0.7823215366833618,
+      "grad_norm": 0.67723548412323,
+      "learning_rate": 6.701362899910053e-06,
+      "loss": 0.6839,
+      "step": 14214
+    },
+    {
+      "epoch": 0.7823765754857175,
+      "grad_norm": 0.7601221203804016,
+      "learning_rate": 6.700955293221614e-06,
+      "loss": 0.7397,
+      "step": 14215
+    },
+    {
+      "epoch": 0.7824316142880731,
+      "grad_norm": 0.6056920289993286,
+      "learning_rate": 6.700547673749249e-06,
+      "loss": 0.7706,
+      "step": 14216
+    },
+    {
+      "epoch": 0.7824866530904288,
+      "grad_norm": 0.6421142816543579,
+      "learning_rate": 6.700140041496024e-06,
+      "loss": 0.7209,
+      "step": 14217
+    },
+    {
+      "epoch": 0.7825416918927844,
+      "grad_norm": 0.6653133034706116,
+      "learning_rate": 6.6997323964650005e-06,
+      "loss": 0.708,
+      "step": 14218
+    },
+    {
+      "epoch": 0.78259673069514,
+      "grad_norm": 0.8854939937591553,
+      "learning_rate": 6.699324738659243e-06,
+      "loss": 0.7658,
+      "step": 14219
+    },
+    {
+      "epoch": 0.7826517694974957,
+      "grad_norm": 0.7130745649337769,
+      "learning_rate": 6.6989170680818175e-06,
+      "loss": 0.7827,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7827068082998514,
+      "grad_norm": 0.953117847442627,
+      "learning_rate": 6.698509384735783e-06,
+      "loss": 0.7852,
+      "step": 14221
+    },
+    {
+      "epoch": 0.7827618471022071,
+      "grad_norm": 0.655768871307373,
+      "learning_rate": 6.698101688624209e-06,
+      "loss": 0.8461,
+      "step": 14222
+    },
+    {
+      "epoch": 0.7828168859045627,
+      "grad_norm": 0.656775951385498,
+      "learning_rate": 6.6976939797501575e-06,
+      "loss": 0.7254,
+      "step": 14223
+    },
+    {
+      "epoch": 0.7828719247069184,
+      "grad_norm": 0.6901991963386536,
+      "learning_rate": 6.697286258116691e-06,
+      "loss": 0.7242,
+      "step": 14224
+    },
+    {
+      "epoch": 0.782926963509274,
+      "grad_norm": 0.8289571404457092,
+      "learning_rate": 6.696878523726875e-06,
+      "loss": 0.8578,
+      "step": 14225
+    },
+    {
+      "epoch": 0.7829820023116297,
+      "grad_norm": 0.6268846392631531,
+      "learning_rate": 6.696470776583775e-06,
+      "loss": 0.737,
+      "step": 14226
+    },
+    {
+      "epoch": 0.7830370411139853,
+      "grad_norm": 0.7026770114898682,
+      "learning_rate": 6.696063016690455e-06,
+      "loss": 0.6771,
+      "step": 14227
+    },
+    {
+      "epoch": 0.783092079916341,
+      "grad_norm": 0.7377839088439941,
+      "learning_rate": 6.69565524404998e-06,
+      "loss": 0.7174,
+      "step": 14228
+    },
+    {
+      "epoch": 0.7831471187186967,
+      "grad_norm": 0.6778523921966553,
+      "learning_rate": 6.695247458665414e-06,
+      "loss": 0.8255,
+      "step": 14229
+    },
+    {
+      "epoch": 0.7832021575210524,
+      "grad_norm": 0.7624330520629883,
+      "learning_rate": 6.69483966053982e-06,
+      "loss": 0.7495,
+      "step": 14230
+    },
+    {
+      "epoch": 0.783257196323408,
+      "grad_norm": 0.8944052457809448,
+      "learning_rate": 6.694431849676267e-06,
+      "loss": 0.868,
+      "step": 14231
+    },
+    {
+      "epoch": 0.7833122351257636,
+      "grad_norm": 0.7391701936721802,
+      "learning_rate": 6.694024026077816e-06,
+      "loss": 0.7032,
+      "step": 14232
+    },
+    {
+      "epoch": 0.7833672739281193,
+      "grad_norm": 0.7548620104789734,
+      "learning_rate": 6.693616189747535e-06,
+      "loss": 0.8272,
+      "step": 14233
+    },
+    {
+      "epoch": 0.783422312730475,
+      "grad_norm": 0.6795994639396667,
+      "learning_rate": 6.693208340688489e-06,
+      "loss": 0.703,
+      "step": 14234
+    },
+    {
+      "epoch": 0.7834773515328306,
+      "grad_norm": 0.6580816507339478,
+      "learning_rate": 6.69280047890374e-06,
+      "loss": 0.7454,
+      "step": 14235
+    },
+    {
+      "epoch": 0.7835323903351863,
+      "grad_norm": 0.7124443650245667,
+      "learning_rate": 6.6923926043963576e-06,
+      "loss": 0.6655,
+      "step": 14236
+    },
+    {
+      "epoch": 0.783587429137542,
+      "grad_norm": 0.6730241179466248,
+      "learning_rate": 6.691984717169404e-06,
+      "loss": 0.7522,
+      "step": 14237
+    },
+    {
+      "epoch": 0.7836424679398977,
+      "grad_norm": 0.8156033158302307,
+      "learning_rate": 6.6915768172259466e-06,
+      "loss": 0.8955,
+      "step": 14238
+    },
+    {
+      "epoch": 0.7836975067422532,
+      "grad_norm": 0.8041443228721619,
+      "learning_rate": 6.6911689045690506e-06,
+      "loss": 0.8019,
+      "step": 14239
+    },
+    {
+      "epoch": 0.7837525455446089,
+      "grad_norm": 0.7252053618431091,
+      "learning_rate": 6.690760979201782e-06,
+      "loss": 0.7014,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7838075843469646,
+      "grad_norm": 0.6969071626663208,
+      "learning_rate": 6.690353041127208e-06,
+      "loss": 0.7304,
+      "step": 14241
+    },
+    {
+      "epoch": 0.7838626231493203,
+      "grad_norm": 0.8254885673522949,
+      "learning_rate": 6.6899450903483906e-06,
+      "loss": 0.7193,
+      "step": 14242
+    },
+    {
+      "epoch": 0.7839176619516759,
+      "grad_norm": 0.7426590323448181,
+      "learning_rate": 6.6895371268684e-06,
+      "loss": 0.697,
+      "step": 14243
+    },
+    {
+      "epoch": 0.7839727007540316,
+      "grad_norm": 0.6744338274002075,
+      "learning_rate": 6.6891291506903e-06,
+      "loss": 0.8363,
+      "step": 14244
+    },
+    {
+      "epoch": 0.7840277395563873,
+      "grad_norm": 0.6609839797019958,
+      "learning_rate": 6.688721161817156e-06,
+      "loss": 0.7756,
+      "step": 14245
+    },
+    {
+      "epoch": 0.784082778358743,
+      "grad_norm": 0.8377131223678589,
+      "learning_rate": 6.688313160252038e-06,
+      "loss": 0.8355,
+      "step": 14246
+    },
+    {
+      "epoch": 0.7841378171610985,
+      "grad_norm": 0.6922308802604675,
+      "learning_rate": 6.687905145998009e-06,
+      "loss": 0.756,
+      "step": 14247
+    },
+    {
+      "epoch": 0.7841928559634542,
+      "grad_norm": 0.7217739820480347,
+      "learning_rate": 6.687497119058137e-06,
+      "loss": 0.7309,
+      "step": 14248
+    },
+    {
+      "epoch": 0.7842478947658099,
+      "grad_norm": 0.6906038522720337,
+      "learning_rate": 6.687089079435488e-06,
+      "loss": 0.6645,
+      "step": 14249
+    },
+    {
+      "epoch": 0.7843029335681656,
+      "grad_norm": 0.6800183057785034,
+      "learning_rate": 6.6866810271331305e-06,
+      "loss": 0.6791,
+      "step": 14250
+    },
+    {
+      "epoch": 0.7843579723705212,
+      "grad_norm": 0.6835503578186035,
+      "learning_rate": 6.686272962154129e-06,
+      "loss": 0.699,
+      "step": 14251
+    },
+    {
+      "epoch": 0.7844130111728769,
+      "grad_norm": 0.6643723845481873,
+      "learning_rate": 6.685864884501552e-06,
+      "loss": 0.7808,
+      "step": 14252
+    },
+    {
+      "epoch": 0.7844680499752326,
+      "grad_norm": 0.6742954850196838,
+      "learning_rate": 6.685456794178464e-06,
+      "loss": 0.7704,
+      "step": 14253
+    },
+    {
+      "epoch": 0.7845230887775883,
+      "grad_norm": 0.6374711990356445,
+      "learning_rate": 6.6850486911879355e-06,
+      "loss": 0.7557,
+      "step": 14254
+    },
+    {
+      "epoch": 0.7845781275799438,
+      "grad_norm": 0.7354347109794617,
+      "learning_rate": 6.684640575533031e-06,
+      "loss": 0.7928,
+      "step": 14255
+    },
+    {
+      "epoch": 0.7846331663822995,
+      "grad_norm": 0.6694937348365784,
+      "learning_rate": 6.684232447216821e-06,
+      "loss": 0.7247,
+      "step": 14256
+    },
+    {
+      "epoch": 0.7846882051846552,
+      "grad_norm": 0.716623842716217,
+      "learning_rate": 6.683824306242368e-06,
+      "loss": 0.8638,
+      "step": 14257
+    },
+    {
+      "epoch": 0.7847432439870109,
+      "grad_norm": 0.667164146900177,
+      "learning_rate": 6.683416152612743e-06,
+      "loss": 0.7455,
+      "step": 14258
+    },
+    {
+      "epoch": 0.7847982827893665,
+      "grad_norm": 0.7302100658416748,
+      "learning_rate": 6.683007986331014e-06,
+      "loss": 0.707,
+      "step": 14259
+    },
+    {
+      "epoch": 0.7848533215917222,
+      "grad_norm": 0.7605045437812805,
+      "learning_rate": 6.682599807400246e-06,
+      "loss": 0.7727,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7849083603940779,
+      "grad_norm": 0.6819437146186829,
+      "learning_rate": 6.682191615823508e-06,
+      "loss": 0.7538,
+      "step": 14261
+    },
+    {
+      "epoch": 0.7849633991964334,
+      "grad_norm": 0.7399439811706543,
+      "learning_rate": 6.6817834116038695e-06,
+      "loss": 0.7499,
+      "step": 14262
+    },
+    {
+      "epoch": 0.7850184379987891,
+      "grad_norm": 0.7864901423454285,
+      "learning_rate": 6.681375194744397e-06,
+      "loss": 0.7128,
+      "step": 14263
+    },
+    {
+      "epoch": 0.7850734768011448,
+      "grad_norm": 0.7308626174926758,
+      "learning_rate": 6.680966965248159e-06,
+      "loss": 0.7239,
+      "step": 14264
+    },
+    {
+      "epoch": 0.7851285156035005,
+      "grad_norm": 0.6553478837013245,
+      "learning_rate": 6.680558723118222e-06,
+      "loss": 0.6984,
+      "step": 14265
+    },
+    {
+      "epoch": 0.7851835544058561,
+      "grad_norm": 0.621415376663208,
+      "learning_rate": 6.680150468357656e-06,
+      "loss": 0.6428,
+      "step": 14266
+    },
+    {
+      "epoch": 0.7852385932082118,
+      "grad_norm": 1.0505764484405518,
+      "learning_rate": 6.679742200969529e-06,
+      "loss": 0.8073,
+      "step": 14267
+    },
+    {
+      "epoch": 0.7852936320105675,
+      "grad_norm": 0.7393355369567871,
+      "learning_rate": 6.67933392095691e-06,
+      "loss": 0.7396,
+      "step": 14268
+    },
+    {
+      "epoch": 0.7853486708129231,
+      "grad_norm": 0.7346563935279846,
+      "learning_rate": 6.678925628322864e-06,
+      "loss": 0.7398,
+      "step": 14269
+    },
+    {
+      "epoch": 0.7854037096152787,
+      "grad_norm": 0.6694674491882324,
+      "learning_rate": 6.678517323070465e-06,
+      "loss": 0.7346,
+      "step": 14270
+    },
+    {
+      "epoch": 0.7854587484176344,
+      "grad_norm": 0.6907033920288086,
+      "learning_rate": 6.678109005202779e-06,
+      "loss": 0.7617,
+      "step": 14271
+    },
+    {
+      "epoch": 0.7855137872199901,
+      "grad_norm": 0.6588131189346313,
+      "learning_rate": 6.677700674722873e-06,
+      "loss": 0.7514,
+      "step": 14272
+    },
+    {
+      "epoch": 0.7855688260223458,
+      "grad_norm": 0.6535136699676514,
+      "learning_rate": 6.677292331633819e-06,
+      "loss": 0.7154,
+      "step": 14273
+    },
+    {
+      "epoch": 0.7856238648247014,
+      "grad_norm": 0.7013682723045349,
+      "learning_rate": 6.676883975938685e-06,
+      "loss": 0.8506,
+      "step": 14274
+    },
+    {
+      "epoch": 0.7856789036270571,
+      "grad_norm": 0.7128416895866394,
+      "learning_rate": 6.67647560764054e-06,
+      "loss": 0.7669,
+      "step": 14275
+    },
+    {
+      "epoch": 0.7857339424294127,
+      "grad_norm": 0.7021318674087524,
+      "learning_rate": 6.676067226742453e-06,
+      "loss": 0.8236,
+      "step": 14276
+    },
+    {
+      "epoch": 0.7857889812317684,
+      "grad_norm": 0.7067561745643616,
+      "learning_rate": 6.675658833247493e-06,
+      "loss": 0.6848,
+      "step": 14277
+    },
+    {
+      "epoch": 0.785844020034124,
+      "grad_norm": 0.6488254070281982,
+      "learning_rate": 6.675250427158731e-06,
+      "loss": 0.7877,
+      "step": 14278
+    },
+    {
+      "epoch": 0.7858990588364797,
+      "grad_norm": 0.7153946757316589,
+      "learning_rate": 6.674842008479234e-06,
+      "loss": 0.7994,
+      "step": 14279
+    },
+    {
+      "epoch": 0.7859540976388354,
+      "grad_norm": 0.7290914058685303,
+      "learning_rate": 6.6744335772120735e-06,
+      "loss": 0.8074,
+      "step": 14280
+    },
+    {
+      "epoch": 0.7860091364411911,
+      "grad_norm": 0.726309061050415,
+      "learning_rate": 6.674025133360316e-06,
+      "loss": 0.7789,
+      "step": 14281
+    },
+    {
+      "epoch": 0.7860641752435467,
+      "grad_norm": 0.6294347047805786,
+      "learning_rate": 6.673616676927037e-06,
+      "loss": 0.6405,
+      "step": 14282
+    },
+    {
+      "epoch": 0.7861192140459023,
+      "grad_norm": 0.654400646686554,
+      "learning_rate": 6.673208207915302e-06,
+      "loss": 0.7876,
+      "step": 14283
+    },
+    {
+      "epoch": 0.786174252848258,
+      "grad_norm": 0.6729328632354736,
+      "learning_rate": 6.672799726328182e-06,
+      "loss": 0.7773,
+      "step": 14284
+    },
+    {
+      "epoch": 0.7862292916506137,
+      "grad_norm": 0.7607905268669128,
+      "learning_rate": 6.672391232168745e-06,
+      "loss": 0.8262,
+      "step": 14285
+    },
+    {
+      "epoch": 0.7862843304529693,
+      "grad_norm": 0.6475018858909607,
+      "learning_rate": 6.671982725440065e-06,
+      "loss": 0.7383,
+      "step": 14286
+    },
+    {
+      "epoch": 0.786339369255325,
+      "grad_norm": 0.8290789723396301,
+      "learning_rate": 6.671574206145211e-06,
+      "loss": 0.7968,
+      "step": 14287
+    },
+    {
+      "epoch": 0.7863944080576807,
+      "grad_norm": 0.7462177872657776,
+      "learning_rate": 6.671165674287252e-06,
+      "loss": 0.7465,
+      "step": 14288
+    },
+    {
+      "epoch": 0.7864494468600364,
+      "grad_norm": 0.7029373049736023,
+      "learning_rate": 6.6707571298692595e-06,
+      "loss": 0.7342,
+      "step": 14289
+    },
+    {
+      "epoch": 0.786504485662392,
+      "grad_norm": 0.8253761529922485,
+      "learning_rate": 6.670348572894303e-06,
+      "loss": 0.8196,
+      "step": 14290
+    },
+    {
+      "epoch": 0.7865595244647476,
+      "grad_norm": 0.7234970331192017,
+      "learning_rate": 6.669940003365455e-06,
+      "loss": 0.7966,
+      "step": 14291
+    },
+    {
+      "epoch": 0.7866145632671033,
+      "grad_norm": 0.8699348568916321,
+      "learning_rate": 6.6695314212857845e-06,
+      "loss": 0.8761,
+      "step": 14292
+    },
+    {
+      "epoch": 0.786669602069459,
+      "grad_norm": 0.6620158553123474,
+      "learning_rate": 6.66912282665836e-06,
+      "loss": 0.7534,
+      "step": 14293
+    },
+    {
+      "epoch": 0.7867246408718146,
+      "grad_norm": 0.6469776630401611,
+      "learning_rate": 6.668714219486259e-06,
+      "loss": 0.7812,
+      "step": 14294
+    },
+    {
+      "epoch": 0.7867796796741703,
+      "grad_norm": 0.6477407813072205,
+      "learning_rate": 6.668305599772546e-06,
+      "loss": 0.7144,
+      "step": 14295
+    },
+    {
+      "epoch": 0.786834718476526,
+      "grad_norm": 0.6626473665237427,
+      "learning_rate": 6.667896967520297e-06,
+      "loss": 0.7283,
+      "step": 14296
+    },
+    {
+      "epoch": 0.7868897572788817,
+      "grad_norm": 0.6214945316314697,
+      "learning_rate": 6.667488322732578e-06,
+      "loss": 0.6835,
+      "step": 14297
+    },
+    {
+      "epoch": 0.7869447960812372,
+      "grad_norm": 0.6199555397033691,
+      "learning_rate": 6.667079665412465e-06,
+      "loss": 0.706,
+      "step": 14298
+    },
+    {
+      "epoch": 0.7869998348835929,
+      "grad_norm": 0.8127612471580505,
+      "learning_rate": 6.666670995563027e-06,
+      "loss": 0.7099,
+      "step": 14299
+    },
+    {
+      "epoch": 0.7870548736859486,
+      "grad_norm": 0.6241362690925598,
+      "learning_rate": 6.6662623131873374e-06,
+      "loss": 0.7076,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7871099124883043,
+      "grad_norm": 0.7260692715644836,
+      "learning_rate": 6.665853618288465e-06,
+      "loss": 0.7842,
+      "step": 14301
+    },
+    {
+      "epoch": 0.7871649512906599,
+      "grad_norm": 0.6644107103347778,
+      "learning_rate": 6.665444910869482e-06,
+      "loss": 0.6515,
+      "step": 14302
+    },
+    {
+      "epoch": 0.7872199900930156,
+      "grad_norm": 0.6629641056060791,
+      "learning_rate": 6.6650361909334616e-06,
+      "loss": 0.7062,
+      "step": 14303
+    },
+    {
+      "epoch": 0.7872750288953713,
+      "grad_norm": 0.6616516709327698,
+      "learning_rate": 6.6646274584834745e-06,
+      "loss": 0.8195,
+      "step": 14304
+    },
+    {
+      "epoch": 0.7873300676977268,
+      "grad_norm": 0.7184805870056152,
+      "learning_rate": 6.664218713522593e-06,
+      "loss": 0.8699,
+      "step": 14305
+    },
+    {
+      "epoch": 0.7873851065000825,
+      "grad_norm": 0.6567219495773315,
+      "learning_rate": 6.6638099560538905e-06,
+      "loss": 0.7679,
+      "step": 14306
+    },
+    {
+      "epoch": 0.7874401453024382,
+      "grad_norm": 0.6952399611473083,
+      "learning_rate": 6.663401186080436e-06,
+      "loss": 0.603,
+      "step": 14307
+    },
+    {
+      "epoch": 0.7874951841047939,
+      "grad_norm": 0.7298767566680908,
+      "learning_rate": 6.662992403605304e-06,
+      "loss": 0.7655,
+      "step": 14308
+    },
+    {
+      "epoch": 0.7875502229071495,
+      "grad_norm": 0.7162219882011414,
+      "learning_rate": 6.662583608631567e-06,
+      "loss": 0.7797,
+      "step": 14309
+    },
+    {
+      "epoch": 0.7876052617095052,
+      "grad_norm": 0.6489827036857605,
+      "learning_rate": 6.662174801162296e-06,
+      "loss": 0.8165,
+      "step": 14310
+    },
+    {
+      "epoch": 0.7876603005118609,
+      "grad_norm": 0.7893611192703247,
+      "learning_rate": 6.6617659812005635e-06,
+      "loss": 0.8082,
+      "step": 14311
+    },
+    {
+      "epoch": 0.7877153393142166,
+      "grad_norm": 0.6709675192832947,
+      "learning_rate": 6.661357148749443e-06,
+      "loss": 0.7549,
+      "step": 14312
+    },
+    {
+      "epoch": 0.7877703781165721,
+      "grad_norm": 0.6166689991950989,
+      "learning_rate": 6.660948303812009e-06,
+      "loss": 0.7116,
+      "step": 14313
+    },
+    {
+      "epoch": 0.7878254169189278,
+      "grad_norm": 0.7941738367080688,
+      "learning_rate": 6.660539446391329e-06,
+      "loss": 0.7981,
+      "step": 14314
+    },
+    {
+      "epoch": 0.7878804557212835,
+      "grad_norm": 0.6339346170425415,
+      "learning_rate": 6.660130576490481e-06,
+      "loss": 0.7306,
+      "step": 14315
+    },
+    {
+      "epoch": 0.7879354945236392,
+      "grad_norm": 0.7044192552566528,
+      "learning_rate": 6.659721694112535e-06,
+      "loss": 0.7811,
+      "step": 14316
+    },
+    {
+      "epoch": 0.7879905333259948,
+      "grad_norm": 0.7853406071662903,
+      "learning_rate": 6.659312799260565e-06,
+      "loss": 0.7652,
+      "step": 14317
+    },
+    {
+      "epoch": 0.7880455721283505,
+      "grad_norm": 0.7076637148857117,
+      "learning_rate": 6.658903891937645e-06,
+      "loss": 0.7672,
+      "step": 14318
+    },
+    {
+      "epoch": 0.7881006109307062,
+      "grad_norm": 0.7043278813362122,
+      "learning_rate": 6.658494972146847e-06,
+      "loss": 0.726,
+      "step": 14319
+    },
+    {
+      "epoch": 0.7881556497330618,
+      "grad_norm": 0.8903809785842896,
+      "learning_rate": 6.658086039891245e-06,
+      "loss": 0.8,
+      "step": 14320
+    },
+    {
+      "epoch": 0.7882106885354174,
+      "grad_norm": 0.8239984512329102,
+      "learning_rate": 6.657677095173911e-06,
+      "loss": 0.7283,
+      "step": 14321
+    },
+    {
+      "epoch": 0.7882657273377731,
+      "grad_norm": 0.7221176028251648,
+      "learning_rate": 6.6572681379979206e-06,
+      "loss": 0.8058,
+      "step": 14322
+    },
+    {
+      "epoch": 0.7883207661401288,
+      "grad_norm": 0.8297285437583923,
+      "learning_rate": 6.6568591683663475e-06,
+      "loss": 0.8064,
+      "step": 14323
+    },
+    {
+      "epoch": 0.7883758049424845,
+      "grad_norm": 0.680659294128418,
+      "learning_rate": 6.656450186282264e-06,
+      "loss": 0.7259,
+      "step": 14324
+    },
+    {
+      "epoch": 0.7884308437448401,
+      "grad_norm": 0.7067807912826538,
+      "learning_rate": 6.656041191748744e-06,
+      "loss": 0.8414,
+      "step": 14325
+    },
+    {
+      "epoch": 0.7884858825471958,
+      "grad_norm": 0.6053900718688965,
+      "learning_rate": 6.655632184768861e-06,
+      "loss": 0.6762,
+      "step": 14326
+    },
+    {
+      "epoch": 0.7885409213495514,
+      "grad_norm": 0.7123621106147766,
+      "learning_rate": 6.65522316534569e-06,
+      "loss": 0.6968,
+      "step": 14327
+    },
+    {
+      "epoch": 0.7885959601519071,
+      "grad_norm": 0.7308228015899658,
+      "learning_rate": 6.6548141334823045e-06,
+      "loss": 0.6715,
+      "step": 14328
+    },
+    {
+      "epoch": 0.7886509989542627,
+      "grad_norm": 0.7508199214935303,
+      "learning_rate": 6.654405089181779e-06,
+      "loss": 0.7884,
+      "step": 14329
+    },
+    {
+      "epoch": 0.7887060377566184,
+      "grad_norm": 0.7317141890525818,
+      "learning_rate": 6.653996032447188e-06,
+      "loss": 0.7319,
+      "step": 14330
+    },
+    {
+      "epoch": 0.7887610765589741,
+      "grad_norm": 0.6797091364860535,
+      "learning_rate": 6.653586963281607e-06,
+      "loss": 0.7898,
+      "step": 14331
+    },
+    {
+      "epoch": 0.7888161153613298,
+      "grad_norm": 0.6293582320213318,
+      "learning_rate": 6.6531778816881065e-06,
+      "loss": 0.6784,
+      "step": 14332
+    },
+    {
+      "epoch": 0.7888711541636854,
+      "grad_norm": 0.7604238986968994,
+      "learning_rate": 6.652768787669763e-06,
+      "loss": 0.7226,
+      "step": 14333
+    },
+    {
+      "epoch": 0.788926192966041,
+      "grad_norm": 0.6921128034591675,
+      "learning_rate": 6.652359681229654e-06,
+      "loss": 0.7375,
+      "step": 14334
+    },
+    {
+      "epoch": 0.7889812317683967,
+      "grad_norm": 0.6532993316650391,
+      "learning_rate": 6.651950562370851e-06,
+      "loss": 0.703,
+      "step": 14335
+    },
+    {
+      "epoch": 0.7890362705707524,
+      "grad_norm": 0.6739360094070435,
+      "learning_rate": 6.651541431096431e-06,
+      "loss": 0.7488,
+      "step": 14336
+    },
+    {
+      "epoch": 0.789091309373108,
+      "grad_norm": 0.7503200173377991,
+      "learning_rate": 6.651132287409466e-06,
+      "loss": 0.7492,
+      "step": 14337
+    },
+    {
+      "epoch": 0.7891463481754637,
+      "grad_norm": 0.6537551879882812,
+      "learning_rate": 6.650723131313035e-06,
+      "loss": 0.723,
+      "step": 14338
+    },
+    {
+      "epoch": 0.7892013869778194,
+      "grad_norm": 0.6378511786460876,
+      "learning_rate": 6.650313962810208e-06,
+      "loss": 0.7764,
+      "step": 14339
+    },
+    {
+      "epoch": 0.7892564257801751,
+      "grad_norm": 0.7948685884475708,
+      "learning_rate": 6.649904781904065e-06,
+      "loss": 0.7996,
+      "step": 14340
+    },
+    {
+      "epoch": 0.7893114645825307,
+      "grad_norm": 0.7558071613311768,
+      "learning_rate": 6.649495588597678e-06,
+      "loss": 0.8249,
+      "step": 14341
+    },
+    {
+      "epoch": 0.7893665033848863,
+      "grad_norm": 0.7158063054084778,
+      "learning_rate": 6.649086382894124e-06,
+      "loss": 0.815,
+      "step": 14342
+    },
+    {
+      "epoch": 0.789421542187242,
+      "grad_norm": 0.7551599144935608,
+      "learning_rate": 6.648677164796479e-06,
+      "loss": 0.7151,
+      "step": 14343
+    },
+    {
+      "epoch": 0.7894765809895977,
+      "grad_norm": 0.6966339349746704,
+      "learning_rate": 6.648267934307817e-06,
+      "loss": 0.8057,
+      "step": 14344
+    },
+    {
+      "epoch": 0.7895316197919533,
+      "grad_norm": 0.6863396167755127,
+      "learning_rate": 6.647858691431214e-06,
+      "loss": 0.7819,
+      "step": 14345
+    },
+    {
+      "epoch": 0.789586658594309,
+      "grad_norm": 0.7352383136749268,
+      "learning_rate": 6.647449436169747e-06,
+      "loss": 0.8101,
+      "step": 14346
+    },
+    {
+      "epoch": 0.7896416973966647,
+      "grad_norm": 0.7630855441093445,
+      "learning_rate": 6.64704016852649e-06,
+      "loss": 0.7155,
+      "step": 14347
+    },
+    {
+      "epoch": 0.7896967361990203,
+      "grad_norm": 0.6740198135375977,
+      "learning_rate": 6.646630888504522e-06,
+      "loss": 0.7255,
+      "step": 14348
+    },
+    {
+      "epoch": 0.7897517750013759,
+      "grad_norm": 0.7095367908477783,
+      "learning_rate": 6.646221596106917e-06,
+      "loss": 0.7527,
+      "step": 14349
+    },
+    {
+      "epoch": 0.7898068138037316,
+      "grad_norm": 0.6096131801605225,
+      "learning_rate": 6.645812291336749e-06,
+      "loss": 0.7116,
+      "step": 14350
+    },
+    {
+      "epoch": 0.7898618526060873,
+      "grad_norm": 0.7212585210800171,
+      "learning_rate": 6.645402974197097e-06,
+      "loss": 0.7647,
+      "step": 14351
+    },
+    {
+      "epoch": 0.7899168914084429,
+      "grad_norm": 0.7145454287528992,
+      "learning_rate": 6.6449936446910376e-06,
+      "loss": 0.7988,
+      "step": 14352
+    },
+    {
+      "epoch": 0.7899719302107986,
+      "grad_norm": 0.668269693851471,
+      "learning_rate": 6.644584302821646e-06,
+      "loss": 0.8453,
+      "step": 14353
+    },
+    {
+      "epoch": 0.7900269690131543,
+      "grad_norm": 0.7431649565696716,
+      "learning_rate": 6.644174948591998e-06,
+      "loss": 0.6981,
+      "step": 14354
+    },
+    {
+      "epoch": 0.79008200781551,
+      "grad_norm": 0.6727485060691833,
+      "learning_rate": 6.643765582005172e-06,
+      "loss": 0.792,
+      "step": 14355
+    },
+    {
+      "epoch": 0.7901370466178655,
+      "grad_norm": 0.7102059721946716,
+      "learning_rate": 6.643356203064244e-06,
+      "loss": 0.7469,
+      "step": 14356
+    },
+    {
+      "epoch": 0.7901920854202212,
+      "grad_norm": 0.6719706654548645,
+      "learning_rate": 6.642946811772291e-06,
+      "loss": 0.7542,
+      "step": 14357
+    },
+    {
+      "epoch": 0.7902471242225769,
+      "grad_norm": 0.7044880986213684,
+      "learning_rate": 6.6425374081323875e-06,
+      "loss": 0.7884,
+      "step": 14358
+    },
+    {
+      "epoch": 0.7903021630249326,
+      "grad_norm": 0.656411349773407,
+      "learning_rate": 6.642127992147614e-06,
+      "loss": 0.7596,
+      "step": 14359
+    },
+    {
+      "epoch": 0.7903572018272882,
+      "grad_norm": 0.6256445050239563,
+      "learning_rate": 6.641718563821047e-06,
+      "loss": 0.6257,
+      "step": 14360
+    },
+    {
+      "epoch": 0.7904122406296439,
+      "grad_norm": 0.6761715412139893,
+      "learning_rate": 6.641309123155761e-06,
+      "loss": 0.7024,
+      "step": 14361
+    },
+    {
+      "epoch": 0.7904672794319996,
+      "grad_norm": 0.7567794322967529,
+      "learning_rate": 6.640899670154837e-06,
+      "loss": 0.7948,
+      "step": 14362
+    },
+    {
+      "epoch": 0.7905223182343553,
+      "grad_norm": 0.6192977428436279,
+      "learning_rate": 6.640490204821349e-06,
+      "loss": 0.7307,
+      "step": 14363
+    },
+    {
+      "epoch": 0.7905773570367108,
+      "grad_norm": 0.8120929002761841,
+      "learning_rate": 6.640080727158376e-06,
+      "loss": 0.7173,
+      "step": 14364
+    },
+    {
+      "epoch": 0.7906323958390665,
+      "grad_norm": 0.7303271293640137,
+      "learning_rate": 6.639671237168996e-06,
+      "loss": 0.8118,
+      "step": 14365
+    },
+    {
+      "epoch": 0.7906874346414222,
+      "grad_norm": 0.6731529831886292,
+      "learning_rate": 6.639261734856284e-06,
+      "loss": 0.76,
+      "step": 14366
+    },
+    {
+      "epoch": 0.7907424734437779,
+      "grad_norm": 0.6909935474395752,
+      "learning_rate": 6.638852220223321e-06,
+      "loss": 0.7732,
+      "step": 14367
+    },
+    {
+      "epoch": 0.7907975122461335,
+      "grad_norm": 0.6543979048728943,
+      "learning_rate": 6.638442693273183e-06,
+      "loss": 0.7408,
+      "step": 14368
+    },
+    {
+      "epoch": 0.7908525510484892,
+      "grad_norm": 0.6411511301994324,
+      "learning_rate": 6.6380331540089485e-06,
+      "loss": 0.6963,
+      "step": 14369
+    },
+    {
+      "epoch": 0.7909075898508449,
+      "grad_norm": 0.6657214164733887,
+      "learning_rate": 6.637623602433694e-06,
+      "loss": 0.7417,
+      "step": 14370
+    },
+    {
+      "epoch": 0.7909626286532006,
+      "grad_norm": 0.6852405071258545,
+      "learning_rate": 6.6372140385505e-06,
+      "loss": 0.7176,
+      "step": 14371
+    },
+    {
+      "epoch": 0.7910176674555561,
+      "grad_norm": 0.6453777551651001,
+      "learning_rate": 6.636804462362444e-06,
+      "loss": 0.7791,
+      "step": 14372
+    },
+    {
+      "epoch": 0.7910727062579118,
+      "grad_norm": 0.6806328296661377,
+      "learning_rate": 6.636394873872603e-06,
+      "loss": 0.7856,
+      "step": 14373
+    },
+    {
+      "epoch": 0.7911277450602675,
+      "grad_norm": 0.6819495558738708,
+      "learning_rate": 6.635985273084058e-06,
+      "loss": 0.7865,
+      "step": 14374
+    },
+    {
+      "epoch": 0.7911827838626232,
+      "grad_norm": 0.7372999787330627,
+      "learning_rate": 6.635575659999883e-06,
+      "loss": 0.8549,
+      "step": 14375
+    },
+    {
+      "epoch": 0.7912378226649788,
+      "grad_norm": 0.8146817684173584,
+      "learning_rate": 6.635166034623162e-06,
+      "loss": 0.7253,
+      "step": 14376
+    },
+    {
+      "epoch": 0.7912928614673345,
+      "grad_norm": 0.8205630779266357,
+      "learning_rate": 6.634756396956969e-06,
+      "loss": 0.6915,
+      "step": 14377
+    },
+    {
+      "epoch": 0.7913479002696902,
+      "grad_norm": 0.7168713808059692,
+      "learning_rate": 6.634346747004383e-06,
+      "loss": 0.7495,
+      "step": 14378
+    },
+    {
+      "epoch": 0.7914029390720458,
+      "grad_norm": 0.7210709452629089,
+      "learning_rate": 6.6339370847684854e-06,
+      "loss": 0.7323,
+      "step": 14379
+    },
+    {
+      "epoch": 0.7914579778744014,
+      "grad_norm": 0.9042065143585205,
+      "learning_rate": 6.633527410252355e-06,
+      "loss": 0.847,
+      "step": 14380
+    },
+    {
+      "epoch": 0.7915130166767571,
+      "grad_norm": 0.6700118184089661,
+      "learning_rate": 6.633117723459071e-06,
+      "loss": 0.7975,
+      "step": 14381
+    },
+    {
+      "epoch": 0.7915680554791128,
+      "grad_norm": 0.6355725526809692,
+      "learning_rate": 6.632708024391707e-06,
+      "loss": 0.7398,
+      "step": 14382
+    },
+    {
+      "epoch": 0.7916230942814685,
+      "grad_norm": 0.8274535536766052,
+      "learning_rate": 6.6322983130533505e-06,
+      "loss": 0.8641,
+      "step": 14383
+    },
+    {
+      "epoch": 0.7916781330838241,
+      "grad_norm": 0.5835573077201843,
+      "learning_rate": 6.631888589447075e-06,
+      "loss": 0.636,
+      "step": 14384
+    },
+    {
+      "epoch": 0.7917331718861798,
+      "grad_norm": 0.6933130621910095,
+      "learning_rate": 6.631478853575963e-06,
+      "loss": 0.7874,
+      "step": 14385
+    },
+    {
+      "epoch": 0.7917882106885354,
+      "grad_norm": 0.8125241994857788,
+      "learning_rate": 6.631069105443092e-06,
+      "loss": 0.7961,
+      "step": 14386
+    },
+    {
+      "epoch": 0.7918432494908911,
+      "grad_norm": 0.6661116480827332,
+      "learning_rate": 6.630659345051542e-06,
+      "loss": 0.6498,
+      "step": 14387
+    },
+    {
+      "epoch": 0.7918982882932467,
+      "grad_norm": 0.6807548403739929,
+      "learning_rate": 6.630249572404393e-06,
+      "loss": 0.6952,
+      "step": 14388
+    },
+    {
+      "epoch": 0.7919533270956024,
+      "grad_norm": 0.6886214017868042,
+      "learning_rate": 6.629839787504726e-06,
+      "loss": 0.7416,
+      "step": 14389
+    },
+    {
+      "epoch": 0.7920083658979581,
+      "grad_norm": 0.7633732557296753,
+      "learning_rate": 6.629429990355617e-06,
+      "loss": 0.8008,
+      "step": 14390
+    },
+    {
+      "epoch": 0.7920634047003137,
+      "grad_norm": 0.8401023745536804,
+      "learning_rate": 6.6290201809601494e-06,
+      "loss": 0.8312,
+      "step": 14391
+    },
+    {
+      "epoch": 0.7921184435026694,
+      "grad_norm": 0.6608526706695557,
+      "learning_rate": 6.628610359321403e-06,
+      "loss": 0.563,
+      "step": 14392
+    },
+    {
+      "epoch": 0.792173482305025,
+      "grad_norm": 0.687045156955719,
+      "learning_rate": 6.6282005254424566e-06,
+      "loss": 0.7451,
+      "step": 14393
+    },
+    {
+      "epoch": 0.7922285211073807,
+      "grad_norm": 0.7129287123680115,
+      "learning_rate": 6.627790679326389e-06,
+      "loss": 0.8495,
+      "step": 14394
+    },
+    {
+      "epoch": 0.7922835599097363,
+      "grad_norm": 0.6951952576637268,
+      "learning_rate": 6.627380820976283e-06,
+      "loss": 0.7895,
+      "step": 14395
+    },
+    {
+      "epoch": 0.792338598712092,
+      "grad_norm": 0.8020780086517334,
+      "learning_rate": 6.626970950395221e-06,
+      "loss": 0.7136,
+      "step": 14396
+    },
+    {
+      "epoch": 0.7923936375144477,
+      "grad_norm": 0.6654007434844971,
+      "learning_rate": 6.626561067586279e-06,
+      "loss": 0.7865,
+      "step": 14397
+    },
+    {
+      "epoch": 0.7924486763168034,
+      "grad_norm": 0.844744861125946,
+      "learning_rate": 6.62615117255254e-06,
+      "loss": 0.7856,
+      "step": 14398
+    },
+    {
+      "epoch": 0.792503715119159,
+      "grad_norm": 0.6890879273414612,
+      "learning_rate": 6.625741265297083e-06,
+      "loss": 0.7574,
+      "step": 14399
+    },
+    {
+      "epoch": 0.7925587539215146,
+      "grad_norm": 0.7559735774993896,
+      "learning_rate": 6.625331345822992e-06,
+      "loss": 0.634,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7926137927238703,
+      "grad_norm": 0.6918107867240906,
+      "learning_rate": 6.624921414133344e-06,
+      "loss": 0.6935,
+      "step": 14401
+    },
+    {
+      "epoch": 0.792668831526226,
+      "grad_norm": 0.7468792200088501,
+      "learning_rate": 6.624511470231221e-06,
+      "loss": 0.7301,
+      "step": 14402
+    },
+    {
+      "epoch": 0.7927238703285816,
+      "grad_norm": 0.6749486327171326,
+      "learning_rate": 6.624101514119705e-06,
+      "loss": 0.7143,
+      "step": 14403
+    },
+    {
+      "epoch": 0.7927789091309373,
+      "grad_norm": 0.7765836119651794,
+      "learning_rate": 6.623691545801878e-06,
+      "loss": 0.7201,
+      "step": 14404
+    },
+    {
+      "epoch": 0.792833947933293,
+      "grad_norm": 0.6263312697410583,
+      "learning_rate": 6.623281565280819e-06,
+      "loss": 0.5866,
+      "step": 14405
+    },
+    {
+      "epoch": 0.7928889867356487,
+      "grad_norm": 0.6325232982635498,
+      "learning_rate": 6.62287157255961e-06,
+      "loss": 0.7389,
+      "step": 14406
+    },
+    {
+      "epoch": 0.7929440255380042,
+      "grad_norm": 0.7165958881378174,
+      "learning_rate": 6.622461567641333e-06,
+      "loss": 0.7378,
+      "step": 14407
+    },
+    {
+      "epoch": 0.7929990643403599,
+      "grad_norm": 0.7611519694328308,
+      "learning_rate": 6.62205155052907e-06,
+      "loss": 0.7146,
+      "step": 14408
+    },
+    {
+      "epoch": 0.7930541031427156,
+      "grad_norm": 0.6764969825744629,
+      "learning_rate": 6.6216415212259e-06,
+      "loss": 0.7802,
+      "step": 14409
+    },
+    {
+      "epoch": 0.7931091419450713,
+      "grad_norm": 0.7266956567764282,
+      "learning_rate": 6.621231479734908e-06,
+      "loss": 0.7065,
+      "step": 14410
+    },
+    {
+      "epoch": 0.7931641807474269,
+      "grad_norm": 0.7540454268455505,
+      "learning_rate": 6.620821426059174e-06,
+      "loss": 0.7327,
+      "step": 14411
+    },
+    {
+      "epoch": 0.7932192195497826,
+      "grad_norm": 0.7931423783302307,
+      "learning_rate": 6.620411360201779e-06,
+      "loss": 0.8032,
+      "step": 14412
+    },
+    {
+      "epoch": 0.7932742583521383,
+      "grad_norm": 1.2976648807525635,
+      "learning_rate": 6.620001282165808e-06,
+      "loss": 0.7422,
+      "step": 14413
+    },
+    {
+      "epoch": 0.793329297154494,
+      "grad_norm": 0.6525906920433044,
+      "learning_rate": 6.619591191954338e-06,
+      "loss": 0.6857,
+      "step": 14414
+    },
+    {
+      "epoch": 0.7933843359568495,
+      "grad_norm": 0.6153263449668884,
+      "learning_rate": 6.619181089570456e-06,
+      "loss": 0.6117,
+      "step": 14415
+    },
+    {
+      "epoch": 0.7934393747592052,
+      "grad_norm": 0.7076815962791443,
+      "learning_rate": 6.6187709750172425e-06,
+      "loss": 0.8053,
+      "step": 14416
+    },
+    {
+      "epoch": 0.7934944135615609,
+      "grad_norm": 0.6999046802520752,
+      "learning_rate": 6.618360848297779e-06,
+      "loss": 0.6275,
+      "step": 14417
+    },
+    {
+      "epoch": 0.7935494523639166,
+      "grad_norm": 0.7043859958648682,
+      "learning_rate": 6.6179507094151484e-06,
+      "loss": 0.8273,
+      "step": 14418
+    },
+    {
+      "epoch": 0.7936044911662722,
+      "grad_norm": 0.6295393705368042,
+      "learning_rate": 6.617540558372434e-06,
+      "loss": 0.6394,
+      "step": 14419
+    },
+    {
+      "epoch": 0.7936595299686279,
+      "grad_norm": 0.8165664076805115,
+      "learning_rate": 6.617130395172718e-06,
+      "loss": 0.8473,
+      "step": 14420
+    },
+    {
+      "epoch": 0.7937145687709836,
+      "grad_norm": 0.7598135471343994,
+      "learning_rate": 6.616720219819082e-06,
+      "loss": 0.729,
+      "step": 14421
+    },
+    {
+      "epoch": 0.7937696075733393,
+      "grad_norm": 0.7222034335136414,
+      "learning_rate": 6.6163100323146105e-06,
+      "loss": 0.7526,
+      "step": 14422
+    },
+    {
+      "epoch": 0.7938246463756948,
+      "grad_norm": 0.7994693517684937,
+      "learning_rate": 6.615899832662385e-06,
+      "loss": 0.8346,
+      "step": 14423
+    },
+    {
+      "epoch": 0.7938796851780505,
+      "grad_norm": 0.6603162884712219,
+      "learning_rate": 6.615489620865489e-06,
+      "loss": 0.7546,
+      "step": 14424
+    },
+    {
+      "epoch": 0.7939347239804062,
+      "grad_norm": 0.6525929570198059,
+      "learning_rate": 6.615079396927005e-06,
+      "loss": 0.7344,
+      "step": 14425
+    },
+    {
+      "epoch": 0.7939897627827619,
+      "grad_norm": 0.6144835948944092,
+      "learning_rate": 6.614669160850016e-06,
+      "loss": 0.6776,
+      "step": 14426
+    },
+    {
+      "epoch": 0.7940448015851175,
+      "grad_norm": 0.7205507159233093,
+      "learning_rate": 6.614258912637607e-06,
+      "loss": 0.809,
+      "step": 14427
+    },
+    {
+      "epoch": 0.7940998403874732,
+      "grad_norm": 0.6757732629776001,
+      "learning_rate": 6.61384865229286e-06,
+      "loss": 0.7403,
+      "step": 14428
+    },
+    {
+      "epoch": 0.7941548791898289,
+      "grad_norm": 0.6392103433609009,
+      "learning_rate": 6.6134383798188586e-06,
+      "loss": 0.7689,
+      "step": 14429
+    },
+    {
+      "epoch": 0.7942099179921845,
+      "grad_norm": 0.6647289395332336,
+      "learning_rate": 6.613028095218685e-06,
+      "loss": 0.6611,
+      "step": 14430
+    },
+    {
+      "epoch": 0.7942649567945401,
+      "grad_norm": 0.6961668133735657,
+      "learning_rate": 6.612617798495426e-06,
+      "loss": 0.7784,
+      "step": 14431
+    },
+    {
+      "epoch": 0.7943199955968958,
+      "grad_norm": 1.1188037395477295,
+      "learning_rate": 6.6122074896521615e-06,
+      "loss": 0.6518,
+      "step": 14432
+    },
+    {
+      "epoch": 0.7943750343992515,
+      "grad_norm": 0.6382507085800171,
+      "learning_rate": 6.611797168691978e-06,
+      "loss": 0.6954,
+      "step": 14433
+    },
+    {
+      "epoch": 0.7944300732016071,
+      "grad_norm": 0.6720117330551147,
+      "learning_rate": 6.6113868356179585e-06,
+      "loss": 0.7267,
+      "step": 14434
+    },
+    {
+      "epoch": 0.7944851120039628,
+      "grad_norm": 0.6667274832725525,
+      "learning_rate": 6.610976490433186e-06,
+      "loss": 0.6867,
+      "step": 14435
+    },
+    {
+      "epoch": 0.7945401508063185,
+      "grad_norm": 0.658217191696167,
+      "learning_rate": 6.610566133140747e-06,
+      "loss": 0.66,
+      "step": 14436
+    },
+    {
+      "epoch": 0.7945951896086741,
+      "grad_norm": 0.6820386648178101,
+      "learning_rate": 6.610155763743723e-06,
+      "loss": 0.7352,
+      "step": 14437
+    },
+    {
+      "epoch": 0.7946502284110297,
+      "grad_norm": 0.788696825504303,
+      "learning_rate": 6.609745382245198e-06,
+      "loss": 0.6822,
+      "step": 14438
+    },
+    {
+      "epoch": 0.7947052672133854,
+      "grad_norm": 0.6485540270805359,
+      "learning_rate": 6.6093349886482596e-06,
+      "loss": 0.718,
+      "step": 14439
+    },
+    {
+      "epoch": 0.7947603060157411,
+      "grad_norm": 0.717659056186676,
+      "learning_rate": 6.60892458295599e-06,
+      "loss": 0.7898,
+      "step": 14440
+    },
+    {
+      "epoch": 0.7948153448180968,
+      "grad_norm": 0.6576352119445801,
+      "learning_rate": 6.608514165171473e-06,
+      "loss": 0.8041,
+      "step": 14441
+    },
+    {
+      "epoch": 0.7948703836204524,
+      "grad_norm": 0.7034726738929749,
+      "learning_rate": 6.608103735297795e-06,
+      "loss": 0.7901,
+      "step": 14442
+    },
+    {
+      "epoch": 0.7949254224228081,
+      "grad_norm": 0.7001451253890991,
+      "learning_rate": 6.6076932933380386e-06,
+      "loss": 0.6814,
+      "step": 14443
+    },
+    {
+      "epoch": 0.7949804612251637,
+      "grad_norm": 0.789359450340271,
+      "learning_rate": 6.607282839295291e-06,
+      "loss": 0.744,
+      "step": 14444
+    },
+    {
+      "epoch": 0.7950355000275194,
+      "grad_norm": 0.7830412983894348,
+      "learning_rate": 6.606872373172636e-06,
+      "loss": 0.8161,
+      "step": 14445
+    },
+    {
+      "epoch": 0.795090538829875,
+      "grad_norm": 0.6462455987930298,
+      "learning_rate": 6.606461894973157e-06,
+      "loss": 0.7723,
+      "step": 14446
+    },
+    {
+      "epoch": 0.7951455776322307,
+      "grad_norm": 0.6232526898384094,
+      "learning_rate": 6.606051404699943e-06,
+      "loss": 0.6723,
+      "step": 14447
+    },
+    {
+      "epoch": 0.7952006164345864,
+      "grad_norm": 0.7790026068687439,
+      "learning_rate": 6.605640902356074e-06,
+      "loss": 0.7687,
+      "step": 14448
+    },
+    {
+      "epoch": 0.7952556552369421,
+      "grad_norm": 0.7281851768493652,
+      "learning_rate": 6.605230387944639e-06,
+      "loss": 0.827,
+      "step": 14449
+    },
+    {
+      "epoch": 0.7953106940392977,
+      "grad_norm": 0.6519556045532227,
+      "learning_rate": 6.604819861468721e-06,
+      "loss": 0.7039,
+      "step": 14450
+    },
+    {
+      "epoch": 0.7953657328416534,
+      "grad_norm": 0.6768763661384583,
+      "learning_rate": 6.604409322931406e-06,
+      "loss": 0.7288,
+      "step": 14451
+    },
+    {
+      "epoch": 0.795420771644009,
+      "grad_norm": 0.7457320094108582,
+      "learning_rate": 6.6039987723357825e-06,
+      "loss": 0.8386,
+      "step": 14452
+    },
+    {
+      "epoch": 0.7954758104463647,
+      "grad_norm": 0.9579072594642639,
+      "learning_rate": 6.6035882096849325e-06,
+      "loss": 0.7552,
+      "step": 14453
+    },
+    {
+      "epoch": 0.7955308492487203,
+      "grad_norm": 0.6709916591644287,
+      "learning_rate": 6.603177634981941e-06,
+      "loss": 0.724,
+      "step": 14454
+    },
+    {
+      "epoch": 0.795585888051076,
+      "grad_norm": 0.6097317934036255,
+      "learning_rate": 6.602767048229897e-06,
+      "loss": 0.6866,
+      "step": 14455
+    },
+    {
+      "epoch": 0.7956409268534317,
+      "grad_norm": 0.7303394675254822,
+      "learning_rate": 6.602356449431885e-06,
+      "loss": 0.682,
+      "step": 14456
+    },
+    {
+      "epoch": 0.7956959656557874,
+      "grad_norm": 0.775979220867157,
+      "learning_rate": 6.601945838590991e-06,
+      "loss": 0.7784,
+      "step": 14457
+    },
+    {
+      "epoch": 0.795751004458143,
+      "grad_norm": 0.7016483545303345,
+      "learning_rate": 6.6015352157103e-06,
+      "loss": 0.7557,
+      "step": 14458
+    },
+    {
+      "epoch": 0.7958060432604986,
+      "grad_norm": 0.688946545124054,
+      "learning_rate": 6.6011245807929e-06,
+      "loss": 0.707,
+      "step": 14459
+    },
+    {
+      "epoch": 0.7958610820628543,
+      "grad_norm": 0.7286174297332764,
+      "learning_rate": 6.600713933841877e-06,
+      "loss": 0.784,
+      "step": 14460
+    },
+    {
+      "epoch": 0.79591612086521,
+      "grad_norm": 0.7604749798774719,
+      "learning_rate": 6.600303274860316e-06,
+      "loss": 0.7099,
+      "step": 14461
+    },
+    {
+      "epoch": 0.7959711596675656,
+      "grad_norm": 0.6626706123352051,
+      "learning_rate": 6.599892603851301e-06,
+      "loss": 0.7137,
+      "step": 14462
+    },
+    {
+      "epoch": 0.7960261984699213,
+      "grad_norm": 0.7692080736160278,
+      "learning_rate": 6.599481920817925e-06,
+      "loss": 0.847,
+      "step": 14463
+    },
+    {
+      "epoch": 0.796081237272277,
+      "grad_norm": 0.6811042428016663,
+      "learning_rate": 6.599071225763269e-06,
+      "loss": 0.7888,
+      "step": 14464
+    },
+    {
+      "epoch": 0.7961362760746327,
+      "grad_norm": 0.654481053352356,
+      "learning_rate": 6.598660518690424e-06,
+      "loss": 0.6973,
+      "step": 14465
+    },
+    {
+      "epoch": 0.7961913148769882,
+      "grad_norm": 0.7332738637924194,
+      "learning_rate": 6.598249799602472e-06,
+      "loss": 0.8311,
+      "step": 14466
+    },
+    {
+      "epoch": 0.7962463536793439,
+      "grad_norm": 0.7098381519317627,
+      "learning_rate": 6.597839068502503e-06,
+      "loss": 0.8265,
+      "step": 14467
+    },
+    {
+      "epoch": 0.7963013924816996,
+      "grad_norm": 0.6338212490081787,
+      "learning_rate": 6.597428325393604e-06,
+      "loss": 0.6889,
+      "step": 14468
+    },
+    {
+      "epoch": 0.7963564312840553,
+      "grad_norm": 0.7001339197158813,
+      "learning_rate": 6.597017570278861e-06,
+      "loss": 0.7613,
+      "step": 14469
+    },
+    {
+      "epoch": 0.7964114700864109,
+      "grad_norm": 0.6565783619880676,
+      "learning_rate": 6.596606803161361e-06,
+      "loss": 0.6284,
+      "step": 14470
+    },
+    {
+      "epoch": 0.7964665088887666,
+      "grad_norm": 0.6638015508651733,
+      "learning_rate": 6.5961960240441935e-06,
+      "loss": 0.6635,
+      "step": 14471
+    },
+    {
+      "epoch": 0.7965215476911223,
+      "grad_norm": 0.6389575600624084,
+      "learning_rate": 6.595785232930443e-06,
+      "loss": 0.6588,
+      "step": 14472
+    },
+    {
+      "epoch": 0.796576586493478,
+      "grad_norm": 0.9486858248710632,
+      "learning_rate": 6.595374429823197e-06,
+      "loss": 0.8314,
+      "step": 14473
+    },
+    {
+      "epoch": 0.7966316252958335,
+      "grad_norm": 0.7555649280548096,
+      "learning_rate": 6.594963614725544e-06,
+      "loss": 0.8173,
+      "step": 14474
+    },
+    {
+      "epoch": 0.7966866640981892,
+      "grad_norm": 0.63021320104599,
+      "learning_rate": 6.5945527876405715e-06,
+      "loss": 0.7038,
+      "step": 14475
+    },
+    {
+      "epoch": 0.7967417029005449,
+      "grad_norm": 0.802980899810791,
+      "learning_rate": 6.594141948571366e-06,
+      "loss": 0.8031,
+      "step": 14476
+    },
+    {
+      "epoch": 0.7967967417029005,
+      "grad_norm": 0.7204614281654358,
+      "learning_rate": 6.593731097521019e-06,
+      "loss": 0.827,
+      "step": 14477
+    },
+    {
+      "epoch": 0.7968517805052562,
+      "grad_norm": 0.6805211305618286,
+      "learning_rate": 6.593320234492613e-06,
+      "loss": 0.7405,
+      "step": 14478
+    },
+    {
+      "epoch": 0.7969068193076119,
+      "grad_norm": 0.7011345028877258,
+      "learning_rate": 6.59290935948924e-06,
+      "loss": 0.7241,
+      "step": 14479
+    },
+    {
+      "epoch": 0.7969618581099676,
+      "grad_norm": 0.8995540738105774,
+      "learning_rate": 6.592498472513986e-06,
+      "loss": 0.6864,
+      "step": 14480
+    },
+    {
+      "epoch": 0.7970168969123231,
+      "grad_norm": 0.7518284320831299,
+      "learning_rate": 6.592087573569941e-06,
+      "loss": 0.7561,
+      "step": 14481
+    },
+    {
+      "epoch": 0.7970719357146788,
+      "grad_norm": 0.6359231472015381,
+      "learning_rate": 6.591676662660191e-06,
+      "loss": 0.6402,
+      "step": 14482
+    },
+    {
+      "epoch": 0.7971269745170345,
+      "grad_norm": 0.6610120534896851,
+      "learning_rate": 6.5912657397878264e-06,
+      "loss": 0.6419,
+      "step": 14483
+    },
+    {
+      "epoch": 0.7971820133193902,
+      "grad_norm": 0.7054341435432434,
+      "learning_rate": 6.590854804955934e-06,
+      "loss": 0.7252,
+      "step": 14484
+    },
+    {
+      "epoch": 0.7972370521217458,
+      "grad_norm": 0.6929903626441956,
+      "learning_rate": 6.5904438581676025e-06,
+      "loss": 0.6566,
+      "step": 14485
+    },
+    {
+      "epoch": 0.7972920909241015,
+      "grad_norm": 0.7354124188423157,
+      "learning_rate": 6.59003289942592e-06,
+      "loss": 0.763,
+      "step": 14486
+    },
+    {
+      "epoch": 0.7973471297264572,
+      "grad_norm": 0.6366610527038574,
+      "learning_rate": 6.5896219287339755e-06,
+      "loss": 0.6601,
+      "step": 14487
+    },
+    {
+      "epoch": 0.7974021685288128,
+      "grad_norm": 0.6916924715042114,
+      "learning_rate": 6.589210946094859e-06,
+      "loss": 0.7683,
+      "step": 14488
+    },
+    {
+      "epoch": 0.7974572073311684,
+      "grad_norm": 0.6567399501800537,
+      "learning_rate": 6.5887999515116586e-06,
+      "loss": 0.7487,
+      "step": 14489
+    },
+    {
+      "epoch": 0.7975122461335241,
+      "grad_norm": 0.8082888722419739,
+      "learning_rate": 6.5883889449874626e-06,
+      "loss": 0.7579,
+      "step": 14490
+    },
+    {
+      "epoch": 0.7975672849358798,
+      "grad_norm": 0.7138401865959167,
+      "learning_rate": 6.58797792652536e-06,
+      "loss": 0.7256,
+      "step": 14491
+    },
+    {
+      "epoch": 0.7976223237382355,
+      "grad_norm": 0.6514482498168945,
+      "learning_rate": 6.587566896128441e-06,
+      "loss": 0.6612,
+      "step": 14492
+    },
+    {
+      "epoch": 0.7976773625405911,
+      "grad_norm": 0.6770455837249756,
+      "learning_rate": 6.587155853799795e-06,
+      "loss": 0.677,
+      "step": 14493
+    },
+    {
+      "epoch": 0.7977324013429468,
+      "grad_norm": 0.6956327557563782,
+      "learning_rate": 6.586744799542511e-06,
+      "loss": 0.7824,
+      "step": 14494
+    },
+    {
+      "epoch": 0.7977874401453025,
+      "grad_norm": 0.6565653085708618,
+      "learning_rate": 6.586333733359676e-06,
+      "loss": 0.7496,
+      "step": 14495
+    },
+    {
+      "epoch": 0.7978424789476581,
+      "grad_norm": 0.6353399157524109,
+      "learning_rate": 6.585922655254382e-06,
+      "loss": 0.7264,
+      "step": 14496
+    },
+    {
+      "epoch": 0.7978975177500137,
+      "grad_norm": 1.037051796913147,
+      "learning_rate": 6.585511565229717e-06,
+      "loss": 0.7562,
+      "step": 14497
+    },
+    {
+      "epoch": 0.7979525565523694,
+      "grad_norm": 0.6447896957397461,
+      "learning_rate": 6.5851004632887725e-06,
+      "loss": 0.7509,
+      "step": 14498
+    },
+    {
+      "epoch": 0.7980075953547251,
+      "grad_norm": 0.7022401690483093,
+      "learning_rate": 6.584689349434636e-06,
+      "loss": 0.7752,
+      "step": 14499
+    },
+    {
+      "epoch": 0.7980626341570808,
+      "grad_norm": 0.7033591270446777,
+      "learning_rate": 6.5842782236703996e-06,
+      "loss": 0.7693,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7981176729594364,
+      "grad_norm": 0.7061769962310791,
+      "learning_rate": 6.583867085999151e-06,
+      "loss": 0.6833,
+      "step": 14501
+    },
+    {
+      "epoch": 0.798172711761792,
+      "grad_norm": 0.7934882640838623,
+      "learning_rate": 6.583455936423984e-06,
+      "loss": 0.799,
+      "step": 14502
+    },
+    {
+      "epoch": 0.7982277505641477,
+      "grad_norm": 0.6968011260032654,
+      "learning_rate": 6.5830447749479835e-06,
+      "loss": 0.7132,
+      "step": 14503
+    },
+    {
+      "epoch": 0.7982827893665034,
+      "grad_norm": 1.7348299026489258,
+      "learning_rate": 6.582633601574243e-06,
+      "loss": 0.8996,
+      "step": 14504
+    },
+    {
+      "epoch": 0.798337828168859,
+      "grad_norm": 0.6822964549064636,
+      "learning_rate": 6.582222416305852e-06,
+      "loss": 0.7381,
+      "step": 14505
+    },
+    {
+      "epoch": 0.7983928669712147,
+      "grad_norm": 0.6600543856620789,
+      "learning_rate": 6.581811219145902e-06,
+      "loss": 0.711,
+      "step": 14506
+    },
+    {
+      "epoch": 0.7984479057735704,
+      "grad_norm": 0.8719834089279175,
+      "learning_rate": 6.581400010097481e-06,
+      "loss": 0.7567,
+      "step": 14507
+    },
+    {
+      "epoch": 0.7985029445759261,
+      "grad_norm": 0.7221046090126038,
+      "learning_rate": 6.580988789163681e-06,
+      "loss": 0.7417,
+      "step": 14508
+    },
+    {
+      "epoch": 0.7985579833782817,
+      "grad_norm": 0.6720401048660278,
+      "learning_rate": 6.580577556347592e-06,
+      "loss": 0.7467,
+      "step": 14509
+    },
+    {
+      "epoch": 0.7986130221806373,
+      "grad_norm": 0.7007263898849487,
+      "learning_rate": 6.580166311652306e-06,
+      "loss": 0.7356,
+      "step": 14510
+    },
+    {
+      "epoch": 0.798668060982993,
+      "grad_norm": 0.7384739518165588,
+      "learning_rate": 6.579755055080912e-06,
+      "loss": 0.7807,
+      "step": 14511
+    },
+    {
+      "epoch": 0.7987230997853487,
+      "grad_norm": 0.8054519295692444,
+      "learning_rate": 6.579343786636503e-06,
+      "loss": 0.7737,
+      "step": 14512
+    },
+    {
+      "epoch": 0.7987781385877043,
+      "grad_norm": 1.042319655418396,
+      "learning_rate": 6.578932506322169e-06,
+      "loss": 0.8708,
+      "step": 14513
+    },
+    {
+      "epoch": 0.79883317739006,
+      "grad_norm": 0.7122198343276978,
+      "learning_rate": 6.578521214141e-06,
+      "loss": 0.7818,
+      "step": 14514
+    },
+    {
+      "epoch": 0.7988882161924157,
+      "grad_norm": 0.9158271551132202,
+      "learning_rate": 6.578109910096088e-06,
+      "loss": 0.7439,
+      "step": 14515
+    },
+    {
+      "epoch": 0.7989432549947714,
+      "grad_norm": 0.7280082106590271,
+      "learning_rate": 6.577698594190524e-06,
+      "loss": 0.7888,
+      "step": 14516
+    },
+    {
+      "epoch": 0.798998293797127,
+      "grad_norm": 0.8203748464584351,
+      "learning_rate": 6.577287266427401e-06,
+      "loss": 0.7669,
+      "step": 14517
+    },
+    {
+      "epoch": 0.7990533325994826,
+      "grad_norm": 0.6998257637023926,
+      "learning_rate": 6.576875926809809e-06,
+      "loss": 0.7819,
+      "step": 14518
+    },
+    {
+      "epoch": 0.7991083714018383,
+      "grad_norm": 0.672575831413269,
+      "learning_rate": 6.57646457534084e-06,
+      "loss": 0.7359,
+      "step": 14519
+    },
+    {
+      "epoch": 0.7991634102041939,
+      "grad_norm": 0.931996762752533,
+      "learning_rate": 6.5760532120235845e-06,
+      "loss": 0.8816,
+      "step": 14520
+    },
+    {
+      "epoch": 0.7992184490065496,
+      "grad_norm": 0.7250553369522095,
+      "learning_rate": 6.575641836861134e-06,
+      "loss": 0.7924,
+      "step": 14521
+    },
+    {
+      "epoch": 0.7992734878089053,
+      "grad_norm": 0.6658768057823181,
+      "learning_rate": 6.575230449856582e-06,
+      "loss": 0.7064,
+      "step": 14522
+    },
+    {
+      "epoch": 0.799328526611261,
+      "grad_norm": 0.6901206374168396,
+      "learning_rate": 6.57481905101302e-06,
+      "loss": 0.7826,
+      "step": 14523
+    },
+    {
+      "epoch": 0.7993835654136165,
+      "grad_norm": 0.6772152781486511,
+      "learning_rate": 6.5744076403335386e-06,
+      "loss": 0.8143,
+      "step": 14524
+    },
+    {
+      "epoch": 0.7994386042159722,
+      "grad_norm": 0.6718147397041321,
+      "learning_rate": 6.5739962178212325e-06,
+      "loss": 0.765,
+      "step": 14525
+    },
+    {
+      "epoch": 0.7994936430183279,
+      "grad_norm": 0.7435488700866699,
+      "learning_rate": 6.573584783479191e-06,
+      "loss": 0.8685,
+      "step": 14526
+    },
+    {
+      "epoch": 0.7995486818206836,
+      "grad_norm": 0.7146314382553101,
+      "learning_rate": 6.573173337310506e-06,
+      "loss": 0.7605,
+      "step": 14527
+    },
+    {
+      "epoch": 0.7996037206230392,
+      "grad_norm": 0.6808409690856934,
+      "learning_rate": 6.572761879318274e-06,
+      "loss": 0.6996,
+      "step": 14528
+    },
+    {
+      "epoch": 0.7996587594253949,
+      "grad_norm": 1.1303905248641968,
+      "learning_rate": 6.572350409505584e-06,
+      "loss": 0.6107,
+      "step": 14529
+    },
+    {
+      "epoch": 0.7997137982277506,
+      "grad_norm": 0.7584583163261414,
+      "learning_rate": 6.571938927875529e-06,
+      "loss": 0.771,
+      "step": 14530
+    },
+    {
+      "epoch": 0.7997688370301063,
+      "grad_norm": 0.808233916759491,
+      "learning_rate": 6.5715274344312015e-06,
+      "loss": 0.7179,
+      "step": 14531
+    },
+    {
+      "epoch": 0.7998238758324618,
+      "grad_norm": 0.7067314386367798,
+      "learning_rate": 6.571115929175695e-06,
+      "loss": 0.7519,
+      "step": 14532
+    },
+    {
+      "epoch": 0.7998789146348175,
+      "grad_norm": 0.7611628174781799,
+      "learning_rate": 6.570704412112101e-06,
+      "loss": 0.8727,
+      "step": 14533
+    },
+    {
+      "epoch": 0.7999339534371732,
+      "grad_norm": 0.6485727429389954,
+      "learning_rate": 6.5702928832435145e-06,
+      "loss": 0.8455,
+      "step": 14534
+    },
+    {
+      "epoch": 0.7999889922395289,
+      "grad_norm": 1.5309134721755981,
+      "learning_rate": 6.569881342573024e-06,
+      "loss": 0.8362,
+      "step": 14535
+    },
+    {
+      "epoch": 0.8000440310418845,
+      "grad_norm": 0.7068225145339966,
+      "learning_rate": 6.569469790103729e-06,
+      "loss": 0.7924,
+      "step": 14536
+    },
+    {
+      "epoch": 0.8000990698442402,
+      "grad_norm": 0.7326669692993164,
+      "learning_rate": 6.569058225838717e-06,
+      "loss": 0.7594,
+      "step": 14537
+    },
+    {
+      "epoch": 0.8001541086465959,
+      "grad_norm": 0.6705706119537354,
+      "learning_rate": 6.568646649781085e-06,
+      "loss": 0.7331,
+      "step": 14538
+    },
+    {
+      "epoch": 0.8002091474489516,
+      "grad_norm": 0.7303051948547363,
+      "learning_rate": 6.568235061933923e-06,
+      "loss": 0.7274,
+      "step": 14539
+    },
+    {
+      "epoch": 0.8002641862513071,
+      "grad_norm": 0.6334550380706787,
+      "learning_rate": 6.567823462300326e-06,
+      "loss": 0.7105,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8003192250536628,
+      "grad_norm": 0.7183839678764343,
+      "learning_rate": 6.56741185088339e-06,
+      "loss": 0.657,
+      "step": 14541
+    },
+    {
+      "epoch": 0.8003742638560185,
+      "grad_norm": 0.6896400451660156,
+      "learning_rate": 6.567000227686204e-06,
+      "loss": 0.7752,
+      "step": 14542
+    },
+    {
+      "epoch": 0.8004293026583742,
+      "grad_norm": 0.7214651703834534,
+      "learning_rate": 6.566588592711864e-06,
+      "loss": 0.753,
+      "step": 14543
+    },
+    {
+      "epoch": 0.8004843414607298,
+      "grad_norm": 0.7064470648765564,
+      "learning_rate": 6.566176945963464e-06,
+      "loss": 0.744,
+      "step": 14544
     }
   ],
   "logging_steps": 1,
@@ -95471,7 +101834,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.023772071182991e+19,
+  "total_flos": 4.292023542595191e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null