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": 0.650393527436843,
+  "epoch": 0.7504540701194342,
   "eval_steps": 500,
-  "global_step": 11817,
+  "global_step": 13635,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -82726,6 +82726,12732 @@
       "learning_rate": 7.634761785456671e-06,
       "loss": 0.7012,
       "step": 11817
+    },
+    {
+      "epoch": 0.6504485662391987,
+      "grad_norm": 0.7214943170547485,
+      "learning_rate": 7.634393374521478e-06,
+      "loss": 0.7386,
+      "step": 11818
+    },
+    {
+      "epoch": 0.6505036050415542,
+      "grad_norm": 0.7026216387748718,
+      "learning_rate": 7.63402494378689e-06,
+      "loss": 0.7444,
+      "step": 11819
+    },
+    {
+      "epoch": 0.6505586438439099,
+      "grad_norm": 0.6271201372146606,
+      "learning_rate": 7.633656493255677e-06,
+      "loss": 0.6567,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6506136826462656,
+      "grad_norm": 0.8359349370002747,
+      "learning_rate": 7.633288022930606e-06,
+      "loss": 0.7081,
+      "step": 11821
+    },
+    {
+      "epoch": 0.6506687214486213,
+      "grad_norm": 0.7009666562080383,
+      "learning_rate": 7.632919532814444e-06,
+      "loss": 0.6892,
+      "step": 11822
+    },
+    {
+      "epoch": 0.6507237602509769,
+      "grad_norm": 0.7445069551467896,
+      "learning_rate": 7.632551022909966e-06,
+      "loss": 0.7854,
+      "step": 11823
+    },
+    {
+      "epoch": 0.6507787990533326,
+      "grad_norm": 0.7204466462135315,
+      "learning_rate": 7.63218249321994e-06,
+      "loss": 0.8065,
+      "step": 11824
+    },
+    {
+      "epoch": 0.6508338378556883,
+      "grad_norm": 0.7058166265487671,
+      "learning_rate": 7.631813943747135e-06,
+      "loss": 0.6668,
+      "step": 11825
+    },
+    {
+      "epoch": 0.650888876658044,
+      "grad_norm": 0.739919126033783,
+      "learning_rate": 7.631445374494319e-06,
+      "loss": 0.8657,
+      "step": 11826
+    },
+    {
+      "epoch": 0.6509439154603995,
+      "grad_norm": 1.0444670915603638,
+      "learning_rate": 7.631076785464263e-06,
+      "loss": 0.7226,
+      "step": 11827
+    },
+    {
+      "epoch": 0.6509989542627552,
+      "grad_norm": 0.7146627306938171,
+      "learning_rate": 7.630708176659743e-06,
+      "loss": 0.7567,
+      "step": 11828
+    },
+    {
+      "epoch": 0.6510539930651109,
+      "grad_norm": 0.6981074810028076,
+      "learning_rate": 7.630339548083521e-06,
+      "loss": 0.7158,
+      "step": 11829
+    },
+    {
+      "epoch": 0.6511090318674666,
+      "grad_norm": 0.7620309591293335,
+      "learning_rate": 7.629970899738372e-06,
+      "loss": 0.811,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6511640706698222,
+      "grad_norm": 0.7017341256141663,
+      "learning_rate": 7.629602231627066e-06,
+      "loss": 0.7092,
+      "step": 11831
+    },
+    {
+      "epoch": 0.6512191094721779,
+      "grad_norm": 0.733524739742279,
+      "learning_rate": 7.629233543752373e-06,
+      "loss": 0.859,
+      "step": 11832
+    },
+    {
+      "epoch": 0.6512741482745336,
+      "grad_norm": 0.7246975898742676,
+      "learning_rate": 7.628864836117065e-06,
+      "loss": 0.7732,
+      "step": 11833
+    },
+    {
+      "epoch": 0.6513291870768892,
+      "grad_norm": 0.5763251185417175,
+      "learning_rate": 7.628496108723911e-06,
+      "loss": 0.6632,
+      "step": 11834
+    },
+    {
+      "epoch": 0.6513842258792448,
+      "grad_norm": 0.6120070815086365,
+      "learning_rate": 7.628127361575685e-06,
+      "loss": 0.6809,
+      "step": 11835
+    },
+    {
+      "epoch": 0.6514392646816005,
+      "grad_norm": 0.8650742769241333,
+      "learning_rate": 7.627758594675157e-06,
+      "loss": 0.6388,
+      "step": 11836
+    },
+    {
+      "epoch": 0.6514943034839562,
+      "grad_norm": 0.8650027513504028,
+      "learning_rate": 7.627389808025099e-06,
+      "loss": 0.7622,
+      "step": 11837
+    },
+    {
+      "epoch": 0.6515493422863119,
+      "grad_norm": 0.6683071851730347,
+      "learning_rate": 7.627021001628283e-06,
+      "loss": 0.7424,
+      "step": 11838
+    },
+    {
+      "epoch": 0.6516043810886675,
+      "grad_norm": 0.6821237206459045,
+      "learning_rate": 7.626652175487479e-06,
+      "loss": 0.7844,
+      "step": 11839
+    },
+    {
+      "epoch": 0.6516594198910232,
+      "grad_norm": 0.7142770886421204,
+      "learning_rate": 7.626283329605462e-06,
+      "loss": 0.7706,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6517144586933789,
+      "grad_norm": 0.7870625257492065,
+      "learning_rate": 7.625914463985002e-06,
+      "loss": 0.7673,
+      "step": 11841
+    },
+    {
+      "epoch": 0.6517694974957345,
+      "grad_norm": 0.7386491894721985,
+      "learning_rate": 7.62554557862887e-06,
+      "loss": 0.7562,
+      "step": 11842
+    },
+    {
+      "epoch": 0.6518245362980901,
+      "grad_norm": 0.6529993414878845,
+      "learning_rate": 7.625176673539843e-06,
+      "loss": 0.8258,
+      "step": 11843
+    },
+    {
+      "epoch": 0.6518795751004458,
+      "grad_norm": 0.7010294795036316,
+      "learning_rate": 7.6248077487206895e-06,
+      "loss": 0.7773,
+      "step": 11844
+    },
+    {
+      "epoch": 0.6519346139028015,
+      "grad_norm": 0.6699075698852539,
+      "learning_rate": 7.624438804174184e-06,
+      "loss": 0.7163,
+      "step": 11845
+    },
+    {
+      "epoch": 0.6519896527051572,
+      "grad_norm": 0.6600161790847778,
+      "learning_rate": 7.624069839903099e-06,
+      "loss": 0.7355,
+      "step": 11846
+    },
+    {
+      "epoch": 0.6520446915075128,
+      "grad_norm": 0.6556873321533203,
+      "learning_rate": 7.623700855910205e-06,
+      "loss": 0.627,
+      "step": 11847
+    },
+    {
+      "epoch": 0.6520997303098685,
+      "grad_norm": 0.6867008805274963,
+      "learning_rate": 7.623331852198281e-06,
+      "loss": 0.8228,
+      "step": 11848
+    },
+    {
+      "epoch": 0.6521547691122241,
+      "grad_norm": 0.6885474324226379,
+      "learning_rate": 7.622962828770095e-06,
+      "loss": 0.6804,
+      "step": 11849
+    },
+    {
+      "epoch": 0.6522098079145798,
+      "grad_norm": 0.6903913021087646,
+      "learning_rate": 7.622593785628425e-06,
+      "loss": 0.6553,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6522648467169354,
+      "grad_norm": 0.6581684947013855,
+      "learning_rate": 7.622224722776039e-06,
+      "loss": 0.7102,
+      "step": 11851
+    },
+    {
+      "epoch": 0.6523198855192911,
+      "grad_norm": 0.8261715769767761,
+      "learning_rate": 7.621855640215716e-06,
+      "loss": 0.676,
+      "step": 11852
+    },
+    {
+      "epoch": 0.6523749243216468,
+      "grad_norm": 0.6238247752189636,
+      "learning_rate": 7.6214865379502265e-06,
+      "loss": 0.7065,
+      "step": 11853
+    },
+    {
+      "epoch": 0.6524299631240025,
+      "grad_norm": 0.7350416779518127,
+      "learning_rate": 7.621117415982346e-06,
+      "loss": 0.7512,
+      "step": 11854
+    },
+    {
+      "epoch": 0.652485001926358,
+      "grad_norm": 0.7337208390235901,
+      "learning_rate": 7.620748274314851e-06,
+      "loss": 0.7593,
+      "step": 11855
+    },
+    {
+      "epoch": 0.6525400407287137,
+      "grad_norm": 0.6568214297294617,
+      "learning_rate": 7.620379112950511e-06,
+      "loss": 0.7363,
+      "step": 11856
+    },
+    {
+      "epoch": 0.6525950795310694,
+      "grad_norm": 0.7099055647850037,
+      "learning_rate": 7.620009931892105e-06,
+      "loss": 0.6631,
+      "step": 11857
+    },
+    {
+      "epoch": 0.652650118333425,
+      "grad_norm": 0.6563010215759277,
+      "learning_rate": 7.6196407311424035e-06,
+      "loss": 0.6617,
+      "step": 11858
+    },
+    {
+      "epoch": 0.6527051571357807,
+      "grad_norm": 0.6664251685142517,
+      "learning_rate": 7.6192715107041845e-06,
+      "loss": 0.7898,
+      "step": 11859
+    },
+    {
+      "epoch": 0.6527601959381364,
+      "grad_norm": 0.6524507403373718,
+      "learning_rate": 7.618902270580222e-06,
+      "loss": 0.767,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6528152347404921,
+      "grad_norm": 0.7391313910484314,
+      "learning_rate": 7.61853301077329e-06,
+      "loss": 0.6015,
+      "step": 11861
+    },
+    {
+      "epoch": 0.6528702735428477,
+      "grad_norm": 0.7691878080368042,
+      "learning_rate": 7.618163731286167e-06,
+      "loss": 0.718,
+      "step": 11862
+    },
+    {
+      "epoch": 0.6529253123452033,
+      "grad_norm": 0.6524633765220642,
+      "learning_rate": 7.617794432121625e-06,
+      "loss": 0.6841,
+      "step": 11863
+    },
+    {
+      "epoch": 0.652980351147559,
+      "grad_norm": 0.7125405073165894,
+      "learning_rate": 7.61742511328244e-06,
+      "loss": 0.7654,
+      "step": 11864
+    },
+    {
+      "epoch": 0.6530353899499147,
+      "grad_norm": 0.7123568058013916,
+      "learning_rate": 7.617055774771389e-06,
+      "loss": 0.7189,
+      "step": 11865
+    },
+    {
+      "epoch": 0.6530904287522703,
+      "grad_norm": 0.6968240141868591,
+      "learning_rate": 7.616686416591248e-06,
+      "loss": 0.7201,
+      "step": 11866
+    },
+    {
+      "epoch": 0.653145467554626,
+      "grad_norm": 0.7208551168441772,
+      "learning_rate": 7.616317038744792e-06,
+      "loss": 0.6644,
+      "step": 11867
+    },
+    {
+      "epoch": 0.6532005063569817,
+      "grad_norm": 0.7320911884307861,
+      "learning_rate": 7.615947641234798e-06,
+      "loss": 0.7118,
+      "step": 11868
+    },
+    {
+      "epoch": 0.6532555451593374,
+      "grad_norm": 0.7762041687965393,
+      "learning_rate": 7.615578224064041e-06,
+      "loss": 0.7501,
+      "step": 11869
+    },
+    {
+      "epoch": 0.653310583961693,
+      "grad_norm": 0.7455989718437195,
+      "learning_rate": 7.6152087872352975e-06,
+      "loss": 0.8058,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6533656227640486,
+      "grad_norm": 0.736044704914093,
+      "learning_rate": 7.614839330751347e-06,
+      "loss": 0.727,
+      "step": 11871
+    },
+    {
+      "epoch": 0.6534206615664043,
+      "grad_norm": 0.680171012878418,
+      "learning_rate": 7.614469854614961e-06,
+      "loss": 0.6722,
+      "step": 11872
+    },
+    {
+      "epoch": 0.65347570036876,
+      "grad_norm": 0.7598134279251099,
+      "learning_rate": 7.614100358828922e-06,
+      "loss": 0.7472,
+      "step": 11873
+    },
+    {
+      "epoch": 0.6535307391711156,
+      "grad_norm": 0.8288099765777588,
+      "learning_rate": 7.613730843396003e-06,
+      "loss": 0.7493,
+      "step": 11874
+    },
+    {
+      "epoch": 0.6535857779734713,
+      "grad_norm": 0.6436724066734314,
+      "learning_rate": 7.613361308318984e-06,
+      "loss": 0.7103,
+      "step": 11875
+    },
+    {
+      "epoch": 0.653640816775827,
+      "grad_norm": 0.671334981918335,
+      "learning_rate": 7.612991753600639e-06,
+      "loss": 0.6949,
+      "step": 11876
+    },
+    {
+      "epoch": 0.6536958555781827,
+      "grad_norm": 0.6019170880317688,
+      "learning_rate": 7.61262217924375e-06,
+      "loss": 0.6116,
+      "step": 11877
+    },
+    {
+      "epoch": 0.6537508943805382,
+      "grad_norm": 1.4682546854019165,
+      "learning_rate": 7.61225258525109e-06,
+      "loss": 0.9343,
+      "step": 11878
+    },
+    {
+      "epoch": 0.6538059331828939,
+      "grad_norm": 0.656822681427002,
+      "learning_rate": 7.611882971625439e-06,
+      "loss": 0.7357,
+      "step": 11879
+    },
+    {
+      "epoch": 0.6538609719852496,
+      "grad_norm": 0.635734498500824,
+      "learning_rate": 7.611513338369576e-06,
+      "loss": 0.6263,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6539160107876053,
+      "grad_norm": 0.7123430967330933,
+      "learning_rate": 7.611143685486277e-06,
+      "loss": 0.8446,
+      "step": 11881
+    },
+    {
+      "epoch": 0.6539710495899609,
+      "grad_norm": 0.7597065567970276,
+      "learning_rate": 7.610774012978322e-06,
+      "loss": 0.7449,
+      "step": 11882
+    },
+    {
+      "epoch": 0.6540260883923166,
+      "grad_norm": 0.7555896043777466,
+      "learning_rate": 7.610404320848486e-06,
+      "loss": 0.7575,
+      "step": 11883
+    },
+    {
+      "epoch": 0.6540811271946723,
+      "grad_norm": 0.7572906613349915,
+      "learning_rate": 7.6100346090995506e-06,
+      "loss": 0.7547,
+      "step": 11884
+    },
+    {
+      "epoch": 0.654136165997028,
+      "grad_norm": 0.6663275957107544,
+      "learning_rate": 7.609664877734295e-06,
+      "loss": 0.7038,
+      "step": 11885
+    },
+    {
+      "epoch": 0.6541912047993835,
+      "grad_norm": 0.7346611618995667,
+      "learning_rate": 7.609295126755496e-06,
+      "loss": 0.7902,
+      "step": 11886
+    },
+    {
+      "epoch": 0.6542462436017392,
+      "grad_norm": 0.6846545338630676,
+      "learning_rate": 7.608925356165934e-06,
+      "loss": 0.7334,
+      "step": 11887
+    },
+    {
+      "epoch": 0.6543012824040949,
+      "grad_norm": 0.6714815497398376,
+      "learning_rate": 7.608555565968385e-06,
+      "loss": 0.7204,
+      "step": 11888
+    },
+    {
+      "epoch": 0.6543563212064506,
+      "grad_norm": 0.805095374584198,
+      "learning_rate": 7.608185756165634e-06,
+      "loss": 0.8521,
+      "step": 11889
+    },
+    {
+      "epoch": 0.6544113600088062,
+      "grad_norm": 0.8415316343307495,
+      "learning_rate": 7.607815926760456e-06,
+      "loss": 0.7076,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6544663988111619,
+      "grad_norm": 0.7665743231773376,
+      "learning_rate": 7.607446077755632e-06,
+      "loss": 0.8072,
+      "step": 11891
+    },
+    {
+      "epoch": 0.6545214376135176,
+      "grad_norm": 0.6705248355865479,
+      "learning_rate": 7.607076209153939e-06,
+      "loss": 0.6607,
+      "step": 11892
+    },
+    {
+      "epoch": 0.6545764764158732,
+      "grad_norm": 0.6791796684265137,
+      "learning_rate": 7.606706320958159e-06,
+      "loss": 0.773,
+      "step": 11893
+    },
+    {
+      "epoch": 0.6546315152182288,
+      "grad_norm": 0.8177357316017151,
+      "learning_rate": 7.606336413171075e-06,
+      "loss": 0.8114,
+      "step": 11894
+    },
+    {
+      "epoch": 0.6546865540205845,
+      "grad_norm": 0.9491637945175171,
+      "learning_rate": 7.605966485795462e-06,
+      "loss": 0.7424,
+      "step": 11895
+    },
+    {
+      "epoch": 0.6547415928229402,
+      "grad_norm": 0.7326256036758423,
+      "learning_rate": 7.605596538834103e-06,
+      "loss": 0.8176,
+      "step": 11896
+    },
+    {
+      "epoch": 0.6547966316252959,
+      "grad_norm": 0.6081808805465698,
+      "learning_rate": 7.6052265722897775e-06,
+      "loss": 0.6827,
+      "step": 11897
+    },
+    {
+      "epoch": 0.6548516704276515,
+      "grad_norm": 0.7165681719779968,
+      "learning_rate": 7.604856586165268e-06,
+      "loss": 0.7854,
+      "step": 11898
+    },
+    {
+      "epoch": 0.6549067092300072,
+      "grad_norm": 0.8777725100517273,
+      "learning_rate": 7.604486580463353e-06,
+      "loss": 0.8084,
+      "step": 11899
+    },
+    {
+      "epoch": 0.6549617480323628,
+      "grad_norm": 0.6814439296722412,
+      "learning_rate": 7.604116555186811e-06,
+      "loss": 0.6869,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6550167868347184,
+      "grad_norm": 0.7060914635658264,
+      "learning_rate": 7.60374651033843e-06,
+      "loss": 0.7066,
+      "step": 11901
+    },
+    {
+      "epoch": 0.6550718256370741,
+      "grad_norm": 0.6823089718818665,
+      "learning_rate": 7.603376445920987e-06,
+      "loss": 0.6095,
+      "step": 11902
+    },
+    {
+      "epoch": 0.6551268644394298,
+      "grad_norm": 0.7099863290786743,
+      "learning_rate": 7.603006361937262e-06,
+      "loss": 0.8037,
+      "step": 11903
+    },
+    {
+      "epoch": 0.6551819032417855,
+      "grad_norm": 0.6479066610336304,
+      "learning_rate": 7.602636258390037e-06,
+      "loss": 0.6844,
+      "step": 11904
+    },
+    {
+      "epoch": 0.6552369420441411,
+      "grad_norm": 0.6663268804550171,
+      "learning_rate": 7.602266135282097e-06,
+      "loss": 0.735,
+      "step": 11905
+    },
+    {
+      "epoch": 0.6552919808464968,
+      "grad_norm": 0.8670598268508911,
+      "learning_rate": 7.60189599261622e-06,
+      "loss": 0.779,
+      "step": 11906
+    },
+    {
+      "epoch": 0.6553470196488524,
+      "grad_norm": 0.607631504535675,
+      "learning_rate": 7.601525830395189e-06,
+      "loss": 0.6288,
+      "step": 11907
+    },
+    {
+      "epoch": 0.6554020584512081,
+      "grad_norm": 0.9054927229881287,
+      "learning_rate": 7.601155648621786e-06,
+      "loss": 0.8562,
+      "step": 11908
+    },
+    {
+      "epoch": 0.6554570972535637,
+      "grad_norm": 0.8069004416465759,
+      "learning_rate": 7.6007854472987955e-06,
+      "loss": 0.88,
+      "step": 11909
+    },
+    {
+      "epoch": 0.6555121360559194,
+      "grad_norm": 0.6393092274665833,
+      "learning_rate": 7.600415226428995e-06,
+      "loss": 0.6908,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6555671748582751,
+      "grad_norm": 0.7533125281333923,
+      "learning_rate": 7.600044986015172e-06,
+      "loss": 0.8061,
+      "step": 11911
+    },
+    {
+      "epoch": 0.6556222136606308,
+      "grad_norm": 0.6859326958656311,
+      "learning_rate": 7.599674726060105e-06,
+      "loss": 0.7603,
+      "step": 11912
+    },
+    {
+      "epoch": 0.6556772524629864,
+      "grad_norm": 0.7284619808197021,
+      "learning_rate": 7.59930444656658e-06,
+      "loss": 0.7698,
+      "step": 11913
+    },
+    {
+      "epoch": 0.655732291265342,
+      "grad_norm": 1.074234127998352,
+      "learning_rate": 7.598934147537378e-06,
+      "loss": 0.8252,
+      "step": 11914
+    },
+    {
+      "epoch": 0.6557873300676977,
+      "grad_norm": 0.6899133920669556,
+      "learning_rate": 7.598563828975283e-06,
+      "loss": 0.6023,
+      "step": 11915
+    },
+    {
+      "epoch": 0.6558423688700534,
+      "grad_norm": 0.6736464500427246,
+      "learning_rate": 7.598193490883077e-06,
+      "loss": 0.788,
+      "step": 11916
+    },
+    {
+      "epoch": 0.655897407672409,
+      "grad_norm": 0.7646307349205017,
+      "learning_rate": 7.597823133263545e-06,
+      "loss": 0.7607,
+      "step": 11917
+    },
+    {
+      "epoch": 0.6559524464747647,
+      "grad_norm": 0.6413717865943909,
+      "learning_rate": 7.59745275611947e-06,
+      "loss": 0.6415,
+      "step": 11918
+    },
+    {
+      "epoch": 0.6560074852771204,
+      "grad_norm": 0.6605532169342041,
+      "learning_rate": 7.597082359453636e-06,
+      "loss": 0.6655,
+      "step": 11919
+    },
+    {
+      "epoch": 0.6560625240794761,
+      "grad_norm": 0.6573199033737183,
+      "learning_rate": 7.596711943268824e-06,
+      "loss": 0.624,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6561175628818317,
+      "grad_norm": 0.8312102556228638,
+      "learning_rate": 7.596341507567822e-06,
+      "loss": 0.6803,
+      "step": 11921
+    },
+    {
+      "epoch": 0.6561726016841873,
+      "grad_norm": 0.6915873289108276,
+      "learning_rate": 7.59597105235341e-06,
+      "loss": 0.6897,
+      "step": 11922
+    },
+    {
+      "epoch": 0.656227640486543,
+      "grad_norm": 0.6916965842247009,
+      "learning_rate": 7.595600577628377e-06,
+      "loss": 0.7154,
+      "step": 11923
+    },
+    {
+      "epoch": 0.6562826792888987,
+      "grad_norm": 0.6712722182273865,
+      "learning_rate": 7.595230083395501e-06,
+      "loss": 0.7236,
+      "step": 11924
+    },
+    {
+      "epoch": 0.6563377180912543,
+      "grad_norm": 0.6514019966125488,
+      "learning_rate": 7.594859569657575e-06,
+      "loss": 0.6895,
+      "step": 11925
+    },
+    {
+      "epoch": 0.65639275689361,
+      "grad_norm": 0.7300555109977722,
+      "learning_rate": 7.594489036417378e-06,
+      "loss": 0.7563,
+      "step": 11926
+    },
+    {
+      "epoch": 0.6564477956959657,
+      "grad_norm": 0.8076907396316528,
+      "learning_rate": 7.594118483677695e-06,
+      "loss": 0.8883,
+      "step": 11927
+    },
+    {
+      "epoch": 0.6565028344983214,
+      "grad_norm": 0.666466236114502,
+      "learning_rate": 7.5937479114413114e-06,
+      "loss": 0.7641,
+      "step": 11928
+    },
+    {
+      "epoch": 0.6565578733006769,
+      "grad_norm": 0.6621832251548767,
+      "learning_rate": 7.593377319711013e-06,
+      "loss": 0.6687,
+      "step": 11929
+    },
+    {
+      "epoch": 0.6566129121030326,
+      "grad_norm": 0.8757139444351196,
+      "learning_rate": 7.593006708489585e-06,
+      "loss": 0.7746,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6566679509053883,
+      "grad_norm": 0.646801769733429,
+      "learning_rate": 7.5926360777798135e-06,
+      "loss": 0.6884,
+      "step": 11931
+    },
+    {
+      "epoch": 0.656722989707744,
+      "grad_norm": 0.6703395843505859,
+      "learning_rate": 7.592265427584482e-06,
+      "loss": 0.6822,
+      "step": 11932
+    },
+    {
+      "epoch": 0.6567780285100996,
+      "grad_norm": 0.7653201222419739,
+      "learning_rate": 7.591894757906378e-06,
+      "loss": 0.7999,
+      "step": 11933
+    },
+    {
+      "epoch": 0.6568330673124553,
+      "grad_norm": 0.6921548247337341,
+      "learning_rate": 7.591524068748288e-06,
+      "loss": 0.7177,
+      "step": 11934
+    },
+    {
+      "epoch": 0.656888106114811,
+      "grad_norm": 0.7085320353507996,
+      "learning_rate": 7.591153360112995e-06,
+      "loss": 0.8395,
+      "step": 11935
+    },
+    {
+      "epoch": 0.6569431449171667,
+      "grad_norm": 0.6565294861793518,
+      "learning_rate": 7.590782632003287e-06,
+      "loss": 0.6969,
+      "step": 11936
+    },
+    {
+      "epoch": 0.6569981837195222,
+      "grad_norm": 0.7023206353187561,
+      "learning_rate": 7.590411884421952e-06,
+      "loss": 0.7321,
+      "step": 11937
+    },
+    {
+      "epoch": 0.6570532225218779,
+      "grad_norm": 0.7848044633865356,
+      "learning_rate": 7.590041117371774e-06,
+      "loss": 0.8857,
+      "step": 11938
+    },
+    {
+      "epoch": 0.6571082613242336,
+      "grad_norm": 1.004591703414917,
+      "learning_rate": 7.589670330855541e-06,
+      "loss": 0.8267,
+      "step": 11939
+    },
+    {
+      "epoch": 0.6571633001265893,
+      "grad_norm": 0.7525139451026917,
+      "learning_rate": 7.589299524876036e-06,
+      "loss": 0.6857,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6572183389289449,
+      "grad_norm": 0.746224582195282,
+      "learning_rate": 7.588928699436051e-06,
+      "loss": 0.805,
+      "step": 11941
+    },
+    {
+      "epoch": 0.6572733777313006,
+      "grad_norm": 0.6304495930671692,
+      "learning_rate": 7.588557854538371e-06,
+      "loss": 0.652,
+      "step": 11942
+    },
+    {
+      "epoch": 0.6573284165336563,
+      "grad_norm": 0.761688768863678,
+      "learning_rate": 7.588186990185783e-06,
+      "loss": 0.7954,
+      "step": 11943
+    },
+    {
+      "epoch": 0.6573834553360118,
+      "grad_norm": 0.7735103368759155,
+      "learning_rate": 7.587816106381073e-06,
+      "loss": 0.7584,
+      "step": 11944
+    },
+    {
+      "epoch": 0.6574384941383675,
+      "grad_norm": 0.7351566553115845,
+      "learning_rate": 7.5874452031270305e-06,
+      "loss": 0.7984,
+      "step": 11945
+    },
+    {
+      "epoch": 0.6574935329407232,
+      "grad_norm": 0.7054993510246277,
+      "learning_rate": 7.587074280426443e-06,
+      "loss": 0.7057,
+      "step": 11946
+    },
+    {
+      "epoch": 0.6575485717430789,
+      "grad_norm": 0.7444368004798889,
+      "learning_rate": 7.586703338282099e-06,
+      "loss": 0.7476,
+      "step": 11947
+    },
+    {
+      "epoch": 0.6576036105454345,
+      "grad_norm": 0.6944568157196045,
+      "learning_rate": 7.586332376696782e-06,
+      "loss": 0.6874,
+      "step": 11948
+    },
+    {
+      "epoch": 0.6576586493477902,
+      "grad_norm": 0.6595578193664551,
+      "learning_rate": 7.585961395673287e-06,
+      "loss": 0.7541,
+      "step": 11949
+    },
+    {
+      "epoch": 0.6577136881501459,
+      "grad_norm": 0.6669502258300781,
+      "learning_rate": 7.585590395214396e-06,
+      "loss": 0.7515,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6577687269525015,
+      "grad_norm": 0.7254583835601807,
+      "learning_rate": 7.585219375322901e-06,
+      "loss": 0.8089,
+      "step": 11951
+    },
+    {
+      "epoch": 0.6578237657548571,
+      "grad_norm": 1.0479141473770142,
+      "learning_rate": 7.584848336001587e-06,
+      "loss": 0.8108,
+      "step": 11952
+    },
+    {
+      "epoch": 0.6578788045572128,
+      "grad_norm": 0.6928718686103821,
+      "learning_rate": 7.584477277253246e-06,
+      "loss": 0.6325,
+      "step": 11953
+    },
+    {
+      "epoch": 0.6579338433595685,
+      "grad_norm": 0.8926869630813599,
+      "learning_rate": 7.584106199080666e-06,
+      "loss": 0.7294,
+      "step": 11954
+    },
+    {
+      "epoch": 0.6579888821619242,
+      "grad_norm": 0.7209964394569397,
+      "learning_rate": 7.583735101486635e-06,
+      "loss": 0.7646,
+      "step": 11955
+    },
+    {
+      "epoch": 0.6580439209642798,
+      "grad_norm": 0.7619316577911377,
+      "learning_rate": 7.583363984473941e-06,
+      "loss": 0.7756,
+      "step": 11956
+    },
+    {
+      "epoch": 0.6580989597666355,
+      "grad_norm": 0.6974903345108032,
+      "learning_rate": 7.582992848045378e-06,
+      "loss": 0.6497,
+      "step": 11957
+    },
+    {
+      "epoch": 0.6581539985689911,
+      "grad_norm": 0.8338617086410522,
+      "learning_rate": 7.582621692203731e-06,
+      "loss": 0.6619,
+      "step": 11958
+    },
+    {
+      "epoch": 0.6582090373713468,
+      "grad_norm": 0.9330396056175232,
+      "learning_rate": 7.5822505169517905e-06,
+      "loss": 0.8219,
+      "step": 11959
+    },
+    {
+      "epoch": 0.6582640761737024,
+      "grad_norm": 0.7725355625152588,
+      "learning_rate": 7.5818793222923445e-06,
+      "loss": 0.7262,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6583191149760581,
+      "grad_norm": 0.7049654722213745,
+      "learning_rate": 7.5815081082281885e-06,
+      "loss": 0.7917,
+      "step": 11961
+    },
+    {
+      "epoch": 0.6583741537784138,
+      "grad_norm": 0.6801711916923523,
+      "learning_rate": 7.581136874762105e-06,
+      "loss": 0.6984,
+      "step": 11962
+    },
+    {
+      "epoch": 0.6584291925807695,
+      "grad_norm": 0.7774253487586975,
+      "learning_rate": 7.58076562189689e-06,
+      "loss": 0.7615,
+      "step": 11963
+    },
+    {
+      "epoch": 0.6584842313831251,
+      "grad_norm": 0.7436443567276001,
+      "learning_rate": 7.58039434963533e-06,
+      "loss": 0.7419,
+      "step": 11964
+    },
+    {
+      "epoch": 0.6585392701854808,
+      "grad_norm": 0.6857719421386719,
+      "learning_rate": 7.580023057980217e-06,
+      "loss": 0.8009,
+      "step": 11965
+    },
+    {
+      "epoch": 0.6585943089878364,
+      "grad_norm": 0.7194758653640747,
+      "learning_rate": 7.579651746934342e-06,
+      "loss": 0.7338,
+      "step": 11966
+    },
+    {
+      "epoch": 0.6586493477901921,
+      "grad_norm": 0.7248701453208923,
+      "learning_rate": 7.579280416500495e-06,
+      "loss": 0.6972,
+      "step": 11967
+    },
+    {
+      "epoch": 0.6587043865925477,
+      "grad_norm": 0.6719415783882141,
+      "learning_rate": 7.578909066681466e-06,
+      "loss": 0.7552,
+      "step": 11968
+    },
+    {
+      "epoch": 0.6587594253949034,
+      "grad_norm": 0.728338897228241,
+      "learning_rate": 7.578537697480046e-06,
+      "loss": 0.8386,
+      "step": 11969
+    },
+    {
+      "epoch": 0.6588144641972591,
+      "grad_norm": 0.7151786684989929,
+      "learning_rate": 7.578166308899029e-06,
+      "loss": 0.7186,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6588695029996148,
+      "grad_norm": 0.664412260055542,
+      "learning_rate": 7.577794900941205e-06,
+      "loss": 0.6672,
+      "step": 11971
+    },
+    {
+      "epoch": 0.6589245418019704,
+      "grad_norm": 0.6915827989578247,
+      "learning_rate": 7.577423473609361e-06,
+      "loss": 0.7427,
+      "step": 11972
+    },
+    {
+      "epoch": 0.658979580604326,
+      "grad_norm": 0.705243706703186,
+      "learning_rate": 7.577052026906295e-06,
+      "loss": 0.7526,
+      "step": 11973
+    },
+    {
+      "epoch": 0.6590346194066817,
+      "grad_norm": 0.6559640169143677,
+      "learning_rate": 7.576680560834795e-06,
+      "loss": 0.8187,
+      "step": 11974
+    },
+    {
+      "epoch": 0.6590896582090374,
+      "grad_norm": 0.7359572649002075,
+      "learning_rate": 7.576309075397653e-06,
+      "loss": 0.8127,
+      "step": 11975
+    },
+    {
+      "epoch": 0.659144697011393,
+      "grad_norm": 0.6581039428710938,
+      "learning_rate": 7.575937570597661e-06,
+      "loss": 0.7066,
+      "step": 11976
+    },
+    {
+      "epoch": 0.6591997358137487,
+      "grad_norm": 0.8360844254493713,
+      "learning_rate": 7.5755660464376134e-06,
+      "loss": 0.7998,
+      "step": 11977
+    },
+    {
+      "epoch": 0.6592547746161044,
+      "grad_norm": 0.7201453447341919,
+      "learning_rate": 7.5751945029203e-06,
+      "loss": 0.7884,
+      "step": 11978
+    },
+    {
+      "epoch": 0.6593098134184601,
+      "grad_norm": 0.6985270977020264,
+      "learning_rate": 7.574822940048514e-06,
+      "loss": 0.7268,
+      "step": 11979
+    },
+    {
+      "epoch": 0.6593648522208156,
+      "grad_norm": 0.6405925154685974,
+      "learning_rate": 7.574451357825048e-06,
+      "loss": 0.6848,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6594198910231713,
+      "grad_norm": 0.6656618714332581,
+      "learning_rate": 7.574079756252694e-06,
+      "loss": 0.7755,
+      "step": 11981
+    },
+    {
+      "epoch": 0.659474929825527,
+      "grad_norm": 0.8461045622825623,
+      "learning_rate": 7.573708135334248e-06,
+      "loss": 0.7171,
+      "step": 11982
+    },
+    {
+      "epoch": 0.6595299686278827,
+      "grad_norm": 0.5527384877204895,
+      "learning_rate": 7.573336495072498e-06,
+      "loss": 0.6668,
+      "step": 11983
+    },
+    {
+      "epoch": 0.6595850074302383,
+      "grad_norm": 0.6703749299049377,
+      "learning_rate": 7.572964835470241e-06,
+      "loss": 0.7128,
+      "step": 11984
+    },
+    {
+      "epoch": 0.659640046232594,
+      "grad_norm": 0.6824783682823181,
+      "learning_rate": 7.57259315653027e-06,
+      "loss": 0.8007,
+      "step": 11985
+    },
+    {
+      "epoch": 0.6596950850349497,
+      "grad_norm": 0.7369599938392639,
+      "learning_rate": 7.572221458255377e-06,
+      "loss": 0.7507,
+      "step": 11986
+    },
+    {
+      "epoch": 0.6597501238373052,
+      "grad_norm": 0.6976807713508606,
+      "learning_rate": 7.571849740648356e-06,
+      "loss": 0.7787,
+      "step": 11987
+    },
+    {
+      "epoch": 0.6598051626396609,
+      "grad_norm": 0.6735848784446716,
+      "learning_rate": 7.571478003711998e-06,
+      "loss": 0.6791,
+      "step": 11988
+    },
+    {
+      "epoch": 0.6598602014420166,
+      "grad_norm": 0.7245956659317017,
+      "learning_rate": 7.5711062474491025e-06,
+      "loss": 0.7999,
+      "step": 11989
+    },
+    {
+      "epoch": 0.6599152402443723,
+      "grad_norm": 0.760748565196991,
+      "learning_rate": 7.5707344718624595e-06,
+      "loss": 0.7904,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6599702790467279,
+      "grad_norm": 0.6745715141296387,
+      "learning_rate": 7.5703626769548654e-06,
+      "loss": 0.6938,
+      "step": 11991
+    },
+    {
+      "epoch": 0.6600253178490836,
+      "grad_norm": 0.7301452159881592,
+      "learning_rate": 7.569990862729113e-06,
+      "loss": 0.7546,
+      "step": 11992
+    },
+    {
+      "epoch": 0.6600803566514393,
+      "grad_norm": 0.68801349401474,
+      "learning_rate": 7.569619029187998e-06,
+      "loss": 0.7592,
+      "step": 11993
+    },
+    {
+      "epoch": 0.660135395453795,
+      "grad_norm": 0.6839548349380493,
+      "learning_rate": 7.569247176334313e-06,
+      "loss": 0.7139,
+      "step": 11994
+    },
+    {
+      "epoch": 0.6601904342561505,
+      "grad_norm": 0.7490861415863037,
+      "learning_rate": 7.568875304170854e-06,
+      "loss": 0.7939,
+      "step": 11995
+    },
+    {
+      "epoch": 0.6602454730585062,
+      "grad_norm": 0.7098836302757263,
+      "learning_rate": 7.568503412700416e-06,
+      "loss": 0.7824,
+      "step": 11996
+    },
+    {
+      "epoch": 0.6603005118608619,
+      "grad_norm": 0.7427988052368164,
+      "learning_rate": 7.568131501925795e-06,
+      "loss": 0.7492,
+      "step": 11997
+    },
+    {
+      "epoch": 0.6603555506632176,
+      "grad_norm": 0.6715356111526489,
+      "learning_rate": 7.567759571849784e-06,
+      "loss": 0.6444,
+      "step": 11998
+    },
+    {
+      "epoch": 0.6604105894655732,
+      "grad_norm": 0.6697829961776733,
+      "learning_rate": 7.5673876224751795e-06,
+      "loss": 0.7064,
+      "step": 11999
+    },
+    {
+      "epoch": 0.6604656282679289,
+      "grad_norm": 0.6778494119644165,
+      "learning_rate": 7.567015653804777e-06,
+      "loss": 0.7517,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6605206670702846,
+      "grad_norm": 0.6423540711402893,
+      "learning_rate": 7.566643665841371e-06,
+      "loss": 0.6321,
+      "step": 12001
+    },
+    {
+      "epoch": 0.6605757058726403,
+      "grad_norm": 0.6874244213104248,
+      "learning_rate": 7.566271658587761e-06,
+      "loss": 0.762,
+      "step": 12002
+    },
+    {
+      "epoch": 0.6606307446749958,
+      "grad_norm": 0.6805301308631897,
+      "learning_rate": 7.565899632046737e-06,
+      "loss": 0.765,
+      "step": 12003
+    },
+    {
+      "epoch": 0.6606857834773515,
+      "grad_norm": 0.7039558291435242,
+      "learning_rate": 7.5655275862211e-06,
+      "loss": 0.728,
+      "step": 12004
+    },
+    {
+      "epoch": 0.6607408222797072,
+      "grad_norm": 0.6513119339942932,
+      "learning_rate": 7.565155521113643e-06,
+      "loss": 0.7711,
+      "step": 12005
+    },
+    {
+      "epoch": 0.6607958610820629,
+      "grad_norm": 0.6483618021011353,
+      "learning_rate": 7.5647834367271655e-06,
+      "loss": 0.7015,
+      "step": 12006
+    },
+    {
+      "epoch": 0.6608508998844185,
+      "grad_norm": 0.7180553674697876,
+      "learning_rate": 7.564411333064461e-06,
+      "loss": 0.812,
+      "step": 12007
+    },
+    {
+      "epoch": 0.6609059386867742,
+      "grad_norm": 0.9036096334457397,
+      "learning_rate": 7.5640392101283285e-06,
+      "loss": 0.7858,
+      "step": 12008
+    },
+    {
+      "epoch": 0.6609609774891299,
+      "grad_norm": 0.7380802035331726,
+      "learning_rate": 7.563667067921563e-06,
+      "loss": 0.6615,
+      "step": 12009
+    },
+    {
+      "epoch": 0.6610160162914855,
+      "grad_norm": 0.6830628514289856,
+      "learning_rate": 7.5632949064469615e-06,
+      "loss": 0.7465,
+      "step": 12010
+    },
+    {
+      "epoch": 0.6610710550938411,
+      "grad_norm": 0.7562816143035889,
+      "learning_rate": 7.562922725707323e-06,
+      "loss": 0.8559,
+      "step": 12011
+    },
+    {
+      "epoch": 0.6611260938961968,
+      "grad_norm": 0.7376649379730225,
+      "learning_rate": 7.562550525705442e-06,
+      "loss": 0.7769,
+      "step": 12012
+    },
+    {
+      "epoch": 0.6611811326985525,
+      "grad_norm": 0.715466320514679,
+      "learning_rate": 7.562178306444116e-06,
+      "loss": 0.8233,
+      "step": 12013
+    },
+    {
+      "epoch": 0.6612361715009082,
+      "grad_norm": 0.6714800596237183,
+      "learning_rate": 7.561806067926147e-06,
+      "loss": 0.6025,
+      "step": 12014
+    },
+    {
+      "epoch": 0.6612912103032638,
+      "grad_norm": 0.7083391547203064,
+      "learning_rate": 7.561433810154328e-06,
+      "loss": 0.7063,
+      "step": 12015
+    },
+    {
+      "epoch": 0.6613462491056195,
+      "grad_norm": 0.8062768578529358,
+      "learning_rate": 7.561061533131457e-06,
+      "loss": 0.7992,
+      "step": 12016
+    },
+    {
+      "epoch": 0.6614012879079751,
+      "grad_norm": 0.741889476776123,
+      "learning_rate": 7.560689236860334e-06,
+      "loss": 0.8149,
+      "step": 12017
+    },
+    {
+      "epoch": 0.6614563267103308,
+      "grad_norm": 0.6834374666213989,
+      "learning_rate": 7.560316921343756e-06,
+      "loss": 0.782,
+      "step": 12018
+    },
+    {
+      "epoch": 0.6615113655126864,
+      "grad_norm": 0.7469872236251831,
+      "learning_rate": 7.559944586584522e-06,
+      "loss": 0.759,
+      "step": 12019
+    },
+    {
+      "epoch": 0.6615664043150421,
+      "grad_norm": 0.8300836086273193,
+      "learning_rate": 7.559572232585428e-06,
+      "loss": 0.8637,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6616214431173978,
+      "grad_norm": 0.6241582632064819,
+      "learning_rate": 7.559199859349276e-06,
+      "loss": 0.7134,
+      "step": 12021
+    },
+    {
+      "epoch": 0.6616764819197535,
+      "grad_norm": 0.6696488261222839,
+      "learning_rate": 7.5588274668788634e-06,
+      "loss": 0.7457,
+      "step": 12022
+    },
+    {
+      "epoch": 0.6617315207221091,
+      "grad_norm": 0.7090815305709839,
+      "learning_rate": 7.558455055176987e-06,
+      "loss": 0.7449,
+      "step": 12023
+    },
+    {
+      "epoch": 0.6617865595244647,
+      "grad_norm": 0.6925215125083923,
+      "learning_rate": 7.558082624246448e-06,
+      "loss": 0.758,
+      "step": 12024
+    },
+    {
+      "epoch": 0.6618415983268204,
+      "grad_norm": 0.6658454537391663,
+      "learning_rate": 7.5577101740900425e-06,
+      "loss": 0.6918,
+      "step": 12025
+    },
+    {
+      "epoch": 0.6618966371291761,
+      "grad_norm": 0.6646405458450317,
+      "learning_rate": 7.557337704710574e-06,
+      "loss": 0.7293,
+      "step": 12026
+    },
+    {
+      "epoch": 0.6619516759315317,
+      "grad_norm": 0.6630399227142334,
+      "learning_rate": 7.556965216110841e-06,
+      "loss": 0.7572,
+      "step": 12027
+    },
+    {
+      "epoch": 0.6620067147338874,
+      "grad_norm": 0.7333918809890747,
+      "learning_rate": 7.556592708293641e-06,
+      "loss": 0.8012,
+      "step": 12028
+    },
+    {
+      "epoch": 0.6620617535362431,
+      "grad_norm": 0.7399254441261292,
+      "learning_rate": 7.556220181261773e-06,
+      "loss": 0.8406,
+      "step": 12029
+    },
+    {
+      "epoch": 0.6621167923385987,
+      "grad_norm": 0.6244909167289734,
+      "learning_rate": 7.55584763501804e-06,
+      "loss": 0.7427,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6621718311409543,
+      "grad_norm": 0.6991485953330994,
+      "learning_rate": 7.55547506956524e-06,
+      "loss": 0.7583,
+      "step": 12031
+    },
+    {
+      "epoch": 0.66222686994331,
+      "grad_norm": 0.7115411162376404,
+      "learning_rate": 7.555102484906174e-06,
+      "loss": 0.7951,
+      "step": 12032
+    },
+    {
+      "epoch": 0.6622819087456657,
+      "grad_norm": 0.7684284448623657,
+      "learning_rate": 7.554729881043641e-06,
+      "loss": 0.717,
+      "step": 12033
+    },
+    {
+      "epoch": 0.6623369475480213,
+      "grad_norm": 0.7705931067466736,
+      "learning_rate": 7.554357257980443e-06,
+      "loss": 0.6903,
+      "step": 12034
+    },
+    {
+      "epoch": 0.662391986350377,
+      "grad_norm": 0.9283333420753479,
+      "learning_rate": 7.553984615719379e-06,
+      "loss": 0.7845,
+      "step": 12035
+    },
+    {
+      "epoch": 0.6624470251527327,
+      "grad_norm": 0.6867572665214539,
+      "learning_rate": 7.553611954263249e-06,
+      "loss": 0.8796,
+      "step": 12036
+    },
+    {
+      "epoch": 0.6625020639550884,
+      "grad_norm": 0.6129451990127563,
+      "learning_rate": 7.553239273614855e-06,
+      "loss": 0.6308,
+      "step": 12037
+    },
+    {
+      "epoch": 0.662557102757444,
+      "grad_norm": 0.749679446220398,
+      "learning_rate": 7.552866573777e-06,
+      "loss": 0.8308,
+      "step": 12038
+    },
+    {
+      "epoch": 0.6626121415597996,
+      "grad_norm": 0.7651422619819641,
+      "learning_rate": 7.552493854752483e-06,
+      "loss": 0.7266,
+      "step": 12039
+    },
+    {
+      "epoch": 0.6626671803621553,
+      "grad_norm": 0.9293195009231567,
+      "learning_rate": 7.552121116544104e-06,
+      "loss": 0.7795,
+      "step": 12040
+    },
+    {
+      "epoch": 0.662722219164511,
+      "grad_norm": 0.7321802973747253,
+      "learning_rate": 7.5517483591546655e-06,
+      "loss": 0.7294,
+      "step": 12041
+    },
+    {
+      "epoch": 0.6627772579668666,
+      "grad_norm": 0.702414333820343,
+      "learning_rate": 7.551375582586971e-06,
+      "loss": 0.7954,
+      "step": 12042
+    },
+    {
+      "epoch": 0.6628322967692223,
+      "grad_norm": 0.7497946619987488,
+      "learning_rate": 7.551002786843819e-06,
+      "loss": 0.7654,
+      "step": 12043
+    },
+    {
+      "epoch": 0.662887335571578,
+      "grad_norm": 0.6125331521034241,
+      "learning_rate": 7.550629971928017e-06,
+      "loss": 0.7299,
+      "step": 12044
+    },
+    {
+      "epoch": 0.6629423743739337,
+      "grad_norm": 0.7252177596092224,
+      "learning_rate": 7.550257137842358e-06,
+      "loss": 0.7553,
+      "step": 12045
+    },
+    {
+      "epoch": 0.6629974131762892,
+      "grad_norm": 0.6463978886604309,
+      "learning_rate": 7.5498842845896515e-06,
+      "loss": 0.7114,
+      "step": 12046
+    },
+    {
+      "epoch": 0.6630524519786449,
+      "grad_norm": 0.7392497062683105,
+      "learning_rate": 7.549511412172696e-06,
+      "loss": 0.6801,
+      "step": 12047
+    },
+    {
+      "epoch": 0.6631074907810006,
+      "grad_norm": 0.8068972229957581,
+      "learning_rate": 7.549138520594297e-06,
+      "loss": 0.8207,
+      "step": 12048
+    },
+    {
+      "epoch": 0.6631625295833563,
+      "grad_norm": 0.7632858753204346,
+      "learning_rate": 7.548765609857254e-06,
+      "loss": 0.7095,
+      "step": 12049
+    },
+    {
+      "epoch": 0.6632175683857119,
+      "grad_norm": 0.7252069115638733,
+      "learning_rate": 7.5483926799643705e-06,
+      "loss": 0.7796,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6632726071880676,
+      "grad_norm": 1.048311471939087,
+      "learning_rate": 7.54801973091845e-06,
+      "loss": 0.7306,
+      "step": 12051
+    },
+    {
+      "epoch": 0.6633276459904233,
+      "grad_norm": 0.7432072758674622,
+      "learning_rate": 7.547646762722296e-06,
+      "loss": 0.8209,
+      "step": 12052
+    },
+    {
+      "epoch": 0.663382684792779,
+      "grad_norm": 0.7191399335861206,
+      "learning_rate": 7.547273775378709e-06,
+      "loss": 0.7011,
+      "step": 12053
+    },
+    {
+      "epoch": 0.6634377235951345,
+      "grad_norm": 0.5776329636573792,
+      "learning_rate": 7.5469007688904975e-06,
+      "loss": 0.6055,
+      "step": 12054
+    },
+    {
+      "epoch": 0.6634927623974902,
+      "grad_norm": 0.9296837449073792,
+      "learning_rate": 7.546527743260459e-06,
+      "loss": 0.7413,
+      "step": 12055
+    },
+    {
+      "epoch": 0.6635478011998459,
+      "grad_norm": 0.7279512286186218,
+      "learning_rate": 7.5461546984914e-06,
+      "loss": 0.7734,
+      "step": 12056
+    },
+    {
+      "epoch": 0.6636028400022016,
+      "grad_norm": 0.7297198176383972,
+      "learning_rate": 7.545781634586125e-06,
+      "loss": 0.7535,
+      "step": 12057
+    },
+    {
+      "epoch": 0.6636578788045572,
+      "grad_norm": 0.7094287872314453,
+      "learning_rate": 7.545408551547435e-06,
+      "loss": 0.7587,
+      "step": 12058
+    },
+    {
+      "epoch": 0.6637129176069129,
+      "grad_norm": 0.7559607028961182,
+      "learning_rate": 7.5450354493781374e-06,
+      "loss": 0.7358,
+      "step": 12059
+    },
+    {
+      "epoch": 0.6637679564092686,
+      "grad_norm": 0.8472892045974731,
+      "learning_rate": 7.544662328081034e-06,
+      "loss": 0.7537,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6638229952116242,
+      "grad_norm": 0.6346176862716675,
+      "learning_rate": 7.544289187658929e-06,
+      "loss": 0.7658,
+      "step": 12061
+    },
+    {
+      "epoch": 0.6638780340139798,
+      "grad_norm": 0.7949367165565491,
+      "learning_rate": 7.543916028114628e-06,
+      "loss": 0.6837,
+      "step": 12062
+    },
+    {
+      "epoch": 0.6639330728163355,
+      "grad_norm": 0.7177689671516418,
+      "learning_rate": 7.5435428494509355e-06,
+      "loss": 0.7218,
+      "step": 12063
+    },
+    {
+      "epoch": 0.6639881116186912,
+      "grad_norm": 0.90680330991745,
+      "learning_rate": 7.5431696516706555e-06,
+      "loss": 0.8274,
+      "step": 12064
+    },
+    {
+      "epoch": 0.6640431504210469,
+      "grad_norm": 0.7799603939056396,
+      "learning_rate": 7.5427964347765916e-06,
+      "loss": 0.7528,
+      "step": 12065
+    },
+    {
+      "epoch": 0.6640981892234025,
+      "grad_norm": 0.7668048739433289,
+      "learning_rate": 7.542423198771553e-06,
+      "loss": 0.746,
+      "step": 12066
+    },
+    {
+      "epoch": 0.6641532280257582,
+      "grad_norm": 1.0042381286621094,
+      "learning_rate": 7.542049943658341e-06,
+      "loss": 0.7836,
+      "step": 12067
+    },
+    {
+      "epoch": 0.6642082668281138,
+      "grad_norm": 0.6915723085403442,
+      "learning_rate": 7.541676669439761e-06,
+      "loss": 0.8042,
+      "step": 12068
+    },
+    {
+      "epoch": 0.6642633056304695,
+      "grad_norm": 0.7268955707550049,
+      "learning_rate": 7.5413033761186215e-06,
+      "loss": 0.689,
+      "step": 12069
+    },
+    {
+      "epoch": 0.6643183444328251,
+      "grad_norm": 0.6418740749359131,
+      "learning_rate": 7.540930063697726e-06,
+      "loss": 0.6302,
+      "step": 12070
+    },
+    {
+      "epoch": 0.6643733832351808,
+      "grad_norm": 0.696384847164154,
+      "learning_rate": 7.540556732179879e-06,
+      "loss": 0.7978,
+      "step": 12071
+    },
+    {
+      "epoch": 0.6644284220375365,
+      "grad_norm": 0.7400668859481812,
+      "learning_rate": 7.540183381567889e-06,
+      "loss": 0.8768,
+      "step": 12072
+    },
+    {
+      "epoch": 0.6644834608398921,
+      "grad_norm": 0.6653871536254883,
+      "learning_rate": 7.539810011864559e-06,
+      "loss": 0.8107,
+      "step": 12073
+    },
+    {
+      "epoch": 0.6645384996422478,
+      "grad_norm": 0.7635810971260071,
+      "learning_rate": 7.539436623072698e-06,
+      "loss": 0.8476,
+      "step": 12074
+    },
+    {
+      "epoch": 0.6645935384446034,
+      "grad_norm": 0.6583054661750793,
+      "learning_rate": 7.53906321519511e-06,
+      "loss": 0.7093,
+      "step": 12075
+    },
+    {
+      "epoch": 0.6646485772469591,
+      "grad_norm": 0.8294859528541565,
+      "learning_rate": 7.538689788234604e-06,
+      "loss": 0.8107,
+      "step": 12076
+    },
+    {
+      "epoch": 0.6647036160493147,
+      "grad_norm": 0.6711081862449646,
+      "learning_rate": 7.538316342193983e-06,
+      "loss": 0.7491,
+      "step": 12077
+    },
+    {
+      "epoch": 0.6647586548516704,
+      "grad_norm": 0.7375408411026001,
+      "learning_rate": 7.5379428770760575e-06,
+      "loss": 0.7853,
+      "step": 12078
+    },
+    {
+      "epoch": 0.6648136936540261,
+      "grad_norm": 0.7322511672973633,
+      "learning_rate": 7.537569392883633e-06,
+      "loss": 0.7568,
+      "step": 12079
+    },
+    {
+      "epoch": 0.6648687324563818,
+      "grad_norm": 0.6390300393104553,
+      "learning_rate": 7.537195889619515e-06,
+      "loss": 0.7191,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6649237712587374,
+      "grad_norm": 0.8155800104141235,
+      "learning_rate": 7.536822367286514e-06,
+      "loss": 0.7499,
+      "step": 12081
+    },
+    {
+      "epoch": 0.664978810061093,
+      "grad_norm": 0.7942230701446533,
+      "learning_rate": 7.536448825887432e-06,
+      "loss": 0.7797,
+      "step": 12082
+    },
+    {
+      "epoch": 0.6650338488634487,
+      "grad_norm": 0.7103378176689148,
+      "learning_rate": 7.536075265425083e-06,
+      "loss": 0.6814,
+      "step": 12083
+    },
+    {
+      "epoch": 0.6650888876658044,
+      "grad_norm": 0.8164991736412048,
+      "learning_rate": 7.535701685902268e-06,
+      "loss": 0.7917,
+      "step": 12084
+    },
+    {
+      "epoch": 0.66514392646816,
+      "grad_norm": 0.6970370411872864,
+      "learning_rate": 7.535328087321799e-06,
+      "loss": 0.7266,
+      "step": 12085
+    },
+    {
+      "epoch": 0.6651989652705157,
+      "grad_norm": 0.6468706130981445,
+      "learning_rate": 7.534954469686484e-06,
+      "loss": 0.7229,
+      "step": 12086
+    },
+    {
+      "epoch": 0.6652540040728714,
+      "grad_norm": 0.6551242470741272,
+      "learning_rate": 7.534580832999128e-06,
+      "loss": 0.6759,
+      "step": 12087
+    },
+    {
+      "epoch": 0.6653090428752271,
+      "grad_norm": 0.670215368270874,
+      "learning_rate": 7.534207177262543e-06,
+      "loss": 0.761,
+      "step": 12088
+    },
+    {
+      "epoch": 0.6653640816775827,
+      "grad_norm": 0.7365970015525818,
+      "learning_rate": 7.533833502479533e-06,
+      "loss": 0.7628,
+      "step": 12089
+    },
+    {
+      "epoch": 0.6654191204799383,
+      "grad_norm": 0.7419471740722656,
+      "learning_rate": 7.53345980865291e-06,
+      "loss": 0.8093,
+      "step": 12090
+    },
+    {
+      "epoch": 0.665474159282294,
+      "grad_norm": 0.6573269963264465,
+      "learning_rate": 7.53308609578548e-06,
+      "loss": 0.6806,
+      "step": 12091
+    },
+    {
+      "epoch": 0.6655291980846497,
+      "grad_norm": 0.9270638227462769,
+      "learning_rate": 7.5327123638800545e-06,
+      "loss": 0.8612,
+      "step": 12092
+    },
+    {
+      "epoch": 0.6655842368870053,
+      "grad_norm": 0.85124671459198,
+      "learning_rate": 7.532338612939441e-06,
+      "loss": 0.6776,
+      "step": 12093
+    },
+    {
+      "epoch": 0.665639275689361,
+      "grad_norm": 0.7791070342063904,
+      "learning_rate": 7.531964842966446e-06,
+      "loss": 0.7571,
+      "step": 12094
+    },
+    {
+      "epoch": 0.6656943144917167,
+      "grad_norm": 0.6604436635971069,
+      "learning_rate": 7.5315910539638825e-06,
+      "loss": 0.781,
+      "step": 12095
+    },
+    {
+      "epoch": 0.6657493532940724,
+      "grad_norm": 0.7567091584205627,
+      "learning_rate": 7.531217245934559e-06,
+      "loss": 0.8005,
+      "step": 12096
+    },
+    {
+      "epoch": 0.6658043920964279,
+      "grad_norm": 0.660637378692627,
+      "learning_rate": 7.530843418881282e-06,
+      "loss": 0.7351,
+      "step": 12097
+    },
+    {
+      "epoch": 0.6658594308987836,
+      "grad_norm": 0.6305738687515259,
+      "learning_rate": 7.530469572806865e-06,
+      "loss": 0.7452,
+      "step": 12098
+    },
+    {
+      "epoch": 0.6659144697011393,
+      "grad_norm": 0.8291265368461609,
+      "learning_rate": 7.5300957077141164e-06,
+      "loss": 0.7799,
+      "step": 12099
+    },
+    {
+      "epoch": 0.665969508503495,
+      "grad_norm": 0.7459661364555359,
+      "learning_rate": 7.5297218236058456e-06,
+      "loss": 0.8273,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6660245473058506,
+      "grad_norm": 0.7570028901100159,
+      "learning_rate": 7.529347920484862e-06,
+      "loss": 0.7622,
+      "step": 12101
+    },
+    {
+      "epoch": 0.6660795861082063,
+      "grad_norm": 0.733403205871582,
+      "learning_rate": 7.528973998353977e-06,
+      "loss": 0.8357,
+      "step": 12102
+    },
+    {
+      "epoch": 0.666134624910562,
+      "grad_norm": 0.8814442753791809,
+      "learning_rate": 7.528600057216e-06,
+      "loss": 0.727,
+      "step": 12103
+    },
+    {
+      "epoch": 0.6661896637129177,
+      "grad_norm": 0.629338800907135,
+      "learning_rate": 7.528226097073742e-06,
+      "loss": 0.6758,
+      "step": 12104
+    },
+    {
+      "epoch": 0.6662447025152732,
+      "grad_norm": 0.7786098122596741,
+      "learning_rate": 7.527852117930014e-06,
+      "loss": 0.7476,
+      "step": 12105
+    },
+    {
+      "epoch": 0.6662997413176289,
+      "grad_norm": 0.6604528427124023,
+      "learning_rate": 7.527478119787626e-06,
+      "loss": 0.7275,
+      "step": 12106
+    },
+    {
+      "epoch": 0.6663547801199846,
+      "grad_norm": 0.6937400698661804,
+      "learning_rate": 7.527104102649387e-06,
+      "loss": 0.7187,
+      "step": 12107
+    },
+    {
+      "epoch": 0.6664098189223403,
+      "grad_norm": 0.6863219738006592,
+      "learning_rate": 7.526730066518113e-06,
+      "loss": 0.7512,
+      "step": 12108
+    },
+    {
+      "epoch": 0.6664648577246959,
+      "grad_norm": 0.7771461606025696,
+      "learning_rate": 7.526356011396609e-06,
+      "loss": 0.8439,
+      "step": 12109
+    },
+    {
+      "epoch": 0.6665198965270516,
+      "grad_norm": 0.7223722338676453,
+      "learning_rate": 7.525981937287692e-06,
+      "loss": 0.6488,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6665749353294073,
+      "grad_norm": 0.8091556429862976,
+      "learning_rate": 7.52560784419417e-06,
+      "loss": 0.6618,
+      "step": 12111
+    },
+    {
+      "epoch": 0.666629974131763,
+      "grad_norm": 0.6435044407844543,
+      "learning_rate": 7.525233732118856e-06,
+      "loss": 0.6994,
+      "step": 12112
+    },
+    {
+      "epoch": 0.6666850129341185,
+      "grad_norm": 0.6933714151382446,
+      "learning_rate": 7.52485960106456e-06,
+      "loss": 0.6917,
+      "step": 12113
+    },
+    {
+      "epoch": 0.6667400517364742,
+      "grad_norm": 0.693192720413208,
+      "learning_rate": 7.524485451034097e-06,
+      "loss": 0.7941,
+      "step": 12114
+    },
+    {
+      "epoch": 0.6667950905388299,
+      "grad_norm": 1.1374844312667847,
+      "learning_rate": 7.524111282030275e-06,
+      "loss": 0.9112,
+      "step": 12115
+    },
+    {
+      "epoch": 0.6668501293411855,
+      "grad_norm": 0.6917465329170227,
+      "learning_rate": 7.523737094055911e-06,
+      "loss": 0.681,
+      "step": 12116
+    },
+    {
+      "epoch": 0.6669051681435412,
+      "grad_norm": 0.8057913184165955,
+      "learning_rate": 7.523362887113812e-06,
+      "loss": 0.8186,
+      "step": 12117
+    },
+    {
+      "epoch": 0.6669602069458969,
+      "grad_norm": 0.7194918394088745,
+      "learning_rate": 7.522988661206795e-06,
+      "loss": 0.7875,
+      "step": 12118
+    },
+    {
+      "epoch": 0.6670152457482525,
+      "grad_norm": 0.6829916834831238,
+      "learning_rate": 7.52261441633767e-06,
+      "loss": 0.6506,
+      "step": 12119
+    },
+    {
+      "epoch": 0.6670702845506081,
+      "grad_norm": 0.7869738936424255,
+      "learning_rate": 7.5222401525092495e-06,
+      "loss": 0.7091,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6671253233529638,
+      "grad_norm": 0.6835895776748657,
+      "learning_rate": 7.5218658697243475e-06,
+      "loss": 0.7839,
+      "step": 12121
+    },
+    {
+      "epoch": 0.6671803621553195,
+      "grad_norm": 0.7462154030799866,
+      "learning_rate": 7.521491567985776e-06,
+      "loss": 0.7073,
+      "step": 12122
+    },
+    {
+      "epoch": 0.6672354009576752,
+      "grad_norm": 0.6413764953613281,
+      "learning_rate": 7.52111724729635e-06,
+      "loss": 0.6472,
+      "step": 12123
+    },
+    {
+      "epoch": 0.6672904397600308,
+      "grad_norm": 0.7085923552513123,
+      "learning_rate": 7.520742907658881e-06,
+      "loss": 0.8167,
+      "step": 12124
+    },
+    {
+      "epoch": 0.6673454785623865,
+      "grad_norm": 0.6490428447723389,
+      "learning_rate": 7.520368549076182e-06,
+      "loss": 0.7693,
+      "step": 12125
+    },
+    {
+      "epoch": 0.6674005173647422,
+      "grad_norm": 0.7082974910736084,
+      "learning_rate": 7.51999417155107e-06,
+      "loss": 0.6707,
+      "step": 12126
+    },
+    {
+      "epoch": 0.6674555561670978,
+      "grad_norm": 0.704335629940033,
+      "learning_rate": 7.519619775086355e-06,
+      "loss": 0.825,
+      "step": 12127
+    },
+    {
+      "epoch": 0.6675105949694534,
+      "grad_norm": 0.6815123558044434,
+      "learning_rate": 7.519245359684852e-06,
+      "loss": 0.762,
+      "step": 12128
+    },
+    {
+      "epoch": 0.6675656337718091,
+      "grad_norm": 0.6497910618782043,
+      "learning_rate": 7.518870925349376e-06,
+      "loss": 0.6934,
+      "step": 12129
+    },
+    {
+      "epoch": 0.6676206725741648,
+      "grad_norm": 0.6699943542480469,
+      "learning_rate": 7.51849647208274e-06,
+      "loss": 0.7816,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6676757113765205,
+      "grad_norm": 0.7139337062835693,
+      "learning_rate": 7.51812199988776e-06,
+      "loss": 0.679,
+      "step": 12131
+    },
+    {
+      "epoch": 0.6677307501788761,
+      "grad_norm": 0.6762346029281616,
+      "learning_rate": 7.517747508767248e-06,
+      "loss": 0.7477,
+      "step": 12132
+    },
+    {
+      "epoch": 0.6677857889812318,
+      "grad_norm": 0.7429338693618774,
+      "learning_rate": 7.517372998724017e-06,
+      "loss": 0.7549,
+      "step": 12133
+    },
+    {
+      "epoch": 0.6678408277835874,
+      "grad_norm": 0.7392850518226624,
+      "learning_rate": 7.516998469760888e-06,
+      "loss": 0.8167,
+      "step": 12134
+    },
+    {
+      "epoch": 0.6678958665859431,
+      "grad_norm": 0.7511306405067444,
+      "learning_rate": 7.516623921880671e-06,
+      "loss": 0.7264,
+      "step": 12135
+    },
+    {
+      "epoch": 0.6679509053882987,
+      "grad_norm": 0.6757550835609436,
+      "learning_rate": 7.516249355086183e-06,
+      "loss": 0.7405,
+      "step": 12136
+    },
+    {
+      "epoch": 0.6680059441906544,
+      "grad_norm": 0.7433735132217407,
+      "learning_rate": 7.515874769380238e-06,
+      "loss": 0.7954,
+      "step": 12137
+    },
+    {
+      "epoch": 0.6680609829930101,
+      "grad_norm": 0.7390886545181274,
+      "learning_rate": 7.51550016476565e-06,
+      "loss": 0.7487,
+      "step": 12138
+    },
+    {
+      "epoch": 0.6681160217953658,
+      "grad_norm": 0.7405929565429688,
+      "learning_rate": 7.5151255412452385e-06,
+      "loss": 0.8127,
+      "step": 12139
+    },
+    {
+      "epoch": 0.6681710605977214,
+      "grad_norm": 0.6628968715667725,
+      "learning_rate": 7.514750898821817e-06,
+      "loss": 0.7009,
+      "step": 12140
+    },
+    {
+      "epoch": 0.668226099400077,
+      "grad_norm": 0.6777421832084656,
+      "learning_rate": 7.514376237498199e-06,
+      "loss": 0.6689,
+      "step": 12141
+    },
+    {
+      "epoch": 0.6682811382024327,
+      "grad_norm": 0.617261528968811,
+      "learning_rate": 7.514001557277202e-06,
+      "loss": 0.7597,
+      "step": 12142
+    },
+    {
+      "epoch": 0.6683361770047884,
+      "grad_norm": 0.6666202545166016,
+      "learning_rate": 7.5136268581616446e-06,
+      "loss": 0.6623,
+      "step": 12143
+    },
+    {
+      "epoch": 0.668391215807144,
+      "grad_norm": 0.7170178890228271,
+      "learning_rate": 7.513252140154339e-06,
+      "loss": 0.8224,
+      "step": 12144
+    },
+    {
+      "epoch": 0.6684462546094997,
+      "grad_norm": 0.6173199415206909,
+      "learning_rate": 7.512877403258103e-06,
+      "loss": 0.6784,
+      "step": 12145
+    },
+    {
+      "epoch": 0.6685012934118554,
+      "grad_norm": 0.6906641125679016,
+      "learning_rate": 7.512502647475753e-06,
+      "loss": 0.6649,
+      "step": 12146
+    },
+    {
+      "epoch": 0.6685563322142111,
+      "grad_norm": 0.6435873508453369,
+      "learning_rate": 7.5121278728101065e-06,
+      "loss": 0.751,
+      "step": 12147
+    },
+    {
+      "epoch": 0.6686113710165666,
+      "grad_norm": 0.8345947861671448,
+      "learning_rate": 7.511753079263978e-06,
+      "loss": 0.7841,
+      "step": 12148
+    },
+    {
+      "epoch": 0.6686664098189223,
+      "grad_norm": 0.6952378153800964,
+      "learning_rate": 7.511378266840187e-06,
+      "loss": 0.8187,
+      "step": 12149
+    },
+    {
+      "epoch": 0.668721448621278,
+      "grad_norm": 0.6878920793533325,
+      "learning_rate": 7.5110034355415484e-06,
+      "loss": 0.6726,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6687764874236337,
+      "grad_norm": 0.7119094729423523,
+      "learning_rate": 7.5106285853708805e-06,
+      "loss": 0.7824,
+      "step": 12151
+    },
+    {
+      "epoch": 0.6688315262259893,
+      "grad_norm": 0.7261053323745728,
+      "learning_rate": 7.5102537163309994e-06,
+      "loss": 0.7122,
+      "step": 12152
+    },
+    {
+      "epoch": 0.668886565028345,
+      "grad_norm": 0.717268168926239,
+      "learning_rate": 7.509878828424725e-06,
+      "loss": 0.7144,
+      "step": 12153
+    },
+    {
+      "epoch": 0.6689416038307007,
+      "grad_norm": 0.8373270630836487,
+      "learning_rate": 7.5095039216548725e-06,
+      "loss": 0.7941,
+      "step": 12154
+    },
+    {
+      "epoch": 0.6689966426330564,
+      "grad_norm": 0.7113829851150513,
+      "learning_rate": 7.509128996024259e-06,
+      "loss": 0.705,
+      "step": 12155
+    },
+    {
+      "epoch": 0.6690516814354119,
+      "grad_norm": 0.7894094586372375,
+      "learning_rate": 7.508754051535705e-06,
+      "loss": 0.8284,
+      "step": 12156
+    },
+    {
+      "epoch": 0.6691067202377676,
+      "grad_norm": 0.6739659905433655,
+      "learning_rate": 7.508379088192028e-06,
+      "loss": 0.7264,
+      "step": 12157
+    },
+    {
+      "epoch": 0.6691617590401233,
+      "grad_norm": 0.735211193561554,
+      "learning_rate": 7.508004105996043e-06,
+      "loss": 0.8187,
+      "step": 12158
+    },
+    {
+      "epoch": 0.6692167978424789,
+      "grad_norm": 0.7438055872917175,
+      "learning_rate": 7.507629104950571e-06,
+      "loss": 0.8949,
+      "step": 12159
+    },
+    {
+      "epoch": 0.6692718366448346,
+      "grad_norm": 1.0734246969223022,
+      "learning_rate": 7.507254085058431e-06,
+      "loss": 0.7687,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6693268754471903,
+      "grad_norm": 0.6719897985458374,
+      "learning_rate": 7.50687904632244e-06,
+      "loss": 0.7522,
+      "step": 12161
+    },
+    {
+      "epoch": 0.669381914249546,
+      "grad_norm": 0.7063966989517212,
+      "learning_rate": 7.506503988745416e-06,
+      "loss": 0.7794,
+      "step": 12162
+    },
+    {
+      "epoch": 0.6694369530519015,
+      "grad_norm": 0.6582265496253967,
+      "learning_rate": 7.506128912330179e-06,
+      "loss": 0.7012,
+      "step": 12163
+    },
+    {
+      "epoch": 0.6694919918542572,
+      "grad_norm": 0.7764506340026855,
+      "learning_rate": 7.50575381707955e-06,
+      "loss": 0.7816,
+      "step": 12164
+    },
+    {
+      "epoch": 0.6695470306566129,
+      "grad_norm": 0.7659780383110046,
+      "learning_rate": 7.505378702996344e-06,
+      "loss": 0.753,
+      "step": 12165
+    },
+    {
+      "epoch": 0.6696020694589686,
+      "grad_norm": 0.9013122916221619,
+      "learning_rate": 7.505003570083385e-06,
+      "loss": 0.8255,
+      "step": 12166
+    },
+    {
+      "epoch": 0.6696571082613242,
+      "grad_norm": 0.6417272686958313,
+      "learning_rate": 7.504628418343487e-06,
+      "loss": 0.6236,
+      "step": 12167
+    },
+    {
+      "epoch": 0.6697121470636799,
+      "grad_norm": 0.7511595487594604,
+      "learning_rate": 7.504253247779474e-06,
+      "loss": 0.7961,
+      "step": 12168
+    },
+    {
+      "epoch": 0.6697671858660356,
+      "grad_norm": 0.7987878918647766,
+      "learning_rate": 7.503878058394163e-06,
+      "loss": 0.7249,
+      "step": 12169
+    },
+    {
+      "epoch": 0.6698222246683913,
+      "grad_norm": 0.6860646605491638,
+      "learning_rate": 7.503502850190374e-06,
+      "loss": 0.7973,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6698772634707468,
+      "grad_norm": 0.7334334850311279,
+      "learning_rate": 7.50312762317093e-06,
+      "loss": 0.8756,
+      "step": 12171
+    },
+    {
+      "epoch": 0.6699323022731025,
+      "grad_norm": 0.7792186737060547,
+      "learning_rate": 7.502752377338647e-06,
+      "loss": 0.8393,
+      "step": 12172
+    },
+    {
+      "epoch": 0.6699873410754582,
+      "grad_norm": 0.6532536149024963,
+      "learning_rate": 7.502377112696346e-06,
+      "loss": 0.6509,
+      "step": 12173
+    },
+    {
+      "epoch": 0.6700423798778139,
+      "grad_norm": 0.6595458984375,
+      "learning_rate": 7.50200182924685e-06,
+      "loss": 0.781,
+      "step": 12174
+    },
+    {
+      "epoch": 0.6700974186801695,
+      "grad_norm": 0.6668636202812195,
+      "learning_rate": 7.501626526992978e-06,
+      "loss": 0.7702,
+      "step": 12175
+    },
+    {
+      "epoch": 0.6701524574825252,
+      "grad_norm": 0.686851441860199,
+      "learning_rate": 7.501251205937551e-06,
+      "loss": 0.8648,
+      "step": 12176
+    },
+    {
+      "epoch": 0.6702074962848809,
+      "grad_norm": 0.7363078594207764,
+      "learning_rate": 7.500875866083388e-06,
+      "loss": 0.7309,
+      "step": 12177
+    },
+    {
+      "epoch": 0.6702625350872365,
+      "grad_norm": 0.6927379369735718,
+      "learning_rate": 7.500500507433312e-06,
+      "loss": 0.7258,
+      "step": 12178
+    },
+    {
+      "epoch": 0.6703175738895921,
+      "grad_norm": 0.6589936017990112,
+      "learning_rate": 7.5001251299901455e-06,
+      "loss": 0.6776,
+      "step": 12179
+    },
+    {
+      "epoch": 0.6703726126919478,
+      "grad_norm": 0.6402539610862732,
+      "learning_rate": 7.499749733756707e-06,
+      "loss": 0.7467,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6704276514943035,
+      "grad_norm": 0.776469886302948,
+      "learning_rate": 7.499374318735817e-06,
+      "loss": 0.7856,
+      "step": 12181
+    },
+    {
+      "epoch": 0.6704826902966592,
+      "grad_norm": 0.7062460780143738,
+      "learning_rate": 7.4989988849303e-06,
+      "loss": 0.8286,
+      "step": 12182
+    },
+    {
+      "epoch": 0.6705377290990148,
+      "grad_norm": 0.6725799441337585,
+      "learning_rate": 7.4986234323429755e-06,
+      "loss": 0.7517,
+      "step": 12183
+    },
+    {
+      "epoch": 0.6705927679013705,
+      "grad_norm": 0.6444042921066284,
+      "learning_rate": 7.498247960976667e-06,
+      "loss": 0.5984,
+      "step": 12184
+    },
+    {
+      "epoch": 0.6706478067037261,
+      "grad_norm": 0.6968628764152527,
+      "learning_rate": 7.497872470834195e-06,
+      "loss": 0.6996,
+      "step": 12185
+    },
+    {
+      "epoch": 0.6707028455060818,
+      "grad_norm": 0.643500030040741,
+      "learning_rate": 7.497496961918381e-06,
+      "loss": 0.6252,
+      "step": 12186
+    },
+    {
+      "epoch": 0.6707578843084374,
+      "grad_norm": 0.7026870846748352,
+      "learning_rate": 7.49712143423205e-06,
+      "loss": 0.7883,
+      "step": 12187
+    },
+    {
+      "epoch": 0.6708129231107931,
+      "grad_norm": 0.8169240951538086,
+      "learning_rate": 7.496745887778022e-06,
+      "loss": 0.6717,
+      "step": 12188
+    },
+    {
+      "epoch": 0.6708679619131488,
+      "grad_norm": 0.6611927151679993,
+      "learning_rate": 7.496370322559121e-06,
+      "loss": 0.6674,
+      "step": 12189
+    },
+    {
+      "epoch": 0.6709230007155045,
+      "grad_norm": 0.7330195307731628,
+      "learning_rate": 7.495994738578169e-06,
+      "loss": 0.7809,
+      "step": 12190
+    },
+    {
+      "epoch": 0.6709780395178601,
+      "grad_norm": 0.6469636559486389,
+      "learning_rate": 7.495619135837988e-06,
+      "loss": 0.6511,
+      "step": 12191
+    },
+    {
+      "epoch": 0.6710330783202157,
+      "grad_norm": 0.6558564901351929,
+      "learning_rate": 7.495243514341402e-06,
+      "loss": 0.7284,
+      "step": 12192
+    },
+    {
+      "epoch": 0.6710881171225714,
+      "grad_norm": 0.6736281514167786,
+      "learning_rate": 7.494867874091233e-06,
+      "loss": 0.7007,
+      "step": 12193
+    },
+    {
+      "epoch": 0.6711431559249271,
+      "grad_norm": 0.7302053570747375,
+      "learning_rate": 7.494492215090304e-06,
+      "loss": 0.77,
+      "step": 12194
+    },
+    {
+      "epoch": 0.6711981947272827,
+      "grad_norm": 0.7368764877319336,
+      "learning_rate": 7.494116537341442e-06,
+      "loss": 0.8478,
+      "step": 12195
+    },
+    {
+      "epoch": 0.6712532335296384,
+      "grad_norm": 0.782767653465271,
+      "learning_rate": 7.493740840847466e-06,
+      "loss": 0.813,
+      "step": 12196
+    },
+    {
+      "epoch": 0.6713082723319941,
+      "grad_norm": 0.6787601113319397,
+      "learning_rate": 7.493365125611202e-06,
+      "loss": 0.7507,
+      "step": 12197
+    },
+    {
+      "epoch": 0.6713633111343498,
+      "grad_norm": 0.6912569999694824,
+      "learning_rate": 7.4929893916354715e-06,
+      "loss": 0.8003,
+      "step": 12198
+    },
+    {
+      "epoch": 0.6714183499367053,
+      "grad_norm": 0.7625328898429871,
+      "learning_rate": 7.4926136389231005e-06,
+      "loss": 0.8021,
+      "step": 12199
+    },
+    {
+      "epoch": 0.671473388739061,
+      "grad_norm": 0.6720984578132629,
+      "learning_rate": 7.4922378674769146e-06,
+      "loss": 0.7757,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6715284275414167,
+      "grad_norm": 0.7816714644432068,
+      "learning_rate": 7.491862077299734e-06,
+      "loss": 0.7086,
+      "step": 12201
+    },
+    {
+      "epoch": 0.6715834663437723,
+      "grad_norm": 0.7546358108520508,
+      "learning_rate": 7.491486268394387e-06,
+      "loss": 0.8365,
+      "step": 12202
+    },
+    {
+      "epoch": 0.671638505146128,
+      "grad_norm": 0.7201979756355286,
+      "learning_rate": 7.491110440763695e-06,
+      "loss": 0.835,
+      "step": 12203
+    },
+    {
+      "epoch": 0.6716935439484837,
+      "grad_norm": 0.8177551031112671,
+      "learning_rate": 7.490734594410484e-06,
+      "loss": 0.8636,
+      "step": 12204
+    },
+    {
+      "epoch": 0.6717485827508394,
+      "grad_norm": 0.7433933019638062,
+      "learning_rate": 7.490358729337578e-06,
+      "loss": 0.745,
+      "step": 12205
+    },
+    {
+      "epoch": 0.671803621553195,
+      "grad_norm": 0.8013591170310974,
+      "learning_rate": 7.489982845547802e-06,
+      "loss": 0.7638,
+      "step": 12206
+    },
+    {
+      "epoch": 0.6718586603555506,
+      "grad_norm": 0.6561495065689087,
+      "learning_rate": 7.489606943043982e-06,
+      "loss": 0.7997,
+      "step": 12207
+    },
+    {
+      "epoch": 0.6719136991579063,
+      "grad_norm": 0.7291023135185242,
+      "learning_rate": 7.489231021828943e-06,
+      "loss": 0.7452,
+      "step": 12208
+    },
+    {
+      "epoch": 0.671968737960262,
+      "grad_norm": 0.6978216171264648,
+      "learning_rate": 7.488855081905511e-06,
+      "loss": 0.7984,
+      "step": 12209
+    },
+    {
+      "epoch": 0.6720237767626176,
+      "grad_norm": 0.701006293296814,
+      "learning_rate": 7.488479123276507e-06,
+      "loss": 0.7218,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6720788155649733,
+      "grad_norm": 0.7275286912918091,
+      "learning_rate": 7.488103145944763e-06,
+      "loss": 0.6872,
+      "step": 12211
+    },
+    {
+      "epoch": 0.672133854367329,
+      "grad_norm": 0.7319645881652832,
+      "learning_rate": 7.487727149913101e-06,
+      "loss": 0.7862,
+      "step": 12212
+    },
+    {
+      "epoch": 0.6721888931696847,
+      "grad_norm": 0.7143612504005432,
+      "learning_rate": 7.487351135184348e-06,
+      "loss": 0.838,
+      "step": 12213
+    },
+    {
+      "epoch": 0.6722439319720402,
+      "grad_norm": 0.7135382294654846,
+      "learning_rate": 7.486975101761329e-06,
+      "loss": 0.7263,
+      "step": 12214
+    },
+    {
+      "epoch": 0.6722989707743959,
+      "grad_norm": 0.6283460259437561,
+      "learning_rate": 7.486599049646872e-06,
+      "loss": 0.7262,
+      "step": 12215
+    },
+    {
+      "epoch": 0.6723540095767516,
+      "grad_norm": 0.7196768522262573,
+      "learning_rate": 7.486222978843801e-06,
+      "loss": 0.6752,
+      "step": 12216
+    },
+    {
+      "epoch": 0.6724090483791073,
+      "grad_norm": 0.5856572389602661,
+      "learning_rate": 7.485846889354944e-06,
+      "loss": 0.6779,
+      "step": 12217
+    },
+    {
+      "epoch": 0.6724640871814629,
+      "grad_norm": 0.7671294808387756,
+      "learning_rate": 7.485470781183126e-06,
+      "loss": 0.766,
+      "step": 12218
+    },
+    {
+      "epoch": 0.6725191259838186,
+      "grad_norm": 0.6780520677566528,
+      "learning_rate": 7.485094654331177e-06,
+      "loss": 0.7474,
+      "step": 12219
+    },
+    {
+      "epoch": 0.6725741647861743,
+      "grad_norm": 0.7537981867790222,
+      "learning_rate": 7.484718508801921e-06,
+      "loss": 0.8347,
+      "step": 12220
+    },
+    {
+      "epoch": 0.67262920358853,
+      "grad_norm": 0.7451551556587219,
+      "learning_rate": 7.484342344598186e-06,
+      "loss": 0.8217,
+      "step": 12221
+    },
+    {
+      "epoch": 0.6726842423908855,
+      "grad_norm": 0.6656951904296875,
+      "learning_rate": 7.483966161722798e-06,
+      "loss": 0.7437,
+      "step": 12222
+    },
+    {
+      "epoch": 0.6727392811932412,
+      "grad_norm": 0.7306267619132996,
+      "learning_rate": 7.483589960178586e-06,
+      "loss": 0.8495,
+      "step": 12223
+    },
+    {
+      "epoch": 0.6727943199955969,
+      "grad_norm": 0.6619658470153809,
+      "learning_rate": 7.483213739968376e-06,
+      "loss": 0.6379,
+      "step": 12224
+    },
+    {
+      "epoch": 0.6728493587979526,
+      "grad_norm": 0.7066444754600525,
+      "learning_rate": 7.4828375010949974e-06,
+      "loss": 0.7307,
+      "step": 12225
+    },
+    {
+      "epoch": 0.6729043976003082,
+      "grad_norm": 0.7356079816818237,
+      "learning_rate": 7.482461243561276e-06,
+      "loss": 0.7781,
+      "step": 12226
+    },
+    {
+      "epoch": 0.6729594364026639,
+      "grad_norm": 0.6759988069534302,
+      "learning_rate": 7.48208496737004e-06,
+      "loss": 0.7808,
+      "step": 12227
+    },
+    {
+      "epoch": 0.6730144752050196,
+      "grad_norm": 0.7519234418869019,
+      "learning_rate": 7.481708672524119e-06,
+      "loss": 0.7948,
+      "step": 12228
+    },
+    {
+      "epoch": 0.6730695140073752,
+      "grad_norm": 0.6387592554092407,
+      "learning_rate": 7.48133235902634e-06,
+      "loss": 0.7423,
+      "step": 12229
+    },
+    {
+      "epoch": 0.6731245528097308,
+      "grad_norm": 1.0615060329437256,
+      "learning_rate": 7.480956026879529e-06,
+      "loss": 0.8668,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6731795916120865,
+      "grad_norm": 0.7578469514846802,
+      "learning_rate": 7.480579676086519e-06,
+      "loss": 0.812,
+      "step": 12231
+    },
+    {
+      "epoch": 0.6732346304144422,
+      "grad_norm": 0.6669226884841919,
+      "learning_rate": 7.480203306650134e-06,
+      "loss": 0.7002,
+      "step": 12232
+    },
+    {
+      "epoch": 0.6732896692167979,
+      "grad_norm": 0.7110459208488464,
+      "learning_rate": 7.479826918573208e-06,
+      "loss": 0.8542,
+      "step": 12233
+    },
+    {
+      "epoch": 0.6733447080191535,
+      "grad_norm": 0.6632254123687744,
+      "learning_rate": 7.479450511858563e-06,
+      "loss": 0.6784,
+      "step": 12234
+    },
+    {
+      "epoch": 0.6733997468215092,
+      "grad_norm": 0.7368438839912415,
+      "learning_rate": 7.479074086509032e-06,
+      "loss": 0.7683,
+      "step": 12235
+    },
+    {
+      "epoch": 0.6734547856238648,
+      "grad_norm": 0.764905571937561,
+      "learning_rate": 7.478697642527447e-06,
+      "loss": 0.7585,
+      "step": 12236
+    },
+    {
+      "epoch": 0.6735098244262205,
+      "grad_norm": 0.7141197323799133,
+      "learning_rate": 7.478321179916632e-06,
+      "loss": 0.7409,
+      "step": 12237
+    },
+    {
+      "epoch": 0.6735648632285761,
+      "grad_norm": 0.6514197587966919,
+      "learning_rate": 7.477944698679419e-06,
+      "loss": 0.7623,
+      "step": 12238
+    },
+    {
+      "epoch": 0.6736199020309318,
+      "grad_norm": 0.7712671160697937,
+      "learning_rate": 7.477568198818636e-06,
+      "loss": 0.777,
+      "step": 12239
+    },
+    {
+      "epoch": 0.6736749408332875,
+      "grad_norm": 0.6690881252288818,
+      "learning_rate": 7.4771916803371145e-06,
+      "loss": 0.7275,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6737299796356432,
+      "grad_norm": 0.7206465601921082,
+      "learning_rate": 7.476815143237683e-06,
+      "loss": 0.853,
+      "step": 12241
+    },
+    {
+      "epoch": 0.6737850184379988,
+      "grad_norm": 0.7052504420280457,
+      "learning_rate": 7.476438587523171e-06,
+      "loss": 0.774,
+      "step": 12242
+    },
+    {
+      "epoch": 0.6738400572403545,
+      "grad_norm": 1.6168169975280762,
+      "learning_rate": 7.476062013196411e-06,
+      "loss": 0.7423,
+      "step": 12243
+    },
+    {
+      "epoch": 0.6738950960427101,
+      "grad_norm": 0.715300977230072,
+      "learning_rate": 7.475685420260232e-06,
+      "loss": 0.78,
+      "step": 12244
+    },
+    {
+      "epoch": 0.6739501348450657,
+      "grad_norm": 0.7774379253387451,
+      "learning_rate": 7.475308808717463e-06,
+      "loss": 0.885,
+      "step": 12245
+    },
+    {
+      "epoch": 0.6740051736474214,
+      "grad_norm": 0.6998060941696167,
+      "learning_rate": 7.474932178570935e-06,
+      "loss": 0.807,
+      "step": 12246
+    },
+    {
+      "epoch": 0.6740602124497771,
+      "grad_norm": 0.6710013747215271,
+      "learning_rate": 7.47455552982348e-06,
+      "loss": 0.7639,
+      "step": 12247
+    },
+    {
+      "epoch": 0.6741152512521328,
+      "grad_norm": 0.707435667514801,
+      "learning_rate": 7.474178862477929e-06,
+      "loss": 0.7914,
+      "step": 12248
+    },
+    {
+      "epoch": 0.6741702900544884,
+      "grad_norm": 0.7344105243682861,
+      "learning_rate": 7.47380217653711e-06,
+      "loss": 0.7464,
+      "step": 12249
+    },
+    {
+      "epoch": 0.674225328856844,
+      "grad_norm": 0.7157585620880127,
+      "learning_rate": 7.473425472003858e-06,
+      "loss": 0.7747,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6742803676591997,
+      "grad_norm": 0.6978434920310974,
+      "learning_rate": 7.473048748881001e-06,
+      "loss": 0.6903,
+      "step": 12251
+    },
+    {
+      "epoch": 0.6743354064615554,
+      "grad_norm": 0.6454086899757385,
+      "learning_rate": 7.472672007171372e-06,
+      "loss": 0.725,
+      "step": 12252
+    },
+    {
+      "epoch": 0.674390445263911,
+      "grad_norm": 0.6729341745376587,
+      "learning_rate": 7.4722952468778035e-06,
+      "loss": 0.7704,
+      "step": 12253
+    },
+    {
+      "epoch": 0.6744454840662667,
+      "grad_norm": 0.7995265126228333,
+      "learning_rate": 7.471918468003122e-06,
+      "loss": 0.7567,
+      "step": 12254
+    },
+    {
+      "epoch": 0.6745005228686224,
+      "grad_norm": 0.729629397392273,
+      "learning_rate": 7.471541670550165e-06,
+      "loss": 0.796,
+      "step": 12255
+    },
+    {
+      "epoch": 0.6745555616709781,
+      "grad_norm": 0.6923666000366211,
+      "learning_rate": 7.471164854521764e-06,
+      "loss": 0.6894,
+      "step": 12256
+    },
+    {
+      "epoch": 0.6746106004733337,
+      "grad_norm": 0.6485042572021484,
+      "learning_rate": 7.470788019920747e-06,
+      "loss": 0.6912,
+      "step": 12257
+    },
+    {
+      "epoch": 0.6746656392756893,
+      "grad_norm": 0.7569034099578857,
+      "learning_rate": 7.470411166749949e-06,
+      "loss": 0.8167,
+      "step": 12258
+    },
+    {
+      "epoch": 0.674720678078045,
+      "grad_norm": 0.6202835440635681,
+      "learning_rate": 7.470034295012203e-06,
+      "loss": 0.6409,
+      "step": 12259
+    },
+    {
+      "epoch": 0.6747757168804007,
+      "grad_norm": 0.6414007544517517,
+      "learning_rate": 7.4696574047103395e-06,
+      "loss": 0.7163,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6748307556827563,
+      "grad_norm": 0.7012181878089905,
+      "learning_rate": 7.469280495847193e-06,
+      "loss": 0.7682,
+      "step": 12261
+    },
+    {
+      "epoch": 0.674885794485112,
+      "grad_norm": 0.7027888298034668,
+      "learning_rate": 7.468903568425596e-06,
+      "loss": 0.7561,
+      "step": 12262
+    },
+    {
+      "epoch": 0.6749408332874677,
+      "grad_norm": 0.7282221913337708,
+      "learning_rate": 7.4685266224483785e-06,
+      "loss": 0.7552,
+      "step": 12263
+    },
+    {
+      "epoch": 0.6749958720898234,
+      "grad_norm": 0.7349117398262024,
+      "learning_rate": 7.468149657918377e-06,
+      "loss": 0.8323,
+      "step": 12264
+    },
+    {
+      "epoch": 0.675050910892179,
+      "grad_norm": 0.8992187976837158,
+      "learning_rate": 7.467772674838424e-06,
+      "loss": 0.7589,
+      "step": 12265
+    },
+    {
+      "epoch": 0.6751059496945346,
+      "grad_norm": 0.6773034930229187,
+      "learning_rate": 7.4673956732113505e-06,
+      "loss": 0.7229,
+      "step": 12266
+    },
+    {
+      "epoch": 0.6751609884968903,
+      "grad_norm": 0.6563699841499329,
+      "learning_rate": 7.467018653039992e-06,
+      "loss": 0.7526,
+      "step": 12267
+    },
+    {
+      "epoch": 0.675216027299246,
+      "grad_norm": 0.7559765577316284,
+      "learning_rate": 7.466641614327181e-06,
+      "loss": 0.708,
+      "step": 12268
+    },
+    {
+      "epoch": 0.6752710661016016,
+      "grad_norm": 0.7077820897102356,
+      "learning_rate": 7.4662645570757545e-06,
+      "loss": 0.6568,
+      "step": 12269
+    },
+    {
+      "epoch": 0.6753261049039573,
+      "grad_norm": 0.8082162141799927,
+      "learning_rate": 7.465887481288541e-06,
+      "loss": 0.8751,
+      "step": 12270
+    },
+    {
+      "epoch": 0.675381143706313,
+      "grad_norm": 0.6940243244171143,
+      "learning_rate": 7.465510386968377e-06,
+      "loss": 0.7826,
+      "step": 12271
+    },
+    {
+      "epoch": 0.6754361825086687,
+      "grad_norm": 0.6634145379066467,
+      "learning_rate": 7.465133274118099e-06,
+      "loss": 0.6816,
+      "step": 12272
+    },
+    {
+      "epoch": 0.6754912213110242,
+      "grad_norm": 0.6797559857368469,
+      "learning_rate": 7.464756142740539e-06,
+      "loss": 0.7101,
+      "step": 12273
+    },
+    {
+      "epoch": 0.6755462601133799,
+      "grad_norm": 0.7696588635444641,
+      "learning_rate": 7.464378992838531e-06,
+      "loss": 0.8114,
+      "step": 12274
+    },
+    {
+      "epoch": 0.6756012989157356,
+      "grad_norm": 0.6733334064483643,
+      "learning_rate": 7.4640018244149105e-06,
+      "loss": 0.7585,
+      "step": 12275
+    },
+    {
+      "epoch": 0.6756563377180913,
+      "grad_norm": 0.7087474465370178,
+      "learning_rate": 7.463624637472512e-06,
+      "loss": 0.6911,
+      "step": 12276
+    },
+    {
+      "epoch": 0.6757113765204469,
+      "grad_norm": 0.6944451928138733,
+      "learning_rate": 7.46324743201417e-06,
+      "loss": 0.7726,
+      "step": 12277
+    },
+    {
+      "epoch": 0.6757664153228026,
+      "grad_norm": 0.7214855551719666,
+      "learning_rate": 7.46287020804272e-06,
+      "loss": 0.7844,
+      "step": 12278
+    },
+    {
+      "epoch": 0.6758214541251583,
+      "grad_norm": 0.7106257677078247,
+      "learning_rate": 7.462492965560995e-06,
+      "loss": 0.7724,
+      "step": 12279
+    },
+    {
+      "epoch": 0.675876492927514,
+      "grad_norm": 0.7403497695922852,
+      "learning_rate": 7.462115704571833e-06,
+      "loss": 0.7558,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6759315317298695,
+      "grad_norm": 0.7157884836196899,
+      "learning_rate": 7.4617384250780685e-06,
+      "loss": 0.6681,
+      "step": 12281
+    },
+    {
+      "epoch": 0.6759865705322252,
+      "grad_norm": 0.6937661170959473,
+      "learning_rate": 7.461361127082538e-06,
+      "loss": 0.7852,
+      "step": 12282
+    },
+    {
+      "epoch": 0.6760416093345809,
+      "grad_norm": 0.7106412053108215,
+      "learning_rate": 7.4609838105880735e-06,
+      "loss": 0.7689,
+      "step": 12283
+    },
+    {
+      "epoch": 0.6760966481369366,
+      "grad_norm": 0.6860619187355042,
+      "learning_rate": 7.460606475597516e-06,
+      "loss": 0.6528,
+      "step": 12284
+    },
+    {
+      "epoch": 0.6761516869392922,
+      "grad_norm": 0.7085865139961243,
+      "learning_rate": 7.460229122113698e-06,
+      "loss": 0.7303,
+      "step": 12285
+    },
+    {
+      "epoch": 0.6762067257416479,
+      "grad_norm": 0.6648178100585938,
+      "learning_rate": 7.459851750139457e-06,
+      "loss": 0.6751,
+      "step": 12286
+    },
+    {
+      "epoch": 0.6762617645440036,
+      "grad_norm": 0.74468594789505,
+      "learning_rate": 7.459474359677629e-06,
+      "loss": 0.756,
+      "step": 12287
+    },
+    {
+      "epoch": 0.6763168033463591,
+      "grad_norm": 0.6408486366271973,
+      "learning_rate": 7.459096950731048e-06,
+      "loss": 0.7737,
+      "step": 12288
+    },
+    {
+      "epoch": 0.6763718421487148,
+      "grad_norm": 0.7204515933990479,
+      "learning_rate": 7.458719523302556e-06,
+      "loss": 0.7845,
+      "step": 12289
+    },
+    {
+      "epoch": 0.6764268809510705,
+      "grad_norm": 0.7373428344726562,
+      "learning_rate": 7.458342077394984e-06,
+      "loss": 0.7245,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6764819197534262,
+      "grad_norm": 0.701654851436615,
+      "learning_rate": 7.45796461301117e-06,
+      "loss": 0.7711,
+      "step": 12291
+    },
+    {
+      "epoch": 0.6765369585557818,
+      "grad_norm": 0.7002573013305664,
+      "learning_rate": 7.4575871301539526e-06,
+      "loss": 0.8138,
+      "step": 12292
+    },
+    {
+      "epoch": 0.6765919973581375,
+      "grad_norm": 0.7460681200027466,
+      "learning_rate": 7.45720962882617e-06,
+      "loss": 0.8012,
+      "step": 12293
+    },
+    {
+      "epoch": 0.6766470361604932,
+      "grad_norm": 0.6478421092033386,
+      "learning_rate": 7.456832109030655e-06,
+      "loss": 0.7161,
+      "step": 12294
+    },
+    {
+      "epoch": 0.6767020749628488,
+      "grad_norm": 0.7101582288742065,
+      "learning_rate": 7.456454570770248e-06,
+      "loss": 0.7348,
+      "step": 12295
+    },
+    {
+      "epoch": 0.6767571137652044,
+      "grad_norm": 0.7735113501548767,
+      "learning_rate": 7.4560770140477865e-06,
+      "loss": 0.7584,
+      "step": 12296
+    },
+    {
+      "epoch": 0.6768121525675601,
+      "grad_norm": 0.6811535358428955,
+      "learning_rate": 7.4556994388661085e-06,
+      "loss": 0.7653,
+      "step": 12297
+    },
+    {
+      "epoch": 0.6768671913699158,
+      "grad_norm": 0.7445605397224426,
+      "learning_rate": 7.455321845228051e-06,
+      "loss": 0.7661,
+      "step": 12298
+    },
+    {
+      "epoch": 0.6769222301722715,
+      "grad_norm": 0.6862059831619263,
+      "learning_rate": 7.4549442331364505e-06,
+      "loss": 0.776,
+      "step": 12299
+    },
+    {
+      "epoch": 0.6769772689746271,
+      "grad_norm": 0.7030314207077026,
+      "learning_rate": 7.4545666025941465e-06,
+      "loss": 0.7393,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6770323077769828,
+      "grad_norm": 0.6718610525131226,
+      "learning_rate": 7.454188953603978e-06,
+      "loss": 0.7375,
+      "step": 12301
+    },
+    {
+      "epoch": 0.6770873465793384,
+      "grad_norm": 0.6716088652610779,
+      "learning_rate": 7.453811286168782e-06,
+      "loss": 0.8021,
+      "step": 12302
+    },
+    {
+      "epoch": 0.6771423853816941,
+      "grad_norm": 0.8916372656822205,
+      "learning_rate": 7.453433600291395e-06,
+      "loss": 0.8274,
+      "step": 12303
+    },
+    {
+      "epoch": 0.6771974241840497,
+      "grad_norm": 0.7396363615989685,
+      "learning_rate": 7.45305589597466e-06,
+      "loss": 0.7892,
+      "step": 12304
+    },
+    {
+      "epoch": 0.6772524629864054,
+      "grad_norm": 0.8074424862861633,
+      "learning_rate": 7.452678173221413e-06,
+      "loss": 0.7586,
+      "step": 12305
+    },
+    {
+      "epoch": 0.6773075017887611,
+      "grad_norm": 0.6928194165229797,
+      "learning_rate": 7.452300432034494e-06,
+      "loss": 0.7914,
+      "step": 12306
+    },
+    {
+      "epoch": 0.6773625405911168,
+      "grad_norm": 0.7064313292503357,
+      "learning_rate": 7.451922672416739e-06,
+      "loss": 0.7948,
+      "step": 12307
+    },
+    {
+      "epoch": 0.6774175793934724,
+      "grad_norm": 0.6828622221946716,
+      "learning_rate": 7.451544894370992e-06,
+      "loss": 0.6723,
+      "step": 12308
+    },
+    {
+      "epoch": 0.677472618195828,
+      "grad_norm": 0.6794914603233337,
+      "learning_rate": 7.45116709790009e-06,
+      "loss": 0.7344,
+      "step": 12309
+    },
+    {
+      "epoch": 0.6775276569981837,
+      "grad_norm": 0.7643330097198486,
+      "learning_rate": 7.45078928300687e-06,
+      "loss": 0.7836,
+      "step": 12310
+    },
+    {
+      "epoch": 0.6775826958005394,
+      "grad_norm": 0.692569375038147,
+      "learning_rate": 7.450411449694176e-06,
+      "loss": 0.7608,
+      "step": 12311
+    },
+    {
+      "epoch": 0.677637734602895,
+      "grad_norm": 0.7718693614006042,
+      "learning_rate": 7.4500335979648455e-06,
+      "loss": 0.7131,
+      "step": 12312
+    },
+    {
+      "epoch": 0.6776927734052507,
+      "grad_norm": 0.6267405152320862,
+      "learning_rate": 7.449655727821716e-06,
+      "loss": 0.7543,
+      "step": 12313
+    },
+    {
+      "epoch": 0.6777478122076064,
+      "grad_norm": 0.8252732157707214,
+      "learning_rate": 7.4492778392676325e-06,
+      "loss": 0.8799,
+      "step": 12314
+    },
+    {
+      "epoch": 0.6778028510099621,
+      "grad_norm": 0.6310145854949951,
+      "learning_rate": 7.448899932305429e-06,
+      "loss": 0.7389,
+      "step": 12315
+    },
+    {
+      "epoch": 0.6778578898123176,
+      "grad_norm": 0.6115848422050476,
+      "learning_rate": 7.448522006937951e-06,
+      "loss": 0.6069,
+      "step": 12316
+    },
+    {
+      "epoch": 0.6779129286146733,
+      "grad_norm": 0.6809090971946716,
+      "learning_rate": 7.448144063168038e-06,
+      "loss": 0.7092,
+      "step": 12317
+    },
+    {
+      "epoch": 0.677967967417029,
+      "grad_norm": 0.7285470366477966,
+      "learning_rate": 7.447766100998529e-06,
+      "loss": 0.714,
+      "step": 12318
+    },
+    {
+      "epoch": 0.6780230062193847,
+      "grad_norm": 0.6637021899223328,
+      "learning_rate": 7.447388120432264e-06,
+      "loss": 0.7247,
+      "step": 12319
+    },
+    {
+      "epoch": 0.6780780450217403,
+      "grad_norm": 0.7735750675201416,
+      "learning_rate": 7.447010121472087e-06,
+      "loss": 0.7616,
+      "step": 12320
+    },
+    {
+      "epoch": 0.678133083824096,
+      "grad_norm": 0.7643262147903442,
+      "learning_rate": 7.446632104120836e-06,
+      "loss": 0.5863,
+      "step": 12321
+    },
+    {
+      "epoch": 0.6781881226264517,
+      "grad_norm": 0.6957301497459412,
+      "learning_rate": 7.446254068381352e-06,
+      "loss": 0.7125,
+      "step": 12322
+    },
+    {
+      "epoch": 0.6782431614288074,
+      "grad_norm": 0.6573877930641174,
+      "learning_rate": 7.445876014256479e-06,
+      "loss": 0.7115,
+      "step": 12323
+    },
+    {
+      "epoch": 0.6782982002311629,
+      "grad_norm": 0.6507790684700012,
+      "learning_rate": 7.445497941749056e-06,
+      "loss": 0.7266,
+      "step": 12324
+    },
+    {
+      "epoch": 0.6783532390335186,
+      "grad_norm": 0.8314819931983948,
+      "learning_rate": 7.4451198508619245e-06,
+      "loss": 0.6902,
+      "step": 12325
+    },
+    {
+      "epoch": 0.6784082778358743,
+      "grad_norm": 0.6907274127006531,
+      "learning_rate": 7.444741741597927e-06,
+      "loss": 0.8253,
+      "step": 12326
+    },
+    {
+      "epoch": 0.67846331663823,
+      "grad_norm": 0.7311725616455078,
+      "learning_rate": 7.444363613959904e-06,
+      "loss": 0.8641,
+      "step": 12327
+    },
+    {
+      "epoch": 0.6785183554405856,
+      "grad_norm": 0.6690121293067932,
+      "learning_rate": 7.443985467950701e-06,
+      "loss": 0.6966,
+      "step": 12328
+    },
+    {
+      "epoch": 0.6785733942429413,
+      "grad_norm": 0.6444346308708191,
+      "learning_rate": 7.443607303573155e-06,
+      "loss": 0.7848,
+      "step": 12329
+    },
+    {
+      "epoch": 0.678628433045297,
+      "grad_norm": 0.7553900480270386,
+      "learning_rate": 7.4432291208301125e-06,
+      "loss": 0.8196,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6786834718476525,
+      "grad_norm": 0.6393183469772339,
+      "learning_rate": 7.442850919724411e-06,
+      "loss": 0.7622,
+      "step": 12331
+    },
+    {
+      "epoch": 0.6787385106500082,
+      "grad_norm": 0.7045423984527588,
+      "learning_rate": 7.442472700258898e-06,
+      "loss": 0.7483,
+      "step": 12332
+    },
+    {
+      "epoch": 0.6787935494523639,
+      "grad_norm": 0.7536678314208984,
+      "learning_rate": 7.442094462436414e-06,
+      "loss": 0.815,
+      "step": 12333
+    },
+    {
+      "epoch": 0.6788485882547196,
+      "grad_norm": 0.645391047000885,
+      "learning_rate": 7.441716206259801e-06,
+      "loss": 0.7394,
+      "step": 12334
+    },
+    {
+      "epoch": 0.6789036270570752,
+      "grad_norm": 0.8870118260383606,
+      "learning_rate": 7.441337931731905e-06,
+      "loss": 0.8076,
+      "step": 12335
+    },
+    {
+      "epoch": 0.6789586658594309,
+      "grad_norm": 0.6672457456588745,
+      "learning_rate": 7.440959638855564e-06,
+      "loss": 0.7573,
+      "step": 12336
+    },
+    {
+      "epoch": 0.6790137046617866,
+      "grad_norm": 0.7104566693305969,
+      "learning_rate": 7.440581327633625e-06,
+      "loss": 0.6855,
+      "step": 12337
+    },
+    {
+      "epoch": 0.6790687434641423,
+      "grad_norm": 0.7201581001281738,
+      "learning_rate": 7.4402029980689294e-06,
+      "loss": 0.7977,
+      "step": 12338
+    },
+    {
+      "epoch": 0.6791237822664978,
+      "grad_norm": 0.6685218811035156,
+      "learning_rate": 7.43982465016432e-06,
+      "loss": 0.8114,
+      "step": 12339
+    },
+    {
+      "epoch": 0.6791788210688535,
+      "grad_norm": 0.6913738250732422,
+      "learning_rate": 7.439446283922645e-06,
+      "loss": 0.7584,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6792338598712092,
+      "grad_norm": 0.7332273721694946,
+      "learning_rate": 7.439067899346742e-06,
+      "loss": 0.7658,
+      "step": 12341
+    },
+    {
+      "epoch": 0.6792888986735649,
+      "grad_norm": 0.777909517288208,
+      "learning_rate": 7.438689496439458e-06,
+      "loss": 0.8064,
+      "step": 12342
+    },
+    {
+      "epoch": 0.6793439374759205,
+      "grad_norm": 0.7444930076599121,
+      "learning_rate": 7.438311075203636e-06,
+      "loss": 0.7896,
+      "step": 12343
+    },
+    {
+      "epoch": 0.6793989762782762,
+      "grad_norm": 0.7678806781768799,
+      "learning_rate": 7.4379326356421224e-06,
+      "loss": 0.8533,
+      "step": 12344
+    },
+    {
+      "epoch": 0.6794540150806319,
+      "grad_norm": 0.6653377413749695,
+      "learning_rate": 7.437554177757759e-06,
+      "loss": 0.7287,
+      "step": 12345
+    },
+    {
+      "epoch": 0.6795090538829875,
+      "grad_norm": 0.6270567178726196,
+      "learning_rate": 7.43717570155339e-06,
+      "loss": 0.6802,
+      "step": 12346
+    },
+    {
+      "epoch": 0.6795640926853431,
+      "grad_norm": 0.7091223001480103,
+      "learning_rate": 7.436797207031861e-06,
+      "loss": 0.7693,
+      "step": 12347
+    },
+    {
+      "epoch": 0.6796191314876988,
+      "grad_norm": 0.6583104133605957,
+      "learning_rate": 7.436418694196018e-06,
+      "loss": 0.7171,
+      "step": 12348
+    },
+    {
+      "epoch": 0.6796741702900545,
+      "grad_norm": 0.6897410750389099,
+      "learning_rate": 7.436040163048703e-06,
+      "loss": 0.7831,
+      "step": 12349
+    },
+    {
+      "epoch": 0.6797292090924102,
+      "grad_norm": 0.6506269574165344,
+      "learning_rate": 7.435661613592763e-06,
+      "loss": 0.8037,
+      "step": 12350
+    },
+    {
+      "epoch": 0.6797842478947658,
+      "grad_norm": 0.6772280931472778,
+      "learning_rate": 7.435283045831041e-06,
+      "loss": 0.8102,
+      "step": 12351
+    },
+    {
+      "epoch": 0.6798392866971215,
+      "grad_norm": 0.8470273017883301,
+      "learning_rate": 7.434904459766384e-06,
+      "loss": 0.7816,
+      "step": 12352
+    },
+    {
+      "epoch": 0.6798943254994771,
+      "grad_norm": 0.6969698071479797,
+      "learning_rate": 7.434525855401638e-06,
+      "loss": 0.6911,
+      "step": 12353
+    },
+    {
+      "epoch": 0.6799493643018328,
+      "grad_norm": 0.9969611763954163,
+      "learning_rate": 7.434147232739646e-06,
+      "loss": 0.7041,
+      "step": 12354
+    },
+    {
+      "epoch": 0.6800044031041884,
+      "grad_norm": 0.6697688698768616,
+      "learning_rate": 7.433768591783255e-06,
+      "loss": 0.6602,
+      "step": 12355
+    },
+    {
+      "epoch": 0.6800594419065441,
+      "grad_norm": 0.9857928156852722,
+      "learning_rate": 7.433389932535311e-06,
+      "loss": 0.6505,
+      "step": 12356
+    },
+    {
+      "epoch": 0.6801144807088998,
+      "grad_norm": 0.8787727355957031,
+      "learning_rate": 7.43301125499866e-06,
+      "loss": 0.7558,
+      "step": 12357
+    },
+    {
+      "epoch": 0.6801695195112555,
+      "grad_norm": 0.6035268306732178,
+      "learning_rate": 7.432632559176147e-06,
+      "loss": 0.6337,
+      "step": 12358
+    },
+    {
+      "epoch": 0.6802245583136111,
+      "grad_norm": 0.7977258563041687,
+      "learning_rate": 7.432253845070621e-06,
+      "loss": 0.7324,
+      "step": 12359
+    },
+    {
+      "epoch": 0.6802795971159667,
+      "grad_norm": 0.5842836499214172,
+      "learning_rate": 7.431875112684923e-06,
+      "loss": 0.677,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6803346359183224,
+      "grad_norm": 0.7134125828742981,
+      "learning_rate": 7.431496362021905e-06,
+      "loss": 0.7034,
+      "step": 12361
+    },
+    {
+      "epoch": 0.6803896747206781,
+      "grad_norm": 0.7101823091506958,
+      "learning_rate": 7.431117593084411e-06,
+      "loss": 0.7526,
+      "step": 12362
+    },
+    {
+      "epoch": 0.6804447135230337,
+      "grad_norm": 0.6543304920196533,
+      "learning_rate": 7.4307388058752865e-06,
+      "loss": 0.7548,
+      "step": 12363
+    },
+    {
+      "epoch": 0.6804997523253894,
+      "grad_norm": 0.6522945761680603,
+      "learning_rate": 7.430360000397381e-06,
+      "loss": 0.7044,
+      "step": 12364
+    },
+    {
+      "epoch": 0.6805547911277451,
+      "grad_norm": 0.7405091524124146,
+      "learning_rate": 7.429981176653539e-06,
+      "loss": 0.8064,
+      "step": 12365
+    },
+    {
+      "epoch": 0.6806098299301008,
+      "grad_norm": 0.6454355716705322,
+      "learning_rate": 7.429602334646611e-06,
+      "loss": 0.7179,
+      "step": 12366
+    },
+    {
+      "epoch": 0.6806648687324564,
+      "grad_norm": 0.8131621479988098,
+      "learning_rate": 7.429223474379439e-06,
+      "loss": 0.7144,
+      "step": 12367
+    },
+    {
+      "epoch": 0.680719907534812,
+      "grad_norm": 0.7203080058097839,
+      "learning_rate": 7.428844595854876e-06,
+      "loss": 0.8189,
+      "step": 12368
+    },
+    {
+      "epoch": 0.6807749463371677,
+      "grad_norm": 0.650414228439331,
+      "learning_rate": 7.428465699075767e-06,
+      "loss": 0.7815,
+      "step": 12369
+    },
+    {
+      "epoch": 0.6808299851395234,
+      "grad_norm": 0.8152775168418884,
+      "learning_rate": 7.42808678404496e-06,
+      "loss": 0.7365,
+      "step": 12370
+    },
+    {
+      "epoch": 0.680885023941879,
+      "grad_norm": 0.5871601700782776,
+      "learning_rate": 7.427707850765302e-06,
+      "loss": 0.6804,
+      "step": 12371
+    },
+    {
+      "epoch": 0.6809400627442347,
+      "grad_norm": 0.7115684747695923,
+      "learning_rate": 7.427328899239643e-06,
+      "loss": 0.728,
+      "step": 12372
+    },
+    {
+      "epoch": 0.6809951015465904,
+      "grad_norm": 0.6575615406036377,
+      "learning_rate": 7.426949929470828e-06,
+      "loss": 0.725,
+      "step": 12373
+    },
+    {
+      "epoch": 0.681050140348946,
+      "grad_norm": 0.7744095325469971,
+      "learning_rate": 7.426570941461708e-06,
+      "loss": 0.7647,
+      "step": 12374
+    },
+    {
+      "epoch": 0.6811051791513016,
+      "grad_norm": 0.6856220364570618,
+      "learning_rate": 7.4261919352151305e-06,
+      "loss": 0.8121,
+      "step": 12375
+    },
+    {
+      "epoch": 0.6811602179536573,
+      "grad_norm": 0.8197830319404602,
+      "learning_rate": 7.425812910733943e-06,
+      "loss": 0.8685,
+      "step": 12376
+    },
+    {
+      "epoch": 0.681215256756013,
+      "grad_norm": 1.240628719329834,
+      "learning_rate": 7.425433868020996e-06,
+      "loss": 0.8063,
+      "step": 12377
+    },
+    {
+      "epoch": 0.6812702955583686,
+      "grad_norm": 0.8716747760772705,
+      "learning_rate": 7.425054807079136e-06,
+      "loss": 0.7384,
+      "step": 12378
+    },
+    {
+      "epoch": 0.6813253343607243,
+      "grad_norm": 0.7512598037719727,
+      "learning_rate": 7.4246757279112135e-06,
+      "loss": 0.7428,
+      "step": 12379
+    },
+    {
+      "epoch": 0.68138037316308,
+      "grad_norm": 0.7002312541007996,
+      "learning_rate": 7.424296630520078e-06,
+      "loss": 0.6066,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6814354119654357,
+      "grad_norm": 0.6422720551490784,
+      "learning_rate": 7.423917514908578e-06,
+      "loss": 0.6645,
+      "step": 12381
+    },
+    {
+      "epoch": 0.6814904507677912,
+      "grad_norm": 0.8667505383491516,
+      "learning_rate": 7.423538381079562e-06,
+      "loss": 0.8663,
+      "step": 12382
+    },
+    {
+      "epoch": 0.6815454895701469,
+      "grad_norm": 0.7045377492904663,
+      "learning_rate": 7.423159229035881e-06,
+      "loss": 0.7684,
+      "step": 12383
+    },
+    {
+      "epoch": 0.6816005283725026,
+      "grad_norm": 0.7663894295692444,
+      "learning_rate": 7.422780058780385e-06,
+      "loss": 0.8051,
+      "step": 12384
+    },
+    {
+      "epoch": 0.6816555671748583,
+      "grad_norm": 0.7612582445144653,
+      "learning_rate": 7.42240087031592e-06,
+      "loss": 0.7771,
+      "step": 12385
+    },
+    {
+      "epoch": 0.6817106059772139,
+      "grad_norm": 0.8682271838188171,
+      "learning_rate": 7.42202166364534e-06,
+      "loss": 0.7761,
+      "step": 12386
+    },
+    {
+      "epoch": 0.6817656447795696,
+      "grad_norm": 0.712204098701477,
+      "learning_rate": 7.421642438771492e-06,
+      "loss": 0.7832,
+      "step": 12387
+    },
+    {
+      "epoch": 0.6818206835819253,
+      "grad_norm": 0.6726338863372803,
+      "learning_rate": 7.42126319569723e-06,
+      "loss": 0.7541,
+      "step": 12388
+    },
+    {
+      "epoch": 0.681875722384281,
+      "grad_norm": 0.647570788860321,
+      "learning_rate": 7.420883934425401e-06,
+      "loss": 0.7281,
+      "step": 12389
+    },
+    {
+      "epoch": 0.6819307611866365,
+      "grad_norm": 0.7058577537536621,
+      "learning_rate": 7.420504654958857e-06,
+      "loss": 0.8315,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6819857999889922,
+      "grad_norm": 0.6683655977249146,
+      "learning_rate": 7.420125357300446e-06,
+      "loss": 0.772,
+      "step": 12391
+    },
+    {
+      "epoch": 0.6820408387913479,
+      "grad_norm": 0.6768681406974792,
+      "learning_rate": 7.419746041453022e-06,
+      "loss": 0.7023,
+      "step": 12392
+    },
+    {
+      "epoch": 0.6820958775937036,
+      "grad_norm": 0.8037514686584473,
+      "learning_rate": 7.419366707419434e-06,
+      "loss": 0.6894,
+      "step": 12393
+    },
+    {
+      "epoch": 0.6821509163960592,
+      "grad_norm": 0.6510934829711914,
+      "learning_rate": 7.418987355202534e-06,
+      "loss": 0.6411,
+      "step": 12394
+    },
+    {
+      "epoch": 0.6822059551984149,
+      "grad_norm": 0.7628617882728577,
+      "learning_rate": 7.418607984805173e-06,
+      "loss": 0.7681,
+      "step": 12395
+    },
+    {
+      "epoch": 0.6822609940007706,
+      "grad_norm": 0.7146260738372803,
+      "learning_rate": 7.418228596230201e-06,
+      "loss": 0.7003,
+      "step": 12396
+    },
+    {
+      "epoch": 0.6823160328031262,
+      "grad_norm": 0.6208338737487793,
+      "learning_rate": 7.41784918948047e-06,
+      "loss": 0.7138,
+      "step": 12397
+    },
+    {
+      "epoch": 0.6823710716054818,
+      "grad_norm": 0.7859066724777222,
+      "learning_rate": 7.417469764558832e-06,
+      "loss": 0.7984,
+      "step": 12398
+    },
+    {
+      "epoch": 0.6824261104078375,
+      "grad_norm": 0.7636224031448364,
+      "learning_rate": 7.417090321468138e-06,
+      "loss": 0.7445,
+      "step": 12399
+    },
+    {
+      "epoch": 0.6824811492101932,
+      "grad_norm": 0.9071671366691589,
+      "learning_rate": 7.41671086021124e-06,
+      "loss": 0.8058,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6825361880125489,
+      "grad_norm": 0.5986278057098389,
+      "learning_rate": 7.416331380790991e-06,
+      "loss": 0.7001,
+      "step": 12401
+    },
+    {
+      "epoch": 0.6825912268149045,
+      "grad_norm": 0.6812893152236938,
+      "learning_rate": 7.415951883210242e-06,
+      "loss": 0.7745,
+      "step": 12402
+    },
+    {
+      "epoch": 0.6826462656172602,
+      "grad_norm": 0.666362464427948,
+      "learning_rate": 7.415572367471844e-06,
+      "loss": 0.7861,
+      "step": 12403
+    },
+    {
+      "epoch": 0.6827013044196159,
+      "grad_norm": 0.6963029503822327,
+      "learning_rate": 7.415192833578653e-06,
+      "loss": 0.7657,
+      "step": 12404
+    },
+    {
+      "epoch": 0.6827563432219715,
+      "grad_norm": 0.669876217842102,
+      "learning_rate": 7.414813281533517e-06,
+      "loss": 0.6441,
+      "step": 12405
+    },
+    {
+      "epoch": 0.6828113820243271,
+      "grad_norm": 0.6608602404594421,
+      "learning_rate": 7.414433711339293e-06,
+      "loss": 0.7203,
+      "step": 12406
+    },
+    {
+      "epoch": 0.6828664208266828,
+      "grad_norm": 0.7262642979621887,
+      "learning_rate": 7.41405412299883e-06,
+      "loss": 0.7842,
+      "step": 12407
+    },
+    {
+      "epoch": 0.6829214596290385,
+      "grad_norm": 0.7728527188301086,
+      "learning_rate": 7.413674516514983e-06,
+      "loss": 0.7551,
+      "step": 12408
+    },
+    {
+      "epoch": 0.6829764984313942,
+      "grad_norm": 0.7970840930938721,
+      "learning_rate": 7.4132948918906035e-06,
+      "loss": 0.8181,
+      "step": 12409
+    },
+    {
+      "epoch": 0.6830315372337498,
+      "grad_norm": 0.6672868728637695,
+      "learning_rate": 7.412915249128546e-06,
+      "loss": 0.7201,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6830865760361055,
+      "grad_norm": 0.8261075019836426,
+      "learning_rate": 7.412535588231664e-06,
+      "loss": 0.6006,
+      "step": 12411
+    },
+    {
+      "epoch": 0.6831416148384611,
+      "grad_norm": 0.6768019795417786,
+      "learning_rate": 7.412155909202809e-06,
+      "loss": 0.7326,
+      "step": 12412
+    },
+    {
+      "epoch": 0.6831966536408168,
+      "grad_norm": 0.7482851147651672,
+      "learning_rate": 7.4117762120448364e-06,
+      "loss": 0.7913,
+      "step": 12413
+    },
+    {
+      "epoch": 0.6832516924431724,
+      "grad_norm": 0.7315956354141235,
+      "learning_rate": 7.411396496760601e-06,
+      "loss": 0.7949,
+      "step": 12414
+    },
+    {
+      "epoch": 0.6833067312455281,
+      "grad_norm": 0.7460561394691467,
+      "learning_rate": 7.411016763352954e-06,
+      "loss": 0.8445,
+      "step": 12415
+    },
+    {
+      "epoch": 0.6833617700478838,
+      "grad_norm": 0.7025588154792786,
+      "learning_rate": 7.410637011824749e-06,
+      "loss": 0.7658,
+      "step": 12416
+    },
+    {
+      "epoch": 0.6834168088502394,
+      "grad_norm": 0.7507885694503784,
+      "learning_rate": 7.410257242178842e-06,
+      "loss": 0.711,
+      "step": 12417
+    },
+    {
+      "epoch": 0.683471847652595,
+      "grad_norm": 0.6935780048370361,
+      "learning_rate": 7.409877454418088e-06,
+      "loss": 0.8376,
+      "step": 12418
+    },
+    {
+      "epoch": 0.6835268864549507,
+      "grad_norm": 0.7747789025306702,
+      "learning_rate": 7.409497648545341e-06,
+      "loss": 0.8173,
+      "step": 12419
+    },
+    {
+      "epoch": 0.6835819252573064,
+      "grad_norm": 0.6559001803398132,
+      "learning_rate": 7.4091178245634525e-06,
+      "loss": 0.7146,
+      "step": 12420
+    },
+    {
+      "epoch": 0.683636964059662,
+      "grad_norm": 0.7123926877975464,
+      "learning_rate": 7.408737982475279e-06,
+      "loss": 0.7544,
+      "step": 12421
+    },
+    {
+      "epoch": 0.6836920028620177,
+      "grad_norm": 0.8163334131240845,
+      "learning_rate": 7.408358122283678e-06,
+      "loss": 0.8008,
+      "step": 12422
+    },
+    {
+      "epoch": 0.6837470416643734,
+      "grad_norm": 0.6837686896324158,
+      "learning_rate": 7.4079782439915e-06,
+      "loss": 0.6595,
+      "step": 12423
+    },
+    {
+      "epoch": 0.6838020804667291,
+      "grad_norm": 0.9385979175567627,
+      "learning_rate": 7.407598347601601e-06,
+      "loss": 0.8135,
+      "step": 12424
+    },
+    {
+      "epoch": 0.6838571192690847,
+      "grad_norm": 0.7197830677032471,
+      "learning_rate": 7.407218433116839e-06,
+      "loss": 0.8401,
+      "step": 12425
+    },
+    {
+      "epoch": 0.6839121580714403,
+      "grad_norm": 0.7165716290473938,
+      "learning_rate": 7.406838500540069e-06,
+      "loss": 0.7864,
+      "step": 12426
+    },
+    {
+      "epoch": 0.683967196873796,
+      "grad_norm": 0.6844950318336487,
+      "learning_rate": 7.4064585498741435e-06,
+      "loss": 0.7409,
+      "step": 12427
+    },
+    {
+      "epoch": 0.6840222356761517,
+      "grad_norm": 0.6237946152687073,
+      "learning_rate": 7.40607858112192e-06,
+      "loss": 0.6915,
+      "step": 12428
+    },
+    {
+      "epoch": 0.6840772744785073,
+      "grad_norm": 0.7437137365341187,
+      "learning_rate": 7.405698594286252e-06,
+      "loss": 0.8191,
+      "step": 12429
+    },
+    {
+      "epoch": 0.684132313280863,
+      "grad_norm": 0.6956225633621216,
+      "learning_rate": 7.4053185893700006e-06,
+      "loss": 0.7662,
+      "step": 12430
+    },
+    {
+      "epoch": 0.6841873520832187,
+      "grad_norm": 0.6508380174636841,
+      "learning_rate": 7.404938566376018e-06,
+      "loss": 0.7758,
+      "step": 12431
+    },
+    {
+      "epoch": 0.6842423908855744,
+      "grad_norm": 0.6759025454521179,
+      "learning_rate": 7.404558525307159e-06,
+      "loss": 0.7713,
+      "step": 12432
+    },
+    {
+      "epoch": 0.68429742968793,
+      "grad_norm": 0.7280172109603882,
+      "learning_rate": 7.404178466166283e-06,
+      "loss": 0.7753,
+      "step": 12433
+    },
+    {
+      "epoch": 0.6843524684902856,
+      "grad_norm": 0.7599073052406311,
+      "learning_rate": 7.403798388956245e-06,
+      "loss": 0.6993,
+      "step": 12434
+    },
+    {
+      "epoch": 0.6844075072926413,
+      "grad_norm": 0.7962353229522705,
+      "learning_rate": 7.403418293679903e-06,
+      "loss": 0.771,
+      "step": 12435
+    },
+    {
+      "epoch": 0.684462546094997,
+      "grad_norm": 0.6714458465576172,
+      "learning_rate": 7.40303818034011e-06,
+      "loss": 0.7077,
+      "step": 12436
+    },
+    {
+      "epoch": 0.6845175848973526,
+      "grad_norm": 0.6770713925361633,
+      "learning_rate": 7.402658048939726e-06,
+      "loss": 0.7695,
+      "step": 12437
+    },
+    {
+      "epoch": 0.6845726236997083,
+      "grad_norm": 0.7337867617607117,
+      "learning_rate": 7.402277899481608e-06,
+      "loss": 0.9453,
+      "step": 12438
+    },
+    {
+      "epoch": 0.684627662502064,
+      "grad_norm": 0.7457698583602905,
+      "learning_rate": 7.401897731968612e-06,
+      "loss": 0.7569,
+      "step": 12439
+    },
+    {
+      "epoch": 0.6846827013044197,
+      "grad_norm": 0.6683285236358643,
+      "learning_rate": 7.401517546403595e-06,
+      "loss": 0.7215,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6847377401067752,
+      "grad_norm": 0.6516628861427307,
+      "learning_rate": 7.401137342789415e-06,
+      "loss": 0.7433,
+      "step": 12441
+    },
+    {
+      "epoch": 0.6847927789091309,
+      "grad_norm": 0.7572295665740967,
+      "learning_rate": 7.400757121128932e-06,
+      "loss": 0.7204,
+      "step": 12442
+    },
+    {
+      "epoch": 0.6848478177114866,
+      "grad_norm": 0.6884106993675232,
+      "learning_rate": 7.400376881425e-06,
+      "loss": 0.6766,
+      "step": 12443
+    },
+    {
+      "epoch": 0.6849028565138423,
+      "grad_norm": 0.798926591873169,
+      "learning_rate": 7.399996623680475e-06,
+      "loss": 0.7673,
+      "step": 12444
+    },
+    {
+      "epoch": 0.6849578953161979,
+      "grad_norm": 0.7200846672058105,
+      "learning_rate": 7.399616347898221e-06,
+      "loss": 0.8032,
+      "step": 12445
+    },
+    {
+      "epoch": 0.6850129341185536,
+      "grad_norm": 0.7085461020469666,
+      "learning_rate": 7.3992360540810915e-06,
+      "loss": 0.7075,
+      "step": 12446
+    },
+    {
+      "epoch": 0.6850679729209093,
+      "grad_norm": 0.6885339021682739,
+      "learning_rate": 7.398855742231947e-06,
+      "loss": 0.7278,
+      "step": 12447
+    },
+    {
+      "epoch": 0.685123011723265,
+      "grad_norm": 0.6693943738937378,
+      "learning_rate": 7.398475412353643e-06,
+      "loss": 0.7134,
+      "step": 12448
+    },
+    {
+      "epoch": 0.6851780505256205,
+      "grad_norm": 0.6908173561096191,
+      "learning_rate": 7.398095064449041e-06,
+      "loss": 0.8054,
+      "step": 12449
+    },
+    {
+      "epoch": 0.6852330893279762,
+      "grad_norm": 0.6207892894744873,
+      "learning_rate": 7.397714698520999e-06,
+      "loss": 0.5789,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6852881281303319,
+      "grad_norm": 0.8367832899093628,
+      "learning_rate": 7.397334314572374e-06,
+      "loss": 0.8186,
+      "step": 12451
+    },
+    {
+      "epoch": 0.6853431669326876,
+      "grad_norm": 0.7005738615989685,
+      "learning_rate": 7.396953912606026e-06,
+      "loss": 0.8177,
+      "step": 12452
+    },
+    {
+      "epoch": 0.6853982057350432,
+      "grad_norm": 0.7189906239509583,
+      "learning_rate": 7.396573492624814e-06,
+      "loss": 0.8387,
+      "step": 12453
+    },
+    {
+      "epoch": 0.6854532445373989,
+      "grad_norm": 1.040576457977295,
+      "learning_rate": 7.3961930546315995e-06,
+      "loss": 0.7165,
+      "step": 12454
+    },
+    {
+      "epoch": 0.6855082833397546,
+      "grad_norm": 0.6417170166969299,
+      "learning_rate": 7.3958125986292385e-06,
+      "loss": 0.6671,
+      "step": 12455
+    },
+    {
+      "epoch": 0.6855633221421102,
+      "grad_norm": 0.6443242430686951,
+      "learning_rate": 7.395432124620589e-06,
+      "loss": 0.6995,
+      "step": 12456
+    },
+    {
+      "epoch": 0.6856183609444658,
+      "grad_norm": 0.5764951705932617,
+      "learning_rate": 7.395051632608516e-06,
+      "loss": 0.6088,
+      "step": 12457
+    },
+    {
+      "epoch": 0.6856733997468215,
+      "grad_norm": 0.6193686127662659,
+      "learning_rate": 7.394671122595873e-06,
+      "loss": 0.7283,
+      "step": 12458
+    },
+    {
+      "epoch": 0.6857284385491772,
+      "grad_norm": 0.6773817539215088,
+      "learning_rate": 7.394290594585525e-06,
+      "loss": 0.8204,
+      "step": 12459
+    },
+    {
+      "epoch": 0.6857834773515328,
+      "grad_norm": 0.7906570434570312,
+      "learning_rate": 7.393910048580328e-06,
+      "loss": 0.7057,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6858385161538885,
+      "grad_norm": 0.7544124126434326,
+      "learning_rate": 7.393529484583145e-06,
+      "loss": 0.8053,
+      "step": 12461
+    },
+    {
+      "epoch": 0.6858935549562442,
+      "grad_norm": 0.6878008842468262,
+      "learning_rate": 7.3931489025968365e-06,
+      "loss": 0.6972,
+      "step": 12462
+    },
+    {
+      "epoch": 0.6859485937585998,
+      "grad_norm": 0.6734861731529236,
+      "learning_rate": 7.392768302624259e-06,
+      "loss": 0.7921,
+      "step": 12463
+    },
+    {
+      "epoch": 0.6860036325609554,
+      "grad_norm": 0.6845618486404419,
+      "learning_rate": 7.392387684668276e-06,
+      "loss": 0.7461,
+      "step": 12464
+    },
+    {
+      "epoch": 0.6860586713633111,
+      "grad_norm": 0.6362663507461548,
+      "learning_rate": 7.392007048731748e-06,
+      "loss": 0.7108,
+      "step": 12465
+    },
+    {
+      "epoch": 0.6861137101656668,
+      "grad_norm": 0.7441046237945557,
+      "learning_rate": 7.391626394817537e-06,
+      "loss": 0.6944,
+      "step": 12466
+    },
+    {
+      "epoch": 0.6861687489680225,
+      "grad_norm": 1.0933935642242432,
+      "learning_rate": 7.391245722928501e-06,
+      "loss": 0.7744,
+      "step": 12467
+    },
+    {
+      "epoch": 0.6862237877703781,
+      "grad_norm": 0.6531348824501038,
+      "learning_rate": 7.3908650330675e-06,
+      "loss": 0.6772,
+      "step": 12468
+    },
+    {
+      "epoch": 0.6862788265727338,
+      "grad_norm": 0.7533715963363647,
+      "learning_rate": 7.390484325237399e-06,
+      "loss": 0.7385,
+      "step": 12469
+    },
+    {
+      "epoch": 0.6863338653750894,
+      "grad_norm": 0.618679940700531,
+      "learning_rate": 7.390103599441058e-06,
+      "loss": 0.6053,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6863889041774451,
+      "grad_norm": 0.7102347612380981,
+      "learning_rate": 7.389722855681338e-06,
+      "loss": 0.7246,
+      "step": 12471
+    },
+    {
+      "epoch": 0.6864439429798007,
+      "grad_norm": 0.8545061945915222,
+      "learning_rate": 7.3893420939611e-06,
+      "loss": 0.7386,
+      "step": 12472
+    },
+    {
+      "epoch": 0.6864989817821564,
+      "grad_norm": 0.6298168897628784,
+      "learning_rate": 7.388961314283207e-06,
+      "loss": 0.6573,
+      "step": 12473
+    },
+    {
+      "epoch": 0.6865540205845121,
+      "grad_norm": 0.6909272074699402,
+      "learning_rate": 7.388580516650521e-06,
+      "loss": 0.7973,
+      "step": 12474
+    },
+    {
+      "epoch": 0.6866090593868678,
+      "grad_norm": 0.6782366037368774,
+      "learning_rate": 7.388199701065904e-06,
+      "loss": 0.7437,
+      "step": 12475
+    },
+    {
+      "epoch": 0.6866640981892234,
+      "grad_norm": 0.6826187372207642,
+      "learning_rate": 7.387818867532213e-06,
+      "loss": 0.6254,
+      "step": 12476
+    },
+    {
+      "epoch": 0.686719136991579,
+      "grad_norm": 0.7471422553062439,
+      "learning_rate": 7.387438016052318e-06,
+      "loss": 0.8668,
+      "step": 12477
+    },
+    {
+      "epoch": 0.6867741757939347,
+      "grad_norm": 0.7987646460533142,
+      "learning_rate": 7.38705714662908e-06,
+      "loss": 0.6759,
+      "step": 12478
+    },
+    {
+      "epoch": 0.6868292145962904,
+      "grad_norm": 0.7318877577781677,
+      "learning_rate": 7.386676259265356e-06,
+      "loss": 0.7167,
+      "step": 12479
+    },
+    {
+      "epoch": 0.686884253398646,
+      "grad_norm": 0.6655439138412476,
+      "learning_rate": 7.386295353964013e-06,
+      "loss": 0.7184,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6869392922010017,
+      "grad_norm": 0.7323878407478333,
+      "learning_rate": 7.385914430727912e-06,
+      "loss": 0.7562,
+      "step": 12481
+    },
+    {
+      "epoch": 0.6869943310033574,
+      "grad_norm": 0.7813006639480591,
+      "learning_rate": 7.385533489559918e-06,
+      "loss": 0.7665,
+      "step": 12482
+    },
+    {
+      "epoch": 0.6870493698057131,
+      "grad_norm": 0.6889718770980835,
+      "learning_rate": 7.385152530462894e-06,
+      "loss": 0.6587,
+      "step": 12483
+    },
+    {
+      "epoch": 0.6871044086080687,
+      "grad_norm": 0.6930332183837891,
+      "learning_rate": 7.384771553439698e-06,
+      "loss": 0.8244,
+      "step": 12484
+    },
+    {
+      "epoch": 0.6871594474104243,
+      "grad_norm": 0.8294679522514343,
+      "learning_rate": 7.384390558493201e-06,
+      "loss": 0.6977,
+      "step": 12485
+    },
+    {
+      "epoch": 0.68721448621278,
+      "grad_norm": 0.7235204577445984,
+      "learning_rate": 7.384009545626262e-06,
+      "loss": 0.7946,
+      "step": 12486
+    },
+    {
+      "epoch": 0.6872695250151357,
+      "grad_norm": 0.6346727609634399,
+      "learning_rate": 7.3836285148417456e-06,
+      "loss": 0.6109,
+      "step": 12487
+    },
+    {
+      "epoch": 0.6873245638174913,
+      "grad_norm": 0.7168872356414795,
+      "learning_rate": 7.383247466142513e-06,
+      "loss": 0.7485,
+      "step": 12488
+    },
+    {
+      "epoch": 0.687379602619847,
+      "grad_norm": 0.6511938571929932,
+      "learning_rate": 7.382866399531434e-06,
+      "loss": 0.8048,
+      "step": 12489
+    },
+    {
+      "epoch": 0.6874346414222027,
+      "grad_norm": 0.7569704651832581,
+      "learning_rate": 7.3824853150113674e-06,
+      "loss": 0.8017,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6874896802245584,
+      "grad_norm": 0.7708210945129395,
+      "learning_rate": 7.382104212585178e-06,
+      "loss": 0.7258,
+      "step": 12491
+    },
+    {
+      "epoch": 0.6875447190269139,
+      "grad_norm": 0.709702730178833,
+      "learning_rate": 7.381723092255731e-06,
+      "loss": 0.7707,
+      "step": 12492
+    },
+    {
+      "epoch": 0.6875997578292696,
+      "grad_norm": 0.6683183908462524,
+      "learning_rate": 7.381341954025892e-06,
+      "loss": 0.702,
+      "step": 12493
+    },
+    {
+      "epoch": 0.6876547966316253,
+      "grad_norm": 0.7639274597167969,
+      "learning_rate": 7.380960797898524e-06,
+      "loss": 0.7027,
+      "step": 12494
+    },
+    {
+      "epoch": 0.687709835433981,
+      "grad_norm": 0.6735698580741882,
+      "learning_rate": 7.380579623876492e-06,
+      "loss": 0.7124,
+      "step": 12495
+    },
+    {
+      "epoch": 0.6877648742363366,
+      "grad_norm": 0.6635340452194214,
+      "learning_rate": 7.38019843196266e-06,
+      "loss": 0.6968,
+      "step": 12496
+    },
+    {
+      "epoch": 0.6878199130386923,
+      "grad_norm": 0.7459729313850403,
+      "learning_rate": 7.379817222159895e-06,
+      "loss": 0.7629,
+      "step": 12497
+    },
+    {
+      "epoch": 0.687874951841048,
+      "grad_norm": 0.7408778667449951,
+      "learning_rate": 7.37943599447106e-06,
+      "loss": 0.8327,
+      "step": 12498
+    },
+    {
+      "epoch": 0.6879299906434037,
+      "grad_norm": 0.659736156463623,
+      "learning_rate": 7.379054748899021e-06,
+      "loss": 0.6746,
+      "step": 12499
+    },
+    {
+      "epoch": 0.6879850294457592,
+      "grad_norm": 0.7429264783859253,
+      "learning_rate": 7.3786734854466435e-06,
+      "loss": 0.8555,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6880400682481149,
+      "grad_norm": 0.7492697834968567,
+      "learning_rate": 7.378292204116793e-06,
+      "loss": 0.7825,
+      "step": 12501
+    },
+    {
+      "epoch": 0.6880951070504706,
+      "grad_norm": 0.6664871573448181,
+      "learning_rate": 7.377910904912336e-06,
+      "loss": 0.7343,
+      "step": 12502
+    },
+    {
+      "epoch": 0.6881501458528262,
+      "grad_norm": 0.8010555505752563,
+      "learning_rate": 7.377529587836135e-06,
+      "loss": 0.6789,
+      "step": 12503
+    },
+    {
+      "epoch": 0.6882051846551819,
+      "grad_norm": 0.6339166164398193,
+      "learning_rate": 7.3771482528910585e-06,
+      "loss": 0.7471,
+      "step": 12504
+    },
+    {
+      "epoch": 0.6882602234575376,
+      "grad_norm": 0.6750906109809875,
+      "learning_rate": 7.376766900079973e-06,
+      "loss": 0.665,
+      "step": 12505
+    },
+    {
+      "epoch": 0.6883152622598933,
+      "grad_norm": 0.6440090537071228,
+      "learning_rate": 7.376385529405743e-06,
+      "loss": 0.6804,
+      "step": 12506
+    },
+    {
+      "epoch": 0.6883703010622488,
+      "grad_norm": 0.7159061431884766,
+      "learning_rate": 7.376004140871236e-06,
+      "loss": 0.7524,
+      "step": 12507
+    },
+    {
+      "epoch": 0.6884253398646045,
+      "grad_norm": 0.7551491260528564,
+      "learning_rate": 7.375622734479316e-06,
+      "loss": 0.891,
+      "step": 12508
+    },
+    {
+      "epoch": 0.6884803786669602,
+      "grad_norm": 0.6584289073944092,
+      "learning_rate": 7.375241310232854e-06,
+      "loss": 0.7313,
+      "step": 12509
+    },
+    {
+      "epoch": 0.6885354174693159,
+      "grad_norm": 0.7616147398948669,
+      "learning_rate": 7.374859868134713e-06,
+      "loss": 0.8351,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6885904562716715,
+      "grad_norm": 0.669541597366333,
+      "learning_rate": 7.374478408187761e-06,
+      "loss": 0.6836,
+      "step": 12511
+    },
+    {
+      "epoch": 0.6886454950740272,
+      "grad_norm": 0.6483158469200134,
+      "learning_rate": 7.374096930394864e-06,
+      "loss": 0.6909,
+      "step": 12512
+    },
+    {
+      "epoch": 0.6887005338763829,
+      "grad_norm": 0.7079604864120483,
+      "learning_rate": 7.3737154347588925e-06,
+      "loss": 0.7151,
+      "step": 12513
+    },
+    {
+      "epoch": 0.6887555726787385,
+      "grad_norm": 0.6805073618888855,
+      "learning_rate": 7.373333921282709e-06,
+      "loss": 0.7761,
+      "step": 12514
+    },
+    {
+      "epoch": 0.6888106114810941,
+      "grad_norm": 0.757008969783783,
+      "learning_rate": 7.372952389969183e-06,
+      "loss": 0.7249,
+      "step": 12515
+    },
+    {
+      "epoch": 0.6888656502834498,
+      "grad_norm": 0.6990587711334229,
+      "learning_rate": 7.372570840821183e-06,
+      "loss": 0.7463,
+      "step": 12516
+    },
+    {
+      "epoch": 0.6889206890858055,
+      "grad_norm": 0.7405683398246765,
+      "learning_rate": 7.3721892738415745e-06,
+      "loss": 0.8039,
+      "step": 12517
+    },
+    {
+      "epoch": 0.6889757278881612,
+      "grad_norm": 0.6736571192741394,
+      "learning_rate": 7.371807689033228e-06,
+      "loss": 0.7084,
+      "step": 12518
+    },
+    {
+      "epoch": 0.6890307666905168,
+      "grad_norm": 0.752955436706543,
+      "learning_rate": 7.3714260863990095e-06,
+      "loss": 0.7951,
+      "step": 12519
+    },
+    {
+      "epoch": 0.6890858054928725,
+      "grad_norm": 0.6810917258262634,
+      "learning_rate": 7.3710444659417855e-06,
+      "loss": 0.7884,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6891408442952281,
+      "grad_norm": 0.727500855922699,
+      "learning_rate": 7.370662827664427e-06,
+      "loss": 0.7617,
+      "step": 12521
+    },
+    {
+      "epoch": 0.6891958830975838,
+      "grad_norm": 0.6739845871925354,
+      "learning_rate": 7.3702811715698016e-06,
+      "loss": 0.6831,
+      "step": 12522
+    },
+    {
+      "epoch": 0.6892509218999394,
+      "grad_norm": 0.850913941860199,
+      "learning_rate": 7.369899497660779e-06,
+      "loss": 0.7658,
+      "step": 12523
+    },
+    {
+      "epoch": 0.6893059607022951,
+      "grad_norm": 0.7352884411811829,
+      "learning_rate": 7.369517805940223e-06,
+      "loss": 0.7748,
+      "step": 12524
+    },
+    {
+      "epoch": 0.6893609995046508,
+      "grad_norm": 0.6702300310134888,
+      "learning_rate": 7.369136096411008e-06,
+      "loss": 0.7557,
+      "step": 12525
+    },
+    {
+      "epoch": 0.6894160383070065,
+      "grad_norm": 0.7117186784744263,
+      "learning_rate": 7.368754369075999e-06,
+      "loss": 0.8147,
+      "step": 12526
+    },
+    {
+      "epoch": 0.6894710771093621,
+      "grad_norm": 0.6896687746047974,
+      "learning_rate": 7.368372623938067e-06,
+      "loss": 0.7753,
+      "step": 12527
+    },
+    {
+      "epoch": 0.6895261159117178,
+      "grad_norm": 0.669207751750946,
+      "learning_rate": 7.367990861000078e-06,
+      "loss": 0.739,
+      "step": 12528
+    },
+    {
+      "epoch": 0.6895811547140734,
+      "grad_norm": 0.7014279961585999,
+      "learning_rate": 7.367609080264906e-06,
+      "loss": 0.7712,
+      "step": 12529
+    },
+    {
+      "epoch": 0.6896361935164291,
+      "grad_norm": 1.0029237270355225,
+      "learning_rate": 7.367227281735418e-06,
+      "loss": 0.7641,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6896912323187847,
+      "grad_norm": 0.6342340707778931,
+      "learning_rate": 7.3668454654144824e-06,
+      "loss": 0.7572,
+      "step": 12531
+    },
+    {
+      "epoch": 0.6897462711211404,
+      "grad_norm": 0.7475802302360535,
+      "learning_rate": 7.3664636313049696e-06,
+      "loss": 0.7969,
+      "step": 12532
+    },
+    {
+      "epoch": 0.6898013099234961,
+      "grad_norm": 0.7478888630867004,
+      "learning_rate": 7.36608177940975e-06,
+      "loss": 0.8299,
+      "step": 12533
+    },
+    {
+      "epoch": 0.6898563487258518,
+      "grad_norm": 0.7017174363136292,
+      "learning_rate": 7.365699909731694e-06,
+      "loss": 0.6608,
+      "step": 12534
+    },
+    {
+      "epoch": 0.6899113875282074,
+      "grad_norm": 0.7259606122970581,
+      "learning_rate": 7.3653180222736695e-06,
+      "loss": 0.7088,
+      "step": 12535
+    },
+    {
+      "epoch": 0.689966426330563,
+      "grad_norm": 0.7049521207809448,
+      "learning_rate": 7.364936117038548e-06,
+      "loss": 0.8177,
+      "step": 12536
+    },
+    {
+      "epoch": 0.6900214651329187,
+      "grad_norm": 0.6557304263114929,
+      "learning_rate": 7.364554194029201e-06,
+      "loss": 0.73,
+      "step": 12537
+    },
+    {
+      "epoch": 0.6900765039352744,
+      "grad_norm": 0.704140305519104,
+      "learning_rate": 7.364172253248497e-06,
+      "loss": 0.7671,
+      "step": 12538
+    },
+    {
+      "epoch": 0.69013154273763,
+      "grad_norm": 0.6879541873931885,
+      "learning_rate": 7.3637902946993064e-06,
+      "loss": 0.6707,
+      "step": 12539
+    },
+    {
+      "epoch": 0.6901865815399857,
+      "grad_norm": 0.7715931534767151,
+      "learning_rate": 7.363408318384501e-06,
+      "loss": 0.7494,
+      "step": 12540
+    },
+    {
+      "epoch": 0.6902416203423414,
+      "grad_norm": 0.7890990972518921,
+      "learning_rate": 7.363026324306952e-06,
+      "loss": 0.7499,
+      "step": 12541
+    },
+    {
+      "epoch": 0.6902966591446971,
+      "grad_norm": 0.7177792191505432,
+      "learning_rate": 7.362644312469529e-06,
+      "loss": 0.8053,
+      "step": 12542
+    },
+    {
+      "epoch": 0.6903516979470526,
+      "grad_norm": 0.7434332370758057,
+      "learning_rate": 7.3622622828751044e-06,
+      "loss": 0.7371,
+      "step": 12543
+    },
+    {
+      "epoch": 0.6904067367494083,
+      "grad_norm": 0.5836912989616394,
+      "learning_rate": 7.361880235526547e-06,
+      "loss": 0.6681,
+      "step": 12544
+    },
+    {
+      "epoch": 0.690461775551764,
+      "grad_norm": 0.6814625263214111,
+      "learning_rate": 7.3614981704267315e-06,
+      "loss": 0.7408,
+      "step": 12545
+    },
+    {
+      "epoch": 0.6905168143541196,
+      "grad_norm": 0.6524162292480469,
+      "learning_rate": 7.361116087578528e-06,
+      "loss": 0.6788,
+      "step": 12546
+    },
+    {
+      "epoch": 0.6905718531564753,
+      "grad_norm": 0.6614788174629211,
+      "learning_rate": 7.360733986984808e-06,
+      "loss": 0.75,
+      "step": 12547
+    },
+    {
+      "epoch": 0.690626891958831,
+      "grad_norm": 1.035152792930603,
+      "learning_rate": 7.360351868648442e-06,
+      "loss": 0.7181,
+      "step": 12548
+    },
+    {
+      "epoch": 0.6906819307611867,
+      "grad_norm": 0.7525657415390015,
+      "learning_rate": 7.359969732572305e-06,
+      "loss": 0.8149,
+      "step": 12549
+    },
+    {
+      "epoch": 0.6907369695635422,
+      "grad_norm": 0.8323431015014648,
+      "learning_rate": 7.359587578759267e-06,
+      "loss": 0.6908,
+      "step": 12550
+    },
+    {
+      "epoch": 0.6907920083658979,
+      "grad_norm": 0.7551344633102417,
+      "learning_rate": 7.3592054072122e-06,
+      "loss": 0.794,
+      "step": 12551
+    },
+    {
+      "epoch": 0.6908470471682536,
+      "grad_norm": 0.5937384366989136,
+      "learning_rate": 7.358823217933977e-06,
+      "loss": 0.6532,
+      "step": 12552
+    },
+    {
+      "epoch": 0.6909020859706093,
+      "grad_norm": 1.5515329837799072,
+      "learning_rate": 7.358441010927468e-06,
+      "loss": 0.7003,
+      "step": 12553
+    },
+    {
+      "epoch": 0.6909571247729649,
+      "grad_norm": 0.6838175654411316,
+      "learning_rate": 7.3580587861955495e-06,
+      "loss": 0.7184,
+      "step": 12554
+    },
+    {
+      "epoch": 0.6910121635753206,
+      "grad_norm": 0.7055354714393616,
+      "learning_rate": 7.357676543741092e-06,
+      "loss": 0.8372,
+      "step": 12555
+    },
+    {
+      "epoch": 0.6910672023776763,
+      "grad_norm": 0.8683249950408936,
+      "learning_rate": 7.3572942835669695e-06,
+      "loss": 0.7594,
+      "step": 12556
+    },
+    {
+      "epoch": 0.691122241180032,
+      "grad_norm": 0.8586179614067078,
+      "learning_rate": 7.3569120056760535e-06,
+      "loss": 0.8422,
+      "step": 12557
+    },
+    {
+      "epoch": 0.6911772799823875,
+      "grad_norm": 0.692132830619812,
+      "learning_rate": 7.356529710071217e-06,
+      "loss": 0.7872,
+      "step": 12558
+    },
+    {
+      "epoch": 0.6912323187847432,
+      "grad_norm": 0.7342404723167419,
+      "learning_rate": 7.356147396755335e-06,
+      "loss": 0.6908,
+      "step": 12559
+    },
+    {
+      "epoch": 0.6912873575870989,
+      "grad_norm": 0.6941357254981995,
+      "learning_rate": 7.35576506573128e-06,
+      "loss": 0.608,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6913423963894546,
+      "grad_norm": 0.648225724697113,
+      "learning_rate": 7.355382717001925e-06,
+      "loss": 0.6923,
+      "step": 12561
+    },
+    {
+      "epoch": 0.6913974351918102,
+      "grad_norm": 0.6735422015190125,
+      "learning_rate": 7.355000350570144e-06,
+      "loss": 0.7502,
+      "step": 12562
+    },
+    {
+      "epoch": 0.6914524739941659,
+      "grad_norm": 0.8507662415504456,
+      "learning_rate": 7.3546179664388105e-06,
+      "loss": 0.7883,
+      "step": 12563
+    },
+    {
+      "epoch": 0.6915075127965216,
+      "grad_norm": 0.7287268042564392,
+      "learning_rate": 7.3542355646108e-06,
+      "loss": 0.8687,
+      "step": 12564
+    },
+    {
+      "epoch": 0.6915625515988773,
+      "grad_norm": 0.6085666418075562,
+      "learning_rate": 7.353853145088983e-06,
+      "loss": 0.6675,
+      "step": 12565
+    },
+    {
+      "epoch": 0.6916175904012328,
+      "grad_norm": 0.727668046951294,
+      "learning_rate": 7.353470707876237e-06,
+      "loss": 0.8591,
+      "step": 12566
+    },
+    {
+      "epoch": 0.6916726292035885,
+      "grad_norm": 0.724846601486206,
+      "learning_rate": 7.353088252975436e-06,
+      "loss": 0.8501,
+      "step": 12567
+    },
+    {
+      "epoch": 0.6917276680059442,
+      "grad_norm": 0.6801046133041382,
+      "learning_rate": 7.352705780389452e-06,
+      "loss": 0.7637,
+      "step": 12568
+    },
+    {
+      "epoch": 0.6917827068082999,
+      "grad_norm": 0.680496335029602,
+      "learning_rate": 7.352323290121161e-06,
+      "loss": 0.7308,
+      "step": 12569
+    },
+    {
+      "epoch": 0.6918377456106555,
+      "grad_norm": 0.7143607139587402,
+      "learning_rate": 7.351940782173439e-06,
+      "loss": 0.7494,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6918927844130112,
+      "grad_norm": 0.679755687713623,
+      "learning_rate": 7.351558256549158e-06,
+      "loss": 0.7731,
+      "step": 12571
+    },
+    {
+      "epoch": 0.6919478232153669,
+      "grad_norm": 0.6626351475715637,
+      "learning_rate": 7.351175713251197e-06,
+      "loss": 0.8593,
+      "step": 12572
+    },
+    {
+      "epoch": 0.6920028620177225,
+      "grad_norm": 0.6830954551696777,
+      "learning_rate": 7.350793152282427e-06,
+      "loss": 0.6327,
+      "step": 12573
+    },
+    {
+      "epoch": 0.6920579008200781,
+      "grad_norm": 0.653810977935791,
+      "learning_rate": 7.350410573645726e-06,
+      "loss": 0.7341,
+      "step": 12574
+    },
+    {
+      "epoch": 0.6921129396224338,
+      "grad_norm": 0.6939566731452942,
+      "learning_rate": 7.3500279773439675e-06,
+      "loss": 0.7823,
+      "step": 12575
+    },
+    {
+      "epoch": 0.6921679784247895,
+      "grad_norm": 0.8212422728538513,
+      "learning_rate": 7.349645363380029e-06,
+      "loss": 0.6388,
+      "step": 12576
+    },
+    {
+      "epoch": 0.6922230172271452,
+      "grad_norm": 0.7703338265419006,
+      "learning_rate": 7.349262731756783e-06,
+      "loss": 0.7476,
+      "step": 12577
+    },
+    {
+      "epoch": 0.6922780560295008,
+      "grad_norm": 0.6710889935493469,
+      "learning_rate": 7.348880082477108e-06,
+      "loss": 0.7869,
+      "step": 12578
+    },
+    {
+      "epoch": 0.6923330948318565,
+      "grad_norm": 0.7384413480758667,
+      "learning_rate": 7.3484974155438795e-06,
+      "loss": 0.6628,
+      "step": 12579
+    },
+    {
+      "epoch": 0.6923881336342121,
+      "grad_norm": 0.7628176212310791,
+      "learning_rate": 7.348114730959973e-06,
+      "loss": 0.7599,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6924431724365678,
+      "grad_norm": 0.683885931968689,
+      "learning_rate": 7.347732028728264e-06,
+      "loss": 0.7134,
+      "step": 12581
+    },
+    {
+      "epoch": 0.6924982112389234,
+      "grad_norm": 0.6710503697395325,
+      "learning_rate": 7.34734930885163e-06,
+      "loss": 0.7147,
+      "step": 12582
+    },
+    {
+      "epoch": 0.6925532500412791,
+      "grad_norm": 0.6984537243843079,
+      "learning_rate": 7.346966571332947e-06,
+      "loss": 0.7517,
+      "step": 12583
+    },
+    {
+      "epoch": 0.6926082888436348,
+      "grad_norm": 0.7563193440437317,
+      "learning_rate": 7.346583816175092e-06,
+      "loss": 0.7971,
+      "step": 12584
+    },
+    {
+      "epoch": 0.6926633276459905,
+      "grad_norm": 0.8407838940620422,
+      "learning_rate": 7.346201043380941e-06,
+      "loss": 0.8227,
+      "step": 12585
+    },
+    {
+      "epoch": 0.6927183664483461,
+      "grad_norm": 0.673098623752594,
+      "learning_rate": 7.345818252953369e-06,
+      "loss": 0.7514,
+      "step": 12586
+    },
+    {
+      "epoch": 0.6927734052507017,
+      "grad_norm": 0.6452111005783081,
+      "learning_rate": 7.345435444895257e-06,
+      "loss": 0.7201,
+      "step": 12587
+    },
+    {
+      "epoch": 0.6928284440530574,
+      "grad_norm": 0.8728383779525757,
+      "learning_rate": 7.345052619209481e-06,
+      "loss": 0.7452,
+      "step": 12588
+    },
+    {
+      "epoch": 0.692883482855413,
+      "grad_norm": 0.7032049298286438,
+      "learning_rate": 7.344669775898914e-06,
+      "loss": 0.8885,
+      "step": 12589
+    },
+    {
+      "epoch": 0.6929385216577687,
+      "grad_norm": 0.7744605541229248,
+      "learning_rate": 7.344286914966438e-06,
+      "loss": 0.8048,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6929935604601244,
+      "grad_norm": 0.7334163784980774,
+      "learning_rate": 7.343904036414931e-06,
+      "loss": 0.8502,
+      "step": 12591
+    },
+    {
+      "epoch": 0.6930485992624801,
+      "grad_norm": 0.6684108376502991,
+      "learning_rate": 7.343521140247266e-06,
+      "loss": 0.8264,
+      "step": 12592
+    },
+    {
+      "epoch": 0.6931036380648357,
+      "grad_norm": 0.6192718744277954,
+      "learning_rate": 7.343138226466324e-06,
+      "loss": 0.6625,
+      "step": 12593
+    },
+    {
+      "epoch": 0.6931586768671913,
+      "grad_norm": 0.6410724520683289,
+      "learning_rate": 7.342755295074984e-06,
+      "loss": 0.717,
+      "step": 12594
+    },
+    {
+      "epoch": 0.693213715669547,
+      "grad_norm": 0.6854361891746521,
+      "learning_rate": 7.342372346076121e-06,
+      "loss": 0.7246,
+      "step": 12595
+    },
+    {
+      "epoch": 0.6932687544719027,
+      "grad_norm": 0.6920250058174133,
+      "learning_rate": 7.341989379472614e-06,
+      "loss": 0.7414,
+      "step": 12596
+    },
+    {
+      "epoch": 0.6933237932742583,
+      "grad_norm": 0.6545842885971069,
+      "learning_rate": 7.341606395267342e-06,
+      "loss": 0.7731,
+      "step": 12597
+    },
+    {
+      "epoch": 0.693378832076614,
+      "grad_norm": 0.6879072785377502,
+      "learning_rate": 7.341223393463184e-06,
+      "loss": 0.7272,
+      "step": 12598
+    },
+    {
+      "epoch": 0.6934338708789697,
+      "grad_norm": 0.7460979223251343,
+      "learning_rate": 7.340840374063018e-06,
+      "loss": 0.771,
+      "step": 12599
+    },
+    {
+      "epoch": 0.6934889096813254,
+      "grad_norm": 0.7836858630180359,
+      "learning_rate": 7.340457337069722e-06,
+      "loss": 0.846,
+      "step": 12600
+    },
+    {
+      "epoch": 0.693543948483681,
+      "grad_norm": 0.958403468132019,
+      "learning_rate": 7.340074282486174e-06,
+      "loss": 0.8913,
+      "step": 12601
+    },
+    {
+      "epoch": 0.6935989872860366,
+      "grad_norm": 0.6614813208580017,
+      "learning_rate": 7.339691210315254e-06,
+      "loss": 0.7129,
+      "step": 12602
+    },
+    {
+      "epoch": 0.6936540260883923,
+      "grad_norm": 0.7303252816200256,
+      "learning_rate": 7.339308120559843e-06,
+      "loss": 0.8395,
+      "step": 12603
+    },
+    {
+      "epoch": 0.693709064890748,
+      "grad_norm": 0.7341620922088623,
+      "learning_rate": 7.338925013222817e-06,
+      "loss": 0.8341,
+      "step": 12604
+    },
+    {
+      "epoch": 0.6937641036931036,
+      "grad_norm": 0.7077179551124573,
+      "learning_rate": 7.338541888307056e-06,
+      "loss": 0.7813,
+      "step": 12605
+    },
+    {
+      "epoch": 0.6938191424954593,
+      "grad_norm": 0.6654969453811646,
+      "learning_rate": 7.338158745815441e-06,
+      "loss": 0.7337,
+      "step": 12606
+    },
+    {
+      "epoch": 0.693874181297815,
+      "grad_norm": 0.6637474894523621,
+      "learning_rate": 7.337775585750852e-06,
+      "loss": 0.8197,
+      "step": 12607
+    },
+    {
+      "epoch": 0.6939292201001707,
+      "grad_norm": 0.654712975025177,
+      "learning_rate": 7.337392408116166e-06,
+      "loss": 0.6991,
+      "step": 12608
+    },
+    {
+      "epoch": 0.6939842589025262,
+      "grad_norm": 0.6698346138000488,
+      "learning_rate": 7.337009212914265e-06,
+      "loss": 0.7991,
+      "step": 12609
+    },
+    {
+      "epoch": 0.6940392977048819,
+      "grad_norm": 0.9616294503211975,
+      "learning_rate": 7.336626000148028e-06,
+      "loss": 0.7326,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6940943365072376,
+      "grad_norm": 0.7749543786048889,
+      "learning_rate": 7.336242769820335e-06,
+      "loss": 0.8015,
+      "step": 12611
+    },
+    {
+      "epoch": 0.6941493753095933,
+      "grad_norm": 0.7263140678405762,
+      "learning_rate": 7.335859521934068e-06,
+      "loss": 0.7538,
+      "step": 12612
+    },
+    {
+      "epoch": 0.6942044141119489,
+      "grad_norm": 0.6383689641952515,
+      "learning_rate": 7.335476256492105e-06,
+      "loss": 0.7611,
+      "step": 12613
+    },
+    {
+      "epoch": 0.6942594529143046,
+      "grad_norm": 0.7464908957481384,
+      "learning_rate": 7.335092973497326e-06,
+      "loss": 0.7904,
+      "step": 12614
+    },
+    {
+      "epoch": 0.6943144917166603,
+      "grad_norm": 1.114864468574524,
+      "learning_rate": 7.334709672952615e-06,
+      "loss": 0.8518,
+      "step": 12615
+    },
+    {
+      "epoch": 0.694369530519016,
+      "grad_norm": 0.6712734699249268,
+      "learning_rate": 7.334326354860852e-06,
+      "loss": 0.7431,
+      "step": 12616
+    },
+    {
+      "epoch": 0.6944245693213715,
+      "grad_norm": 0.7559850811958313,
+      "learning_rate": 7.3339430192249166e-06,
+      "loss": 0.7556,
+      "step": 12617
+    },
+    {
+      "epoch": 0.6944796081237272,
+      "grad_norm": 0.7262033224105835,
+      "learning_rate": 7.333559666047689e-06,
+      "loss": 0.7624,
+      "step": 12618
+    },
+    {
+      "epoch": 0.6945346469260829,
+      "grad_norm": 0.6428695917129517,
+      "learning_rate": 7.333176295332053e-06,
+      "loss": 0.6894,
+      "step": 12619
+    },
+    {
+      "epoch": 0.6945896857284386,
+      "grad_norm": 0.7353672385215759,
+      "learning_rate": 7.3327929070808875e-06,
+      "loss": 0.7611,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6946447245307942,
+      "grad_norm": 0.7063810229301453,
+      "learning_rate": 7.332409501297076e-06,
+      "loss": 0.7428,
+      "step": 12621
+    },
+    {
+      "epoch": 0.6946997633331499,
+      "grad_norm": 0.6552421450614929,
+      "learning_rate": 7.332026077983498e-06,
+      "loss": 0.7046,
+      "step": 12622
+    },
+    {
+      "epoch": 0.6947548021355056,
+      "grad_norm": 0.8843327760696411,
+      "learning_rate": 7.331642637143037e-06,
+      "loss": 0.6952,
+      "step": 12623
+    },
+    {
+      "epoch": 0.6948098409378612,
+      "grad_norm": 0.7279102802276611,
+      "learning_rate": 7.331259178778574e-06,
+      "loss": 0.7911,
+      "step": 12624
+    },
+    {
+      "epoch": 0.6948648797402168,
+      "grad_norm": 0.6585525870323181,
+      "learning_rate": 7.33087570289299e-06,
+      "loss": 0.7684,
+      "step": 12625
+    },
+    {
+      "epoch": 0.6949199185425725,
+      "grad_norm": 0.663185715675354,
+      "learning_rate": 7.3304922094891695e-06,
+      "loss": 0.6753,
+      "step": 12626
+    },
+    {
+      "epoch": 0.6949749573449282,
+      "grad_norm": 0.652765691280365,
+      "learning_rate": 7.330108698569993e-06,
+      "loss": 0.7333,
+      "step": 12627
+    },
+    {
+      "epoch": 0.6950299961472839,
+      "grad_norm": 0.7781688570976257,
+      "learning_rate": 7.329725170138343e-06,
+      "loss": 0.7312,
+      "step": 12628
+    },
+    {
+      "epoch": 0.6950850349496395,
+      "grad_norm": 0.6798241138458252,
+      "learning_rate": 7.329341624197102e-06,
+      "loss": 0.7747,
+      "step": 12629
+    },
+    {
+      "epoch": 0.6951400737519952,
+      "grad_norm": 0.7588373422622681,
+      "learning_rate": 7.328958060749153e-06,
+      "loss": 0.8535,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6951951125543508,
+      "grad_norm": 0.8833348155021667,
+      "learning_rate": 7.328574479797379e-06,
+      "loss": 0.8345,
+      "step": 12631
+    },
+    {
+      "epoch": 0.6952501513567064,
+      "grad_norm": 0.799454927444458,
+      "learning_rate": 7.328190881344663e-06,
+      "loss": 0.7571,
+      "step": 12632
+    },
+    {
+      "epoch": 0.6953051901590621,
+      "grad_norm": 0.8030340671539307,
+      "learning_rate": 7.327807265393887e-06,
+      "loss": 0.7426,
+      "step": 12633
+    },
+    {
+      "epoch": 0.6953602289614178,
+      "grad_norm": 0.6246228218078613,
+      "learning_rate": 7.327423631947934e-06,
+      "loss": 0.6712,
+      "step": 12634
+    },
+    {
+      "epoch": 0.6954152677637735,
+      "grad_norm": 0.7203500866889954,
+      "learning_rate": 7.32703998100969e-06,
+      "loss": 0.8315,
+      "step": 12635
+    },
+    {
+      "epoch": 0.6954703065661291,
+      "grad_norm": 0.6128239035606384,
+      "learning_rate": 7.326656312582035e-06,
+      "loss": 0.6788,
+      "step": 12636
+    },
+    {
+      "epoch": 0.6955253453684848,
+      "grad_norm": 0.8052619695663452,
+      "learning_rate": 7.326272626667852e-06,
+      "loss": 0.8076,
+      "step": 12637
+    },
+    {
+      "epoch": 0.6955803841708404,
+      "grad_norm": 0.9128470420837402,
+      "learning_rate": 7.325888923270029e-06,
+      "loss": 0.7135,
+      "step": 12638
+    },
+    {
+      "epoch": 0.6956354229731961,
+      "grad_norm": 0.6815299391746521,
+      "learning_rate": 7.325505202391447e-06,
+      "loss": 0.7756,
+      "step": 12639
+    },
+    {
+      "epoch": 0.6956904617755517,
+      "grad_norm": 0.6278733611106873,
+      "learning_rate": 7.325121464034991e-06,
+      "loss": 0.6583,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6957455005779074,
+      "grad_norm": 0.7161649465560913,
+      "learning_rate": 7.324737708203543e-06,
+      "loss": 0.7106,
+      "step": 12641
+    },
+    {
+      "epoch": 0.6958005393802631,
+      "grad_norm": 0.6827715635299683,
+      "learning_rate": 7.324353934899989e-06,
+      "loss": 0.7988,
+      "step": 12642
+    },
+    {
+      "epoch": 0.6958555781826188,
+      "grad_norm": 0.9999695420265198,
+      "learning_rate": 7.323970144127215e-06,
+      "loss": 0.8222,
+      "step": 12643
+    },
+    {
+      "epoch": 0.6959106169849744,
+      "grad_norm": 0.8048173785209656,
+      "learning_rate": 7.323586335888102e-06,
+      "loss": 0.7157,
+      "step": 12644
+    },
+    {
+      "epoch": 0.69596565578733,
+      "grad_norm": 0.7403637170791626,
+      "learning_rate": 7.323202510185536e-06,
+      "loss": 0.7516,
+      "step": 12645
+    },
+    {
+      "epoch": 0.6960206945896857,
+      "grad_norm": 0.6660793423652649,
+      "learning_rate": 7.322818667022402e-06,
+      "loss": 0.7081,
+      "step": 12646
+    },
+    {
+      "epoch": 0.6960757333920414,
+      "grad_norm": 0.713985800743103,
+      "learning_rate": 7.322434806401585e-06,
+      "loss": 0.7682,
+      "step": 12647
+    },
+    {
+      "epoch": 0.696130772194397,
+      "grad_norm": 0.739253044128418,
+      "learning_rate": 7.322050928325969e-06,
+      "loss": 0.838,
+      "step": 12648
+    },
+    {
+      "epoch": 0.6961858109967527,
+      "grad_norm": 0.8350489735603333,
+      "learning_rate": 7.32166703279844e-06,
+      "loss": 0.7627,
+      "step": 12649
+    },
+    {
+      "epoch": 0.6962408497991084,
+      "grad_norm": 0.580456018447876,
+      "learning_rate": 7.321283119821883e-06,
+      "loss": 0.6248,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6962958886014641,
+      "grad_norm": 0.8619480729103088,
+      "learning_rate": 7.320899189399183e-06,
+      "loss": 0.848,
+      "step": 12651
+    },
+    {
+      "epoch": 0.6963509274038197,
+      "grad_norm": 0.6201381087303162,
+      "learning_rate": 7.320515241533227e-06,
+      "loss": 0.6506,
+      "step": 12652
+    },
+    {
+      "epoch": 0.6964059662061753,
+      "grad_norm": 0.6956773400306702,
+      "learning_rate": 7.320131276226898e-06,
+      "loss": 0.7561,
+      "step": 12653
+    },
+    {
+      "epoch": 0.696461005008531,
+      "grad_norm": 0.6382080912590027,
+      "learning_rate": 7.319747293483085e-06,
+      "loss": 0.6462,
+      "step": 12654
+    },
+    {
+      "epoch": 0.6965160438108867,
+      "grad_norm": 0.7288708686828613,
+      "learning_rate": 7.319363293304672e-06,
+      "loss": 0.7907,
+      "step": 12655
+    },
+    {
+      "epoch": 0.6965710826132423,
+      "grad_norm": 0.6280390024185181,
+      "learning_rate": 7.318979275694546e-06,
+      "loss": 0.6882,
+      "step": 12656
+    },
+    {
+      "epoch": 0.696626121415598,
+      "grad_norm": 0.7260308861732483,
+      "learning_rate": 7.31859524065559e-06,
+      "loss": 0.756,
+      "step": 12657
+    },
+    {
+      "epoch": 0.6966811602179537,
+      "grad_norm": 0.6715009212493896,
+      "learning_rate": 7.318211188190696e-06,
+      "loss": 0.7194,
+      "step": 12658
+    },
+    {
+      "epoch": 0.6967361990203094,
+      "grad_norm": 0.6770408749580383,
+      "learning_rate": 7.3178271183027465e-06,
+      "loss": 0.808,
+      "step": 12659
+    },
+    {
+      "epoch": 0.6967912378226649,
+      "grad_norm": 0.7209904789924622,
+      "learning_rate": 7.317443030994628e-06,
+      "loss": 0.7242,
+      "step": 12660
+    },
+    {
+      "epoch": 0.6968462766250206,
+      "grad_norm": 0.6943202018737793,
+      "learning_rate": 7.317058926269227e-06,
+      "loss": 0.758,
+      "step": 12661
+    },
+    {
+      "epoch": 0.6969013154273763,
+      "grad_norm": 0.6073412299156189,
+      "learning_rate": 7.316674804129432e-06,
+      "loss": 0.6571,
+      "step": 12662
+    },
+    {
+      "epoch": 0.696956354229732,
+      "grad_norm": 0.7065439224243164,
+      "learning_rate": 7.316290664578129e-06,
+      "loss": 0.7333,
+      "step": 12663
+    },
+    {
+      "epoch": 0.6970113930320876,
+      "grad_norm": 0.6275133490562439,
+      "learning_rate": 7.315906507618207e-06,
+      "loss": 0.6785,
+      "step": 12664
+    },
+    {
+      "epoch": 0.6970664318344433,
+      "grad_norm": 0.6484677791595459,
+      "learning_rate": 7.315522333252551e-06,
+      "loss": 0.7461,
+      "step": 12665
+    },
+    {
+      "epoch": 0.697121470636799,
+      "grad_norm": 0.6815413236618042,
+      "learning_rate": 7.315138141484049e-06,
+      "loss": 0.673,
+      "step": 12666
+    },
+    {
+      "epoch": 0.6971765094391547,
+      "grad_norm": 0.7227872610092163,
+      "learning_rate": 7.314753932315587e-06,
+      "loss": 0.7212,
+      "step": 12667
+    },
+    {
+      "epoch": 0.6972315482415102,
+      "grad_norm": 0.661568284034729,
+      "learning_rate": 7.314369705750055e-06,
+      "loss": 0.7633,
+      "step": 12668
+    },
+    {
+      "epoch": 0.6972865870438659,
+      "grad_norm": 0.5873990654945374,
+      "learning_rate": 7.3139854617903405e-06,
+      "loss": 0.6142,
+      "step": 12669
+    },
+    {
+      "epoch": 0.6973416258462216,
+      "grad_norm": 0.7015652656555176,
+      "learning_rate": 7.313601200439331e-06,
+      "loss": 0.6762,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6973966646485773,
+      "grad_norm": 0.7060853242874146,
+      "learning_rate": 7.313216921699913e-06,
+      "loss": 0.8111,
+      "step": 12671
+    },
+    {
+      "epoch": 0.6974517034509329,
+      "grad_norm": 0.6198092699050903,
+      "learning_rate": 7.312832625574977e-06,
+      "loss": 0.7058,
+      "step": 12672
+    },
+    {
+      "epoch": 0.6975067422532886,
+      "grad_norm": 0.6785464286804199,
+      "learning_rate": 7.312448312067408e-06,
+      "loss": 0.7509,
+      "step": 12673
+    },
+    {
+      "epoch": 0.6975617810556443,
+      "grad_norm": 0.74974524974823,
+      "learning_rate": 7.312063981180097e-06,
+      "loss": 0.7679,
+      "step": 12674
+    },
+    {
+      "epoch": 0.6976168198579998,
+      "grad_norm": 0.6188651919364929,
+      "learning_rate": 7.311679632915934e-06,
+      "loss": 0.663,
+      "step": 12675
+    },
+    {
+      "epoch": 0.6976718586603555,
+      "grad_norm": 0.7458493113517761,
+      "learning_rate": 7.3112952672778044e-06,
+      "loss": 0.7316,
+      "step": 12676
+    },
+    {
+      "epoch": 0.6977268974627112,
+      "grad_norm": 0.7480403780937195,
+      "learning_rate": 7.310910884268597e-06,
+      "loss": 0.8476,
+      "step": 12677
+    },
+    {
+      "epoch": 0.6977819362650669,
+      "grad_norm": 0.6921943426132202,
+      "learning_rate": 7.310526483891204e-06,
+      "loss": 0.7931,
+      "step": 12678
+    },
+    {
+      "epoch": 0.6978369750674225,
+      "grad_norm": 0.7384023666381836,
+      "learning_rate": 7.3101420661485124e-06,
+      "loss": 0.7698,
+      "step": 12679
+    },
+    {
+      "epoch": 0.6978920138697782,
+      "grad_norm": 0.6693310141563416,
+      "learning_rate": 7.3097576310434105e-06,
+      "loss": 0.6838,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6979470526721339,
+      "grad_norm": 0.6888617873191833,
+      "learning_rate": 7.309373178578789e-06,
+      "loss": 0.7196,
+      "step": 12681
+    },
+    {
+      "epoch": 0.6980020914744895,
+      "grad_norm": 0.7608165144920349,
+      "learning_rate": 7.308988708757536e-06,
+      "loss": 0.7483,
+      "step": 12682
+    },
+    {
+      "epoch": 0.6980571302768451,
+      "grad_norm": 0.6969812512397766,
+      "learning_rate": 7.308604221582543e-06,
+      "loss": 0.7415,
+      "step": 12683
+    },
+    {
+      "epoch": 0.6981121690792008,
+      "grad_norm": 0.7440872192382812,
+      "learning_rate": 7.3082197170566996e-06,
+      "loss": 0.7776,
+      "step": 12684
+    },
+    {
+      "epoch": 0.6981672078815565,
+      "grad_norm": 0.7920299768447876,
+      "learning_rate": 7.307835195182892e-06,
+      "loss": 0.746,
+      "step": 12685
+    },
+    {
+      "epoch": 0.6982222466839122,
+      "grad_norm": 0.7002919912338257,
+      "learning_rate": 7.3074506559640134e-06,
+      "loss": 0.7948,
+      "step": 12686
+    },
+    {
+      "epoch": 0.6982772854862678,
+      "grad_norm": 0.7199681997299194,
+      "learning_rate": 7.3070660994029554e-06,
+      "loss": 0.7568,
+      "step": 12687
+    },
+    {
+      "epoch": 0.6983323242886235,
+      "grad_norm": 0.6287575960159302,
+      "learning_rate": 7.306681525502604e-06,
+      "loss": 0.6564,
+      "step": 12688
+    },
+    {
+      "epoch": 0.6983873630909792,
+      "grad_norm": 0.6910778880119324,
+      "learning_rate": 7.306296934265853e-06,
+      "loss": 0.7892,
+      "step": 12689
+    },
+    {
+      "epoch": 0.6984424018933348,
+      "grad_norm": 0.6454603672027588,
+      "learning_rate": 7.30591232569559e-06,
+      "loss": 0.7848,
+      "step": 12690
+    },
+    {
+      "epoch": 0.6984974406956904,
+      "grad_norm": 0.7337101101875305,
+      "learning_rate": 7.305527699794709e-06,
+      "loss": 0.8012,
+      "step": 12691
+    },
+    {
+      "epoch": 0.6985524794980461,
+      "grad_norm": 0.6694337129592896,
+      "learning_rate": 7.305143056566098e-06,
+      "loss": 0.7767,
+      "step": 12692
+    },
+    {
+      "epoch": 0.6986075183004018,
+      "grad_norm": 0.6485214233398438,
+      "learning_rate": 7.30475839601265e-06,
+      "loss": 0.7142,
+      "step": 12693
+    },
+    {
+      "epoch": 0.6986625571027575,
+      "grad_norm": 0.6401854753494263,
+      "learning_rate": 7.304373718137253e-06,
+      "loss": 0.6562,
+      "step": 12694
+    },
+    {
+      "epoch": 0.6987175959051131,
+      "grad_norm": 0.7190635800361633,
+      "learning_rate": 7.303989022942801e-06,
+      "loss": 0.7513,
+      "step": 12695
+    },
+    {
+      "epoch": 0.6987726347074688,
+      "grad_norm": 0.7100299596786499,
+      "learning_rate": 7.3036043104321854e-06,
+      "loss": 0.759,
+      "step": 12696
+    },
+    {
+      "epoch": 0.6988276735098244,
+      "grad_norm": 0.8507145047187805,
+      "learning_rate": 7.303219580608295e-06,
+      "loss": 0.7567,
+      "step": 12697
+    },
+    {
+      "epoch": 0.6988827123121801,
+      "grad_norm": 0.6758378744125366,
+      "learning_rate": 7.302834833474022e-06,
+      "loss": 0.6751,
+      "step": 12698
+    },
+    {
+      "epoch": 0.6989377511145357,
+      "grad_norm": 0.7602974772453308,
+      "learning_rate": 7.30245006903226e-06,
+      "loss": 0.7304,
+      "step": 12699
+    },
+    {
+      "epoch": 0.6989927899168914,
+      "grad_norm": 0.7519045472145081,
+      "learning_rate": 7.3020652872859e-06,
+      "loss": 0.7573,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6990478287192471,
+      "grad_norm": 0.6076456904411316,
+      "learning_rate": 7.301680488237832e-06,
+      "loss": 0.6335,
+      "step": 12701
+    },
+    {
+      "epoch": 0.6991028675216028,
+      "grad_norm": 0.6900685429573059,
+      "learning_rate": 7.30129567189095e-06,
+      "loss": 0.7787,
+      "step": 12702
+    },
+    {
+      "epoch": 0.6991579063239584,
+      "grad_norm": 0.7366316318511963,
+      "learning_rate": 7.300910838248146e-06,
+      "loss": 0.8176,
+      "step": 12703
+    },
+    {
+      "epoch": 0.699212945126314,
+      "grad_norm": 0.6658521890640259,
+      "learning_rate": 7.300525987312312e-06,
+      "loss": 0.6436,
+      "step": 12704
+    },
+    {
+      "epoch": 0.6992679839286697,
+      "grad_norm": 0.7635871171951294,
+      "learning_rate": 7.300141119086341e-06,
+      "loss": 0.8421,
+      "step": 12705
+    },
+    {
+      "epoch": 0.6993230227310254,
+      "grad_norm": 0.7257800698280334,
+      "learning_rate": 7.299756233573125e-06,
+      "loss": 0.6468,
+      "step": 12706
+    },
+    {
+      "epoch": 0.699378061533381,
+      "grad_norm": 0.7536096572875977,
+      "learning_rate": 7.299371330775558e-06,
+      "loss": 0.7782,
+      "step": 12707
+    },
+    {
+      "epoch": 0.6994331003357367,
+      "grad_norm": 0.7504379153251648,
+      "learning_rate": 7.298986410696529e-06,
+      "loss": 0.7097,
+      "step": 12708
+    },
+    {
+      "epoch": 0.6994881391380924,
+      "grad_norm": 0.7340306043624878,
+      "learning_rate": 7.298601473338936e-06,
+      "loss": 0.8165,
+      "step": 12709
+    },
+    {
+      "epoch": 0.6995431779404481,
+      "grad_norm": 0.6928045749664307,
+      "learning_rate": 7.298216518705667e-06,
+      "loss": 0.777,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6995982167428036,
+      "grad_norm": 0.6942496299743652,
+      "learning_rate": 7.29783154679962e-06,
+      "loss": 0.6607,
+      "step": 12711
+    },
+    {
+      "epoch": 0.6996532555451593,
+      "grad_norm": 0.6646896600723267,
+      "learning_rate": 7.297446557623684e-06,
+      "loss": 0.712,
+      "step": 12712
+    },
+    {
+      "epoch": 0.699708294347515,
+      "grad_norm": 0.6828078627586365,
+      "learning_rate": 7.297061551180758e-06,
+      "loss": 0.7251,
+      "step": 12713
+    },
+    {
+      "epoch": 0.6997633331498707,
+      "grad_norm": 0.7554219365119934,
+      "learning_rate": 7.296676527473729e-06,
+      "loss": 0.8279,
+      "step": 12714
+    },
+    {
+      "epoch": 0.6998183719522263,
+      "grad_norm": 0.8122106194496155,
+      "learning_rate": 7.296291486505495e-06,
+      "loss": 0.8039,
+      "step": 12715
+    },
+    {
+      "epoch": 0.699873410754582,
+      "grad_norm": 0.6602222323417664,
+      "learning_rate": 7.295906428278949e-06,
+      "loss": 0.7149,
+      "step": 12716
+    },
+    {
+      "epoch": 0.6999284495569377,
+      "grad_norm": 0.8341954350471497,
+      "learning_rate": 7.2955213527969845e-06,
+      "loss": 0.7868,
+      "step": 12717
+    },
+    {
+      "epoch": 0.6999834883592932,
+      "grad_norm": 0.7157256603240967,
+      "learning_rate": 7.295136260062496e-06,
+      "loss": 0.745,
+      "step": 12718
+    },
+    {
+      "epoch": 0.7000385271616489,
+      "grad_norm": 0.5845672488212585,
+      "learning_rate": 7.294751150078379e-06,
+      "loss": 0.657,
+      "step": 12719
+    },
+    {
+      "epoch": 0.7000935659640046,
+      "grad_norm": 0.7370786070823669,
+      "learning_rate": 7.2943660228475265e-06,
+      "loss": 0.7883,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7001486047663603,
+      "grad_norm": 0.6687451004981995,
+      "learning_rate": 7.293980878372833e-06,
+      "loss": 0.7945,
+      "step": 12721
+    },
+    {
+      "epoch": 0.7002036435687159,
+      "grad_norm": 0.6352105736732483,
+      "learning_rate": 7.293595716657192e-06,
+      "loss": 0.6581,
+      "step": 12722
+    },
+    {
+      "epoch": 0.7002586823710716,
+      "grad_norm": 0.7371370196342468,
+      "learning_rate": 7.293210537703499e-06,
+      "loss": 0.7859,
+      "step": 12723
+    },
+    {
+      "epoch": 0.7003137211734273,
+      "grad_norm": 0.6885504722595215,
+      "learning_rate": 7.292825341514651e-06,
+      "loss": 0.7355,
+      "step": 12724
+    },
+    {
+      "epoch": 0.700368759975783,
+      "grad_norm": 0.6930849552154541,
+      "learning_rate": 7.292440128093542e-06,
+      "loss": 0.8145,
+      "step": 12725
+    },
+    {
+      "epoch": 0.7004237987781385,
+      "grad_norm": 0.6767199635505676,
+      "learning_rate": 7.292054897443065e-06,
+      "loss": 0.7136,
+      "step": 12726
+    },
+    {
+      "epoch": 0.7004788375804942,
+      "grad_norm": 0.6672216653823853,
+      "learning_rate": 7.291669649566117e-06,
+      "loss": 0.6131,
+      "step": 12727
+    },
+    {
+      "epoch": 0.7005338763828499,
+      "grad_norm": 0.6618815064430237,
+      "learning_rate": 7.291284384465595e-06,
+      "loss": 0.7633,
+      "step": 12728
+    },
+    {
+      "epoch": 0.7005889151852056,
+      "grad_norm": 0.6573876142501831,
+      "learning_rate": 7.290899102144392e-06,
+      "loss": 0.7621,
+      "step": 12729
+    },
+    {
+      "epoch": 0.7006439539875612,
+      "grad_norm": 0.7449564337730408,
+      "learning_rate": 7.290513802605405e-06,
+      "loss": 0.6488,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7006989927899169,
+      "grad_norm": 0.7307295203208923,
+      "learning_rate": 7.290128485851529e-06,
+      "loss": 0.7095,
+      "step": 12731
+    },
+    {
+      "epoch": 0.7007540315922726,
+      "grad_norm": 0.698699951171875,
+      "learning_rate": 7.2897431518856596e-06,
+      "loss": 0.7428,
+      "step": 12732
+    },
+    {
+      "epoch": 0.7008090703946283,
+      "grad_norm": 0.6334750056266785,
+      "learning_rate": 7.289357800710695e-06,
+      "loss": 0.6977,
+      "step": 12733
+    },
+    {
+      "epoch": 0.7008641091969838,
+      "grad_norm": 0.6526468396186829,
+      "learning_rate": 7.288972432329529e-06,
+      "loss": 0.6375,
+      "step": 12734
+    },
+    {
+      "epoch": 0.7009191479993395,
+      "grad_norm": 0.7282149791717529,
+      "learning_rate": 7.288587046745059e-06,
+      "loss": 0.7494,
+      "step": 12735
+    },
+    {
+      "epoch": 0.7009741868016952,
+      "grad_norm": 0.8511056900024414,
+      "learning_rate": 7.288201643960182e-06,
+      "loss": 0.7494,
+      "step": 12736
+    },
+    {
+      "epoch": 0.7010292256040509,
+      "grad_norm": 0.6908526420593262,
+      "learning_rate": 7.287816223977793e-06,
+      "loss": 0.6861,
+      "step": 12737
+    },
+    {
+      "epoch": 0.7010842644064065,
+      "grad_norm": 0.7582982182502747,
+      "learning_rate": 7.2874307868007896e-06,
+      "loss": 0.7758,
+      "step": 12738
+    },
+    {
+      "epoch": 0.7011393032087622,
+      "grad_norm": 0.9717779159545898,
+      "learning_rate": 7.2870453324320685e-06,
+      "loss": 0.7221,
+      "step": 12739
+    },
+    {
+      "epoch": 0.7011943420111179,
+      "grad_norm": 0.6532751321792603,
+      "learning_rate": 7.286659860874529e-06,
+      "loss": 0.8009,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7012493808134735,
+      "grad_norm": 0.6708540320396423,
+      "learning_rate": 7.286274372131065e-06,
+      "loss": 0.7177,
+      "step": 12741
+    },
+    {
+      "epoch": 0.7013044196158291,
+      "grad_norm": 0.7624804973602295,
+      "learning_rate": 7.285888866204575e-06,
+      "loss": 0.7878,
+      "step": 12742
+    },
+    {
+      "epoch": 0.7013594584181848,
+      "grad_norm": 0.7167851328849792,
+      "learning_rate": 7.285503343097955e-06,
+      "loss": 0.7276,
+      "step": 12743
+    },
+    {
+      "epoch": 0.7014144972205405,
+      "grad_norm": 0.6592209935188293,
+      "learning_rate": 7.2851178028141045e-06,
+      "loss": 0.7665,
+      "step": 12744
+    },
+    {
+      "epoch": 0.7014695360228962,
+      "grad_norm": 0.684847354888916,
+      "learning_rate": 7.284732245355921e-06,
+      "loss": 0.7358,
+      "step": 12745
+    },
+    {
+      "epoch": 0.7015245748252518,
+      "grad_norm": 0.6852415800094604,
+      "learning_rate": 7.2843466707262985e-06,
+      "loss": 0.7805,
+      "step": 12746
+    },
+    {
+      "epoch": 0.7015796136276075,
+      "grad_norm": 0.6422114968299866,
+      "learning_rate": 7.283961078928141e-06,
+      "loss": 0.7386,
+      "step": 12747
+    },
+    {
+      "epoch": 0.7016346524299631,
+      "grad_norm": 0.7538495659828186,
+      "learning_rate": 7.283575469964343e-06,
+      "loss": 0.798,
+      "step": 12748
+    },
+    {
+      "epoch": 0.7016896912323188,
+      "grad_norm": 0.6646687984466553,
+      "learning_rate": 7.2831898438378025e-06,
+      "loss": 0.7048,
+      "step": 12749
+    },
+    {
+      "epoch": 0.7017447300346744,
+      "grad_norm": 0.8338429927825928,
+      "learning_rate": 7.2828042005514176e-06,
+      "loss": 0.8585,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7017997688370301,
+      "grad_norm": 0.7086663842201233,
+      "learning_rate": 7.282418540108088e-06,
+      "loss": 0.8011,
+      "step": 12751
+    },
+    {
+      "epoch": 0.7018548076393858,
+      "grad_norm": 0.6040074229240417,
+      "learning_rate": 7.282032862510712e-06,
+      "loss": 0.6327,
+      "step": 12752
+    },
+    {
+      "epoch": 0.7019098464417415,
+      "grad_norm": 0.7030978798866272,
+      "learning_rate": 7.281647167762187e-06,
+      "loss": 0.6373,
+      "step": 12753
+    },
+    {
+      "epoch": 0.7019648852440971,
+      "grad_norm": 0.662308394908905,
+      "learning_rate": 7.281261455865414e-06,
+      "loss": 0.7283,
+      "step": 12754
+    },
+    {
+      "epoch": 0.7020199240464527,
+      "grad_norm": 0.7369368672370911,
+      "learning_rate": 7.28087572682329e-06,
+      "loss": 0.7632,
+      "step": 12755
+    },
+    {
+      "epoch": 0.7020749628488084,
+      "grad_norm": 0.6887282729148865,
+      "learning_rate": 7.280489980638714e-06,
+      "loss": 0.7629,
+      "step": 12756
+    },
+    {
+      "epoch": 0.702130001651164,
+      "grad_norm": 0.656512975692749,
+      "learning_rate": 7.280104217314587e-06,
+      "loss": 0.8028,
+      "step": 12757
+    },
+    {
+      "epoch": 0.7021850404535197,
+      "grad_norm": 0.7006264328956604,
+      "learning_rate": 7.279718436853805e-06,
+      "loss": 0.7025,
+      "step": 12758
+    },
+    {
+      "epoch": 0.7022400792558754,
+      "grad_norm": 0.675585925579071,
+      "learning_rate": 7.279332639259271e-06,
+      "loss": 0.8001,
+      "step": 12759
+    },
+    {
+      "epoch": 0.7022951180582311,
+      "grad_norm": 0.7105827331542969,
+      "learning_rate": 7.278946824533883e-06,
+      "loss": 0.7767,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7023501568605867,
+      "grad_norm": 0.8310064673423767,
+      "learning_rate": 7.27856099268054e-06,
+      "loss": 0.7828,
+      "step": 12761
+    },
+    {
+      "epoch": 0.7024051956629423,
+      "grad_norm": 0.6885055899620056,
+      "learning_rate": 7.278175143702142e-06,
+      "loss": 0.7018,
+      "step": 12762
+    },
+    {
+      "epoch": 0.702460234465298,
+      "grad_norm": 0.6542866826057434,
+      "learning_rate": 7.27778927760159e-06,
+      "loss": 0.7118,
+      "step": 12763
+    },
+    {
+      "epoch": 0.7025152732676537,
+      "grad_norm": 0.9102655053138733,
+      "learning_rate": 7.277403394381784e-06,
+      "loss": 0.8381,
+      "step": 12764
+    },
+    {
+      "epoch": 0.7025703120700093,
+      "grad_norm": 0.6538355946540833,
+      "learning_rate": 7.277017494045624e-06,
+      "loss": 0.7766,
+      "step": 12765
+    },
+    {
+      "epoch": 0.702625350872365,
+      "grad_norm": 0.6691237092018127,
+      "learning_rate": 7.27663157659601e-06,
+      "loss": 0.8077,
+      "step": 12766
+    },
+    {
+      "epoch": 0.7026803896747207,
+      "grad_norm": 0.7159995436668396,
+      "learning_rate": 7.2762456420358414e-06,
+      "loss": 0.8333,
+      "step": 12767
+    },
+    {
+      "epoch": 0.7027354284770764,
+      "grad_norm": 0.6518422365188599,
+      "learning_rate": 7.275859690368022e-06,
+      "loss": 0.7634,
+      "step": 12768
+    },
+    {
+      "epoch": 0.702790467279432,
+      "grad_norm": 0.6969057321548462,
+      "learning_rate": 7.275473721595449e-06,
+      "loss": 0.7481,
+      "step": 12769
+    },
+    {
+      "epoch": 0.7028455060817876,
+      "grad_norm": 0.6788915395736694,
+      "learning_rate": 7.2750877357210225e-06,
+      "loss": 0.7402,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7029005448841433,
+      "grad_norm": 0.7323998212814331,
+      "learning_rate": 7.274701732747649e-06,
+      "loss": 0.7122,
+      "step": 12771
+    },
+    {
+      "epoch": 0.702955583686499,
+      "grad_norm": 0.7224077582359314,
+      "learning_rate": 7.274315712678224e-06,
+      "loss": 0.7333,
+      "step": 12772
+    },
+    {
+      "epoch": 0.7030106224888546,
+      "grad_norm": 0.9009444117546082,
+      "learning_rate": 7.273929675515652e-06,
+      "loss": 0.6912,
+      "step": 12773
+    },
+    {
+      "epoch": 0.7030656612912103,
+      "grad_norm": 0.7076312899589539,
+      "learning_rate": 7.273543621262832e-06,
+      "loss": 0.7651,
+      "step": 12774
+    },
+    {
+      "epoch": 0.703120700093566,
+      "grad_norm": 0.78575599193573,
+      "learning_rate": 7.273157549922668e-06,
+      "loss": 0.7443,
+      "step": 12775
+    },
+    {
+      "epoch": 0.7031757388959217,
+      "grad_norm": 0.6957094669342041,
+      "learning_rate": 7.27277146149806e-06,
+      "loss": 0.7684,
+      "step": 12776
+    },
+    {
+      "epoch": 0.7032307776982772,
+      "grad_norm": 1.177878975868225,
+      "learning_rate": 7.27238535599191e-06,
+      "loss": 0.9033,
+      "step": 12777
+    },
+    {
+      "epoch": 0.7032858165006329,
+      "grad_norm": 0.6929007768630981,
+      "learning_rate": 7.27199923340712e-06,
+      "loss": 0.7411,
+      "step": 12778
+    },
+    {
+      "epoch": 0.7033408553029886,
+      "grad_norm": 0.7725315093994141,
+      "learning_rate": 7.2716130937465926e-06,
+      "loss": 0.7833,
+      "step": 12779
+    },
+    {
+      "epoch": 0.7033958941053443,
+      "grad_norm": 0.6512928605079651,
+      "learning_rate": 7.271226937013228e-06,
+      "loss": 0.7918,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7034509329076999,
+      "grad_norm": 0.7033893465995789,
+      "learning_rate": 7.270840763209931e-06,
+      "loss": 0.843,
+      "step": 12781
+    },
+    {
+      "epoch": 0.7035059717100556,
+      "grad_norm": 0.7596432566642761,
+      "learning_rate": 7.2704545723396e-06,
+      "loss": 0.7916,
+      "step": 12782
+    },
+    {
+      "epoch": 0.7035610105124113,
+      "grad_norm": 0.6256046891212463,
+      "learning_rate": 7.270068364405143e-06,
+      "loss": 0.6531,
+      "step": 12783
+    },
+    {
+      "epoch": 0.703616049314767,
+      "grad_norm": 0.8107615113258362,
+      "learning_rate": 7.26968213940946e-06,
+      "loss": 0.7755,
+      "step": 12784
+    },
+    {
+      "epoch": 0.7036710881171225,
+      "grad_norm": 0.6742845177650452,
+      "learning_rate": 7.269295897355451e-06,
+      "loss": 0.834,
+      "step": 12785
+    },
+    {
+      "epoch": 0.7037261269194782,
+      "grad_norm": 0.6665072441101074,
+      "learning_rate": 7.268909638246024e-06,
+      "loss": 0.6864,
+      "step": 12786
+    },
+    {
+      "epoch": 0.7037811657218339,
+      "grad_norm": 0.68357914686203,
+      "learning_rate": 7.268523362084078e-06,
+      "loss": 0.7789,
+      "step": 12787
+    },
+    {
+      "epoch": 0.7038362045241896,
+      "grad_norm": 0.6878114938735962,
+      "learning_rate": 7.268137068872519e-06,
+      "loss": 0.7277,
+      "step": 12788
+    },
+    {
+      "epoch": 0.7038912433265452,
+      "grad_norm": 0.7173313498497009,
+      "learning_rate": 7.267750758614247e-06,
+      "loss": 0.8156,
+      "step": 12789
+    },
+    {
+      "epoch": 0.7039462821289009,
+      "grad_norm": 0.6523084044456482,
+      "learning_rate": 7.267364431312169e-06,
+      "loss": 0.7143,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7040013209312566,
+      "grad_norm": 0.7403815388679504,
+      "learning_rate": 7.2669780869691865e-06,
+      "loss": 0.8196,
+      "step": 12791
+    },
+    {
+      "epoch": 0.7040563597336122,
+      "grad_norm": 0.6411255598068237,
+      "learning_rate": 7.266591725588204e-06,
+      "loss": 0.6645,
+      "step": 12792
+    },
+    {
+      "epoch": 0.7041113985359678,
+      "grad_norm": 0.9094020128250122,
+      "learning_rate": 7.266205347172124e-06,
+      "loss": 0.8023,
+      "step": 12793
+    },
+    {
+      "epoch": 0.7041664373383235,
+      "grad_norm": 1.1041208505630493,
+      "learning_rate": 7.265818951723851e-06,
+      "loss": 0.7011,
+      "step": 12794
+    },
+    {
+      "epoch": 0.7042214761406792,
+      "grad_norm": 0.7339954376220703,
+      "learning_rate": 7.265432539246289e-06,
+      "loss": 0.7467,
+      "step": 12795
+    },
+    {
+      "epoch": 0.7042765149430349,
+      "grad_norm": 0.7055865526199341,
+      "learning_rate": 7.265046109742344e-06,
+      "loss": 0.7364,
+      "step": 12796
+    },
+    {
+      "epoch": 0.7043315537453905,
+      "grad_norm": 0.7052320241928101,
+      "learning_rate": 7.264659663214917e-06,
+      "loss": 0.7611,
+      "step": 12797
+    },
+    {
+      "epoch": 0.7043865925477462,
+      "grad_norm": 0.7374194860458374,
+      "learning_rate": 7.264273199666915e-06,
+      "loss": 0.7612,
+      "step": 12798
+    },
+    {
+      "epoch": 0.7044416313501018,
+      "grad_norm": 0.634986162185669,
+      "learning_rate": 7.263886719101242e-06,
+      "loss": 0.8001,
+      "step": 12799
+    },
+    {
+      "epoch": 0.7044966701524574,
+      "grad_norm": 0.8178644180297852,
+      "learning_rate": 7.2635002215208014e-06,
+      "loss": 0.8404,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7045517089548131,
+      "grad_norm": 0.7743822336196899,
+      "learning_rate": 7.263113706928501e-06,
+      "loss": 0.7297,
+      "step": 12801
+    },
+    {
+      "epoch": 0.7046067477571688,
+      "grad_norm": 0.6558601260185242,
+      "learning_rate": 7.262727175327242e-06,
+      "loss": 0.6933,
+      "step": 12802
+    },
+    {
+      "epoch": 0.7046617865595245,
+      "grad_norm": 1.0608787536621094,
+      "learning_rate": 7.262340626719933e-06,
+      "loss": 0.8792,
+      "step": 12803
+    },
+    {
+      "epoch": 0.7047168253618801,
+      "grad_norm": 0.7488270401954651,
+      "learning_rate": 7.261954061109475e-06,
+      "loss": 0.7755,
+      "step": 12804
+    },
+    {
+      "epoch": 0.7047718641642358,
+      "grad_norm": 0.8960574865341187,
+      "learning_rate": 7.261567478498778e-06,
+      "loss": 0.7274,
+      "step": 12805
+    },
+    {
+      "epoch": 0.7048269029665915,
+      "grad_norm": 0.6289944648742676,
+      "learning_rate": 7.2611808788907436e-06,
+      "loss": 0.6469,
+      "step": 12806
+    },
+    {
+      "epoch": 0.7048819417689471,
+      "grad_norm": 0.6488339900970459,
+      "learning_rate": 7.26079426228828e-06,
+      "loss": 0.7581,
+      "step": 12807
+    },
+    {
+      "epoch": 0.7049369805713027,
+      "grad_norm": 0.7354650497436523,
+      "learning_rate": 7.260407628694292e-06,
+      "loss": 0.7596,
+      "step": 12808
+    },
+    {
+      "epoch": 0.7049920193736584,
+      "grad_norm": 0.8163169026374817,
+      "learning_rate": 7.2600209781116834e-06,
+      "loss": 0.8291,
+      "step": 12809
+    },
+    {
+      "epoch": 0.7050470581760141,
+      "grad_norm": 0.8223916292190552,
+      "learning_rate": 7.259634310543364e-06,
+      "loss": 0.7089,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7051020969783698,
+      "grad_norm": 0.7815924286842346,
+      "learning_rate": 7.2592476259922374e-06,
+      "loss": 0.8098,
+      "step": 12811
+    },
+    {
+      "epoch": 0.7051571357807254,
+      "grad_norm": 0.7027734518051147,
+      "learning_rate": 7.2588609244612105e-06,
+      "loss": 0.7276,
+      "step": 12812
+    },
+    {
+      "epoch": 0.705212174583081,
+      "grad_norm": 0.7345930337905884,
+      "learning_rate": 7.2584742059531894e-06,
+      "loss": 0.803,
+      "step": 12813
+    },
+    {
+      "epoch": 0.7052672133854367,
+      "grad_norm": 0.6998127102851868,
+      "learning_rate": 7.258087470471081e-06,
+      "loss": 0.7938,
+      "step": 12814
+    },
+    {
+      "epoch": 0.7053222521877924,
+      "grad_norm": 0.6418118476867676,
+      "learning_rate": 7.257700718017793e-06,
+      "loss": 0.66,
+      "step": 12815
+    },
+    {
+      "epoch": 0.705377290990148,
+      "grad_norm": 0.6774695515632629,
+      "learning_rate": 7.257313948596228e-06,
+      "loss": 0.7143,
+      "step": 12816
+    },
+    {
+      "epoch": 0.7054323297925037,
+      "grad_norm": 0.7107009291648865,
+      "learning_rate": 7.256927162209298e-06,
+      "loss": 0.8378,
+      "step": 12817
+    },
+    {
+      "epoch": 0.7054873685948594,
+      "grad_norm": 0.7287374138832092,
+      "learning_rate": 7.256540358859906e-06,
+      "loss": 0.88,
+      "step": 12818
+    },
+    {
+      "epoch": 0.7055424073972151,
+      "grad_norm": 0.651221752166748,
+      "learning_rate": 7.256153538550961e-06,
+      "loss": 0.7092,
+      "step": 12819
+    },
+    {
+      "epoch": 0.7055974461995707,
+      "grad_norm": 0.6549085974693298,
+      "learning_rate": 7.255766701285371e-06,
+      "loss": 0.6697,
+      "step": 12820
+    },
+    {
+      "epoch": 0.7056524850019263,
+      "grad_norm": 0.6617292165756226,
+      "learning_rate": 7.255379847066041e-06,
+      "loss": 0.7779,
+      "step": 12821
+    },
+    {
+      "epoch": 0.705707523804282,
+      "grad_norm": 0.6677221655845642,
+      "learning_rate": 7.254992975895879e-06,
+      "loss": 0.7821,
+      "step": 12822
+    },
+    {
+      "epoch": 0.7057625626066377,
+      "grad_norm": 0.8183515667915344,
+      "learning_rate": 7.2546060877777945e-06,
+      "loss": 0.7727,
+      "step": 12823
+    },
+    {
+      "epoch": 0.7058176014089933,
+      "grad_norm": 0.6574132442474365,
+      "learning_rate": 7.2542191827146945e-06,
+      "loss": 0.7118,
+      "step": 12824
+    },
+    {
+      "epoch": 0.705872640211349,
+      "grad_norm": 0.6874130964279175,
+      "learning_rate": 7.253832260709487e-06,
+      "loss": 0.7677,
+      "step": 12825
+    },
+    {
+      "epoch": 0.7059276790137047,
+      "grad_norm": 0.6460297107696533,
+      "learning_rate": 7.253445321765079e-06,
+      "loss": 0.725,
+      "step": 12826
+    },
+    {
+      "epoch": 0.7059827178160604,
+      "grad_norm": 0.6618219614028931,
+      "learning_rate": 7.253058365884379e-06,
+      "loss": 0.7504,
+      "step": 12827
+    },
+    {
+      "epoch": 0.706037756618416,
+      "grad_norm": 0.6519019603729248,
+      "learning_rate": 7.252671393070295e-06,
+      "loss": 0.7382,
+      "step": 12828
+    },
+    {
+      "epoch": 0.7060927954207716,
+      "grad_norm": 0.7114588022232056,
+      "learning_rate": 7.252284403325737e-06,
+      "loss": 0.8364,
+      "step": 12829
+    },
+    {
+      "epoch": 0.7061478342231273,
+      "grad_norm": 0.6304726600646973,
+      "learning_rate": 7.251897396653611e-06,
+      "loss": 0.6972,
+      "step": 12830
+    },
+    {
+      "epoch": 0.706202873025483,
+      "grad_norm": 0.6728807687759399,
+      "learning_rate": 7.251510373056827e-06,
+      "loss": 0.671,
+      "step": 12831
+    },
+    {
+      "epoch": 0.7062579118278386,
+      "grad_norm": 0.690641462802887,
+      "learning_rate": 7.251123332538295e-06,
+      "loss": 0.7381,
+      "step": 12832
+    },
+    {
+      "epoch": 0.7063129506301943,
+      "grad_norm": 0.7018027305603027,
+      "learning_rate": 7.2507362751009226e-06,
+      "loss": 0.7546,
+      "step": 12833
+    },
+    {
+      "epoch": 0.70636798943255,
+      "grad_norm": 0.7203684449195862,
+      "learning_rate": 7.250349200747617e-06,
+      "loss": 0.7534,
+      "step": 12834
+    },
+    {
+      "epoch": 0.7064230282349057,
+      "grad_norm": 0.6936585903167725,
+      "learning_rate": 7.24996210948129e-06,
+      "loss": 0.7716,
+      "step": 12835
+    },
+    {
+      "epoch": 0.7064780670372612,
+      "grad_norm": 0.7421281337738037,
+      "learning_rate": 7.249575001304851e-06,
+      "loss": 0.7517,
+      "step": 12836
+    },
+    {
+      "epoch": 0.7065331058396169,
+      "grad_norm": 0.6622288227081299,
+      "learning_rate": 7.249187876221207e-06,
+      "loss": 0.6799,
+      "step": 12837
+    },
+    {
+      "epoch": 0.7065881446419726,
+      "grad_norm": 0.7267055511474609,
+      "learning_rate": 7.24880073423327e-06,
+      "loss": 0.7871,
+      "step": 12838
+    },
+    {
+      "epoch": 0.7066431834443283,
+      "grad_norm": 0.6978085041046143,
+      "learning_rate": 7.2484135753439485e-06,
+      "loss": 0.7812,
+      "step": 12839
+    },
+    {
+      "epoch": 0.7066982222466839,
+      "grad_norm": 0.8353652358055115,
+      "learning_rate": 7.248026399556153e-06,
+      "loss": 0.7481,
+      "step": 12840
+    },
+    {
+      "epoch": 0.7067532610490396,
+      "grad_norm": 0.8402471542358398,
+      "learning_rate": 7.247639206872792e-06,
+      "loss": 0.783,
+      "step": 12841
+    },
+    {
+      "epoch": 0.7068082998513953,
+      "grad_norm": 0.8279419541358948,
+      "learning_rate": 7.247251997296777e-06,
+      "loss": 0.8177,
+      "step": 12842
+    },
+    {
+      "epoch": 0.7068633386537508,
+      "grad_norm": 0.6850735545158386,
+      "learning_rate": 7.246864770831017e-06,
+      "loss": 0.7586,
+      "step": 12843
+    },
+    {
+      "epoch": 0.7069183774561065,
+      "grad_norm": 0.7327665090560913,
+      "learning_rate": 7.246477527478422e-06,
+      "loss": 0.9327,
+      "step": 12844
+    },
+    {
+      "epoch": 0.7069734162584622,
+      "grad_norm": 0.6343075037002563,
+      "learning_rate": 7.246090267241905e-06,
+      "loss": 0.6957,
+      "step": 12845
+    },
+    {
+      "epoch": 0.7070284550608179,
+      "grad_norm": 0.7028965353965759,
+      "learning_rate": 7.245702990124373e-06,
+      "loss": 0.7524,
+      "step": 12846
+    },
+    {
+      "epoch": 0.7070834938631735,
+      "grad_norm": 0.7578299045562744,
+      "learning_rate": 7.24531569612874e-06,
+      "loss": 0.7302,
+      "step": 12847
+    },
+    {
+      "epoch": 0.7071385326655292,
+      "grad_norm": 0.8113438487052917,
+      "learning_rate": 7.2449283852579146e-06,
+      "loss": 0.7658,
+      "step": 12848
+    },
+    {
+      "epoch": 0.7071935714678849,
+      "grad_norm": 0.6442512273788452,
+      "learning_rate": 7.244541057514809e-06,
+      "loss": 0.6742,
+      "step": 12849
+    },
+    {
+      "epoch": 0.7072486102702406,
+      "grad_norm": 0.8595272898674011,
+      "learning_rate": 7.244153712902333e-06,
+      "loss": 0.7944,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7073036490725961,
+      "grad_norm": 0.6565983891487122,
+      "learning_rate": 7.243766351423398e-06,
+      "loss": 0.7411,
+      "step": 12851
+    },
+    {
+      "epoch": 0.7073586878749518,
+      "grad_norm": 0.7935337424278259,
+      "learning_rate": 7.243378973080917e-06,
+      "loss": 0.8109,
+      "step": 12852
+    },
+    {
+      "epoch": 0.7074137266773075,
+      "grad_norm": 0.7083927392959595,
+      "learning_rate": 7.242991577877799e-06,
+      "loss": 0.8405,
+      "step": 12853
+    },
+    {
+      "epoch": 0.7074687654796632,
+      "grad_norm": 0.7452830672264099,
+      "learning_rate": 7.242604165816958e-06,
+      "loss": 0.7972,
+      "step": 12854
+    },
+    {
+      "epoch": 0.7075238042820188,
+      "grad_norm": 0.6775808334350586,
+      "learning_rate": 7.242216736901302e-06,
+      "loss": 0.7114,
+      "step": 12855
+    },
+    {
+      "epoch": 0.7075788430843745,
+      "grad_norm": 0.8069992661476135,
+      "learning_rate": 7.241829291133748e-06,
+      "loss": 0.6606,
+      "step": 12856
+    },
+    {
+      "epoch": 0.7076338818867302,
+      "grad_norm": 0.6690802574157715,
+      "learning_rate": 7.241441828517203e-06,
+      "loss": 0.742,
+      "step": 12857
+    },
+    {
+      "epoch": 0.7076889206890858,
+      "grad_norm": 0.8077805638313293,
+      "learning_rate": 7.2410543490545814e-06,
+      "loss": 0.7786,
+      "step": 12858
+    },
+    {
+      "epoch": 0.7077439594914414,
+      "grad_norm": 0.6906875967979431,
+      "learning_rate": 7.240666852748795e-06,
+      "loss": 0.7445,
+      "step": 12859
+    },
+    {
+      "epoch": 0.7077989982937971,
+      "grad_norm": 0.6830704808235168,
+      "learning_rate": 7.2402793396027585e-06,
+      "loss": 0.7664,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7078540370961528,
+      "grad_norm": 0.8118640780448914,
+      "learning_rate": 7.23989180961938e-06,
+      "loss": 0.7654,
+      "step": 12861
+    },
+    {
+      "epoch": 0.7079090758985085,
+      "grad_norm": 0.6819882392883301,
+      "learning_rate": 7.2395042628015755e-06,
+      "loss": 0.649,
+      "step": 12862
+    },
+    {
+      "epoch": 0.7079641147008641,
+      "grad_norm": 0.6543441414833069,
+      "learning_rate": 7.239116699152256e-06,
+      "loss": 0.8054,
+      "step": 12863
+    },
+    {
+      "epoch": 0.7080191535032198,
+      "grad_norm": 0.8613989353179932,
+      "learning_rate": 7.238729118674335e-06,
+      "loss": 0.7283,
+      "step": 12864
+    },
+    {
+      "epoch": 0.7080741923055754,
+      "grad_norm": 0.6993124485015869,
+      "learning_rate": 7.238341521370725e-06,
+      "loss": 0.8145,
+      "step": 12865
+    },
+    {
+      "epoch": 0.7081292311079311,
+      "grad_norm": 0.7047560811042786,
+      "learning_rate": 7.237953907244339e-06,
+      "loss": 0.6729,
+      "step": 12866
+    },
+    {
+      "epoch": 0.7081842699102867,
+      "grad_norm": 0.7923689484596252,
+      "learning_rate": 7.237566276298091e-06,
+      "loss": 0.7615,
+      "step": 12867
+    },
+    {
+      "epoch": 0.7082393087126424,
+      "grad_norm": 0.6873850226402283,
+      "learning_rate": 7.237178628534894e-06,
+      "loss": 0.7638,
+      "step": 12868
+    },
+    {
+      "epoch": 0.7082943475149981,
+      "grad_norm": 0.6483134031295776,
+      "learning_rate": 7.236790963957661e-06,
+      "loss": 0.6366,
+      "step": 12869
+    },
+    {
+      "epoch": 0.7083493863173538,
+      "grad_norm": 0.6623784899711609,
+      "learning_rate": 7.236403282569305e-06,
+      "loss": 0.7032,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7084044251197094,
+      "grad_norm": 0.7004366517066956,
+      "learning_rate": 7.236015584372741e-06,
+      "loss": 0.6436,
+      "step": 12871
+    },
+    {
+      "epoch": 0.708459463922065,
+      "grad_norm": 0.5676529407501221,
+      "learning_rate": 7.235627869370883e-06,
+      "loss": 0.6395,
+      "step": 12872
+    },
+    {
+      "epoch": 0.7085145027244207,
+      "grad_norm": 0.6909729838371277,
+      "learning_rate": 7.235240137566644e-06,
+      "loss": 0.7063,
+      "step": 12873
+    },
+    {
+      "epoch": 0.7085695415267764,
+      "grad_norm": 0.7635348439216614,
+      "learning_rate": 7.234852388962939e-06,
+      "loss": 0.7518,
+      "step": 12874
+    },
+    {
+      "epoch": 0.708624580329132,
+      "grad_norm": 0.7217742204666138,
+      "learning_rate": 7.2344646235626815e-06,
+      "loss": 0.7782,
+      "step": 12875
+    },
+    {
+      "epoch": 0.7086796191314877,
+      "grad_norm": 0.6506509184837341,
+      "learning_rate": 7.2340768413687855e-06,
+      "loss": 0.7456,
+      "step": 12876
+    },
+    {
+      "epoch": 0.7087346579338434,
+      "grad_norm": 0.6537386775016785,
+      "learning_rate": 7.2336890423841664e-06,
+      "loss": 0.7395,
+      "step": 12877
+    },
+    {
+      "epoch": 0.7087896967361991,
+      "grad_norm": 0.7759900689125061,
+      "learning_rate": 7.233301226611737e-06,
+      "loss": 0.8098,
+      "step": 12878
+    },
+    {
+      "epoch": 0.7088447355385546,
+      "grad_norm": 0.8476354479789734,
+      "learning_rate": 7.232913394054415e-06,
+      "loss": 0.8241,
+      "step": 12879
+    },
+    {
+      "epoch": 0.7088997743409103,
+      "grad_norm": 0.6770507097244263,
+      "learning_rate": 7.232525544715114e-06,
+      "loss": 0.6966,
+      "step": 12880
+    },
+    {
+      "epoch": 0.708954813143266,
+      "grad_norm": 0.7750027775764465,
+      "learning_rate": 7.232137678596747e-06,
+      "loss": 0.8038,
+      "step": 12881
+    },
+    {
+      "epoch": 0.7090098519456217,
+      "grad_norm": 0.6507213711738586,
+      "learning_rate": 7.231749795702232e-06,
+      "loss": 0.6446,
+      "step": 12882
+    },
+    {
+      "epoch": 0.7090648907479773,
+      "grad_norm": 0.7554625272750854,
+      "learning_rate": 7.231361896034481e-06,
+      "loss": 0.7769,
+      "step": 12883
+    },
+    {
+      "epoch": 0.709119929550333,
+      "grad_norm": 0.8175020813941956,
+      "learning_rate": 7.230973979596414e-06,
+      "loss": 0.8283,
+      "step": 12884
+    },
+    {
+      "epoch": 0.7091749683526887,
+      "grad_norm": 0.7528663873672485,
+      "learning_rate": 7.2305860463909416e-06,
+      "loss": 0.7737,
+      "step": 12885
+    },
+    {
+      "epoch": 0.7092300071550443,
+      "grad_norm": 0.9242768883705139,
+      "learning_rate": 7.230198096420983e-06,
+      "loss": 0.647,
+      "step": 12886
+    },
+    {
+      "epoch": 0.7092850459573999,
+      "grad_norm": 0.899874746799469,
+      "learning_rate": 7.229810129689452e-06,
+      "loss": 0.8952,
+      "step": 12887
+    },
+    {
+      "epoch": 0.7093400847597556,
+      "grad_norm": 0.8221275806427002,
+      "learning_rate": 7.229422146199266e-06,
+      "loss": 0.6845,
+      "step": 12888
+    },
+    {
+      "epoch": 0.7093951235621113,
+      "grad_norm": 0.6964027285575867,
+      "learning_rate": 7.229034145953338e-06,
+      "loss": 0.7153,
+      "step": 12889
+    },
+    {
+      "epoch": 0.7094501623644669,
+      "grad_norm": 0.8018684387207031,
+      "learning_rate": 7.228646128954588e-06,
+      "loss": 0.6421,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7095052011668226,
+      "grad_norm": 0.6874614953994751,
+      "learning_rate": 7.228258095205928e-06,
+      "loss": 0.8024,
+      "step": 12891
+    },
+    {
+      "epoch": 0.7095602399691783,
+      "grad_norm": 0.7141417860984802,
+      "learning_rate": 7.227870044710277e-06,
+      "loss": 0.7746,
+      "step": 12892
+    },
+    {
+      "epoch": 0.709615278771534,
+      "grad_norm": 0.7109399437904358,
+      "learning_rate": 7.227481977470552e-06,
+      "loss": 0.7826,
+      "step": 12893
+    },
+    {
+      "epoch": 0.7096703175738895,
+      "grad_norm": 0.7021867036819458,
+      "learning_rate": 7.227093893489669e-06,
+      "loss": 0.7196,
+      "step": 12894
+    },
+    {
+      "epoch": 0.7097253563762452,
+      "grad_norm": 0.6896560788154602,
+      "learning_rate": 7.226705792770543e-06,
+      "loss": 0.6925,
+      "step": 12895
+    },
+    {
+      "epoch": 0.7097803951786009,
+      "grad_norm": 0.7138262987136841,
+      "learning_rate": 7.226317675316094e-06,
+      "loss": 0.7417,
+      "step": 12896
+    },
+    {
+      "epoch": 0.7098354339809566,
+      "grad_norm": 0.6789212226867676,
+      "learning_rate": 7.225929541129236e-06,
+      "loss": 0.7095,
+      "step": 12897
+    },
+    {
+      "epoch": 0.7098904727833122,
+      "grad_norm": 0.8102045059204102,
+      "learning_rate": 7.225541390212889e-06,
+      "loss": 0.9252,
+      "step": 12898
+    },
+    {
+      "epoch": 0.7099455115856679,
+      "grad_norm": 0.6220358610153198,
+      "learning_rate": 7.2251532225699674e-06,
+      "loss": 0.7205,
+      "step": 12899
+    },
+    {
+      "epoch": 0.7100005503880236,
+      "grad_norm": 0.6375265121459961,
+      "learning_rate": 7.224765038203391e-06,
+      "loss": 0.7974,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7100555891903793,
+      "grad_norm": 0.7457360029220581,
+      "learning_rate": 7.224376837116075e-06,
+      "loss": 0.7083,
+      "step": 12901
+    },
+    {
+      "epoch": 0.7101106279927348,
+      "grad_norm": 0.7012878060340881,
+      "learning_rate": 7.2239886193109374e-06,
+      "loss": 0.7334,
+      "step": 12902
+    },
+    {
+      "epoch": 0.7101656667950905,
+      "grad_norm": 0.7437683343887329,
+      "learning_rate": 7.223600384790898e-06,
+      "loss": 0.82,
+      "step": 12903
+    },
+    {
+      "epoch": 0.7102207055974462,
+      "grad_norm": 0.6727370619773865,
+      "learning_rate": 7.223212133558872e-06,
+      "loss": 0.7339,
+      "step": 12904
+    },
+    {
+      "epoch": 0.7102757443998019,
+      "grad_norm": 0.9253849983215332,
+      "learning_rate": 7.222823865617781e-06,
+      "loss": 0.7398,
+      "step": 12905
+    },
+    {
+      "epoch": 0.7103307832021575,
+      "grad_norm": 0.6664100885391235,
+      "learning_rate": 7.222435580970539e-06,
+      "loss": 0.7519,
+      "step": 12906
+    },
+    {
+      "epoch": 0.7103858220045132,
+      "grad_norm": 0.7452943325042725,
+      "learning_rate": 7.222047279620066e-06,
+      "loss": 0.7382,
+      "step": 12907
+    },
+    {
+      "epoch": 0.7104408608068689,
+      "grad_norm": 0.7235015630722046,
+      "learning_rate": 7.22165896156928e-06,
+      "loss": 0.7726,
+      "step": 12908
+    },
+    {
+      "epoch": 0.7104958996092245,
+      "grad_norm": 0.6324653029441833,
+      "learning_rate": 7.221270626821102e-06,
+      "loss": 0.7451,
+      "step": 12909
+    },
+    {
+      "epoch": 0.7105509384115801,
+      "grad_norm": 0.789829432964325,
+      "learning_rate": 7.220882275378447e-06,
+      "loss": 0.7375,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7106059772139358,
+      "grad_norm": 0.9090244174003601,
+      "learning_rate": 7.220493907244236e-06,
+      "loss": 0.8935,
+      "step": 12911
+    },
+    {
+      "epoch": 0.7106610160162915,
+      "grad_norm": 0.6570677757263184,
+      "learning_rate": 7.220105522421388e-06,
+      "loss": 0.7259,
+      "step": 12912
+    },
+    {
+      "epoch": 0.7107160548186472,
+      "grad_norm": 0.7142132520675659,
+      "learning_rate": 7.219717120912819e-06,
+      "loss": 0.7862,
+      "step": 12913
+    },
+    {
+      "epoch": 0.7107710936210028,
+      "grad_norm": 0.7359404563903809,
+      "learning_rate": 7.219328702721452e-06,
+      "loss": 0.7074,
+      "step": 12914
+    },
+    {
+      "epoch": 0.7108261324233585,
+      "grad_norm": 0.7118046283721924,
+      "learning_rate": 7.218940267850203e-06,
+      "loss": 0.8151,
+      "step": 12915
+    },
+    {
+      "epoch": 0.7108811712257141,
+      "grad_norm": 0.8301580548286438,
+      "learning_rate": 7.218551816301994e-06,
+      "loss": 0.7031,
+      "step": 12916
+    },
+    {
+      "epoch": 0.7109362100280698,
+      "grad_norm": 0.6647501587867737,
+      "learning_rate": 7.218163348079743e-06,
+      "loss": 0.8309,
+      "step": 12917
+    },
+    {
+      "epoch": 0.7109912488304254,
+      "grad_norm": 0.6546997427940369,
+      "learning_rate": 7.217774863186371e-06,
+      "loss": 0.717,
+      "step": 12918
+    },
+    {
+      "epoch": 0.7110462876327811,
+      "grad_norm": 0.6639735102653503,
+      "learning_rate": 7.217386361624795e-06,
+      "loss": 0.7308,
+      "step": 12919
+    },
+    {
+      "epoch": 0.7111013264351368,
+      "grad_norm": 0.724433183670044,
+      "learning_rate": 7.216997843397938e-06,
+      "loss": 0.7576,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7111563652374925,
+      "grad_norm": 0.750253438949585,
+      "learning_rate": 7.216609308508719e-06,
+      "loss": 0.7014,
+      "step": 12921
+    },
+    {
+      "epoch": 0.7112114040398481,
+      "grad_norm": 0.7010897397994995,
+      "learning_rate": 7.216220756960058e-06,
+      "loss": 0.6951,
+      "step": 12922
+    },
+    {
+      "epoch": 0.7112664428422037,
+      "grad_norm": 0.7739251852035522,
+      "learning_rate": 7.215832188754873e-06,
+      "loss": 0.7392,
+      "step": 12923
+    },
+    {
+      "epoch": 0.7113214816445594,
+      "grad_norm": 0.6893059015274048,
+      "learning_rate": 7.215443603896088e-06,
+      "loss": 0.7029,
+      "step": 12924
+    },
+    {
+      "epoch": 0.7113765204469151,
+      "grad_norm": 0.8061872124671936,
+      "learning_rate": 7.215055002386622e-06,
+      "loss": 0.7557,
+      "step": 12925
+    },
+    {
+      "epoch": 0.7114315592492707,
+      "grad_norm": 1.089525580406189,
+      "learning_rate": 7.214666384229395e-06,
+      "loss": 0.6701,
+      "step": 12926
+    },
+    {
+      "epoch": 0.7114865980516264,
+      "grad_norm": 0.7601733207702637,
+      "learning_rate": 7.2142777494273275e-06,
+      "loss": 0.8113,
+      "step": 12927
+    },
+    {
+      "epoch": 0.7115416368539821,
+      "grad_norm": 0.7863540649414062,
+      "learning_rate": 7.213889097983342e-06,
+      "loss": 0.7945,
+      "step": 12928
+    },
+    {
+      "epoch": 0.7115966756563377,
+      "grad_norm": 0.7722556591033936,
+      "learning_rate": 7.21350042990036e-06,
+      "loss": 0.9492,
+      "step": 12929
+    },
+    {
+      "epoch": 0.7116517144586934,
+      "grad_norm": 0.6834682822227478,
+      "learning_rate": 7.213111745181299e-06,
+      "loss": 0.7138,
+      "step": 12930
+    },
+    {
+      "epoch": 0.711706753261049,
+      "grad_norm": 0.6974432468414307,
+      "learning_rate": 7.212723043829083e-06,
+      "loss": 0.7654,
+      "step": 12931
+    },
+    {
+      "epoch": 0.7117617920634047,
+      "grad_norm": 0.9797543883323669,
+      "learning_rate": 7.2123343258466334e-06,
+      "loss": 0.7786,
+      "step": 12932
+    },
+    {
+      "epoch": 0.7118168308657603,
+      "grad_norm": 0.6337804794311523,
+      "learning_rate": 7.211945591236872e-06,
+      "loss": 0.7147,
+      "step": 12933
+    },
+    {
+      "epoch": 0.711871869668116,
+      "grad_norm": 0.7450474500656128,
+      "learning_rate": 7.211556840002718e-06,
+      "loss": 0.8516,
+      "step": 12934
+    },
+    {
+      "epoch": 0.7119269084704717,
+      "grad_norm": 0.7786532640457153,
+      "learning_rate": 7.2111680721470965e-06,
+      "loss": 0.837,
+      "step": 12935
+    },
+    {
+      "epoch": 0.7119819472728274,
+      "grad_norm": 0.666020393371582,
+      "learning_rate": 7.210779287672927e-06,
+      "loss": 0.7646,
+      "step": 12936
+    },
+    {
+      "epoch": 0.712036986075183,
+      "grad_norm": 0.622648298740387,
+      "learning_rate": 7.210390486583132e-06,
+      "loss": 0.7102,
+      "step": 12937
+    },
+    {
+      "epoch": 0.7120920248775386,
+      "grad_norm": 0.7175952792167664,
+      "learning_rate": 7.210001668880634e-06,
+      "loss": 0.7043,
+      "step": 12938
+    },
+    {
+      "epoch": 0.7121470636798943,
+      "grad_norm": 0.8019681572914124,
+      "learning_rate": 7.209612834568353e-06,
+      "loss": 0.8166,
+      "step": 12939
+    },
+    {
+      "epoch": 0.71220210248225,
+      "grad_norm": 0.804457426071167,
+      "learning_rate": 7.209223983649216e-06,
+      "loss": 0.7182,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7122571412846056,
+      "grad_norm": 0.7261730432510376,
+      "learning_rate": 7.208835116126143e-06,
+      "loss": 0.6634,
+      "step": 12941
+    },
+    {
+      "epoch": 0.7123121800869613,
+      "grad_norm": 0.7461307644844055,
+      "learning_rate": 7.208446232002055e-06,
+      "loss": 0.709,
+      "step": 12942
+    },
+    {
+      "epoch": 0.712367218889317,
+      "grad_norm": 0.6730383634567261,
+      "learning_rate": 7.208057331279877e-06,
+      "loss": 0.7111,
+      "step": 12943
+    },
+    {
+      "epoch": 0.7124222576916727,
+      "grad_norm": 0.829530656337738,
+      "learning_rate": 7.207668413962531e-06,
+      "loss": 0.729,
+      "step": 12944
+    },
+    {
+      "epoch": 0.7124772964940282,
+      "grad_norm": 0.5997991561889648,
+      "learning_rate": 7.20727948005294e-06,
+      "loss": 0.6385,
+      "step": 12945
+    },
+    {
+      "epoch": 0.7125323352963839,
+      "grad_norm": 0.9590086936950684,
+      "learning_rate": 7.206890529554027e-06,
+      "loss": 0.7217,
+      "step": 12946
+    },
+    {
+      "epoch": 0.7125873740987396,
+      "grad_norm": 0.7818330526351929,
+      "learning_rate": 7.206501562468717e-06,
+      "loss": 0.7276,
+      "step": 12947
+    },
+    {
+      "epoch": 0.7126424129010953,
+      "grad_norm": 0.6033679842948914,
+      "learning_rate": 7.206112578799931e-06,
+      "loss": 0.5935,
+      "step": 12948
+    },
+    {
+      "epoch": 0.7126974517034509,
+      "grad_norm": 0.7431650757789612,
+      "learning_rate": 7.205723578550593e-06,
+      "loss": 0.8649,
+      "step": 12949
+    },
+    {
+      "epoch": 0.7127524905058066,
+      "grad_norm": 0.7026848793029785,
+      "learning_rate": 7.205334561723627e-06,
+      "loss": 0.7484,
+      "step": 12950
+    },
+    {
+      "epoch": 0.7128075293081623,
+      "grad_norm": 0.6328058242797852,
+      "learning_rate": 7.204945528321956e-06,
+      "loss": 0.6994,
+      "step": 12951
+    },
+    {
+      "epoch": 0.712862568110518,
+      "grad_norm": 0.6806536912918091,
+      "learning_rate": 7.204556478348507e-06,
+      "loss": 0.7461,
+      "step": 12952
+    },
+    {
+      "epoch": 0.7129176069128735,
+      "grad_norm": 0.6822162866592407,
+      "learning_rate": 7.2041674118062e-06,
+      "loss": 0.7947,
+      "step": 12953
+    },
+    {
+      "epoch": 0.7129726457152292,
+      "grad_norm": 0.7283263802528381,
+      "learning_rate": 7.203778328697962e-06,
+      "loss": 0.7559,
+      "step": 12954
+    },
+    {
+      "epoch": 0.7130276845175849,
+      "grad_norm": 0.663564920425415,
+      "learning_rate": 7.203389229026714e-06,
+      "loss": 0.6898,
+      "step": 12955
+    },
+    {
+      "epoch": 0.7130827233199406,
+      "grad_norm": 0.7218708395957947,
+      "learning_rate": 7.203000112795383e-06,
+      "loss": 0.8095,
+      "step": 12956
+    },
+    {
+      "epoch": 0.7131377621222962,
+      "grad_norm": 0.6931518912315369,
+      "learning_rate": 7.202610980006893e-06,
+      "loss": 0.7591,
+      "step": 12957
+    },
+    {
+      "epoch": 0.7131928009246519,
+      "grad_norm": 0.6982918381690979,
+      "learning_rate": 7.2022218306641704e-06,
+      "loss": 0.7651,
+      "step": 12958
+    },
+    {
+      "epoch": 0.7132478397270076,
+      "grad_norm": 0.8033974170684814,
+      "learning_rate": 7.201832664770135e-06,
+      "loss": 0.8857,
+      "step": 12959
+    },
+    {
+      "epoch": 0.7133028785293632,
+      "grad_norm": 0.6625493764877319,
+      "learning_rate": 7.201443482327717e-06,
+      "loss": 0.752,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7133579173317188,
+      "grad_norm": 0.8149683475494385,
+      "learning_rate": 7.201054283339838e-06,
+      "loss": 0.8528,
+      "step": 12961
+    },
+    {
+      "epoch": 0.7134129561340745,
+      "grad_norm": 0.7894958257675171,
+      "learning_rate": 7.200665067809425e-06,
+      "loss": 0.8554,
+      "step": 12962
+    },
+    {
+      "epoch": 0.7134679949364302,
+      "grad_norm": 0.7613523602485657,
+      "learning_rate": 7.200275835739401e-06,
+      "loss": 0.7435,
+      "step": 12963
+    },
+    {
+      "epoch": 0.7135230337387859,
+      "grad_norm": 0.665985643863678,
+      "learning_rate": 7.199886587132693e-06,
+      "loss": 0.7072,
+      "step": 12964
+    },
+    {
+      "epoch": 0.7135780725411415,
+      "grad_norm": 0.7523592710494995,
+      "learning_rate": 7.199497321992227e-06,
+      "loss": 0.7945,
+      "step": 12965
+    },
+    {
+      "epoch": 0.7136331113434972,
+      "grad_norm": 0.8894450664520264,
+      "learning_rate": 7.199108040320928e-06,
+      "loss": 0.7885,
+      "step": 12966
+    },
+    {
+      "epoch": 0.7136881501458529,
+      "grad_norm": 0.639108419418335,
+      "learning_rate": 7.198718742121722e-06,
+      "loss": 0.6975,
+      "step": 12967
+    },
+    {
+      "epoch": 0.7137431889482085,
+      "grad_norm": 0.670013964176178,
+      "learning_rate": 7.198329427397532e-06,
+      "loss": 0.7441,
+      "step": 12968
+    },
+    {
+      "epoch": 0.7137982277505641,
+      "grad_norm": 0.7695425748825073,
+      "learning_rate": 7.197940096151289e-06,
+      "loss": 0.7616,
+      "step": 12969
+    },
+    {
+      "epoch": 0.7138532665529198,
+      "grad_norm": 0.9098057150840759,
+      "learning_rate": 7.197550748385917e-06,
+      "loss": 0.9028,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7139083053552755,
+      "grad_norm": 0.7677769660949707,
+      "learning_rate": 7.197161384104341e-06,
+      "loss": 0.7926,
+      "step": 12971
+    },
+    {
+      "epoch": 0.7139633441576311,
+      "grad_norm": 0.7020674347877502,
+      "learning_rate": 7.196772003309487e-06,
+      "loss": 0.7248,
+      "step": 12972
+    },
+    {
+      "epoch": 0.7140183829599868,
+      "grad_norm": 0.6616366505622864,
+      "learning_rate": 7.196382606004283e-06,
+      "loss": 0.7137,
+      "step": 12973
+    },
+    {
+      "epoch": 0.7140734217623425,
+      "grad_norm": 0.7174738645553589,
+      "learning_rate": 7.195993192191656e-06,
+      "loss": 0.8167,
+      "step": 12974
+    },
+    {
+      "epoch": 0.7141284605646981,
+      "grad_norm": 0.6672176122665405,
+      "learning_rate": 7.1956037618745325e-06,
+      "loss": 0.6516,
+      "step": 12975
+    },
+    {
+      "epoch": 0.7141834993670537,
+      "grad_norm": 0.714790403842926,
+      "learning_rate": 7.195214315055837e-06,
+      "loss": 0.865,
+      "step": 12976
+    },
+    {
+      "epoch": 0.7142385381694094,
+      "grad_norm": 0.6637690663337708,
+      "learning_rate": 7.1948248517385e-06,
+      "loss": 0.7328,
+      "step": 12977
+    },
+    {
+      "epoch": 0.7142935769717651,
+      "grad_norm": 0.8998367786407471,
+      "learning_rate": 7.194435371925446e-06,
+      "loss": 0.7097,
+      "step": 12978
+    },
+    {
+      "epoch": 0.7143486157741208,
+      "grad_norm": 0.7472445964813232,
+      "learning_rate": 7.194045875619604e-06,
+      "loss": 0.7556,
+      "step": 12979
+    },
+    {
+      "epoch": 0.7144036545764764,
+      "grad_norm": 0.7897135019302368,
+      "learning_rate": 7.1936563628239e-06,
+      "loss": 0.8728,
+      "step": 12980
+    },
+    {
+      "epoch": 0.714458693378832,
+      "grad_norm": 0.6520817279815674,
+      "learning_rate": 7.193266833541261e-06,
+      "loss": 0.6824,
+      "step": 12981
+    },
+    {
+      "epoch": 0.7145137321811877,
+      "grad_norm": 0.833849310874939,
+      "learning_rate": 7.192877287774618e-06,
+      "loss": 0.8877,
+      "step": 12982
+    },
+    {
+      "epoch": 0.7145687709835434,
+      "grad_norm": 0.7105151414871216,
+      "learning_rate": 7.192487725526896e-06,
+      "loss": 0.7799,
+      "step": 12983
+    },
+    {
+      "epoch": 0.714623809785899,
+      "grad_norm": 0.7515869140625,
+      "learning_rate": 7.192098146801021e-06,
+      "loss": 0.7012,
+      "step": 12984
+    },
+    {
+      "epoch": 0.7146788485882547,
+      "grad_norm": 0.7447199821472168,
+      "learning_rate": 7.191708551599923e-06,
+      "loss": 0.7545,
+      "step": 12985
+    },
+    {
+      "epoch": 0.7147338873906104,
+      "grad_norm": 0.8502823114395142,
+      "learning_rate": 7.191318939926532e-06,
+      "loss": 0.7232,
+      "step": 12986
+    },
+    {
+      "epoch": 0.7147889261929661,
+      "grad_norm": 0.7193031907081604,
+      "learning_rate": 7.190929311783774e-06,
+      "loss": 0.762,
+      "step": 12987
+    },
+    {
+      "epoch": 0.7148439649953217,
+      "grad_norm": 0.8479939699172974,
+      "learning_rate": 7.190539667174576e-06,
+      "loss": 0.7238,
+      "step": 12988
+    },
+    {
+      "epoch": 0.7148990037976773,
+      "grad_norm": 0.8313719630241394,
+      "learning_rate": 7.1901500061018704e-06,
+      "loss": 0.8145,
+      "step": 12989
+    },
+    {
+      "epoch": 0.714954042600033,
+      "grad_norm": 0.7019978165626526,
+      "learning_rate": 7.189760328568584e-06,
+      "loss": 0.6461,
+      "step": 12990
+    },
+    {
+      "epoch": 0.7150090814023887,
+      "grad_norm": 0.897280216217041,
+      "learning_rate": 7.1893706345776436e-06,
+      "loss": 0.818,
+      "step": 12991
+    },
+    {
+      "epoch": 0.7150641202047443,
+      "grad_norm": 0.7495617866516113,
+      "learning_rate": 7.1889809241319795e-06,
+      "loss": 0.7533,
+      "step": 12992
+    },
+    {
+      "epoch": 0.7151191590071,
+      "grad_norm": 0.733496904373169,
+      "learning_rate": 7.188591197234522e-06,
+      "loss": 0.7405,
+      "step": 12993
+    },
+    {
+      "epoch": 0.7151741978094557,
+      "grad_norm": 0.8873284459114075,
+      "learning_rate": 7.1882014538882e-06,
+      "loss": 0.7525,
+      "step": 12994
+    },
+    {
+      "epoch": 0.7152292366118114,
+      "grad_norm": 0.6693230271339417,
+      "learning_rate": 7.187811694095939e-06,
+      "loss": 0.7509,
+      "step": 12995
+    },
+    {
+      "epoch": 0.715284275414167,
+      "grad_norm": 0.8513357043266296,
+      "learning_rate": 7.187421917860671e-06,
+      "loss": 0.8111,
+      "step": 12996
+    },
+    {
+      "epoch": 0.7153393142165226,
+      "grad_norm": 0.6986566185951233,
+      "learning_rate": 7.187032125185326e-06,
+      "loss": 0.8013,
+      "step": 12997
+    },
+    {
+      "epoch": 0.7153943530188783,
+      "grad_norm": 0.7062557339668274,
+      "learning_rate": 7.1866423160728335e-06,
+      "loss": 0.7266,
+      "step": 12998
+    },
+    {
+      "epoch": 0.715449391821234,
+      "grad_norm": 0.6329573392868042,
+      "learning_rate": 7.186252490526122e-06,
+      "loss": 0.6753,
+      "step": 12999
+    },
+    {
+      "epoch": 0.7155044306235896,
+      "grad_norm": 0.6740719079971313,
+      "learning_rate": 7.185862648548122e-06,
+      "loss": 0.7197,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7155594694259453,
+      "grad_norm": 0.7911732196807861,
+      "learning_rate": 7.185472790141764e-06,
+      "loss": 0.6939,
+      "step": 13001
+    },
+    {
+      "epoch": 0.715614508228301,
+      "grad_norm": 0.7368680238723755,
+      "learning_rate": 7.185082915309978e-06,
+      "loss": 0.6919,
+      "step": 13002
+    },
+    {
+      "epoch": 0.7156695470306567,
+      "grad_norm": 0.6374472975730896,
+      "learning_rate": 7.1846930240556925e-06,
+      "loss": 0.6645,
+      "step": 13003
+    },
+    {
+      "epoch": 0.7157245858330122,
+      "grad_norm": 0.6727073192596436,
+      "learning_rate": 7.184303116381839e-06,
+      "loss": 0.5995,
+      "step": 13004
+    },
+    {
+      "epoch": 0.7157796246353679,
+      "grad_norm": 0.6122208833694458,
+      "learning_rate": 7.183913192291348e-06,
+      "loss": 0.6755,
+      "step": 13005
+    },
+    {
+      "epoch": 0.7158346634377236,
+      "grad_norm": 0.7095892429351807,
+      "learning_rate": 7.1835232517871525e-06,
+      "loss": 0.8009,
+      "step": 13006
+    },
+    {
+      "epoch": 0.7158897022400793,
+      "grad_norm": 0.6828192472457886,
+      "learning_rate": 7.1831332948721786e-06,
+      "loss": 0.7755,
+      "step": 13007
+    },
+    {
+      "epoch": 0.7159447410424349,
+      "grad_norm": 0.7997334003448486,
+      "learning_rate": 7.182743321549359e-06,
+      "loss": 0.7259,
+      "step": 13008
+    },
+    {
+      "epoch": 0.7159997798447906,
+      "grad_norm": 0.7431252002716064,
+      "learning_rate": 7.182353331821626e-06,
+      "loss": 0.7765,
+      "step": 13009
+    },
+    {
+      "epoch": 0.7160548186471463,
+      "grad_norm": 0.7202625870704651,
+      "learning_rate": 7.181963325691907e-06,
+      "loss": 0.7638,
+      "step": 13010
+    },
+    {
+      "epoch": 0.716109857449502,
+      "grad_norm": 0.7617568373680115,
+      "learning_rate": 7.181573303163139e-06,
+      "loss": 0.825,
+      "step": 13011
+    },
+    {
+      "epoch": 0.7161648962518575,
+      "grad_norm": 0.7382665276527405,
+      "learning_rate": 7.181183264238247e-06,
+      "loss": 0.8005,
+      "step": 13012
+    },
+    {
+      "epoch": 0.7162199350542132,
+      "grad_norm": 0.7782611846923828,
+      "learning_rate": 7.180793208920167e-06,
+      "loss": 0.7044,
+      "step": 13013
+    },
+    {
+      "epoch": 0.7162749738565689,
+      "grad_norm": 0.7020898461341858,
+      "learning_rate": 7.18040313721183e-06,
+      "loss": 0.8059,
+      "step": 13014
+    },
+    {
+      "epoch": 0.7163300126589245,
+      "grad_norm": 1.2005099058151245,
+      "learning_rate": 7.1800130491161656e-06,
+      "loss": 0.6663,
+      "step": 13015
+    },
+    {
+      "epoch": 0.7163850514612802,
+      "grad_norm": 0.6663569211959839,
+      "learning_rate": 7.1796229446361066e-06,
+      "loss": 0.7046,
+      "step": 13016
+    },
+    {
+      "epoch": 0.7164400902636359,
+      "grad_norm": 0.7010110020637512,
+      "learning_rate": 7.1792328237745845e-06,
+      "loss": 0.6433,
+      "step": 13017
+    },
+    {
+      "epoch": 0.7164951290659916,
+      "grad_norm": 0.6447514891624451,
+      "learning_rate": 7.178842686534534e-06,
+      "loss": 0.7794,
+      "step": 13018
+    },
+    {
+      "epoch": 0.7165501678683471,
+      "grad_norm": 0.6813021302223206,
+      "learning_rate": 7.1784525329188835e-06,
+      "loss": 0.7413,
+      "step": 13019
+    },
+    {
+      "epoch": 0.7166052066707028,
+      "grad_norm": 0.6894733905792236,
+      "learning_rate": 7.178062362930567e-06,
+      "loss": 0.7896,
+      "step": 13020
+    },
+    {
+      "epoch": 0.7166602454730585,
+      "grad_norm": 0.6717034578323364,
+      "learning_rate": 7.177672176572517e-06,
+      "loss": 0.7599,
+      "step": 13021
+    },
+    {
+      "epoch": 0.7167152842754142,
+      "grad_norm": 0.7861666083335876,
+      "learning_rate": 7.177281973847665e-06,
+      "loss": 0.9068,
+      "step": 13022
+    },
+    {
+      "epoch": 0.7167703230777698,
+      "grad_norm": 0.6784214973449707,
+      "learning_rate": 7.176891754758946e-06,
+      "loss": 0.8319,
+      "step": 13023
+    },
+    {
+      "epoch": 0.7168253618801255,
+      "grad_norm": 0.7053580284118652,
+      "learning_rate": 7.176501519309289e-06,
+      "loss": 0.8085,
+      "step": 13024
+    },
+    {
+      "epoch": 0.7168804006824812,
+      "grad_norm": 0.9643208980560303,
+      "learning_rate": 7.176111267501631e-06,
+      "loss": 0.7799,
+      "step": 13025
+    },
+    {
+      "epoch": 0.7169354394848368,
+      "grad_norm": 0.8921111822128296,
+      "learning_rate": 7.175720999338902e-06,
+      "loss": 0.6465,
+      "step": 13026
+    },
+    {
+      "epoch": 0.7169904782871924,
+      "grad_norm": 0.7356166839599609,
+      "learning_rate": 7.1753307148240385e-06,
+      "loss": 0.7862,
+      "step": 13027
+    },
+    {
+      "epoch": 0.7170455170895481,
+      "grad_norm": 0.6906836628913879,
+      "learning_rate": 7.174940413959968e-06,
+      "loss": 0.7341,
+      "step": 13028
+    },
+    {
+      "epoch": 0.7171005558919038,
+      "grad_norm": 0.6229632496833801,
+      "learning_rate": 7.174550096749632e-06,
+      "loss": 0.721,
+      "step": 13029
+    },
+    {
+      "epoch": 0.7171555946942595,
+      "grad_norm": 0.6832499504089355,
+      "learning_rate": 7.174159763195958e-06,
+      "loss": 0.6733,
+      "step": 13030
+    },
+    {
+      "epoch": 0.7172106334966151,
+      "grad_norm": 0.8304060697555542,
+      "learning_rate": 7.1737694133018806e-06,
+      "loss": 0.7732,
+      "step": 13031
+    },
+    {
+      "epoch": 0.7172656722989708,
+      "grad_norm": 0.6813186407089233,
+      "learning_rate": 7.173379047070333e-06,
+      "loss": 0.7742,
+      "step": 13032
+    },
+    {
+      "epoch": 0.7173207111013264,
+      "grad_norm": 0.6671963930130005,
+      "learning_rate": 7.172988664504252e-06,
+      "loss": 0.6516,
+      "step": 13033
+    },
+    {
+      "epoch": 0.7173757499036821,
+      "grad_norm": 0.661108136177063,
+      "learning_rate": 7.172598265606569e-06,
+      "loss": 0.7361,
+      "step": 13034
+    },
+    {
+      "epoch": 0.7174307887060377,
+      "grad_norm": 0.7097620368003845,
+      "learning_rate": 7.1722078503802196e-06,
+      "loss": 0.8142,
+      "step": 13035
+    },
+    {
+      "epoch": 0.7174858275083934,
+      "grad_norm": 0.7663383483886719,
+      "learning_rate": 7.1718174188281365e-06,
+      "loss": 0.8149,
+      "step": 13036
+    },
+    {
+      "epoch": 0.7175408663107491,
+      "grad_norm": 0.7142401337623596,
+      "learning_rate": 7.171426970953256e-06,
+      "loss": 0.7539,
+      "step": 13037
+    },
+    {
+      "epoch": 0.7175959051131048,
+      "grad_norm": 0.667346715927124,
+      "learning_rate": 7.171036506758512e-06,
+      "loss": 0.7517,
+      "step": 13038
+    },
+    {
+      "epoch": 0.7176509439154604,
+      "grad_norm": 0.5933231711387634,
+      "learning_rate": 7.170646026246838e-06,
+      "loss": 0.6852,
+      "step": 13039
+    },
+    {
+      "epoch": 0.717705982717816,
+      "grad_norm": 0.730015218257904,
+      "learning_rate": 7.170255529421168e-06,
+      "loss": 0.7316,
+      "step": 13040
+    },
+    {
+      "epoch": 0.7177610215201717,
+      "grad_norm": 0.6146146059036255,
+      "learning_rate": 7.169865016284442e-06,
+      "loss": 0.6715,
+      "step": 13041
+    },
+    {
+      "epoch": 0.7178160603225274,
+      "grad_norm": 0.694131076335907,
+      "learning_rate": 7.16947448683959e-06,
+      "loss": 0.7944,
+      "step": 13042
+    },
+    {
+      "epoch": 0.717871099124883,
+      "grad_norm": 0.6736807823181152,
+      "learning_rate": 7.169083941089547e-06,
+      "loss": 0.7922,
+      "step": 13043
+    },
+    {
+      "epoch": 0.7179261379272387,
+      "grad_norm": 0.6748425364494324,
+      "learning_rate": 7.16869337903725e-06,
+      "loss": 0.6738,
+      "step": 13044
+    },
+    {
+      "epoch": 0.7179811767295944,
+      "grad_norm": 0.6807510852813721,
+      "learning_rate": 7.168302800685635e-06,
+      "loss": 0.7291,
+      "step": 13045
+    },
+    {
+      "epoch": 0.7180362155319501,
+      "grad_norm": 0.6613160371780396,
+      "learning_rate": 7.167912206037637e-06,
+      "loss": 0.6839,
+      "step": 13046
+    },
+    {
+      "epoch": 0.7180912543343057,
+      "grad_norm": 0.7184692621231079,
+      "learning_rate": 7.16752159509619e-06,
+      "loss": 0.6748,
+      "step": 13047
+    },
+    {
+      "epoch": 0.7181462931366613,
+      "grad_norm": 0.6938989758491516,
+      "learning_rate": 7.167130967864231e-06,
+      "loss": 0.7926,
+      "step": 13048
+    },
+    {
+      "epoch": 0.718201331939017,
+      "grad_norm": 0.6871020793914795,
+      "learning_rate": 7.166740324344696e-06,
+      "loss": 0.8229,
+      "step": 13049
+    },
+    {
+      "epoch": 0.7182563707413727,
+      "grad_norm": 0.8003624081611633,
+      "learning_rate": 7.166349664540521e-06,
+      "loss": 0.8488,
+      "step": 13050
+    },
+    {
+      "epoch": 0.7183114095437283,
+      "grad_norm": 0.7309357523918152,
+      "learning_rate": 7.165958988454642e-06,
+      "loss": 0.7442,
+      "step": 13051
+    },
+    {
+      "epoch": 0.718366448346084,
+      "grad_norm": 0.7462141513824463,
+      "learning_rate": 7.165568296089993e-06,
+      "loss": 0.8014,
+      "step": 13052
+    },
+    {
+      "epoch": 0.7184214871484397,
+      "grad_norm": 0.8335661292076111,
+      "learning_rate": 7.165177587449516e-06,
+      "loss": 0.6773,
+      "step": 13053
+    },
+    {
+      "epoch": 0.7184765259507954,
+      "grad_norm": 0.6996884346008301,
+      "learning_rate": 7.164786862536142e-06,
+      "loss": 0.7491,
+      "step": 13054
+    },
+    {
+      "epoch": 0.7185315647531509,
+      "grad_norm": 0.7203043103218079,
+      "learning_rate": 7.164396121352809e-06,
+      "loss": 0.7196,
+      "step": 13055
+    },
+    {
+      "epoch": 0.7185866035555066,
+      "grad_norm": 0.7109461426734924,
+      "learning_rate": 7.164005363902453e-06,
+      "loss": 0.7336,
+      "step": 13056
+    },
+    {
+      "epoch": 0.7186416423578623,
+      "grad_norm": 0.7057282328605652,
+      "learning_rate": 7.1636145901880135e-06,
+      "loss": 0.734,
+      "step": 13057
+    },
+    {
+      "epoch": 0.7186966811602179,
+      "grad_norm": 0.7288782000541687,
+      "learning_rate": 7.163223800212427e-06,
+      "loss": 0.8141,
+      "step": 13058
+    },
+    {
+      "epoch": 0.7187517199625736,
+      "grad_norm": 0.6812320947647095,
+      "learning_rate": 7.162832993978628e-06,
+      "loss": 0.7525,
+      "step": 13059
+    },
+    {
+      "epoch": 0.7188067587649293,
+      "grad_norm": 0.6782627105712891,
+      "learning_rate": 7.1624421714895546e-06,
+      "loss": 0.7647,
+      "step": 13060
+    },
+    {
+      "epoch": 0.718861797567285,
+      "grad_norm": 0.7361965775489807,
+      "learning_rate": 7.162051332748146e-06,
+      "loss": 0.7774,
+      "step": 13061
+    },
+    {
+      "epoch": 0.7189168363696405,
+      "grad_norm": 0.68894362449646,
+      "learning_rate": 7.161660477757337e-06,
+      "loss": 0.767,
+      "step": 13062
+    },
+    {
+      "epoch": 0.7189718751719962,
+      "grad_norm": 0.6440854668617249,
+      "learning_rate": 7.161269606520067e-06,
+      "loss": 0.7062,
+      "step": 13063
+    },
+    {
+      "epoch": 0.7190269139743519,
+      "grad_norm": 0.8411546945571899,
+      "learning_rate": 7.160878719039273e-06,
+      "loss": 0.728,
+      "step": 13064
+    },
+    {
+      "epoch": 0.7190819527767076,
+      "grad_norm": 0.6895145177841187,
+      "learning_rate": 7.160487815317895e-06,
+      "loss": 0.6667,
+      "step": 13065
+    },
+    {
+      "epoch": 0.7191369915790632,
+      "grad_norm": 0.6943626403808594,
+      "learning_rate": 7.160096895358866e-06,
+      "loss": 0.7579,
+      "step": 13066
+    },
+    {
+      "epoch": 0.7191920303814189,
+      "grad_norm": 0.7940205335617065,
+      "learning_rate": 7.1597059591651294e-06,
+      "loss": 0.7286,
+      "step": 13067
+    },
+    {
+      "epoch": 0.7192470691837746,
+      "grad_norm": 0.7350896000862122,
+      "learning_rate": 7.159315006739619e-06,
+      "loss": 0.7174,
+      "step": 13068
+    },
+    {
+      "epoch": 0.7193021079861303,
+      "grad_norm": 0.7663372159004211,
+      "learning_rate": 7.158924038085275e-06,
+      "loss": 0.7871,
+      "step": 13069
+    },
+    {
+      "epoch": 0.7193571467884858,
+      "grad_norm": 0.7368965744972229,
+      "learning_rate": 7.1585330532050375e-06,
+      "loss": 0.7356,
+      "step": 13070
+    },
+    {
+      "epoch": 0.7194121855908415,
+      "grad_norm": 0.7345212697982788,
+      "learning_rate": 7.158142052101843e-06,
+      "loss": 0.7784,
+      "step": 13071
+    },
+    {
+      "epoch": 0.7194672243931972,
+      "grad_norm": 0.7847188711166382,
+      "learning_rate": 7.157751034778629e-06,
+      "loss": 0.7899,
+      "step": 13072
+    },
+    {
+      "epoch": 0.7195222631955529,
+      "grad_norm": 0.757514476776123,
+      "learning_rate": 7.157360001238337e-06,
+      "loss": 0.8899,
+      "step": 13073
+    },
+    {
+      "epoch": 0.7195773019979085,
+      "grad_norm": 0.73405522108078,
+      "learning_rate": 7.156968951483905e-06,
+      "loss": 0.7283,
+      "step": 13074
+    },
+    {
+      "epoch": 0.7196323408002642,
+      "grad_norm": 0.7950206398963928,
+      "learning_rate": 7.156577885518271e-06,
+      "loss": 0.7338,
+      "step": 13075
+    },
+    {
+      "epoch": 0.7196873796026199,
+      "grad_norm": 0.8082411289215088,
+      "learning_rate": 7.156186803344374e-06,
+      "loss": 0.711,
+      "step": 13076
+    },
+    {
+      "epoch": 0.7197424184049755,
+      "grad_norm": 0.6868693828582764,
+      "learning_rate": 7.1557957049651574e-06,
+      "loss": 0.7583,
+      "step": 13077
+    },
+    {
+      "epoch": 0.7197974572073311,
+      "grad_norm": 0.7226251363754272,
+      "learning_rate": 7.155404590383554e-06,
+      "loss": 0.746,
+      "step": 13078
+    },
+    {
+      "epoch": 0.7198524960096868,
+      "grad_norm": 0.7437220811843872,
+      "learning_rate": 7.155013459602509e-06,
+      "loss": 0.6884,
+      "step": 13079
+    },
+    {
+      "epoch": 0.7199075348120425,
+      "grad_norm": 0.7486164569854736,
+      "learning_rate": 7.154622312624958e-06,
+      "loss": 0.6968,
+      "step": 13080
+    },
+    {
+      "epoch": 0.7199625736143982,
+      "grad_norm": 0.7709106802940369,
+      "learning_rate": 7.154231149453843e-06,
+      "loss": 0.838,
+      "step": 13081
+    },
+    {
+      "epoch": 0.7200176124167538,
+      "grad_norm": 0.6962981224060059,
+      "learning_rate": 7.153839970092104e-06,
+      "loss": 0.7186,
+      "step": 13082
+    },
+    {
+      "epoch": 0.7200726512191095,
+      "grad_norm": 0.8195380568504333,
+      "learning_rate": 7.15344877454268e-06,
+      "loss": 0.7949,
+      "step": 13083
+    },
+    {
+      "epoch": 0.7201276900214651,
+      "grad_norm": 0.735285758972168,
+      "learning_rate": 7.15305756280851e-06,
+      "loss": 0.7477,
+      "step": 13084
+    },
+    {
+      "epoch": 0.7201827288238208,
+      "grad_norm": 0.6121101379394531,
+      "learning_rate": 7.1526663348925375e-06,
+      "loss": 0.6686,
+      "step": 13085
+    },
+    {
+      "epoch": 0.7202377676261764,
+      "grad_norm": 0.7204885482788086,
+      "learning_rate": 7.1522750907977e-06,
+      "loss": 0.8013,
+      "step": 13086
+    },
+    {
+      "epoch": 0.7202928064285321,
+      "grad_norm": 0.6808584332466125,
+      "learning_rate": 7.15188383052694e-06,
+      "loss": 0.7847,
+      "step": 13087
+    },
+    {
+      "epoch": 0.7203478452308878,
+      "grad_norm": 0.7049086093902588,
+      "learning_rate": 7.151492554083195e-06,
+      "loss": 0.7563,
+      "step": 13088
+    },
+    {
+      "epoch": 0.7204028840332435,
+      "grad_norm": 0.765708327293396,
+      "learning_rate": 7.151101261469411e-06,
+      "loss": 0.7648,
+      "step": 13089
+    },
+    {
+      "epoch": 0.7204579228355991,
+      "grad_norm": 0.6810007095336914,
+      "learning_rate": 7.150709952688525e-06,
+      "loss": 0.731,
+      "step": 13090
+    },
+    {
+      "epoch": 0.7205129616379548,
+      "grad_norm": 0.7242745757102966,
+      "learning_rate": 7.150318627743478e-06,
+      "loss": 0.8027,
+      "step": 13091
+    },
+    {
+      "epoch": 0.7205680004403104,
+      "grad_norm": 0.7452220916748047,
+      "learning_rate": 7.14992728663721e-06,
+      "loss": 0.7848,
+      "step": 13092
+    },
+    {
+      "epoch": 0.7206230392426661,
+      "grad_norm": 0.6333943605422974,
+      "learning_rate": 7.149535929372667e-06,
+      "loss": 0.7105,
+      "step": 13093
+    },
+    {
+      "epoch": 0.7206780780450217,
+      "grad_norm": 0.7565333247184753,
+      "learning_rate": 7.149144555952785e-06,
+      "loss": 0.8006,
+      "step": 13094
+    },
+    {
+      "epoch": 0.7207331168473774,
+      "grad_norm": 0.7703632712364197,
+      "learning_rate": 7.14875316638051e-06,
+      "loss": 0.7323,
+      "step": 13095
+    },
+    {
+      "epoch": 0.7207881556497331,
+      "grad_norm": 0.6275011301040649,
+      "learning_rate": 7.148361760658779e-06,
+      "loss": 0.6817,
+      "step": 13096
+    },
+    {
+      "epoch": 0.7208431944520888,
+      "grad_norm": 0.7363598942756653,
+      "learning_rate": 7.147970338790537e-06,
+      "loss": 0.7641,
+      "step": 13097
+    },
+    {
+      "epoch": 0.7208982332544444,
+      "grad_norm": 0.6284294724464417,
+      "learning_rate": 7.147578900778727e-06,
+      "loss": 0.7117,
+      "step": 13098
+    },
+    {
+      "epoch": 0.7209532720568,
+      "grad_norm": 0.7878503203392029,
+      "learning_rate": 7.147187446626287e-06,
+      "loss": 0.8184,
+      "step": 13099
+    },
+    {
+      "epoch": 0.7210083108591557,
+      "grad_norm": 0.6973691582679749,
+      "learning_rate": 7.146795976336159e-06,
+      "loss": 0.7815,
+      "step": 13100
+    },
+    {
+      "epoch": 0.7210633496615113,
+      "grad_norm": 0.7018479704856873,
+      "learning_rate": 7.146404489911291e-06,
+      "loss": 0.7305,
+      "step": 13101
+    },
+    {
+      "epoch": 0.721118388463867,
+      "grad_norm": 0.6903830766677856,
+      "learning_rate": 7.14601298735462e-06,
+      "loss": 0.7074,
+      "step": 13102
+    },
+    {
+      "epoch": 0.7211734272662227,
+      "grad_norm": 0.7612621188163757,
+      "learning_rate": 7.145621468669089e-06,
+      "loss": 0.8189,
+      "step": 13103
+    },
+    {
+      "epoch": 0.7212284660685784,
+      "grad_norm": 0.7256856560707092,
+      "learning_rate": 7.145229933857643e-06,
+      "loss": 0.5959,
+      "step": 13104
+    },
+    {
+      "epoch": 0.721283504870934,
+      "grad_norm": 0.6632323265075684,
+      "learning_rate": 7.1448383829232205e-06,
+      "loss": 0.7519,
+      "step": 13105
+    },
+    {
+      "epoch": 0.7213385436732896,
+      "grad_norm": 0.6320651769638062,
+      "learning_rate": 7.144446815868768e-06,
+      "loss": 0.7259,
+      "step": 13106
+    },
+    {
+      "epoch": 0.7213935824756453,
+      "grad_norm": 0.6883212924003601,
+      "learning_rate": 7.144055232697227e-06,
+      "loss": 0.7776,
+      "step": 13107
+    },
+    {
+      "epoch": 0.721448621278001,
+      "grad_norm": 0.7159759402275085,
+      "learning_rate": 7.1436636334115415e-06,
+      "loss": 0.6915,
+      "step": 13108
+    },
+    {
+      "epoch": 0.7215036600803566,
+      "grad_norm": 0.7108080983161926,
+      "learning_rate": 7.1432720180146535e-06,
+      "loss": 0.731,
+      "step": 13109
+    },
+    {
+      "epoch": 0.7215586988827123,
+      "grad_norm": 0.7765033841133118,
+      "learning_rate": 7.142880386509506e-06,
+      "loss": 0.6965,
+      "step": 13110
+    },
+    {
+      "epoch": 0.721613737685068,
+      "grad_norm": 0.7205119132995605,
+      "learning_rate": 7.142488738899045e-06,
+      "loss": 0.7262,
+      "step": 13111
+    },
+    {
+      "epoch": 0.7216687764874237,
+      "grad_norm": 0.6786921620368958,
+      "learning_rate": 7.142097075186212e-06,
+      "loss": 0.805,
+      "step": 13112
+    },
+    {
+      "epoch": 0.7217238152897792,
+      "grad_norm": 0.7947409152984619,
+      "learning_rate": 7.141705395373949e-06,
+      "loss": 0.7701,
+      "step": 13113
+    },
+    {
+      "epoch": 0.7217788540921349,
+      "grad_norm": 0.6672971844673157,
+      "learning_rate": 7.141313699465204e-06,
+      "loss": 0.7325,
+      "step": 13114
+    },
+    {
+      "epoch": 0.7218338928944906,
+      "grad_norm": 0.641765296459198,
+      "learning_rate": 7.140921987462916e-06,
+      "loss": 0.7902,
+      "step": 13115
+    },
+    {
+      "epoch": 0.7218889316968463,
+      "grad_norm": 0.6675699353218079,
+      "learning_rate": 7.140530259370032e-06,
+      "loss": 0.7422,
+      "step": 13116
+    },
+    {
+      "epoch": 0.7219439704992019,
+      "grad_norm": 0.6940729022026062,
+      "learning_rate": 7.140138515189495e-06,
+      "loss": 0.6978,
+      "step": 13117
+    },
+    {
+      "epoch": 0.7219990093015576,
+      "grad_norm": 0.6805779337882996,
+      "learning_rate": 7.1397467549242514e-06,
+      "loss": 0.7498,
+      "step": 13118
+    },
+    {
+      "epoch": 0.7220540481039133,
+      "grad_norm": 0.6231662631034851,
+      "learning_rate": 7.139354978577243e-06,
+      "loss": 0.7344,
+      "step": 13119
+    },
+    {
+      "epoch": 0.722109086906269,
+      "grad_norm": 0.6883575916290283,
+      "learning_rate": 7.138963186151416e-06,
+      "loss": 0.835,
+      "step": 13120
+    },
+    {
+      "epoch": 0.7221641257086245,
+      "grad_norm": 0.6902666687965393,
+      "learning_rate": 7.138571377649712e-06,
+      "loss": 0.7427,
+      "step": 13121
+    },
+    {
+      "epoch": 0.7222191645109802,
+      "grad_norm": 0.7156440019607544,
+      "learning_rate": 7.1381795530750805e-06,
+      "loss": 0.7661,
+      "step": 13122
+    },
+    {
+      "epoch": 0.7222742033133359,
+      "grad_norm": 0.6727150678634644,
+      "learning_rate": 7.137787712430464e-06,
+      "loss": 0.7872,
+      "step": 13123
+    },
+    {
+      "epoch": 0.7223292421156916,
+      "grad_norm": 0.6200405359268188,
+      "learning_rate": 7.137395855718806e-06,
+      "loss": 0.6108,
+      "step": 13124
+    },
+    {
+      "epoch": 0.7223842809180472,
+      "grad_norm": 0.6384756565093994,
+      "learning_rate": 7.137003982943054e-06,
+      "loss": 0.698,
+      "step": 13125
+    },
+    {
+      "epoch": 0.7224393197204029,
+      "grad_norm": 0.7212089896202087,
+      "learning_rate": 7.1366120941061515e-06,
+      "loss": 0.7679,
+      "step": 13126
+    },
+    {
+      "epoch": 0.7224943585227586,
+      "grad_norm": 0.737352192401886,
+      "learning_rate": 7.136220189211044e-06,
+      "loss": 0.8173,
+      "step": 13127
+    },
+    {
+      "epoch": 0.7225493973251143,
+      "grad_norm": 0.6244099736213684,
+      "learning_rate": 7.135828268260679e-06,
+      "loss": 0.7224,
+      "step": 13128
+    },
+    {
+      "epoch": 0.7226044361274698,
+      "grad_norm": 0.8191885948181152,
+      "learning_rate": 7.135436331257997e-06,
+      "loss": 0.8122,
+      "step": 13129
+    },
+    {
+      "epoch": 0.7226594749298255,
+      "grad_norm": 0.7069095373153687,
+      "learning_rate": 7.135044378205949e-06,
+      "loss": 0.7844,
+      "step": 13130
+    },
+    {
+      "epoch": 0.7227145137321812,
+      "grad_norm": 0.6094380021095276,
+      "learning_rate": 7.13465240910748e-06,
+      "loss": 0.7093,
+      "step": 13131
+    },
+    {
+      "epoch": 0.7227695525345369,
+      "grad_norm": 0.7075843811035156,
+      "learning_rate": 7.134260423965534e-06,
+      "loss": 0.8109,
+      "step": 13132
+    },
+    {
+      "epoch": 0.7228245913368925,
+      "grad_norm": 0.6684398651123047,
+      "learning_rate": 7.133868422783057e-06,
+      "loss": 0.7224,
+      "step": 13133
+    },
+    {
+      "epoch": 0.7228796301392482,
+      "grad_norm": 0.6574007272720337,
+      "learning_rate": 7.133476405562998e-06,
+      "loss": 0.6763,
+      "step": 13134
+    },
+    {
+      "epoch": 0.7229346689416039,
+      "grad_norm": 0.7124022841453552,
+      "learning_rate": 7.133084372308301e-06,
+      "loss": 0.8047,
+      "step": 13135
+    },
+    {
+      "epoch": 0.7229897077439595,
+      "grad_norm": 0.7035976648330688,
+      "learning_rate": 7.1326923230219124e-06,
+      "loss": 0.7544,
+      "step": 13136
+    },
+    {
+      "epoch": 0.7230447465463151,
+      "grad_norm": 0.7007604241371155,
+      "learning_rate": 7.132300257706779e-06,
+      "loss": 0.7584,
+      "step": 13137
+    },
+    {
+      "epoch": 0.7230997853486708,
+      "grad_norm": 0.6917324066162109,
+      "learning_rate": 7.131908176365848e-06,
+      "loss": 0.6846,
+      "step": 13138
+    },
+    {
+      "epoch": 0.7231548241510265,
+      "grad_norm": 0.6857448816299438,
+      "learning_rate": 7.1315160790020666e-06,
+      "loss": 0.8142,
+      "step": 13139
+    },
+    {
+      "epoch": 0.7232098629533822,
+      "grad_norm": 0.8381820321083069,
+      "learning_rate": 7.13112396561838e-06,
+      "loss": 0.8132,
+      "step": 13140
+    },
+    {
+      "epoch": 0.7232649017557378,
+      "grad_norm": 0.7024879455566406,
+      "learning_rate": 7.130731836217735e-06,
+      "loss": 0.7157,
+      "step": 13141
+    },
+    {
+      "epoch": 0.7233199405580935,
+      "grad_norm": 0.7313332557678223,
+      "learning_rate": 7.130339690803081e-06,
+      "loss": 0.7623,
+      "step": 13142
+    },
+    {
+      "epoch": 0.7233749793604491,
+      "grad_norm": 0.697536051273346,
+      "learning_rate": 7.129947529377364e-06,
+      "loss": 0.7202,
+      "step": 13143
+    },
+    {
+      "epoch": 0.7234300181628047,
+      "grad_norm": 0.6946722865104675,
+      "learning_rate": 7.129555351943533e-06,
+      "loss": 0.7862,
+      "step": 13144
+    },
+    {
+      "epoch": 0.7234850569651604,
+      "grad_norm": 0.6643924117088318,
+      "learning_rate": 7.129163158504532e-06,
+      "loss": 0.7055,
+      "step": 13145
+    },
+    {
+      "epoch": 0.7235400957675161,
+      "grad_norm": 0.7285693287849426,
+      "learning_rate": 7.1287709490633104e-06,
+      "loss": 0.6815,
+      "step": 13146
+    },
+    {
+      "epoch": 0.7235951345698718,
+      "grad_norm": 1.2701799869537354,
+      "learning_rate": 7.128378723622818e-06,
+      "loss": 0.8596,
+      "step": 13147
+    },
+    {
+      "epoch": 0.7236501733722274,
+      "grad_norm": 0.7067306041717529,
+      "learning_rate": 7.127986482186e-06,
+      "loss": 0.7077,
+      "step": 13148
+    },
+    {
+      "epoch": 0.7237052121745831,
+      "grad_norm": 0.8863486051559448,
+      "learning_rate": 7.127594224755805e-06,
+      "loss": 0.8961,
+      "step": 13149
+    },
+    {
+      "epoch": 0.7237602509769387,
+      "grad_norm": 0.7286190986633301,
+      "learning_rate": 7.127201951335182e-06,
+      "loss": 0.7941,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7238152897792944,
+      "grad_norm": 0.8756779432296753,
+      "learning_rate": 7.126809661927079e-06,
+      "loss": 0.7862,
+      "step": 13151
+    },
+    {
+      "epoch": 0.72387032858165,
+      "grad_norm": 0.7780876755714417,
+      "learning_rate": 7.126417356534443e-06,
+      "loss": 0.7095,
+      "step": 13152
+    },
+    {
+      "epoch": 0.7239253673840057,
+      "grad_norm": 0.6332812905311584,
+      "learning_rate": 7.1260250351602225e-06,
+      "loss": 0.7057,
+      "step": 13153
+    },
+    {
+      "epoch": 0.7239804061863614,
+      "grad_norm": 0.8350435495376587,
+      "learning_rate": 7.125632697807368e-06,
+      "loss": 0.7695,
+      "step": 13154
+    },
+    {
+      "epoch": 0.7240354449887171,
+      "grad_norm": 0.8306411504745483,
+      "learning_rate": 7.125240344478827e-06,
+      "loss": 0.6605,
+      "step": 13155
+    },
+    {
+      "epoch": 0.7240904837910727,
+      "grad_norm": 0.7495117783546448,
+      "learning_rate": 7.124847975177548e-06,
+      "loss": 0.8078,
+      "step": 13156
+    },
+    {
+      "epoch": 0.7241455225934283,
+      "grad_norm": 0.6481010317802429,
+      "learning_rate": 7.12445558990648e-06,
+      "loss": 0.8094,
+      "step": 13157
+    },
+    {
+      "epoch": 0.724200561395784,
+      "grad_norm": 0.7742613554000854,
+      "learning_rate": 7.124063188668573e-06,
+      "loss": 0.78,
+      "step": 13158
+    },
+    {
+      "epoch": 0.7242556001981397,
+      "grad_norm": 0.8394206762313843,
+      "learning_rate": 7.123670771466776e-06,
+      "loss": 0.8983,
+      "step": 13159
+    },
+    {
+      "epoch": 0.7243106390004953,
+      "grad_norm": 0.7196840047836304,
+      "learning_rate": 7.123278338304038e-06,
+      "loss": 0.7203,
+      "step": 13160
+    },
+    {
+      "epoch": 0.724365677802851,
+      "grad_norm": 0.5964440107345581,
+      "learning_rate": 7.122885889183309e-06,
+      "loss": 0.6251,
+      "step": 13161
+    },
+    {
+      "epoch": 0.7244207166052067,
+      "grad_norm": 0.7394048571586609,
+      "learning_rate": 7.1224934241075375e-06,
+      "loss": 0.7755,
+      "step": 13162
+    },
+    {
+      "epoch": 0.7244757554075624,
+      "grad_norm": 0.6427145004272461,
+      "learning_rate": 7.1221009430796724e-06,
+      "loss": 0.74,
+      "step": 13163
+    },
+    {
+      "epoch": 0.724530794209918,
+      "grad_norm": 0.7084387540817261,
+      "learning_rate": 7.121708446102667e-06,
+      "loss": 0.7464,
+      "step": 13164
+    },
+    {
+      "epoch": 0.7245858330122736,
+      "grad_norm": 0.6623230576515198,
+      "learning_rate": 7.121315933179466e-06,
+      "loss": 0.7237,
+      "step": 13165
+    },
+    {
+      "epoch": 0.7246408718146293,
+      "grad_norm": 0.9234243631362915,
+      "learning_rate": 7.120923404313024e-06,
+      "loss": 0.8238,
+      "step": 13166
+    },
+    {
+      "epoch": 0.724695910616985,
+      "grad_norm": 0.6458896994590759,
+      "learning_rate": 7.120530859506289e-06,
+      "loss": 0.8105,
+      "step": 13167
+    },
+    {
+      "epoch": 0.7247509494193406,
+      "grad_norm": 0.7160854935646057,
+      "learning_rate": 7.1201382987622115e-06,
+      "loss": 0.7954,
+      "step": 13168
+    },
+    {
+      "epoch": 0.7248059882216963,
+      "grad_norm": 0.6896069645881653,
+      "learning_rate": 7.119745722083742e-06,
+      "loss": 0.7281,
+      "step": 13169
+    },
+    {
+      "epoch": 0.724861027024052,
+      "grad_norm": 0.6609574556350708,
+      "learning_rate": 7.119353129473831e-06,
+      "loss": 0.7682,
+      "step": 13170
+    },
+    {
+      "epoch": 0.7249160658264077,
+      "grad_norm": 0.6477035880088806,
+      "learning_rate": 7.118960520935429e-06,
+      "loss": 0.8183,
+      "step": 13171
+    },
+    {
+      "epoch": 0.7249711046287632,
+      "grad_norm": 1.4488556385040283,
+      "learning_rate": 7.1185678964714885e-06,
+      "loss": 0.8321,
+      "step": 13172
+    },
+    {
+      "epoch": 0.7250261434311189,
+      "grad_norm": 0.8502382040023804,
+      "learning_rate": 7.118175256084958e-06,
+      "loss": 0.7881,
+      "step": 13173
+    },
+    {
+      "epoch": 0.7250811822334746,
+      "grad_norm": 0.6969912648200989,
+      "learning_rate": 7.117782599778788e-06,
+      "loss": 0.7598,
+      "step": 13174
+    },
+    {
+      "epoch": 0.7251362210358303,
+      "grad_norm": 0.7254889011383057,
+      "learning_rate": 7.117389927555933e-06,
+      "loss": 0.8473,
+      "step": 13175
+    },
+    {
+      "epoch": 0.7251912598381859,
+      "grad_norm": 0.9958444237709045,
+      "learning_rate": 7.116997239419341e-06,
+      "loss": 0.7558,
+      "step": 13176
+    },
+    {
+      "epoch": 0.7252462986405416,
+      "grad_norm": 0.6694881916046143,
+      "learning_rate": 7.116604535371963e-06,
+      "loss": 0.7072,
+      "step": 13177
+    },
+    {
+      "epoch": 0.7253013374428973,
+      "grad_norm": 1.0730634927749634,
+      "learning_rate": 7.116211815416754e-06,
+      "loss": 0.7607,
+      "step": 13178
+    },
+    {
+      "epoch": 0.725356376245253,
+      "grad_norm": 0.6770226359367371,
+      "learning_rate": 7.115819079556663e-06,
+      "loss": 0.7213,
+      "step": 13179
+    },
+    {
+      "epoch": 0.7254114150476085,
+      "grad_norm": 0.866215705871582,
+      "learning_rate": 7.115426327794642e-06,
+      "loss": 0.7273,
+      "step": 13180
+    },
+    {
+      "epoch": 0.7254664538499642,
+      "grad_norm": 0.7303730845451355,
+      "learning_rate": 7.115033560133642e-06,
+      "loss": 0.764,
+      "step": 13181
+    },
+    {
+      "epoch": 0.7255214926523199,
+      "grad_norm": 0.6900389194488525,
+      "learning_rate": 7.114640776576617e-06,
+      "loss": 0.6958,
+      "step": 13182
+    },
+    {
+      "epoch": 0.7255765314546756,
+      "grad_norm": 0.7255710959434509,
+      "learning_rate": 7.114247977126518e-06,
+      "loss": 0.6507,
+      "step": 13183
+    },
+    {
+      "epoch": 0.7256315702570312,
+      "grad_norm": 0.6848479509353638,
+      "learning_rate": 7.113855161786297e-06,
+      "loss": 0.6848,
+      "step": 13184
+    },
+    {
+      "epoch": 0.7256866090593869,
+      "grad_norm": 0.6800528764724731,
+      "learning_rate": 7.113462330558907e-06,
+      "loss": 0.7354,
+      "step": 13185
+    },
+    {
+      "epoch": 0.7257416478617426,
+      "grad_norm": 0.7271339297294617,
+      "learning_rate": 7.113069483447299e-06,
+      "loss": 0.7695,
+      "step": 13186
+    },
+    {
+      "epoch": 0.7257966866640981,
+      "grad_norm": 0.8212381601333618,
+      "learning_rate": 7.112676620454427e-06,
+      "loss": 0.7348,
+      "step": 13187
+    },
+    {
+      "epoch": 0.7258517254664538,
+      "grad_norm": 0.6714771389961243,
+      "learning_rate": 7.112283741583242e-06,
+      "loss": 0.75,
+      "step": 13188
+    },
+    {
+      "epoch": 0.7259067642688095,
+      "grad_norm": 0.7834941148757935,
+      "learning_rate": 7.111890846836699e-06,
+      "loss": 0.6914,
+      "step": 13189
+    },
+    {
+      "epoch": 0.7259618030711652,
+      "grad_norm": 0.8107824325561523,
+      "learning_rate": 7.111497936217748e-06,
+      "loss": 0.803,
+      "step": 13190
+    },
+    {
+      "epoch": 0.7260168418735208,
+      "grad_norm": 0.6306549906730652,
+      "learning_rate": 7.1111050097293464e-06,
+      "loss": 0.7915,
+      "step": 13191
+    },
+    {
+      "epoch": 0.7260718806758765,
+      "grad_norm": 0.7030252814292908,
+      "learning_rate": 7.110712067374444e-06,
+      "loss": 0.7091,
+      "step": 13192
+    },
+    {
+      "epoch": 0.7261269194782322,
+      "grad_norm": 0.7625641226768494,
+      "learning_rate": 7.110319109155992e-06,
+      "loss": 0.774,
+      "step": 13193
+    },
+    {
+      "epoch": 0.7261819582805878,
+      "grad_norm": 0.6382628083229065,
+      "learning_rate": 7.109926135076949e-06,
+      "loss": 0.6774,
+      "step": 13194
+    },
+    {
+      "epoch": 0.7262369970829434,
+      "grad_norm": 0.6594563722610474,
+      "learning_rate": 7.109533145140265e-06,
+      "loss": 0.7977,
+      "step": 13195
+    },
+    {
+      "epoch": 0.7262920358852991,
+      "grad_norm": 0.7177248001098633,
+      "learning_rate": 7.109140139348895e-06,
+      "loss": 0.6771,
+      "step": 13196
+    },
+    {
+      "epoch": 0.7263470746876548,
+      "grad_norm": 0.6631305813789368,
+      "learning_rate": 7.108747117705792e-06,
+      "loss": 0.6877,
+      "step": 13197
+    },
+    {
+      "epoch": 0.7264021134900105,
+      "grad_norm": 0.6783736944198608,
+      "learning_rate": 7.10835408021391e-06,
+      "loss": 0.8048,
+      "step": 13198
+    },
+    {
+      "epoch": 0.7264571522923661,
+      "grad_norm": 0.7368303537368774,
+      "learning_rate": 7.107961026876204e-06,
+      "loss": 0.7962,
+      "step": 13199
+    },
+    {
+      "epoch": 0.7265121910947218,
+      "grad_norm": 0.7697044014930725,
+      "learning_rate": 7.107567957695627e-06,
+      "loss": 0.769,
+      "step": 13200
+    },
+    {
+      "epoch": 0.7265672298970774,
+      "grad_norm": 0.639934241771698,
+      "learning_rate": 7.1071748726751325e-06,
+      "loss": 0.722,
+      "step": 13201
+    },
+    {
+      "epoch": 0.7266222686994331,
+      "grad_norm": 0.8410669565200806,
+      "learning_rate": 7.106781771817676e-06,
+      "loss": 0.8861,
+      "step": 13202
+    },
+    {
+      "epoch": 0.7266773075017887,
+      "grad_norm": 0.654924213886261,
+      "learning_rate": 7.106388655126212e-06,
+      "loss": 0.7463,
+      "step": 13203
+    },
+    {
+      "epoch": 0.7267323463041444,
+      "grad_norm": 0.719714879989624,
+      "learning_rate": 7.105995522603695e-06,
+      "loss": 0.759,
+      "step": 13204
+    },
+    {
+      "epoch": 0.7267873851065001,
+      "grad_norm": 0.7019139528274536,
+      "learning_rate": 7.105602374253078e-06,
+      "loss": 0.7965,
+      "step": 13205
+    },
+    {
+      "epoch": 0.7268424239088558,
+      "grad_norm": 0.7289487719535828,
+      "learning_rate": 7.105209210077318e-06,
+      "loss": 0.8591,
+      "step": 13206
+    },
+    {
+      "epoch": 0.7268974627112114,
+      "grad_norm": 0.670274019241333,
+      "learning_rate": 7.104816030079369e-06,
+      "loss": 0.7707,
+      "step": 13207
+    },
+    {
+      "epoch": 0.726952501513567,
+      "grad_norm": 0.7156813740730286,
+      "learning_rate": 7.104422834262187e-06,
+      "loss": 0.7724,
+      "step": 13208
+    },
+    {
+      "epoch": 0.7270075403159227,
+      "grad_norm": 0.6776198148727417,
+      "learning_rate": 7.104029622628726e-06,
+      "loss": 0.7331,
+      "step": 13209
+    },
+    {
+      "epoch": 0.7270625791182784,
+      "grad_norm": 0.8008358478546143,
+      "learning_rate": 7.103636395181941e-06,
+      "loss": 0.8279,
+      "step": 13210
+    },
+    {
+      "epoch": 0.727117617920634,
+      "grad_norm": 0.6622886061668396,
+      "learning_rate": 7.1032431519247876e-06,
+      "loss": 0.6646,
+      "step": 13211
+    },
+    {
+      "epoch": 0.7271726567229897,
+      "grad_norm": 0.6834877729415894,
+      "learning_rate": 7.102849892860223e-06,
+      "loss": 0.75,
+      "step": 13212
+    },
+    {
+      "epoch": 0.7272276955253454,
+      "grad_norm": 0.7659596800804138,
+      "learning_rate": 7.1024566179912e-06,
+      "loss": 0.6999,
+      "step": 13213
+    },
+    {
+      "epoch": 0.7272827343277011,
+      "grad_norm": 0.7368002533912659,
+      "learning_rate": 7.102063327320677e-06,
+      "loss": 0.7376,
+      "step": 13214
+    },
+    {
+      "epoch": 0.7273377731300567,
+      "grad_norm": 0.7286058664321899,
+      "learning_rate": 7.101670020851609e-06,
+      "loss": 0.8139,
+      "step": 13215
+    },
+    {
+      "epoch": 0.7273928119324123,
+      "grad_norm": 1.0521546602249146,
+      "learning_rate": 7.101276698586951e-06,
+      "loss": 0.8545,
+      "step": 13216
+    },
+    {
+      "epoch": 0.727447850734768,
+      "grad_norm": 0.6940305233001709,
+      "learning_rate": 7.100883360529659e-06,
+      "loss": 0.7534,
+      "step": 13217
+    },
+    {
+      "epoch": 0.7275028895371237,
+      "grad_norm": 0.8279024362564087,
+      "learning_rate": 7.100490006682691e-06,
+      "loss": 0.852,
+      "step": 13218
+    },
+    {
+      "epoch": 0.7275579283394793,
+      "grad_norm": 0.63093501329422,
+      "learning_rate": 7.100096637049002e-06,
+      "loss": 0.6728,
+      "step": 13219
+    },
+    {
+      "epoch": 0.727612967141835,
+      "grad_norm": 0.7576018571853638,
+      "learning_rate": 7.099703251631549e-06,
+      "loss": 0.6343,
+      "step": 13220
+    },
+    {
+      "epoch": 0.7276680059441907,
+      "grad_norm": 0.9493140578269958,
+      "learning_rate": 7.0993098504332894e-06,
+      "loss": 0.82,
+      "step": 13221
+    },
+    {
+      "epoch": 0.7277230447465464,
+      "grad_norm": 0.7279804944992065,
+      "learning_rate": 7.098916433457177e-06,
+      "loss": 0.8149,
+      "step": 13222
+    },
+    {
+      "epoch": 0.7277780835489019,
+      "grad_norm": 0.7660531401634216,
+      "learning_rate": 7.0985230007061725e-06,
+      "loss": 0.8278,
+      "step": 13223
+    },
+    {
+      "epoch": 0.7278331223512576,
+      "grad_norm": 0.6468318104743958,
+      "learning_rate": 7.09812955218323e-06,
+      "loss": 0.7193,
+      "step": 13224
+    },
+    {
+      "epoch": 0.7278881611536133,
+      "grad_norm": 0.6389151811599731,
+      "learning_rate": 7.097736087891306e-06,
+      "loss": 0.6744,
+      "step": 13225
+    },
+    {
+      "epoch": 0.727943199955969,
+      "grad_norm": 0.6565649509429932,
+      "learning_rate": 7.097342607833361e-06,
+      "loss": 0.7586,
+      "step": 13226
+    },
+    {
+      "epoch": 0.7279982387583246,
+      "grad_norm": 0.6867381930351257,
+      "learning_rate": 7.09694911201235e-06,
+      "loss": 0.684,
+      "step": 13227
+    },
+    {
+      "epoch": 0.7280532775606803,
+      "grad_norm": 0.7509286403656006,
+      "learning_rate": 7.096555600431229e-06,
+      "loss": 0.8242,
+      "step": 13228
+    },
+    {
+      "epoch": 0.728108316363036,
+      "grad_norm": 0.6997731328010559,
+      "learning_rate": 7.096162073092959e-06,
+      "loss": 0.8182,
+      "step": 13229
+    },
+    {
+      "epoch": 0.7281633551653915,
+      "grad_norm": 0.6698907017707825,
+      "learning_rate": 7.095768530000496e-06,
+      "loss": 0.7752,
+      "step": 13230
+    },
+    {
+      "epoch": 0.7282183939677472,
+      "grad_norm": 0.7219094634056091,
+      "learning_rate": 7.095374971156799e-06,
+      "loss": 0.792,
+      "step": 13231
+    },
+    {
+      "epoch": 0.7282734327701029,
+      "grad_norm": 0.6479744911193848,
+      "learning_rate": 7.094981396564822e-06,
+      "loss": 0.7556,
+      "step": 13232
+    },
+    {
+      "epoch": 0.7283284715724586,
+      "grad_norm": 0.6795497536659241,
+      "learning_rate": 7.094587806227527e-06,
+      "loss": 0.7611,
+      "step": 13233
+    },
+    {
+      "epoch": 0.7283835103748142,
+      "grad_norm": 0.7145074605941772,
+      "learning_rate": 7.094194200147871e-06,
+      "loss": 0.8064,
+      "step": 13234
+    },
+    {
+      "epoch": 0.7284385491771699,
+      "grad_norm": 0.6750605702400208,
+      "learning_rate": 7.093800578328811e-06,
+      "loss": 0.7054,
+      "step": 13235
+    },
+    {
+      "epoch": 0.7284935879795256,
+      "grad_norm": 0.7574751377105713,
+      "learning_rate": 7.093406940773307e-06,
+      "loss": 0.7878,
+      "step": 13236
+    },
+    {
+      "epoch": 0.7285486267818813,
+      "grad_norm": 0.7836418747901917,
+      "learning_rate": 7.093013287484316e-06,
+      "loss": 0.7445,
+      "step": 13237
+    },
+    {
+      "epoch": 0.7286036655842368,
+      "grad_norm": 0.7658870220184326,
+      "learning_rate": 7.092619618464799e-06,
+      "loss": 0.7513,
+      "step": 13238
+    },
+    {
+      "epoch": 0.7286587043865925,
+      "grad_norm": 1.1127573251724243,
+      "learning_rate": 7.092225933717711e-06,
+      "loss": 0.7601,
+      "step": 13239
+    },
+    {
+      "epoch": 0.7287137431889482,
+      "grad_norm": 0.7003853917121887,
+      "learning_rate": 7.091832233246015e-06,
+      "loss": 0.8533,
+      "step": 13240
+    },
+    {
+      "epoch": 0.7287687819913039,
+      "grad_norm": 0.6513979434967041,
+      "learning_rate": 7.091438517052667e-06,
+      "loss": 0.7285,
+      "step": 13241
+    },
+    {
+      "epoch": 0.7288238207936595,
+      "grad_norm": 0.7072234153747559,
+      "learning_rate": 7.091044785140626e-06,
+      "loss": 0.7741,
+      "step": 13242
+    },
+    {
+      "epoch": 0.7288788595960152,
+      "grad_norm": 0.8117190599441528,
+      "learning_rate": 7.090651037512854e-06,
+      "loss": 0.6851,
+      "step": 13243
+    },
+    {
+      "epoch": 0.7289338983983709,
+      "grad_norm": 0.6876427531242371,
+      "learning_rate": 7.090257274172306e-06,
+      "loss": 0.7162,
+      "step": 13244
+    },
+    {
+      "epoch": 0.7289889372007266,
+      "grad_norm": 0.7128324508666992,
+      "learning_rate": 7.0898634951219455e-06,
+      "loss": 0.7302,
+      "step": 13245
+    },
+    {
+      "epoch": 0.7290439760030821,
+      "grad_norm": 0.6918201446533203,
+      "learning_rate": 7.089469700364731e-06,
+      "loss": 0.8582,
+      "step": 13246
+    },
+    {
+      "epoch": 0.7290990148054378,
+      "grad_norm": 0.6172242164611816,
+      "learning_rate": 7.08907588990362e-06,
+      "loss": 0.6846,
+      "step": 13247
+    },
+    {
+      "epoch": 0.7291540536077935,
+      "grad_norm": 0.6799596548080444,
+      "learning_rate": 7.088682063741575e-06,
+      "loss": 0.7174,
+      "step": 13248
+    },
+    {
+      "epoch": 0.7292090924101492,
+      "grad_norm": 0.6663293838500977,
+      "learning_rate": 7.088288221881554e-06,
+      "loss": 0.7237,
+      "step": 13249
+    },
+    {
+      "epoch": 0.7292641312125048,
+      "grad_norm": 0.6758549213409424,
+      "learning_rate": 7.0878943643265175e-06,
+      "loss": 0.7912,
+      "step": 13250
+    },
+    {
+      "epoch": 0.7293191700148605,
+      "grad_norm": 0.6937153339385986,
+      "learning_rate": 7.087500491079427e-06,
+      "loss": 0.742,
+      "step": 13251
+    },
+    {
+      "epoch": 0.7293742088172162,
+      "grad_norm": 0.6441238522529602,
+      "learning_rate": 7.087106602143241e-06,
+      "loss": 0.7676,
+      "step": 13252
+    },
+    {
+      "epoch": 0.7294292476195718,
+      "grad_norm": 0.6615588068962097,
+      "learning_rate": 7.08671269752092e-06,
+      "loss": 0.7069,
+      "step": 13253
+    },
+    {
+      "epoch": 0.7294842864219274,
+      "grad_norm": 0.8052160739898682,
+      "learning_rate": 7.086318777215424e-06,
+      "loss": 0.811,
+      "step": 13254
+    },
+    {
+      "epoch": 0.7295393252242831,
+      "grad_norm": 0.7293280363082886,
+      "learning_rate": 7.085924841229716e-06,
+      "loss": 0.7127,
+      "step": 13255
+    },
+    {
+      "epoch": 0.7295943640266388,
+      "grad_norm": 0.7104617953300476,
+      "learning_rate": 7.085530889566756e-06,
+      "loss": 0.716,
+      "step": 13256
+    },
+    {
+      "epoch": 0.7296494028289945,
+      "grad_norm": 0.72947758436203,
+      "learning_rate": 7.085136922229503e-06,
+      "loss": 0.8144,
+      "step": 13257
+    },
+    {
+      "epoch": 0.7297044416313501,
+      "grad_norm": 0.7993913292884827,
+      "learning_rate": 7.08474293922092e-06,
+      "loss": 0.7609,
+      "step": 13258
+    },
+    {
+      "epoch": 0.7297594804337058,
+      "grad_norm": 0.7810680270195007,
+      "learning_rate": 7.0843489405439656e-06,
+      "loss": 0.8107,
+      "step": 13259
+    },
+    {
+      "epoch": 0.7298145192360614,
+      "grad_norm": 0.6383776664733887,
+      "learning_rate": 7.083954926201604e-06,
+      "loss": 0.7842,
+      "step": 13260
+    },
+    {
+      "epoch": 0.7298695580384171,
+      "grad_norm": 0.7653967142105103,
+      "learning_rate": 7.083560896196795e-06,
+      "loss": 0.729,
+      "step": 13261
+    },
+    {
+      "epoch": 0.7299245968407727,
+      "grad_norm": 0.6693821549415588,
+      "learning_rate": 7.083166850532498e-06,
+      "loss": 0.6901,
+      "step": 13262
+    },
+    {
+      "epoch": 0.7299796356431284,
+      "grad_norm": 0.7408621907234192,
+      "learning_rate": 7.082772789211678e-06,
+      "loss": 0.7415,
+      "step": 13263
+    },
+    {
+      "epoch": 0.7300346744454841,
+      "grad_norm": 0.6693123579025269,
+      "learning_rate": 7.082378712237295e-06,
+      "loss": 0.8102,
+      "step": 13264
+    },
+    {
+      "epoch": 0.7300897132478398,
+      "grad_norm": 0.6572727560997009,
+      "learning_rate": 7.081984619612311e-06,
+      "loss": 0.6595,
+      "step": 13265
+    },
+    {
+      "epoch": 0.7301447520501954,
+      "grad_norm": 0.7934693694114685,
+      "learning_rate": 7.081590511339687e-06,
+      "loss": 0.8024,
+      "step": 13266
+    },
+    {
+      "epoch": 0.730199790852551,
+      "grad_norm": 1.0663061141967773,
+      "learning_rate": 7.081196387422388e-06,
+      "loss": 0.7844,
+      "step": 13267
+    },
+    {
+      "epoch": 0.7302548296549067,
+      "grad_norm": 0.8005035519599915,
+      "learning_rate": 7.080802247863372e-06,
+      "loss": 0.751,
+      "step": 13268
+    },
+    {
+      "epoch": 0.7303098684572624,
+      "grad_norm": 0.6480177044868469,
+      "learning_rate": 7.0804080926656046e-06,
+      "loss": 0.7745,
+      "step": 13269
+    },
+    {
+      "epoch": 0.730364907259618,
+      "grad_norm": 0.7026820182800293,
+      "learning_rate": 7.080013921832047e-06,
+      "loss": 0.7545,
+      "step": 13270
+    },
+    {
+      "epoch": 0.7304199460619737,
+      "grad_norm": 0.673954427242279,
+      "learning_rate": 7.079619735365662e-06,
+      "loss": 0.7142,
+      "step": 13271
+    },
+    {
+      "epoch": 0.7304749848643294,
+      "grad_norm": 0.7296637296676636,
+      "learning_rate": 7.079225533269411e-06,
+      "loss": 0.8493,
+      "step": 13272
+    },
+    {
+      "epoch": 0.730530023666685,
+      "grad_norm": 0.7147308588027954,
+      "learning_rate": 7.0788313155462576e-06,
+      "loss": 0.7638,
+      "step": 13273
+    },
+    {
+      "epoch": 0.7305850624690406,
+      "grad_norm": 0.7531922459602356,
+      "learning_rate": 7.078437082199163e-06,
+      "loss": 0.8644,
+      "step": 13274
+    },
+    {
+      "epoch": 0.7306401012713963,
+      "grad_norm": 0.6581404805183411,
+      "learning_rate": 7.078042833231092e-06,
+      "loss": 0.7555,
+      "step": 13275
+    },
+    {
+      "epoch": 0.730695140073752,
+      "grad_norm": 0.6781187057495117,
+      "learning_rate": 7.0776485686450095e-06,
+      "loss": 0.7536,
+      "step": 13276
+    },
+    {
+      "epoch": 0.7307501788761076,
+      "grad_norm": 0.7164949774742126,
+      "learning_rate": 7.077254288443874e-06,
+      "loss": 0.7275,
+      "step": 13277
+    },
+    {
+      "epoch": 0.7308052176784633,
+      "grad_norm": 0.8158305287361145,
+      "learning_rate": 7.076859992630652e-06,
+      "loss": 0.6821,
+      "step": 13278
+    },
+    {
+      "epoch": 0.730860256480819,
+      "grad_norm": 0.7101448178291321,
+      "learning_rate": 7.076465681208307e-06,
+      "loss": 0.69,
+      "step": 13279
+    },
+    {
+      "epoch": 0.7309152952831747,
+      "grad_norm": 0.6844518780708313,
+      "learning_rate": 7.076071354179802e-06,
+      "loss": 0.7577,
+      "step": 13280
+    },
+    {
+      "epoch": 0.7309703340855302,
+      "grad_norm": 0.6564158797264099,
+      "learning_rate": 7.0756770115481e-06,
+      "loss": 0.6752,
+      "step": 13281
+    },
+    {
+      "epoch": 0.7310253728878859,
+      "grad_norm": 0.7444283962249756,
+      "learning_rate": 7.0752826533161655e-06,
+      "loss": 0.8118,
+      "step": 13282
+    },
+    {
+      "epoch": 0.7310804116902416,
+      "grad_norm": 0.7657533884048462,
+      "learning_rate": 7.074888279486962e-06,
+      "loss": 0.8819,
+      "step": 13283
+    },
+    {
+      "epoch": 0.7311354504925973,
+      "grad_norm": 0.6924453973770142,
+      "learning_rate": 7.074493890063453e-06,
+      "loss": 0.7674,
+      "step": 13284
+    },
+    {
+      "epoch": 0.7311904892949529,
+      "grad_norm": 0.676188588142395,
+      "learning_rate": 7.074099485048603e-06,
+      "loss": 0.7266,
+      "step": 13285
+    },
+    {
+      "epoch": 0.7312455280973086,
+      "grad_norm": 0.6325914263725281,
+      "learning_rate": 7.073705064445378e-06,
+      "loss": 0.6856,
+      "step": 13286
+    },
+    {
+      "epoch": 0.7313005668996643,
+      "grad_norm": 0.662558913230896,
+      "learning_rate": 7.073310628256739e-06,
+      "loss": 0.751,
+      "step": 13287
+    },
+    {
+      "epoch": 0.73135560570202,
+      "grad_norm": 0.8313137292861938,
+      "learning_rate": 7.072916176485654e-06,
+      "loss": 0.7187,
+      "step": 13288
+    },
+    {
+      "epoch": 0.7314106445043755,
+      "grad_norm": 0.7033550143241882,
+      "learning_rate": 7.072521709135084e-06,
+      "loss": 0.8132,
+      "step": 13289
+    },
+    {
+      "epoch": 0.7314656833067312,
+      "grad_norm": 0.715242862701416,
+      "learning_rate": 7.0721272262079965e-06,
+      "loss": 0.8551,
+      "step": 13290
+    },
+    {
+      "epoch": 0.7315207221090869,
+      "grad_norm": 0.7545164227485657,
+      "learning_rate": 7.071732727707356e-06,
+      "loss": 0.7772,
+      "step": 13291
+    },
+    {
+      "epoch": 0.7315757609114426,
+      "grad_norm": 0.7181825637817383,
+      "learning_rate": 7.071338213636126e-06,
+      "loss": 0.7378,
+      "step": 13292
+    },
+    {
+      "epoch": 0.7316307997137982,
+      "grad_norm": 0.7793779969215393,
+      "learning_rate": 7.070943683997273e-06,
+      "loss": 0.7801,
+      "step": 13293
+    },
+    {
+      "epoch": 0.7316858385161539,
+      "grad_norm": 0.7456476092338562,
+      "learning_rate": 7.070549138793762e-06,
+      "loss": 0.8038,
+      "step": 13294
+    },
+    {
+      "epoch": 0.7317408773185096,
+      "grad_norm": 0.652519702911377,
+      "learning_rate": 7.0701545780285576e-06,
+      "loss": 0.746,
+      "step": 13295
+    },
+    {
+      "epoch": 0.7317959161208653,
+      "grad_norm": 0.784450888633728,
+      "learning_rate": 7.069760001704625e-06,
+      "loss": 0.8065,
+      "step": 13296
+    },
+    {
+      "epoch": 0.7318509549232208,
+      "grad_norm": 0.8052587509155273,
+      "learning_rate": 7.069365409824931e-06,
+      "loss": 0.8098,
+      "step": 13297
+    },
+    {
+      "epoch": 0.7319059937255765,
+      "grad_norm": 0.6890794038772583,
+      "learning_rate": 7.06897080239244e-06,
+      "loss": 0.783,
+      "step": 13298
+    },
+    {
+      "epoch": 0.7319610325279322,
+      "grad_norm": 0.7470653057098389,
+      "learning_rate": 7.068576179410119e-06,
+      "loss": 0.7658,
+      "step": 13299
+    },
+    {
+      "epoch": 0.7320160713302879,
+      "grad_norm": 0.6831437945365906,
+      "learning_rate": 7.068181540880932e-06,
+      "loss": 0.7864,
+      "step": 13300
+    },
+    {
+      "epoch": 0.7320711101326435,
+      "grad_norm": 0.7058265209197998,
+      "learning_rate": 7.067786886807847e-06,
+      "loss": 0.8254,
+      "step": 13301
+    },
+    {
+      "epoch": 0.7321261489349992,
+      "grad_norm": 0.7938248515129089,
+      "learning_rate": 7.067392217193828e-06,
+      "loss": 0.7291,
+      "step": 13302
+    },
+    {
+      "epoch": 0.7321811877373549,
+      "grad_norm": 0.7261865735054016,
+      "learning_rate": 7.066997532041844e-06,
+      "loss": 0.8115,
+      "step": 13303
+    },
+    {
+      "epoch": 0.7322362265397105,
+      "grad_norm": 0.6971743702888489,
+      "learning_rate": 7.0666028313548586e-06,
+      "loss": 0.7504,
+      "step": 13304
+    },
+    {
+      "epoch": 0.7322912653420661,
+      "grad_norm": 0.844879150390625,
+      "learning_rate": 7.0662081151358405e-06,
+      "loss": 0.7903,
+      "step": 13305
+    },
+    {
+      "epoch": 0.7323463041444218,
+      "grad_norm": 0.6670572757720947,
+      "learning_rate": 7.065813383387755e-06,
+      "loss": 0.7597,
+      "step": 13306
+    },
+    {
+      "epoch": 0.7324013429467775,
+      "grad_norm": 0.669711172580719,
+      "learning_rate": 7.06541863611357e-06,
+      "loss": 0.7179,
+      "step": 13307
+    },
+    {
+      "epoch": 0.7324563817491332,
+      "grad_norm": 0.7176600098609924,
+      "learning_rate": 7.0650238733162506e-06,
+      "loss": 0.8157,
+      "step": 13308
+    },
+    {
+      "epoch": 0.7325114205514888,
+      "grad_norm": 0.7230100631713867,
+      "learning_rate": 7.064629094998765e-06,
+      "loss": 0.7902,
+      "step": 13309
+    },
+    {
+      "epoch": 0.7325664593538445,
+      "grad_norm": 0.8811234831809998,
+      "learning_rate": 7.064234301164078e-06,
+      "loss": 0.7746,
+      "step": 13310
+    },
+    {
+      "epoch": 0.7326214981562001,
+      "grad_norm": 0.6777653098106384,
+      "learning_rate": 7.06383949181516e-06,
+      "loss": 0.7708,
+      "step": 13311
+    },
+    {
+      "epoch": 0.7326765369585558,
+      "grad_norm": 0.6692547798156738,
+      "learning_rate": 7.063444666954977e-06,
+      "loss": 0.7103,
+      "step": 13312
+    },
+    {
+      "epoch": 0.7327315757609114,
+      "grad_norm": 1.2304950952529907,
+      "learning_rate": 7.063049826586496e-06,
+      "loss": 0.7878,
+      "step": 13313
+    },
+    {
+      "epoch": 0.7327866145632671,
+      "grad_norm": 0.7073930501937866,
+      "learning_rate": 7.0626549707126834e-06,
+      "loss": 0.7546,
+      "step": 13314
+    },
+    {
+      "epoch": 0.7328416533656228,
+      "grad_norm": 0.7184866070747375,
+      "learning_rate": 7.06226009933651e-06,
+      "loss": 0.7207,
+      "step": 13315
+    },
+    {
+      "epoch": 0.7328966921679784,
+      "grad_norm": 0.7098046541213989,
+      "learning_rate": 7.061865212460941e-06,
+      "loss": 0.6415,
+      "step": 13316
+    },
+    {
+      "epoch": 0.7329517309703341,
+      "grad_norm": 0.714379608631134,
+      "learning_rate": 7.0614703100889445e-06,
+      "loss": 0.7305,
+      "step": 13317
+    },
+    {
+      "epoch": 0.7330067697726897,
+      "grad_norm": 0.655060887336731,
+      "learning_rate": 7.061075392223491e-06,
+      "loss": 0.6125,
+      "step": 13318
+    },
+    {
+      "epoch": 0.7330618085750454,
+      "grad_norm": 0.6481055617332458,
+      "learning_rate": 7.060680458867545e-06,
+      "loss": 0.7059,
+      "step": 13319
+    },
+    {
+      "epoch": 0.733116847377401,
+      "grad_norm": 0.7123916745185852,
+      "learning_rate": 7.060285510024076e-06,
+      "loss": 0.8007,
+      "step": 13320
+    },
+    {
+      "epoch": 0.7331718861797567,
+      "grad_norm": 0.7231262922286987,
+      "learning_rate": 7.059890545696053e-06,
+      "loss": 0.7781,
+      "step": 13321
+    },
+    {
+      "epoch": 0.7332269249821124,
+      "grad_norm": 0.8415369391441345,
+      "learning_rate": 7.0594955658864435e-06,
+      "loss": 0.6649,
+      "step": 13322
+    },
+    {
+      "epoch": 0.7332819637844681,
+      "grad_norm": 0.7243070006370544,
+      "learning_rate": 7.059100570598217e-06,
+      "loss": 0.6588,
+      "step": 13323
+    },
+    {
+      "epoch": 0.7333370025868237,
+      "grad_norm": 0.6581026315689087,
+      "learning_rate": 7.058705559834342e-06,
+      "loss": 0.7938,
+      "step": 13324
+    },
+    {
+      "epoch": 0.7333920413891793,
+      "grad_norm": 0.6213739514350891,
+      "learning_rate": 7.058310533597787e-06,
+      "loss": 0.7092,
+      "step": 13325
+    },
+    {
+      "epoch": 0.733447080191535,
+      "grad_norm": 0.6857954859733582,
+      "learning_rate": 7.057915491891522e-06,
+      "loss": 0.698,
+      "step": 13326
+    },
+    {
+      "epoch": 0.7335021189938907,
+      "grad_norm": 0.7528544068336487,
+      "learning_rate": 7.0575204347185135e-06,
+      "loss": 0.7234,
+      "step": 13327
+    },
+    {
+      "epoch": 0.7335571577962463,
+      "grad_norm": 0.6449099779129028,
+      "learning_rate": 7.057125362081733e-06,
+      "loss": 0.7391,
+      "step": 13328
+    },
+    {
+      "epoch": 0.733612196598602,
+      "grad_norm": 0.640689492225647,
+      "learning_rate": 7.0567302739841495e-06,
+      "loss": 0.5316,
+      "step": 13329
+    },
+    {
+      "epoch": 0.7336672354009577,
+      "grad_norm": 0.6686868071556091,
+      "learning_rate": 7.056335170428731e-06,
+      "loss": 0.7713,
+      "step": 13330
+    },
+    {
+      "epoch": 0.7337222742033134,
+      "grad_norm": 0.7627772688865662,
+      "learning_rate": 7.055940051418447e-06,
+      "loss": 0.7706,
+      "step": 13331
+    },
+    {
+      "epoch": 0.733777313005669,
+      "grad_norm": 0.7421852350234985,
+      "learning_rate": 7.055544916956269e-06,
+      "loss": 0.6418,
+      "step": 13332
+    },
+    {
+      "epoch": 0.7338323518080246,
+      "grad_norm": 0.7414699196815491,
+      "learning_rate": 7.0551497670451666e-06,
+      "loss": 0.811,
+      "step": 13333
+    },
+    {
+      "epoch": 0.7338873906103803,
+      "grad_norm": 0.7054136991500854,
+      "learning_rate": 7.0547546016881064e-06,
+      "loss": 0.8005,
+      "step": 13334
+    },
+    {
+      "epoch": 0.733942429412736,
+      "grad_norm": 0.670174241065979,
+      "learning_rate": 7.054359420888062e-06,
+      "loss": 0.6136,
+      "step": 13335
+    },
+    {
+      "epoch": 0.7339974682150916,
+      "grad_norm": 0.728255033493042,
+      "learning_rate": 7.053964224648001e-06,
+      "loss": 0.848,
+      "step": 13336
+    },
+    {
+      "epoch": 0.7340525070174473,
+      "grad_norm": 0.729815661907196,
+      "learning_rate": 7.053569012970896e-06,
+      "loss": 0.6985,
+      "step": 13337
+    },
+    {
+      "epoch": 0.734107545819803,
+      "grad_norm": 0.7564244866371155,
+      "learning_rate": 7.053173785859715e-06,
+      "loss": 0.7995,
+      "step": 13338
+    },
+    {
+      "epoch": 0.7341625846221587,
+      "grad_norm": 0.7746061682701111,
+      "learning_rate": 7.05277854331743e-06,
+      "loss": 0.7663,
+      "step": 13339
+    },
+    {
+      "epoch": 0.7342176234245142,
+      "grad_norm": 0.6878651976585388,
+      "learning_rate": 7.052383285347011e-06,
+      "loss": 0.8624,
+      "step": 13340
+    },
+    {
+      "epoch": 0.7342726622268699,
+      "grad_norm": 0.6989734768867493,
+      "learning_rate": 7.051988011951428e-06,
+      "loss": 0.7221,
+      "step": 13341
+    },
+    {
+      "epoch": 0.7343277010292256,
+      "grad_norm": 0.6854223012924194,
+      "learning_rate": 7.051592723133654e-06,
+      "loss": 0.7878,
+      "step": 13342
+    },
+    {
+      "epoch": 0.7343827398315813,
+      "grad_norm": 0.746696949005127,
+      "learning_rate": 7.051197418896657e-06,
+      "loss": 0.7074,
+      "step": 13343
+    },
+    {
+      "epoch": 0.7344377786339369,
+      "grad_norm": 0.6933150887489319,
+      "learning_rate": 7.050802099243409e-06,
+      "loss": 0.7587,
+      "step": 13344
+    },
+    {
+      "epoch": 0.7344928174362926,
+      "grad_norm": 0.7285788655281067,
+      "learning_rate": 7.050406764176882e-06,
+      "loss": 0.6589,
+      "step": 13345
+    },
+    {
+      "epoch": 0.7345478562386483,
+      "grad_norm": 0.6834994554519653,
+      "learning_rate": 7.050011413700046e-06,
+      "loss": 0.7196,
+      "step": 13346
+    },
+    {
+      "epoch": 0.734602895041004,
+      "grad_norm": 0.6504353880882263,
+      "learning_rate": 7.049616047815873e-06,
+      "loss": 0.7675,
+      "step": 13347
+    },
+    {
+      "epoch": 0.7346579338433595,
+      "grad_norm": 0.7009296417236328,
+      "learning_rate": 7.049220666527335e-06,
+      "loss": 0.7638,
+      "step": 13348
+    },
+    {
+      "epoch": 0.7347129726457152,
+      "grad_norm": 0.6210034489631653,
+      "learning_rate": 7.0488252698374024e-06,
+      "loss": 0.6872,
+      "step": 13349
+    },
+    {
+      "epoch": 0.7347680114480709,
+      "grad_norm": 0.6280165910720825,
+      "learning_rate": 7.0484298577490485e-06,
+      "loss": 0.7084,
+      "step": 13350
+    },
+    {
+      "epoch": 0.7348230502504266,
+      "grad_norm": 0.8055418133735657,
+      "learning_rate": 7.048034430265242e-06,
+      "loss": 0.8202,
+      "step": 13351
+    },
+    {
+      "epoch": 0.7348780890527822,
+      "grad_norm": 0.6674166917800903,
+      "learning_rate": 7.047638987388959e-06,
+      "loss": 0.6368,
+      "step": 13352
+    },
+    {
+      "epoch": 0.7349331278551379,
+      "grad_norm": 0.9182783961296082,
+      "learning_rate": 7.04724352912317e-06,
+      "loss": 0.6734,
+      "step": 13353
+    },
+    {
+      "epoch": 0.7349881666574936,
+      "grad_norm": 0.6371243596076965,
+      "learning_rate": 7.046848055470845e-06,
+      "loss": 0.7308,
+      "step": 13354
+    },
+    {
+      "epoch": 0.7350432054598492,
+      "grad_norm": 0.6454519033432007,
+      "learning_rate": 7.046452566434959e-06,
+      "loss": 0.6882,
+      "step": 13355
+    },
+    {
+      "epoch": 0.7350982442622048,
+      "grad_norm": 0.648970365524292,
+      "learning_rate": 7.046057062018483e-06,
+      "loss": 0.7247,
+      "step": 13356
+    },
+    {
+      "epoch": 0.7351532830645605,
+      "grad_norm": 0.668886661529541,
+      "learning_rate": 7.04566154222439e-06,
+      "loss": 0.7379,
+      "step": 13357
+    },
+    {
+      "epoch": 0.7352083218669162,
+      "grad_norm": 0.6593654751777649,
+      "learning_rate": 7.045266007055651e-06,
+      "loss": 0.7473,
+      "step": 13358
+    },
+    {
+      "epoch": 0.7352633606692718,
+      "grad_norm": 0.8418927192687988,
+      "learning_rate": 7.044870456515241e-06,
+      "loss": 0.7949,
+      "step": 13359
+    },
+    {
+      "epoch": 0.7353183994716275,
+      "grad_norm": 0.7350470423698425,
+      "learning_rate": 7.044474890606132e-06,
+      "loss": 0.7545,
+      "step": 13360
+    },
+    {
+      "epoch": 0.7353734382739832,
+      "grad_norm": 0.7786250114440918,
+      "learning_rate": 7.044079309331298e-06,
+      "loss": 0.8587,
+      "step": 13361
+    },
+    {
+      "epoch": 0.7354284770763388,
+      "grad_norm": 0.6345693469047546,
+      "learning_rate": 7.04368371269371e-06,
+      "loss": 0.77,
+      "step": 13362
+    },
+    {
+      "epoch": 0.7354835158786944,
+      "grad_norm": 0.7030417919158936,
+      "learning_rate": 7.043288100696343e-06,
+      "loss": 0.7624,
+      "step": 13363
+    },
+    {
+      "epoch": 0.7355385546810501,
+      "grad_norm": 0.7526041865348816,
+      "learning_rate": 7.042892473342169e-06,
+      "loss": 0.8018,
+      "step": 13364
+    },
+    {
+      "epoch": 0.7355935934834058,
+      "grad_norm": 0.6419941782951355,
+      "learning_rate": 7.042496830634162e-06,
+      "loss": 0.6788,
+      "step": 13365
+    },
+    {
+      "epoch": 0.7356486322857615,
+      "grad_norm": 0.6952203512191772,
+      "learning_rate": 7.042101172575297e-06,
+      "loss": 0.7747,
+      "step": 13366
+    },
+    {
+      "epoch": 0.7357036710881171,
+      "grad_norm": 0.8046327829360962,
+      "learning_rate": 7.041705499168544e-06,
+      "loss": 0.8216,
+      "step": 13367
+    },
+    {
+      "epoch": 0.7357587098904728,
+      "grad_norm": 0.6641537547111511,
+      "learning_rate": 7.041309810416881e-06,
+      "loss": 0.7313,
+      "step": 13368
+    },
+    {
+      "epoch": 0.7358137486928285,
+      "grad_norm": 0.6824444532394409,
+      "learning_rate": 7.040914106323278e-06,
+      "loss": 0.7179,
+      "step": 13369
+    },
+    {
+      "epoch": 0.7358687874951841,
+      "grad_norm": 0.6469557285308838,
+      "learning_rate": 7.040518386890711e-06,
+      "loss": 0.7671,
+      "step": 13370
+    },
+    {
+      "epoch": 0.7359238262975397,
+      "grad_norm": 0.6826488971710205,
+      "learning_rate": 7.040122652122156e-06,
+      "loss": 0.7,
+      "step": 13371
+    },
+    {
+      "epoch": 0.7359788650998954,
+      "grad_norm": 0.6931618452072144,
+      "learning_rate": 7.039726902020583e-06,
+      "loss": 0.7641,
+      "step": 13372
+    },
+    {
+      "epoch": 0.7360339039022511,
+      "grad_norm": 0.7445465922355652,
+      "learning_rate": 7.039331136588971e-06,
+      "loss": 0.7458,
+      "step": 13373
+    },
+    {
+      "epoch": 0.7360889427046068,
+      "grad_norm": 0.6358756422996521,
+      "learning_rate": 7.038935355830289e-06,
+      "loss": 0.6125,
+      "step": 13374
+    },
+    {
+      "epoch": 0.7361439815069624,
+      "grad_norm": 0.6966063380241394,
+      "learning_rate": 7.038539559747517e-06,
+      "loss": 0.6812,
+      "step": 13375
+    },
+    {
+      "epoch": 0.736199020309318,
+      "grad_norm": 0.9898090362548828,
+      "learning_rate": 7.038143748343626e-06,
+      "loss": 0.707,
+      "step": 13376
+    },
+    {
+      "epoch": 0.7362540591116737,
+      "grad_norm": 0.685951828956604,
+      "learning_rate": 7.0377479216215935e-06,
+      "loss": 0.7932,
+      "step": 13377
+    },
+    {
+      "epoch": 0.7363090979140294,
+      "grad_norm": 0.7056856751441956,
+      "learning_rate": 7.037352079584392e-06,
+      "loss": 0.7432,
+      "step": 13378
+    },
+    {
+      "epoch": 0.736364136716385,
+      "grad_norm": 0.7802489995956421,
+      "learning_rate": 7.036956222234999e-06,
+      "loss": 0.8275,
+      "step": 13379
+    },
+    {
+      "epoch": 0.7364191755187407,
+      "grad_norm": 0.7990192770957947,
+      "learning_rate": 7.036560349576387e-06,
+      "loss": 0.893,
+      "step": 13380
+    },
+    {
+      "epoch": 0.7364742143210964,
+      "grad_norm": 0.6454586386680603,
+      "learning_rate": 7.0361644616115334e-06,
+      "loss": 0.751,
+      "step": 13381
+    },
+    {
+      "epoch": 0.7365292531234521,
+      "grad_norm": 0.7071009278297424,
+      "learning_rate": 7.035768558343412e-06,
+      "loss": 0.7771,
+      "step": 13382
+    },
+    {
+      "epoch": 0.7365842919258077,
+      "grad_norm": 0.6530466079711914,
+      "learning_rate": 7.035372639774999e-06,
+      "loss": 0.7529,
+      "step": 13383
+    },
+    {
+      "epoch": 0.7366393307281633,
+      "grad_norm": 0.728689968585968,
+      "learning_rate": 7.03497670590927e-06,
+      "loss": 0.7862,
+      "step": 13384
+    },
+    {
+      "epoch": 0.736694369530519,
+      "grad_norm": 0.6640015244483948,
+      "learning_rate": 7.034580756749202e-06,
+      "loss": 0.6876,
+      "step": 13385
+    },
+    {
+      "epoch": 0.7367494083328747,
+      "grad_norm": 0.7388426661491394,
+      "learning_rate": 7.034184792297769e-06,
+      "loss": 0.8168,
+      "step": 13386
+    },
+    {
+      "epoch": 0.7368044471352303,
+      "grad_norm": 0.6543731093406677,
+      "learning_rate": 7.0337888125579465e-06,
+      "loss": 0.7555,
+      "step": 13387
+    },
+    {
+      "epoch": 0.736859485937586,
+      "grad_norm": 0.7783555388450623,
+      "learning_rate": 7.0333928175327125e-06,
+      "loss": 0.755,
+      "step": 13388
+    },
+    {
+      "epoch": 0.7369145247399417,
+      "grad_norm": 0.6275887489318848,
+      "learning_rate": 7.032996807225043e-06,
+      "loss": 0.7187,
+      "step": 13389
+    },
+    {
+      "epoch": 0.7369695635422974,
+      "grad_norm": 0.7007517218589783,
+      "learning_rate": 7.032600781637913e-06,
+      "loss": 0.6993,
+      "step": 13390
+    },
+    {
+      "epoch": 0.737024602344653,
+      "grad_norm": 0.6322247385978699,
+      "learning_rate": 7.0322047407743e-06,
+      "loss": 0.7178,
+      "step": 13391
+    },
+    {
+      "epoch": 0.7370796411470086,
+      "grad_norm": 0.7160976529121399,
+      "learning_rate": 7.0318086846371804e-06,
+      "loss": 0.6884,
+      "step": 13392
+    },
+    {
+      "epoch": 0.7371346799493643,
+      "grad_norm": 0.6056101322174072,
+      "learning_rate": 7.03141261322953e-06,
+      "loss": 0.6672,
+      "step": 13393
+    },
+    {
+      "epoch": 0.73718971875172,
+      "grad_norm": 0.8779410123825073,
+      "learning_rate": 7.0310165265543264e-06,
+      "loss": 0.7564,
+      "step": 13394
+    },
+    {
+      "epoch": 0.7372447575540756,
+      "grad_norm": 0.6868176460266113,
+      "learning_rate": 7.030620424614546e-06,
+      "loss": 0.7658,
+      "step": 13395
+    },
+    {
+      "epoch": 0.7372997963564313,
+      "grad_norm": 0.7611618041992188,
+      "learning_rate": 7.030224307413166e-06,
+      "loss": 0.6445,
+      "step": 13396
+    },
+    {
+      "epoch": 0.737354835158787,
+      "grad_norm": 0.7688242793083191,
+      "learning_rate": 7.0298281749531636e-06,
+      "loss": 0.8061,
+      "step": 13397
+    },
+    {
+      "epoch": 0.7374098739611427,
+      "grad_norm": 0.6781700849533081,
+      "learning_rate": 7.029432027237518e-06,
+      "loss": 0.6374,
+      "step": 13398
+    },
+    {
+      "epoch": 0.7374649127634982,
+      "grad_norm": 0.6719028353691101,
+      "learning_rate": 7.0290358642692e-06,
+      "loss": 0.7585,
+      "step": 13399
+    },
+    {
+      "epoch": 0.7375199515658539,
+      "grad_norm": 0.704429030418396,
+      "learning_rate": 7.028639686051195e-06,
+      "loss": 0.7052,
+      "step": 13400
+    },
+    {
+      "epoch": 0.7375749903682096,
+      "grad_norm": 0.714914083480835,
+      "learning_rate": 7.028243492586478e-06,
+      "loss": 0.7785,
+      "step": 13401
+    },
+    {
+      "epoch": 0.7376300291705652,
+      "grad_norm": 0.7732700705528259,
+      "learning_rate": 7.027847283878023e-06,
+      "loss": 0.7812,
+      "step": 13402
+    },
+    {
+      "epoch": 0.7376850679729209,
+      "grad_norm": 0.6849464178085327,
+      "learning_rate": 7.027451059928813e-06,
+      "loss": 0.7657,
+      "step": 13403
+    },
+    {
+      "epoch": 0.7377401067752766,
+      "grad_norm": 0.6924402117729187,
+      "learning_rate": 7.027054820741822e-06,
+      "loss": 0.677,
+      "step": 13404
+    },
+    {
+      "epoch": 0.7377951455776323,
+      "grad_norm": 0.7142716646194458,
+      "learning_rate": 7.02665856632003e-06,
+      "loss": 0.7071,
+      "step": 13405
+    },
+    {
+      "epoch": 0.7378501843799878,
+      "grad_norm": 0.7227265238761902,
+      "learning_rate": 7.0262622966664154e-06,
+      "loss": 0.6986,
+      "step": 13406
+    },
+    {
+      "epoch": 0.7379052231823435,
+      "grad_norm": 0.6387726664543152,
+      "learning_rate": 7.025866011783954e-06,
+      "loss": 0.6563,
+      "step": 13407
+    },
+    {
+      "epoch": 0.7379602619846992,
+      "grad_norm": 0.6411992311477661,
+      "learning_rate": 7.025469711675628e-06,
+      "loss": 0.5842,
+      "step": 13408
+    },
+    {
+      "epoch": 0.7380153007870549,
+      "grad_norm": 0.6811027526855469,
+      "learning_rate": 7.025073396344413e-06,
+      "loss": 0.6746,
+      "step": 13409
+    },
+    {
+      "epoch": 0.7380703395894105,
+      "grad_norm": 1.0705479383468628,
+      "learning_rate": 7.024677065793289e-06,
+      "loss": 0.7457,
+      "step": 13410
+    },
+    {
+      "epoch": 0.7381253783917662,
+      "grad_norm": 0.6920849084854126,
+      "learning_rate": 7.024280720025232e-06,
+      "loss": 0.6838,
+      "step": 13411
+    },
+    {
+      "epoch": 0.7381804171941219,
+      "grad_norm": 0.8089182376861572,
+      "learning_rate": 7.0238843590432236e-06,
+      "loss": 0.6682,
+      "step": 13412
+    },
+    {
+      "epoch": 0.7382354559964776,
+      "grad_norm": 0.6140334010124207,
+      "learning_rate": 7.023487982850244e-06,
+      "loss": 0.6992,
+      "step": 13413
+    },
+    {
+      "epoch": 0.7382904947988331,
+      "grad_norm": 0.8564643263816833,
+      "learning_rate": 7.023091591449269e-06,
+      "loss": 0.8512,
+      "step": 13414
+    },
+    {
+      "epoch": 0.7383455336011888,
+      "grad_norm": 0.655516505241394,
+      "learning_rate": 7.02269518484328e-06,
+      "loss": 0.7291,
+      "step": 13415
+    },
+    {
+      "epoch": 0.7384005724035445,
+      "grad_norm": 0.6373177766799927,
+      "learning_rate": 7.022298763035255e-06,
+      "loss": 0.7553,
+      "step": 13416
+    },
+    {
+      "epoch": 0.7384556112059002,
+      "grad_norm": 0.7023805379867554,
+      "learning_rate": 7.021902326028174e-06,
+      "loss": 0.7562,
+      "step": 13417
+    },
+    {
+      "epoch": 0.7385106500082558,
+      "grad_norm": 0.654181182384491,
+      "learning_rate": 7.021505873825016e-06,
+      "loss": 0.7153,
+      "step": 13418
+    },
+    {
+      "epoch": 0.7385656888106115,
+      "grad_norm": 0.6633459329605103,
+      "learning_rate": 7.02110940642876e-06,
+      "loss": 0.6779,
+      "step": 13419
+    },
+    {
+      "epoch": 0.7386207276129672,
+      "grad_norm": 0.7050659656524658,
+      "learning_rate": 7.020712923842388e-06,
+      "loss": 0.741,
+      "step": 13420
+    },
+    {
+      "epoch": 0.7386757664153228,
+      "grad_norm": 0.7241182327270508,
+      "learning_rate": 7.020316426068879e-06,
+      "loss": 0.7479,
+      "step": 13421
+    },
+    {
+      "epoch": 0.7387308052176784,
+      "grad_norm": 1.0262155532836914,
+      "learning_rate": 7.019919913111212e-06,
+      "loss": 0.8418,
+      "step": 13422
+    },
+    {
+      "epoch": 0.7387858440200341,
+      "grad_norm": 0.6765457391738892,
+      "learning_rate": 7.019523384972366e-06,
+      "loss": 0.727,
+      "step": 13423
+    },
+    {
+      "epoch": 0.7388408828223898,
+      "grad_norm": 0.6871724724769592,
+      "learning_rate": 7.0191268416553245e-06,
+      "loss": 0.8273,
+      "step": 13424
+    },
+    {
+      "epoch": 0.7388959216247455,
+      "grad_norm": 0.8085252046585083,
+      "learning_rate": 7.018730283163067e-06,
+      "loss": 0.7306,
+      "step": 13425
+    },
+    {
+      "epoch": 0.7389509604271011,
+      "grad_norm": 0.6822873950004578,
+      "learning_rate": 7.018333709498572e-06,
+      "loss": 0.7454,
+      "step": 13426
+    },
+    {
+      "epoch": 0.7390059992294568,
+      "grad_norm": 0.7210521697998047,
+      "learning_rate": 7.01793712066482e-06,
+      "loss": 0.8306,
+      "step": 13427
+    },
+    {
+      "epoch": 0.7390610380318124,
+      "grad_norm": 0.6404997110366821,
+      "learning_rate": 7.017540516664795e-06,
+      "loss": 0.7151,
+      "step": 13428
+    },
+    {
+      "epoch": 0.7391160768341681,
+      "grad_norm": 0.6662821769714355,
+      "learning_rate": 7.017143897501475e-06,
+      "loss": 0.7446,
+      "step": 13429
+    },
+    {
+      "epoch": 0.7391711156365237,
+      "grad_norm": 0.8048129081726074,
+      "learning_rate": 7.0167472631778415e-06,
+      "loss": 0.7953,
+      "step": 13430
+    },
+    {
+      "epoch": 0.7392261544388794,
+      "grad_norm": 0.7215000987052917,
+      "learning_rate": 7.016350613696873e-06,
+      "loss": 0.8373,
+      "step": 13431
+    },
+    {
+      "epoch": 0.7392811932412351,
+      "grad_norm": 0.7309150099754333,
+      "learning_rate": 7.015953949061555e-06,
+      "loss": 0.7654,
+      "step": 13432
+    },
+    {
+      "epoch": 0.7393362320435908,
+      "grad_norm": 0.6487464904785156,
+      "learning_rate": 7.0155572692748665e-06,
+      "loss": 0.6473,
+      "step": 13433
+    },
+    {
+      "epoch": 0.7393912708459464,
+      "grad_norm": 0.6172077059745789,
+      "learning_rate": 7.01516057433979e-06,
+      "loss": 0.6672,
+      "step": 13434
+    },
+    {
+      "epoch": 0.739446309648302,
+      "grad_norm": 0.7569651007652283,
+      "learning_rate": 7.014763864259304e-06,
+      "loss": 0.8501,
+      "step": 13435
+    },
+    {
+      "epoch": 0.7395013484506577,
+      "grad_norm": 0.824669599533081,
+      "learning_rate": 7.014367139036393e-06,
+      "loss": 0.8596,
+      "step": 13436
+    },
+    {
+      "epoch": 0.7395563872530134,
+      "grad_norm": 0.6904401183128357,
+      "learning_rate": 7.013970398674038e-06,
+      "loss": 0.7403,
+      "step": 13437
+    },
+    {
+      "epoch": 0.739611426055369,
+      "grad_norm": 0.7999581098556519,
+      "learning_rate": 7.013573643175221e-06,
+      "loss": 0.8879,
+      "step": 13438
+    },
+    {
+      "epoch": 0.7396664648577247,
+      "grad_norm": 0.6600533723831177,
+      "learning_rate": 7.0131768725429236e-06,
+      "loss": 0.7324,
+      "step": 13439
+    },
+    {
+      "epoch": 0.7397215036600804,
+      "grad_norm": 0.7174191474914551,
+      "learning_rate": 7.0127800867801275e-06,
+      "loss": 0.7474,
+      "step": 13440
+    },
+    {
+      "epoch": 0.7397765424624361,
+      "grad_norm": 0.7023884654045105,
+      "learning_rate": 7.012383285889814e-06,
+      "loss": 0.7826,
+      "step": 13441
+    },
+    {
+      "epoch": 0.7398315812647916,
+      "grad_norm": 0.6486913561820984,
+      "learning_rate": 7.011986469874969e-06,
+      "loss": 0.6553,
+      "step": 13442
+    },
+    {
+      "epoch": 0.7398866200671473,
+      "grad_norm": 0.7238486409187317,
+      "learning_rate": 7.011589638738569e-06,
+      "loss": 0.6759,
+      "step": 13443
+    },
+    {
+      "epoch": 0.739941658869503,
+      "grad_norm": 0.7879656553268433,
+      "learning_rate": 7.011192792483601e-06,
+      "loss": 0.886,
+      "step": 13444
+    },
+    {
+      "epoch": 0.7399966976718586,
+      "grad_norm": 0.6592407822608948,
+      "learning_rate": 7.010795931113047e-06,
+      "loss": 0.7746,
+      "step": 13445
+    },
+    {
+      "epoch": 0.7400517364742143,
+      "grad_norm": 0.8274507522583008,
+      "learning_rate": 7.010399054629889e-06,
+      "loss": 0.7615,
+      "step": 13446
+    },
+    {
+      "epoch": 0.74010677527657,
+      "grad_norm": 0.6233614087104797,
+      "learning_rate": 7.010002163037109e-06,
+      "loss": 0.695,
+      "step": 13447
+    },
+    {
+      "epoch": 0.7401618140789257,
+      "grad_norm": 0.7082701921463013,
+      "learning_rate": 7.00960525633769e-06,
+      "loss": 0.6677,
+      "step": 13448
+    },
+    {
+      "epoch": 0.7402168528812813,
+      "grad_norm": 1.0694652795791626,
+      "learning_rate": 7.009208334534618e-06,
+      "loss": 0.7792,
+      "step": 13449
+    },
+    {
+      "epoch": 0.7402718916836369,
+      "grad_norm": 0.7189109325408936,
+      "learning_rate": 7.008811397630874e-06,
+      "loss": 0.8606,
+      "step": 13450
+    },
+    {
+      "epoch": 0.7403269304859926,
+      "grad_norm": 0.7136901617050171,
+      "learning_rate": 7.00841444562944e-06,
+      "loss": 0.7142,
+      "step": 13451
+    },
+    {
+      "epoch": 0.7403819692883483,
+      "grad_norm": 0.6508508920669556,
+      "learning_rate": 7.008017478533301e-06,
+      "loss": 0.6748,
+      "step": 13452
+    },
+    {
+      "epoch": 0.7404370080907039,
+      "grad_norm": 0.6560903191566467,
+      "learning_rate": 7.007620496345441e-06,
+      "loss": 0.7929,
+      "step": 13453
+    },
+    {
+      "epoch": 0.7404920468930596,
+      "grad_norm": 0.6909067034721375,
+      "learning_rate": 7.007223499068841e-06,
+      "loss": 0.6118,
+      "step": 13454
+    },
+    {
+      "epoch": 0.7405470856954153,
+      "grad_norm": 0.6554582715034485,
+      "learning_rate": 7.0068264867064874e-06,
+      "loss": 0.7687,
+      "step": 13455
+    },
+    {
+      "epoch": 0.740602124497771,
+      "grad_norm": 0.7788346409797668,
+      "learning_rate": 7.006429459261363e-06,
+      "loss": 0.7535,
+      "step": 13456
+    },
+    {
+      "epoch": 0.7406571633001265,
+      "grad_norm": 0.7702943682670593,
+      "learning_rate": 7.006032416736452e-06,
+      "loss": 0.833,
+      "step": 13457
+    },
+    {
+      "epoch": 0.7407122021024822,
+      "grad_norm": 0.6860190033912659,
+      "learning_rate": 7.005635359134738e-06,
+      "loss": 0.6643,
+      "step": 13458
+    },
+    {
+      "epoch": 0.7407672409048379,
+      "grad_norm": 0.7470136880874634,
+      "learning_rate": 7.005238286459205e-06,
+      "loss": 0.7811,
+      "step": 13459
+    },
+    {
+      "epoch": 0.7408222797071936,
+      "grad_norm": 0.6769132614135742,
+      "learning_rate": 7.004841198712839e-06,
+      "loss": 0.7322,
+      "step": 13460
+    },
+    {
+      "epoch": 0.7408773185095492,
+      "grad_norm": 0.7865259647369385,
+      "learning_rate": 7.004444095898623e-06,
+      "loss": 0.817,
+      "step": 13461
+    },
+    {
+      "epoch": 0.7409323573119049,
+      "grad_norm": 0.7352784276008606,
+      "learning_rate": 7.004046978019542e-06,
+      "loss": 0.7373,
+      "step": 13462
+    },
+    {
+      "epoch": 0.7409873961142606,
+      "grad_norm": 0.7647448182106018,
+      "learning_rate": 7.00364984507858e-06,
+      "loss": 0.7129,
+      "step": 13463
+    },
+    {
+      "epoch": 0.7410424349166163,
+      "grad_norm": 0.6979989409446716,
+      "learning_rate": 7.003252697078722e-06,
+      "loss": 0.7833,
+      "step": 13464
+    },
+    {
+      "epoch": 0.7410974737189718,
+      "grad_norm": 0.6117465496063232,
+      "learning_rate": 7.002855534022953e-06,
+      "loss": 0.6732,
+      "step": 13465
+    },
+    {
+      "epoch": 0.7411525125213275,
+      "grad_norm": 0.6754159331321716,
+      "learning_rate": 7.002458355914258e-06,
+      "loss": 0.6939,
+      "step": 13466
+    },
+    {
+      "epoch": 0.7412075513236832,
+      "grad_norm": 0.6713566184043884,
+      "learning_rate": 7.002061162755621e-06,
+      "loss": 0.7459,
+      "step": 13467
+    },
+    {
+      "epoch": 0.7412625901260389,
+      "grad_norm": 0.6475394368171692,
+      "learning_rate": 7.001663954550029e-06,
+      "loss": 0.7912,
+      "step": 13468
+    },
+    {
+      "epoch": 0.7413176289283945,
+      "grad_norm": 0.6577908992767334,
+      "learning_rate": 7.001266731300467e-06,
+      "loss": 0.6903,
+      "step": 13469
+    },
+    {
+      "epoch": 0.7413726677307502,
+      "grad_norm": 0.8129748106002808,
+      "learning_rate": 7.00086949300992e-06,
+      "loss": 0.8277,
+      "step": 13470
+    },
+    {
+      "epoch": 0.7414277065331059,
+      "grad_norm": 0.6730444431304932,
+      "learning_rate": 7.000472239681372e-06,
+      "loss": 0.7357,
+      "step": 13471
+    },
+    {
+      "epoch": 0.7414827453354615,
+      "grad_norm": 0.7166460156440735,
+      "learning_rate": 7.000074971317812e-06,
+      "loss": 0.7544,
+      "step": 13472
+    },
+    {
+      "epoch": 0.7415377841378171,
+      "grad_norm": 0.6668731570243835,
+      "learning_rate": 6.9996776879222225e-06,
+      "loss": 0.7073,
+      "step": 13473
+    },
+    {
+      "epoch": 0.7415928229401728,
+      "grad_norm": 0.7031315565109253,
+      "learning_rate": 6.999280389497591e-06,
+      "loss": 0.7262,
+      "step": 13474
+    },
+    {
+      "epoch": 0.7416478617425285,
+      "grad_norm": 0.7426775693893433,
+      "learning_rate": 6.998883076046904e-06,
+      "loss": 0.7394,
+      "step": 13475
+    },
+    {
+      "epoch": 0.7417029005448842,
+      "grad_norm": 0.665226399898529,
+      "learning_rate": 6.9984857475731475e-06,
+      "loss": 0.7365,
+      "step": 13476
+    },
+    {
+      "epoch": 0.7417579393472398,
+      "grad_norm": 0.7762128114700317,
+      "learning_rate": 6.998088404079306e-06,
+      "loss": 0.8551,
+      "step": 13477
+    },
+    {
+      "epoch": 0.7418129781495955,
+      "grad_norm": 0.7129524350166321,
+      "learning_rate": 6.997691045568366e-06,
+      "loss": 0.7646,
+      "step": 13478
+    },
+    {
+      "epoch": 0.7418680169519511,
+      "grad_norm": 0.7199442386627197,
+      "learning_rate": 6.997293672043316e-06,
+      "loss": 0.6879,
+      "step": 13479
+    },
+    {
+      "epoch": 0.7419230557543068,
+      "grad_norm": 0.6559237241744995,
+      "learning_rate": 6.9968962835071415e-06,
+      "loss": 0.6965,
+      "step": 13480
+    },
+    {
+      "epoch": 0.7419780945566624,
+      "grad_norm": 0.7428768277168274,
+      "learning_rate": 6.996498879962829e-06,
+      "loss": 0.7748,
+      "step": 13481
+    },
+    {
+      "epoch": 0.7420331333590181,
+      "grad_norm": 0.7344076633453369,
+      "learning_rate": 6.996101461413365e-06,
+      "loss": 0.6554,
+      "step": 13482
+    },
+    {
+      "epoch": 0.7420881721613738,
+      "grad_norm": 0.7080272436141968,
+      "learning_rate": 6.995704027861736e-06,
+      "loss": 0.7335,
+      "step": 13483
+    },
+    {
+      "epoch": 0.7421432109637295,
+      "grad_norm": 0.6296887397766113,
+      "learning_rate": 6.9953065793109306e-06,
+      "loss": 0.6411,
+      "step": 13484
+    },
+    {
+      "epoch": 0.7421982497660851,
+      "grad_norm": 0.7597532868385315,
+      "learning_rate": 6.994909115763935e-06,
+      "loss": 0.8281,
+      "step": 13485
+    },
+    {
+      "epoch": 0.7422532885684407,
+      "grad_norm": 0.7059680819511414,
+      "learning_rate": 6.994511637223737e-06,
+      "loss": 0.8075,
+      "step": 13486
+    },
+    {
+      "epoch": 0.7423083273707964,
+      "grad_norm": 0.8097653388977051,
+      "learning_rate": 6.994114143693323e-06,
+      "loss": 0.772,
+      "step": 13487
+    },
+    {
+      "epoch": 0.742363366173152,
+      "grad_norm": 0.7609913945198059,
+      "learning_rate": 6.993716635175681e-06,
+      "loss": 0.8265,
+      "step": 13488
+    },
+    {
+      "epoch": 0.7424184049755077,
+      "grad_norm": 0.6209948062896729,
+      "learning_rate": 6.993319111673799e-06,
+      "loss": 0.6266,
+      "step": 13489
+    },
+    {
+      "epoch": 0.7424734437778634,
+      "grad_norm": 0.6655107140541077,
+      "learning_rate": 6.992921573190663e-06,
+      "loss": 0.7519,
+      "step": 13490
+    },
+    {
+      "epoch": 0.7425284825802191,
+      "grad_norm": 1.1243617534637451,
+      "learning_rate": 6.992524019729262e-06,
+      "loss": 0.7707,
+      "step": 13491
+    },
+    {
+      "epoch": 0.7425835213825747,
+      "grad_norm": 0.6680326461791992,
+      "learning_rate": 6.9921264512925845e-06,
+      "loss": 0.7344,
+      "step": 13492
+    },
+    {
+      "epoch": 0.7426385601849304,
+      "grad_norm": 0.7689213156700134,
+      "learning_rate": 6.991728867883618e-06,
+      "loss": 0.7591,
+      "step": 13493
+    },
+    {
+      "epoch": 0.742693598987286,
+      "grad_norm": 0.8587394952774048,
+      "learning_rate": 6.99133126950535e-06,
+      "loss": 0.6991,
+      "step": 13494
+    },
+    {
+      "epoch": 0.7427486377896417,
+      "grad_norm": 0.6736756563186646,
+      "learning_rate": 6.990933656160768e-06,
+      "loss": 0.7604,
+      "step": 13495
+    },
+    {
+      "epoch": 0.7428036765919973,
+      "grad_norm": 0.6538887023925781,
+      "learning_rate": 6.990536027852864e-06,
+      "loss": 0.7332,
+      "step": 13496
+    },
+    {
+      "epoch": 0.742858715394353,
+      "grad_norm": 0.6578357815742493,
+      "learning_rate": 6.990138384584623e-06,
+      "loss": 0.7238,
+      "step": 13497
+    },
+    {
+      "epoch": 0.7429137541967087,
+      "grad_norm": 0.6865534782409668,
+      "learning_rate": 6.989740726359035e-06,
+      "loss": 0.7012,
+      "step": 13498
+    },
+    {
+      "epoch": 0.7429687929990644,
+      "grad_norm": 0.6198129057884216,
+      "learning_rate": 6.989343053179088e-06,
+      "loss": 0.7391,
+      "step": 13499
+    },
+    {
+      "epoch": 0.74302383180142,
+      "grad_norm": 0.6929547786712646,
+      "learning_rate": 6.98894536504777e-06,
+      "loss": 0.8498,
+      "step": 13500
+    },
+    {
+      "epoch": 0.7430788706037756,
+      "grad_norm": 0.6863006353378296,
+      "learning_rate": 6.988547661968072e-06,
+      "loss": 0.6589,
+      "step": 13501
+    },
+    {
+      "epoch": 0.7431339094061313,
+      "grad_norm": 0.7490457892417908,
+      "learning_rate": 6.988149943942982e-06,
+      "loss": 0.8145,
+      "step": 13502
+    },
+    {
+      "epoch": 0.743188948208487,
+      "grad_norm": 0.6597211360931396,
+      "learning_rate": 6.987752210975489e-06,
+      "loss": 0.7786,
+      "step": 13503
+    },
+    {
+      "epoch": 0.7432439870108426,
+      "grad_norm": 0.7211003303527832,
+      "learning_rate": 6.987354463068583e-06,
+      "loss": 0.7668,
+      "step": 13504
+    },
+    {
+      "epoch": 0.7432990258131983,
+      "grad_norm": 0.6257827877998352,
+      "learning_rate": 6.9869567002252526e-06,
+      "loss": 0.7378,
+      "step": 13505
+    },
+    {
+      "epoch": 0.743354064615554,
+      "grad_norm": 0.656944751739502,
+      "learning_rate": 6.986558922448488e-06,
+      "loss": 0.6408,
+      "step": 13506
+    },
+    {
+      "epoch": 0.7434091034179097,
+      "grad_norm": 0.6862110495567322,
+      "learning_rate": 6.986161129741276e-06,
+      "loss": 0.7648,
+      "step": 13507
+    },
+    {
+      "epoch": 0.7434641422202652,
+      "grad_norm": 0.6216374039649963,
+      "learning_rate": 6.985763322106612e-06,
+      "loss": 0.6826,
+      "step": 13508
+    },
+    {
+      "epoch": 0.7435191810226209,
+      "grad_norm": 0.7959128618240356,
+      "learning_rate": 6.985365499547479e-06,
+      "loss": 0.7554,
+      "step": 13509
+    },
+    {
+      "epoch": 0.7435742198249766,
+      "grad_norm": 0.5882300734519958,
+      "learning_rate": 6.984967662066875e-06,
+      "loss": 0.6523,
+      "step": 13510
+    },
+    {
+      "epoch": 0.7436292586273323,
+      "grad_norm": 0.8529833555221558,
+      "learning_rate": 6.9845698096677805e-06,
+      "loss": 0.7871,
+      "step": 13511
+    },
+    {
+      "epoch": 0.7436842974296879,
+      "grad_norm": 1.2988953590393066,
+      "learning_rate": 6.9841719423531925e-06,
+      "loss": 0.708,
+      "step": 13512
+    },
+    {
+      "epoch": 0.7437393362320436,
+      "grad_norm": 0.6735696792602539,
+      "learning_rate": 6.983774060126101e-06,
+      "loss": 0.7962,
+      "step": 13513
+    },
+    {
+      "epoch": 0.7437943750343993,
+      "grad_norm": 0.8145982623100281,
+      "learning_rate": 6.9833761629894925e-06,
+      "loss": 0.9067,
+      "step": 13514
+    },
+    {
+      "epoch": 0.743849413836755,
+      "grad_norm": 0.7107387781143188,
+      "learning_rate": 6.98297825094636e-06,
+      "loss": 0.7986,
+      "step": 13515
+    },
+    {
+      "epoch": 0.7439044526391105,
+      "grad_norm": 0.7350436449050903,
+      "learning_rate": 6.9825803239996934e-06,
+      "loss": 0.7724,
+      "step": 13516
+    },
+    {
+      "epoch": 0.7439594914414662,
+      "grad_norm": 0.7300962805747986,
+      "learning_rate": 6.982182382152485e-06,
+      "loss": 0.734,
+      "step": 13517
+    },
+    {
+      "epoch": 0.7440145302438219,
+      "grad_norm": 0.7088475823402405,
+      "learning_rate": 6.981784425407724e-06,
+      "loss": 0.818,
+      "step": 13518
+    },
+    {
+      "epoch": 0.7440695690461776,
+      "grad_norm": 0.6911785006523132,
+      "learning_rate": 6.981386453768402e-06,
+      "loss": 0.6857,
+      "step": 13519
+    },
+    {
+      "epoch": 0.7441246078485332,
+      "grad_norm": 0.794143795967102,
+      "learning_rate": 6.980988467237508e-06,
+      "loss": 0.7496,
+      "step": 13520
+    },
+    {
+      "epoch": 0.7441796466508889,
+      "grad_norm": 0.7116371989250183,
+      "learning_rate": 6.980590465818037e-06,
+      "loss": 0.7082,
+      "step": 13521
+    },
+    {
+      "epoch": 0.7442346854532446,
+      "grad_norm": 0.6306180953979492,
+      "learning_rate": 6.980192449512978e-06,
+      "loss": 0.7227,
+      "step": 13522
+    },
+    {
+      "epoch": 0.7442897242556002,
+      "grad_norm": 0.6662481427192688,
+      "learning_rate": 6.979794418325323e-06,
+      "loss": 0.7323,
+      "step": 13523
+    },
+    {
+      "epoch": 0.7443447630579558,
+      "grad_norm": 0.6824387907981873,
+      "learning_rate": 6.97939637225806e-06,
+      "loss": 0.7188,
+      "step": 13524
+    },
+    {
+      "epoch": 0.7443998018603115,
+      "grad_norm": 0.7429190278053284,
+      "learning_rate": 6.9789983113141865e-06,
+      "loss": 0.7818,
+      "step": 13525
+    },
+    {
+      "epoch": 0.7444548406626672,
+      "grad_norm": 0.7148364782333374,
+      "learning_rate": 6.978600235496692e-06,
+      "loss": 0.7665,
+      "step": 13526
+    },
+    {
+      "epoch": 0.7445098794650229,
+      "grad_norm": 0.711482584476471,
+      "learning_rate": 6.978202144808567e-06,
+      "loss": 0.7865,
+      "step": 13527
+    },
+    {
+      "epoch": 0.7445649182673785,
+      "grad_norm": 0.6913465857505798,
+      "learning_rate": 6.977804039252802e-06,
+      "loss": 0.8206,
+      "step": 13528
+    },
+    {
+      "epoch": 0.7446199570697342,
+      "grad_norm": 0.9090713858604431,
+      "learning_rate": 6.977405918832394e-06,
+      "loss": 0.7243,
+      "step": 13529
+    },
+    {
+      "epoch": 0.7446749958720899,
+      "grad_norm": 0.7680408954620361,
+      "learning_rate": 6.977007783550331e-06,
+      "loss": 0.847,
+      "step": 13530
+    },
+    {
+      "epoch": 0.7447300346744454,
+      "grad_norm": 0.6486232876777649,
+      "learning_rate": 6.976609633409608e-06,
+      "loss": 0.7258,
+      "step": 13531
+    },
+    {
+      "epoch": 0.7447850734768011,
+      "grad_norm": 0.7612336277961731,
+      "learning_rate": 6.976211468413214e-06,
+      "loss": 0.7452,
+      "step": 13532
+    },
+    {
+      "epoch": 0.7448401122791568,
+      "grad_norm": 0.7539309859275818,
+      "learning_rate": 6.975813288564146e-06,
+      "loss": 0.8292,
+      "step": 13533
+    },
+    {
+      "epoch": 0.7448951510815125,
+      "grad_norm": 0.64984530210495,
+      "learning_rate": 6.975415093865394e-06,
+      "loss": 0.6818,
+      "step": 13534
+    },
+    {
+      "epoch": 0.7449501898838681,
+      "grad_norm": 0.6415309309959412,
+      "learning_rate": 6.9750168843199506e-06,
+      "loss": 0.7369,
+      "step": 13535
+    },
+    {
+      "epoch": 0.7450052286862238,
+      "grad_norm": 0.7107319235801697,
+      "learning_rate": 6.974618659930807e-06,
+      "loss": 0.7364,
+      "step": 13536
+    },
+    {
+      "epoch": 0.7450602674885795,
+      "grad_norm": 0.7358448505401611,
+      "learning_rate": 6.9742204207009605e-06,
+      "loss": 0.7784,
+      "step": 13537
+    },
+    {
+      "epoch": 0.7451153062909351,
+      "grad_norm": 0.6950068473815918,
+      "learning_rate": 6.9738221666334e-06,
+      "loss": 0.792,
+      "step": 13538
+    },
+    {
+      "epoch": 0.7451703450932907,
+      "grad_norm": 0.7355311512947083,
+      "learning_rate": 6.973423897731122e-06,
+      "loss": 0.7631,
+      "step": 13539
+    },
+    {
+      "epoch": 0.7452253838956464,
+      "grad_norm": 0.6813983917236328,
+      "learning_rate": 6.9730256139971175e-06,
+      "loss": 0.7397,
+      "step": 13540
+    },
+    {
+      "epoch": 0.7452804226980021,
+      "grad_norm": 0.7698497772216797,
+      "learning_rate": 6.9726273154343806e-06,
+      "loss": 0.7769,
+      "step": 13541
+    },
+    {
+      "epoch": 0.7453354615003578,
+      "grad_norm": 0.7406428456306458,
+      "learning_rate": 6.972229002045905e-06,
+      "loss": 0.6502,
+      "step": 13542
+    },
+    {
+      "epoch": 0.7453905003027134,
+      "grad_norm": 0.6976667046546936,
+      "learning_rate": 6.9718306738346846e-06,
+      "loss": 0.773,
+      "step": 13543
+    },
+    {
+      "epoch": 0.745445539105069,
+      "grad_norm": 0.6932592391967773,
+      "learning_rate": 6.9714323308037115e-06,
+      "loss": 0.7315,
+      "step": 13544
+    },
+    {
+      "epoch": 0.7455005779074247,
+      "grad_norm": 0.7329851984977722,
+      "learning_rate": 6.971033972955981e-06,
+      "loss": 0.7432,
+      "step": 13545
+    },
+    {
+      "epoch": 0.7455556167097804,
+      "grad_norm": 0.6262860298156738,
+      "learning_rate": 6.970635600294489e-06,
+      "loss": 0.6368,
+      "step": 13546
+    },
+    {
+      "epoch": 0.745610655512136,
+      "grad_norm": 0.7157273292541504,
+      "learning_rate": 6.970237212822225e-06,
+      "loss": 0.7209,
+      "step": 13547
+    },
+    {
+      "epoch": 0.7456656943144917,
+      "grad_norm": 0.7256374955177307,
+      "learning_rate": 6.9698388105421855e-06,
+      "loss": 0.794,
+      "step": 13548
+    },
+    {
+      "epoch": 0.7457207331168474,
+      "grad_norm": 0.7763124704360962,
+      "learning_rate": 6.969440393457365e-06,
+      "loss": 0.7211,
+      "step": 13549
+    },
+    {
+      "epoch": 0.7457757719192031,
+      "grad_norm": 0.7139148712158203,
+      "learning_rate": 6.9690419615707585e-06,
+      "loss": 0.6612,
+      "step": 13550
+    },
+    {
+      "epoch": 0.7458308107215587,
+      "grad_norm": 0.7532974481582642,
+      "learning_rate": 6.968643514885359e-06,
+      "loss": 0.6952,
+      "step": 13551
+    },
+    {
+      "epoch": 0.7458858495239143,
+      "grad_norm": 0.6845714449882507,
+      "learning_rate": 6.968245053404161e-06,
+      "loss": 0.6972,
+      "step": 13552
+    },
+    {
+      "epoch": 0.74594088832627,
+      "grad_norm": 0.7445462346076965,
+      "learning_rate": 6.967846577130162e-06,
+      "loss": 0.7826,
+      "step": 13553
+    },
+    {
+      "epoch": 0.7459959271286257,
+      "grad_norm": 0.7269366383552551,
+      "learning_rate": 6.967448086066353e-06,
+      "loss": 0.7353,
+      "step": 13554
+    },
+    {
+      "epoch": 0.7460509659309813,
+      "grad_norm": 0.7366362810134888,
+      "learning_rate": 6.967049580215732e-06,
+      "loss": 0.7955,
+      "step": 13555
+    },
+    {
+      "epoch": 0.746106004733337,
+      "grad_norm": 0.6456870436668396,
+      "learning_rate": 6.966651059581292e-06,
+      "loss": 0.7467,
+      "step": 13556
+    },
+    {
+      "epoch": 0.7461610435356927,
+      "grad_norm": 0.7196624279022217,
+      "learning_rate": 6.966252524166031e-06,
+      "loss": 0.6621,
+      "step": 13557
+    },
+    {
+      "epoch": 0.7462160823380484,
+      "grad_norm": 0.6776413917541504,
+      "learning_rate": 6.965853973972941e-06,
+      "loss": 0.7647,
+      "step": 13558
+    },
+    {
+      "epoch": 0.746271121140404,
+      "grad_norm": 0.7319629192352295,
+      "learning_rate": 6.9654554090050195e-06,
+      "loss": 0.8172,
+      "step": 13559
+    },
+    {
+      "epoch": 0.7463261599427596,
+      "grad_norm": 0.6995210647583008,
+      "learning_rate": 6.96505682926526e-06,
+      "loss": 0.7252,
+      "step": 13560
+    },
+    {
+      "epoch": 0.7463811987451153,
+      "grad_norm": 0.6520518064498901,
+      "learning_rate": 6.964658234756659e-06,
+      "loss": 0.6856,
+      "step": 13561
+    },
+    {
+      "epoch": 0.746436237547471,
+      "grad_norm": 0.7562724947929382,
+      "learning_rate": 6.964259625482215e-06,
+      "loss": 0.7088,
+      "step": 13562
+    },
+    {
+      "epoch": 0.7464912763498266,
+      "grad_norm": 0.788045346736908,
+      "learning_rate": 6.963861001444919e-06,
+      "loss": 0.7183,
+      "step": 13563
+    },
+    {
+      "epoch": 0.7465463151521823,
+      "grad_norm": 0.7461729049682617,
+      "learning_rate": 6.96346236264777e-06,
+      "loss": 0.6725,
+      "step": 13564
+    },
+    {
+      "epoch": 0.746601353954538,
+      "grad_norm": 0.7283952832221985,
+      "learning_rate": 6.963063709093764e-06,
+      "loss": 0.7765,
+      "step": 13565
+    },
+    {
+      "epoch": 0.7466563927568937,
+      "grad_norm": 0.7947741150856018,
+      "learning_rate": 6.962665040785896e-06,
+      "loss": 0.8423,
+      "step": 13566
+    },
+    {
+      "epoch": 0.7467114315592492,
+      "grad_norm": 0.7964398264884949,
+      "learning_rate": 6.962266357727164e-06,
+      "loss": 0.7589,
+      "step": 13567
+    },
+    {
+      "epoch": 0.7467664703616049,
+      "grad_norm": 0.7807595133781433,
+      "learning_rate": 6.961867659920563e-06,
+      "loss": 0.7843,
+      "step": 13568
+    },
+    {
+      "epoch": 0.7468215091639606,
+      "grad_norm": 0.678011417388916,
+      "learning_rate": 6.961468947369089e-06,
+      "loss": 0.6664,
+      "step": 13569
+    },
+    {
+      "epoch": 0.7468765479663163,
+      "grad_norm": 0.6768447756767273,
+      "learning_rate": 6.961070220075741e-06,
+      "loss": 0.7531,
+      "step": 13570
+    },
+    {
+      "epoch": 0.7469315867686719,
+      "grad_norm": 0.7405245304107666,
+      "learning_rate": 6.960671478043514e-06,
+      "loss": 0.8278,
+      "step": 13571
+    },
+    {
+      "epoch": 0.7469866255710276,
+      "grad_norm": 0.605675458908081,
+      "learning_rate": 6.960272721275403e-06,
+      "loss": 0.7167,
+      "step": 13572
+    },
+    {
+      "epoch": 0.7470416643733833,
+      "grad_norm": 0.7406657338142395,
+      "learning_rate": 6.959873949774409e-06,
+      "loss": 0.8191,
+      "step": 13573
+    },
+    {
+      "epoch": 0.7470967031757388,
+      "grad_norm": 0.6163522601127625,
+      "learning_rate": 6.959475163543526e-06,
+      "loss": 0.6711,
+      "step": 13574
+    },
+    {
+      "epoch": 0.7471517419780945,
+      "grad_norm": 0.6036590337753296,
+      "learning_rate": 6.9590763625857525e-06,
+      "loss": 0.7029,
+      "step": 13575
+    },
+    {
+      "epoch": 0.7472067807804502,
+      "grad_norm": 0.8638957738876343,
+      "learning_rate": 6.9586775469040845e-06,
+      "loss": 0.6288,
+      "step": 13576
+    },
+    {
+      "epoch": 0.7472618195828059,
+      "grad_norm": 0.7490845322608948,
+      "learning_rate": 6.958278716501521e-06,
+      "loss": 0.7375,
+      "step": 13577
+    },
+    {
+      "epoch": 0.7473168583851615,
+      "grad_norm": 0.7788114547729492,
+      "learning_rate": 6.957879871381059e-06,
+      "loss": 0.814,
+      "step": 13578
+    },
+    {
+      "epoch": 0.7473718971875172,
+      "grad_norm": 0.7247292995452881,
+      "learning_rate": 6.957481011545697e-06,
+      "loss": 0.6187,
+      "step": 13579
+    },
+    {
+      "epoch": 0.7474269359898729,
+      "grad_norm": 0.9642785787582397,
+      "learning_rate": 6.95708213699843e-06,
+      "loss": 0.8745,
+      "step": 13580
+    },
+    {
+      "epoch": 0.7474819747922286,
+      "grad_norm": 0.701675295829773,
+      "learning_rate": 6.956683247742259e-06,
+      "loss": 0.8474,
+      "step": 13581
+    },
+    {
+      "epoch": 0.7475370135945841,
+      "grad_norm": 0.6338050961494446,
+      "learning_rate": 6.9562843437801795e-06,
+      "loss": 0.7346,
+      "step": 13582
+    },
+    {
+      "epoch": 0.7475920523969398,
+      "grad_norm": 0.6954126358032227,
+      "learning_rate": 6.955885425115191e-06,
+      "loss": 0.8083,
+      "step": 13583
+    },
+    {
+      "epoch": 0.7476470911992955,
+      "grad_norm": 0.7316300272941589,
+      "learning_rate": 6.95548649175029e-06,
+      "loss": 0.8009,
+      "step": 13584
+    },
+    {
+      "epoch": 0.7477021300016512,
+      "grad_norm": 0.6314196586608887,
+      "learning_rate": 6.955087543688477e-06,
+      "loss": 0.6375,
+      "step": 13585
+    },
+    {
+      "epoch": 0.7477571688040068,
+      "grad_norm": 0.6604906320571899,
+      "learning_rate": 6.9546885809327495e-06,
+      "loss": 0.7081,
+      "step": 13586
+    },
+    {
+      "epoch": 0.7478122076063625,
+      "grad_norm": 0.8251973986625671,
+      "learning_rate": 6.9542896034861064e-06,
+      "loss": 0.7483,
+      "step": 13587
+    },
+    {
+      "epoch": 0.7478672464087182,
+      "grad_norm": 0.6946399211883545,
+      "learning_rate": 6.953890611351544e-06,
+      "loss": 0.8849,
+      "step": 13588
+    },
+    {
+      "epoch": 0.7479222852110738,
+      "grad_norm": 0.7713609933853149,
+      "learning_rate": 6.953491604532063e-06,
+      "loss": 0.7913,
+      "step": 13589
+    },
+    {
+      "epoch": 0.7479773240134294,
+      "grad_norm": 0.734355092048645,
+      "learning_rate": 6.953092583030664e-06,
+      "loss": 0.7216,
+      "step": 13590
+    },
+    {
+      "epoch": 0.7480323628157851,
+      "grad_norm": 0.6147064566612244,
+      "learning_rate": 6.952693546850342e-06,
+      "loss": 0.6894,
+      "step": 13591
+    },
+    {
+      "epoch": 0.7480874016181408,
+      "grad_norm": 0.7472255229949951,
+      "learning_rate": 6.9522944959940986e-06,
+      "loss": 0.7941,
+      "step": 13592
+    },
+    {
+      "epoch": 0.7481424404204965,
+      "grad_norm": 0.6478431224822998,
+      "learning_rate": 6.951895430464935e-06,
+      "loss": 0.6995,
+      "step": 13593
+    },
+    {
+      "epoch": 0.7481974792228521,
+      "grad_norm": 0.6956225633621216,
+      "learning_rate": 6.951496350265844e-06,
+      "loss": 0.7637,
+      "step": 13594
+    },
+    {
+      "epoch": 0.7482525180252078,
+      "grad_norm": 1.0637938976287842,
+      "learning_rate": 6.95109725539983e-06,
+      "loss": 0.7448,
+      "step": 13595
+    },
+    {
+      "epoch": 0.7483075568275634,
+      "grad_norm": 0.6948299407958984,
+      "learning_rate": 6.9506981458698916e-06,
+      "loss": 0.7343,
+      "step": 13596
+    },
+    {
+      "epoch": 0.7483625956299191,
+      "grad_norm": 0.9034255743026733,
+      "learning_rate": 6.950299021679028e-06,
+      "loss": 0.6481,
+      "step": 13597
+    },
+    {
+      "epoch": 0.7484176344322747,
+      "grad_norm": 0.7901731729507446,
+      "learning_rate": 6.949899882830239e-06,
+      "loss": 0.8368,
+      "step": 13598
+    },
+    {
+      "epoch": 0.7484726732346304,
+      "grad_norm": 0.7791730761528015,
+      "learning_rate": 6.949500729326525e-06,
+      "loss": 0.7912,
+      "step": 13599
+    },
+    {
+      "epoch": 0.7485277120369861,
+      "grad_norm": 0.7678626179695129,
+      "learning_rate": 6.949101561170883e-06,
+      "loss": 0.7514,
+      "step": 13600
+    },
+    {
+      "epoch": 0.7485827508393418,
+      "grad_norm": 0.709762454032898,
+      "learning_rate": 6.948702378366318e-06,
+      "loss": 0.6809,
+      "step": 13601
+    },
+    {
+      "epoch": 0.7486377896416974,
+      "grad_norm": 0.706031084060669,
+      "learning_rate": 6.948303180915827e-06,
+      "loss": 0.7454,
+      "step": 13602
+    },
+    {
+      "epoch": 0.748692828444053,
+      "grad_norm": 0.658869743347168,
+      "learning_rate": 6.9479039688224105e-06,
+      "loss": 0.6498,
+      "step": 13603
+    },
+    {
+      "epoch": 0.7487478672464087,
+      "grad_norm": 0.7253865599632263,
+      "learning_rate": 6.9475047420890685e-06,
+      "loss": 0.8063,
+      "step": 13604
+    },
+    {
+      "epoch": 0.7488029060487644,
+      "grad_norm": 0.752839207649231,
+      "learning_rate": 6.947105500718804e-06,
+      "loss": 0.7708,
+      "step": 13605
+    },
+    {
+      "epoch": 0.74885794485112,
+      "grad_norm": 0.6694571375846863,
+      "learning_rate": 6.946706244714615e-06,
+      "loss": 0.7121,
+      "step": 13606
+    },
+    {
+      "epoch": 0.7489129836534757,
+      "grad_norm": 0.751380443572998,
+      "learning_rate": 6.946306974079503e-06,
+      "loss": 0.8797,
+      "step": 13607
+    },
+    {
+      "epoch": 0.7489680224558314,
+      "grad_norm": 0.8001984357833862,
+      "learning_rate": 6.9459076888164676e-06,
+      "loss": 0.8963,
+      "step": 13608
+    },
+    {
+      "epoch": 0.7490230612581871,
+      "grad_norm": 0.7149432301521301,
+      "learning_rate": 6.945508388928511e-06,
+      "loss": 0.8311,
+      "step": 13609
+    },
+    {
+      "epoch": 0.7490781000605427,
+      "grad_norm": 0.8295183777809143,
+      "learning_rate": 6.945109074418635e-06,
+      "loss": 0.7466,
+      "step": 13610
+    },
+    {
+      "epoch": 0.7491331388628983,
+      "grad_norm": 0.7480556964874268,
+      "learning_rate": 6.94470974528984e-06,
+      "loss": 0.8277,
+      "step": 13611
+    },
+    {
+      "epoch": 0.749188177665254,
+      "grad_norm": 0.7962234616279602,
+      "learning_rate": 6.944310401545127e-06,
+      "loss": 0.7143,
+      "step": 13612
+    },
+    {
+      "epoch": 0.7492432164676097,
+      "grad_norm": 0.7722699642181396,
+      "learning_rate": 6.943911043187497e-06,
+      "loss": 0.6619,
+      "step": 13613
+    },
+    {
+      "epoch": 0.7492982552699653,
+      "grad_norm": 0.8495624661445618,
+      "learning_rate": 6.943511670219952e-06,
+      "loss": 0.8475,
+      "step": 13614
+    },
+    {
+      "epoch": 0.749353294072321,
+      "grad_norm": 0.7702826261520386,
+      "learning_rate": 6.943112282645494e-06,
+      "loss": 0.826,
+      "step": 13615
+    },
+    {
+      "epoch": 0.7494083328746767,
+      "grad_norm": 0.7435297966003418,
+      "learning_rate": 6.942712880467124e-06,
+      "loss": 0.8121,
+      "step": 13616
+    },
+    {
+      "epoch": 0.7494633716770323,
+      "grad_norm": 0.8108325600624084,
+      "learning_rate": 6.942313463687844e-06,
+      "loss": 0.7282,
+      "step": 13617
+    },
+    {
+      "epoch": 0.7495184104793879,
+      "grad_norm": 0.6840381622314453,
+      "learning_rate": 6.9419140323106574e-06,
+      "loss": 0.7446,
+      "step": 13618
+    },
+    {
+      "epoch": 0.7495734492817436,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 6.941514586338562e-06,
+      "loss": 0.7598,
+      "step": 13619
+    },
+    {
+      "epoch": 0.7496284880840993,
+      "grad_norm": 0.7693290114402771,
+      "learning_rate": 6.941115125774564e-06,
+      "loss": 0.7666,
+      "step": 13620
+    },
+    {
+      "epoch": 0.7496835268864549,
+      "grad_norm": 0.6918750405311584,
+      "learning_rate": 6.940715650621665e-06,
+      "loss": 0.6831,
+      "step": 13621
+    },
+    {
+      "epoch": 0.7497385656888106,
+      "grad_norm": 0.8241471648216248,
+      "learning_rate": 6.9403161608828654e-06,
+      "loss": 0.6753,
+      "step": 13622
+    },
+    {
+      "epoch": 0.7497936044911663,
+      "grad_norm": 0.6659193634986877,
+      "learning_rate": 6.93991665656117e-06,
+      "loss": 0.6988,
+      "step": 13623
+    },
+    {
+      "epoch": 0.749848643293522,
+      "grad_norm": 0.8012998700141907,
+      "learning_rate": 6.9395171376595795e-06,
+      "loss": 0.7922,
+      "step": 13624
+    },
+    {
+      "epoch": 0.7499036820958775,
+      "grad_norm": 0.783018946647644,
+      "learning_rate": 6.9391176041810974e-06,
+      "loss": 0.7062,
+      "step": 13625
+    },
+    {
+      "epoch": 0.7499587208982332,
+      "grad_norm": 0.8228014707565308,
+      "learning_rate": 6.938718056128726e-06,
+      "loss": 0.7762,
+      "step": 13626
+    },
+    {
+      "epoch": 0.7500137597005889,
+      "grad_norm": 0.783525288105011,
+      "learning_rate": 6.9383184935054705e-06,
+      "loss": 0.7517,
+      "step": 13627
+    },
+    {
+      "epoch": 0.7500687985029446,
+      "grad_norm": 0.6686612963676453,
+      "learning_rate": 6.93791891631433e-06,
+      "loss": 0.7372,
+      "step": 13628
+    },
+    {
+      "epoch": 0.7501238373053002,
+      "grad_norm": 0.7089647054672241,
+      "learning_rate": 6.937519324558312e-06,
+      "loss": 0.7847,
+      "step": 13629
+    },
+    {
+      "epoch": 0.7501788761076559,
+      "grad_norm": 0.7674399018287659,
+      "learning_rate": 6.937119718240415e-06,
+      "loss": 0.7414,
+      "step": 13630
+    },
+    {
+      "epoch": 0.7502339149100116,
+      "grad_norm": 0.6331565380096436,
+      "learning_rate": 6.936720097363646e-06,
+      "loss": 0.7603,
+      "step": 13631
+    },
+    {
+      "epoch": 0.7502889537123673,
+      "grad_norm": 0.7084798812866211,
+      "learning_rate": 6.9363204619310065e-06,
+      "loss": 0.6844,
+      "step": 13632
+    },
+    {
+      "epoch": 0.7503439925147228,
+      "grad_norm": 0.8624362945556641,
+      "learning_rate": 6.9359208119455015e-06,
+      "loss": 0.7098,
+      "step": 13633
+    },
+    {
+      "epoch": 0.7503990313170785,
+      "grad_norm": 0.7681849598884583,
+      "learning_rate": 6.935521147410134e-06,
+      "loss": 0.7896,
+      "step": 13634
+    },
+    {
+      "epoch": 0.7504540701194342,
+      "grad_norm": 0.7494263052940369,
+      "learning_rate": 6.935121468327907e-06,
+      "loss": 0.7858,
+      "step": 13635
     }
   ],
   "logging_steps": 1,
@@ -82745,7 +95471,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.4872691283585925e+19,
+  "total_flos": 4.023772071182991e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null