diff --git a/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/README.md b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/README.md
new file mode 100644
index 0000000000..1a42db4980
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/README.md
@@ -0,0 +1,53 @@
+# Record: SP4096 + Depth Recurrence + Parallel Residuals + MuonEq-R + Legal TTT — val_bpb 1.0896 (3-seed mean)
+
+**val_bpb = 1.0896** (3-seed mean, std 0.0008) | **~15.99 MB** | 8xH100 SXM
+
+## 3-Seed Results (8xH100 80GB SXM, PyTorch 2.9.1+cu128)
+
+| Seed | Sliding BPB | **TTT BPB** | TTT gain | Artifact |
+|------|-------------|-------------|----------|----------|
+| 42   | 1.0896      | **1.0889**  | -0.0007  | 15,999,165 |
+| 314  | 1.0915      | **1.0906**  | -0.0010  | 15,974,112 |
+| 999  | 1.0901      | **1.0894**  | -0.0007  | 15,996,001 |
+| **Mean** | | **1.0896** | **-0.0008** | |
+
+Merged SOTA (PR #1019): **1.1147 BPB**. Delta: **-0.0251 BPB**.
+
+## Key Techniques
+
+1. **4096-Vocab + MLP 4x + WD 0.090** — PR #1218 @clarkkev, PR #1285 @dexhunter
+2. **Depth Recurrence (layers 4,5)** — PR #1204 @msisovic, PR #1260 @dexhunter
+3. **Parallel Residuals (from layer 7)** — PR #1204 @msisovic, PR #1289 @MatoTeziTanka
+4. **MuonEq-R** — arXiv:2603.28254, PR #1260 @dexhunter
+5. **QK-Gain 5.0** — PR #1217 @bigbag
+6. **Legal Score-First TTT** — score each 32K-token chunk under torch.no_grad before SGD training. Compiled scoring for correctness. PR #461 @Christopher-Lee-McClendon
+7. **Full GPTQ int6 + Brotli + Compressed Wrapper**
+
+## TTT Compliance
+
+Legal score-first per PR #461 framework:
+- Every token scored BEFORE any weight update (enforced by torch.no_grad + compiled scoring)
+- No training data access during evaluation
+- No multi-epoch scoring — each chunk scored exactly once
+- Total eval time: ~600s (sliding ~100s + TTT ~300s)
+
+## Compliance
+
+- Legal score-first TTT (backward-looking only)
+- No SLOT, no n-gram cache
+- GPTQ calibration within training budget
+- All four conditions from Issue #1017 satisfied
+
+## Reproduction
+
+```bash
+pip install brotli
+MATCHED_FINEWEB_REPO_ID=kevclark/parameter-golf python3 data/cached_challenge_fineweb.py --variant sp4096 --skip-manifest
+SEED=42 RECUR_LAYERS=4,5 RECUR_START_STEP=3000 PARALLEL_START_LAYER=7 \
+TTT_ENABLED=1 TTT_LR=0.002 TTT_EPOCHS=3 TTT_CHUNK_TOKENS=32768 TTT_FREEZE_BLOCKS=0 \
+torchrun --standalone --nproc_per_node=8 train_gpt.py
+```
+
+## Credits
+
+PR #1218 @clarkkev, PR #1285 @dexhunter, PR #1204 @msisovic, PR #1289 @MatoTeziTanka, PR #1260 @dexhunter, PR #1019 @abaybektursun, PR #1287 @dentity007, PR #1217 @bigbag, PR #493 @parinzee, PR #461 @Christopher-Lee-McClendon
diff --git a/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/submission.json b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/submission.json
new file mode 100644
index 0000000000..7c7c7330b1
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/submission.json
@@ -0,0 +1,20 @@
+{
+  "author": "aryanbhosale",
+  "github_id": "aryanbhosale",
+  "name": "SP4096 + Depth Recurrence + Parallel Residuals + MuonEq-R + QK-Gain 5.0 + Legal TTT",
+  "date": "2026-04-04",
+  "track": "10min_16mb",
+  "val_bpb": 1.08963553,
+  "val_bpb_std": 0.00083386,
+  "seeds": [42, 314, 999],
+  "seed_results": {
+    "42": {"val_bpb": 1.08894061, "artifact_bytes": 15999165},
+    "314": {"val_bpb": 1.09056017, "artifact_bytes": 15974112},
+    "999": {"val_bpb": 1.08940582, "artifact_bytes": 15996001}
+  },
+  "comparison_baseline_pr": 1019,
+  "delta_vs_pr1019_bpb": -0.02509956,
+  "hardware": "8xH100 80GB SXM",
+  "pytorch_version": "2.9.1+cu128",
+  "technique_summary": "SP4096 + MLP 4x + WD 0.090 + Depth Recurrence + Parallel Residuals + MuonEq-R + QK-Gain 5.0 + Legal Score-First TTT + Full GPTQ int6 + Brotli"
+}
diff --git a/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_gpt.py b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_gpt.py
new file mode 100644
index 0000000000..953c429a21
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_gpt.py
@@ -0,0 +1,2 @@
+import lzma as L,base64 as B
+exec(L.decompress(B.b85decode("{Wp48S^xk9=GL@E0stWa8~^|S5YJf5;YX`WBV7PAn@VT6Qap3bu0*kgCR~YUqB0W9R)iarr*QtEZpesGY3>~CZRiK|6Dwut$nH#N""!RYqQnA}G^`ZsFO;ar92)Xt#3E3Ki5S1}OfSx<=$c<4=h|J{kt$27^CQ01M+lVgZ0tGgX0&I*V@{U&JgYc0U!(4F-btCy*+qzv6D""p~UW!y~6{U*}y$E@2-R}vd?t*s#fnDO{!j>OImt34A(d+9n>hnnvzmd((<?7SwGA+W$Or}R^?z{Ree`d98K#2b?5_cfZ%Lt17{dO""-GN2&5;3ECS=cE#O*hTiZtOgX4XIkBuR{PAJ5$6e;Eel8-VVQFtsWBkSb@w{YAVOHKclDdD}`(Z5P0z`*Zpr!j8psZi3dj0Awg*|""SM}VUSWz|5#6=0zD3_4d4q0&x{n`YrJE1HXz*`$RPocfD=uX!IiV{O9O7|!NUnrgGHr-RntBL^weTkpGrkcD-11{fi$WX(*bn-nb""gD6YkvCezcUG_I`KWJk#hl}8>_D1Cghg~(bQ$Yj)>!Y%{*o9ex8FWa#U)!OI!!5Prl^?bnBX2V(=(Bvc+CvGo!S{LhLn7pSsR!}@""U=OBW0)h6IYneQ1{|$<&k9TS^qGQpb-;#vEPAl%11UF)?6mtC8c04XzR$+h2=j84E2|i`pOEt$uyM`lGs*ejIF-}^SvSRZK$ePh1""`gt+?%1r#=OV<HVjrIqeJYGxh%pSpPlDdQ66RHQe_t;x^R`DvG*6PkO>y3pW`{ofc)6PhQQRP|_h56zl+sQ(le1eJ^&&qZxdGb15""aOb^-R1ouqi-H1_w|H;g(()bKz_!0+L{#HFmtQSw%~n|MX3ij_2{lW(_*6gdIz`XT%tzkhK-k5tAu}`=>u|z+<A5OWlS!1&|7Bsh""La`v33WEyuh5cV6d%c)Zp^vZ`+%XW7Ga$QNikdrw8}teN{lfKFYW8jLKpUE<fS0HI?!#S}z5(PnD*$%*&e)`J?Z?l)B@7}E*EtBQ""<ajK2J2UVwA>|uP4UnMNw^{d&KAm;P6`40&zphh*D=e*8?KGZuo~y*`y#Wg}r(PV<PiL09`OF@Nh{a<w{zds*!NOI%lISf!q#(&1""k8wXKJfU5I>}?J$O<io<ujGSX9NWFVj^&nmc<AiUyr-`}m?om^MsCAZ)88TX4@F=Gr{ba+%8mF)D2K;PS)r}K`0OW1F4MRM)*s7t""_jmGr27SW%DBvhmeslZ(N-kfEgbZN}VIhl%?M7XjWH8wf?lo_X|5_oRRVxUolXyMc*~(GxY5pkyJ-C9J))sNjfO=O;m+FUfndz~U""9#}!)KfJsF`v`a^5o$1$QiNNu!W#ZYR)zDNo?=6w(J-Mr%~Ccd@JvU7j6*UqI1G_U5)xIOVfFefW#-2fZay)3C_@H2WgGU<<j3Y(""=kG=f*7C;k&v(6y7C}?)ki&L~IiRiQ(NiyD1>ae&vv%2(eb|cSn}oTUc1&B%M^B@xlvn10$Ol5{Rc(||g*`0AO8zk14;L<B89tNM""DJH<nBP9kHyG#H@%Y{;9I%*-8=cLIDmul}ETySn#xzQAAJ)H3oDC_09q<t7H;T?vgXJnFe2W2eEM#4@zx^{Q`RI4rmXYqJ`j>7H$""%dzpg>)#q=gM_#uqc-I?vb;GXT^_H=R{GSK>=VuvqSe%#z|2c2YlT&*kjZ2;Kl!jcm$Q-fVo|ict&Ja4ywb`i8dvCk+h?3OCDUXX""&zMsrr>Pa>PMm^gxGuU~bBkga*GhQ$y8PeK(vV5>_veh8WQRi`fT<y2?%qa)RDFxTOGTC_Yl#}yg8C;;?WCn_iwWXDzN03qd_<{4""@#oI{v^Qp5Ta*t1IOY_9!%73?T>)Py)OBaHdkXS7+2I5uiFuTjpn{|fPzQ@LVpRws+Pag!JmkiczrmnPtnMu9kKEb_KkhHbOR{Wt""8%3T$xbVm9!-A4Y!!lg(FQ%bN;xc;)J}M}+SWEaAW4y-yw9R0Q^AK@NSl^zI0|UPwqq?6`rga!kw{>S5i#M*_l%=<QTa4?+=kuK8""?SQ!Dyq|d0>PQ;>8$L|2FxThROI+2n1yh%V%z*SB!)&{G^cNY-1sEn*QxI>3(Mq#-79*&HQx{l(5_G&smxlM^P9n^Rw+snn;2J5K""+nT$`x$au4QlQ;UK9VKSJ^ZsCG*f{fZeE<S6&RgSd`M<O(Q7T9!o_It$&ZW$IhNQeH%l6S^rT2>UIiN6m{@t+`ew5)<bTDfA-be&""NXxje4Rgz}uG(;XBW^Fx?8gPcv2GASs~X<py*Y1_)}=TJ2B(B0U-Apr+zpZCCt^(RZ*%qumqng>LOdk-VMqWlfcBU7?nxLFgfE?x""b6Az{EuCO<ClAIjZWVwj=Tt|n1E``uD8R#hXo+A8F72#VI^Ez1U6YOmP|$ME70x6<gQxlaYdp*A{8*32TF&c8&oL-yf=7t#5m=9X""(F_#aeJ!xQ#3w^ILlZ~cBm}Cx3-pc3!`(}coA`bU@^#@A3|;$ybX(j#%|Afdm3v~C@OL&dVP<9P6m(%tso)YX*@>+S7#4-pxaPAl""(DONH5{t3k?rrWsL@mpCUHPbMPya8454xhhOJwf3(m|sY7eZT;5;fOW8%FFIEqW8&2?)-71^h*Ga9=nxiacg5nW%jY56<NH+lZ_d""-49s!C}rP(AJbw`n!(99k+EPQ;HKTrr4?*2F5A7aC&rie(7^llIC(w#()W@Fi*vlzm{EUY1*FoC6kR>jIq4Am!#Af%Ad7b;fSu8t""$W0?^M^3JvwEwjL7wvBuvv-sQt;=;pA19#M1g=W3%g!++Ta&Pj)}6%(rmy(w1Ya{^5TZ6je7V3lo*%vz3G$IqGSc1Yk<@ueD-^lP""3>l=~HATuJsnUd7lSv1hzR!o!Cqg*vpqo?H!H#F#qeqDew8JLvu8?Ch?xbP*m(b{S0`8GH-3tOgQT7;SaPNV8fm#zUwNRTcE@G@W""+jk9^I`mCeVsjz}lIoTv50U$zCf#Ui4uX%VDo5bs-vLZo_x0r`2|tIB*<d^T5JJ79VkwdLg$xR@Q&ToX44mA1G`dps9bJNJbv=RB""H^+m$hZteDb_FF!n>U}cLyXQ7=78niZCTHmt1$P#TUdM`P1r%XC!Mfom@NI{l{DZy1;89(d%L$BtDZ<k<P54+HSLbOp)yDoP5o4U""h@3Xr*t2FovO1NJOgRuaB{2S@TwT6{rMR~)R7d|2Hc;bd{o||&$qJFH%Ta$%vrZR5Z%~v(OJiK&ZK9L)(D0|se|2*Y1PA4-shr);""!-J%n5^YOWs|oI7K=_)@=ldoP)xVLE-=SFwEl77x^kEeVB8|`}bD@VcH`m@HRy}~ke$)x>FH1NaTdSZcWfMclca35;Qg22-674m;""pFNb#C7XsQB*kC4dE$?&KNO4Qs%Q~HXIRE%U)*?lT{Xj~p!?ka)_Bno=GJBn5M!Zfb>?Eqj}7jL?ct;;1tV1%i%E>in)Cx!WeUF(""<(m7t3ADHh@DgeP#pc}Eeu2?bCz3w^A+AI1(xU2Hqbbe!&Fxs|<x_@6fzr>v14@I;SI5er>_S++RxO|ur%K6xYT(6JaB}s)QlHbv""Z)6Bw+AX&rZptjNW62D6QMGF>PTDD-=uVIFXKGEg`>~}zmCKBwRaC>?s@~W?#0buJ@U`4@l&pe61;H%?;6Q6%<%})hiqO<F=QiZa"")onkHLj?^svGhE*oSrA3=0W^UpN_)uVFgSN_29_`bun%Fu0DD|-C(MV7c@H{j(a!6@d7l=24g!a=TuXWXY@(1N+!f#xUv$kNTB!W""V!G7@|9%yqX&)?yGjr`A&E7Sdm;ku4C|{;IONjpf#j^0G?KWJOGhU2F-m8pfbW;1RF6LA{;2rWkJUb}V$8F2^1jc1gkkI+W{Axlj""%dgwyADi!2e6HA>2AWA>0v<WQ2(yH{h}0<FZ4totRh&2=YvS78&p}wsx2f91SM3!+?rC`%_K#WIH9#D<UUhuq^WK`^!ZQ+<#cP=i""mNu@BFw-3h3ZuEk^=5<UI=_cnbV4u*M{W13b<)erQz6fv_EP+|I05!viv@(&#K=uU>O6s%^XlaGwMC@uO5^K8dY|}XfV|*YTjvMf""^@u$1$x6v0XJ|vHxw#$4ip?Awj`#mhj}NwB`myCgJoy;q`7y#}mkqeC7SdtKKRAU3ZU7;w<pXL1UKUT}6ASCa|4w9I{an^5qcJ-{""aF~BuvJU<Pe0M^AzNuCkCWT!_;l!A?5?Gh^bDSQW2h`rsqQfa<fK}+xLhi?_DwD-AAX57_cVr`&sx@A&O%I8oh)c6gCr*7=1wdau""foJdP9zcv!KM4M{x2n^hauFVqM@@;wkTR<yGVn@hVNi%05IqCHJhDdnURH00cWHHwD)2;SuUg`T;D7+N$(E1fTd&@nEidDAKSzz!""OB}d~F<RKd%=m>p5hW0zyiDIOE*VW{9UXqA$mN)lVbH6dhCF*>HaJ>tNE)-@TssUii}Y*CD{I4_G%`eqw#b0%fyt|;*i@e0cz=;@""5>?^i@R4Ck?NQyTO+R4Xm)XAqsI4O(u8t}56&!qffK?uZhzjPSzgY^w!+3jjeYG>Jrnj2Oq6ns@F|=M$2<WTo7?eJ>v^JTkd7g@C""M!d$)Fs^r4e|;n)DLzd}`BG1LKaonOUt%k9>o8xDr?#e`$lIbNs==FD>W{W(n$jL{;Xfkmp6DWf$U8&ND9y^j5|rhhi>Y%r<Xym~""e7rWkVq>&i(GMQU4%}Sol)2_maQ+b*(nZ9)@$bk0G0y@(GL?Zhk%>MHXV*)(OXx#s!*Q-aja3QbuK9f4p-F93>F4xs+ZTRG_@*Ox""(Tw(zGwQzlTvONhN*6L?P;rgmivbLx%mE>!68)lZxUPaWAdP6(JWyEqUi7PAab6M3Q1k^ryv#jSKMXM4ES(rB9@NC*BPnz|UVtBH""OT1m7zs3BNQReVFG!KarK*nL)BE<-R-gTg&c>4Q9)JpEMVLJi_B29>mSHPm3Poc3o!gPjs3fQD~Eu5S|t__-z`AU~AE1e51)*Qd+""9NIi=^&Y;!iv7d4C|(QPjI*`7UwTO9^AF%{n&lD1e?|U2LGb3iPE#Yx$6Ag=B%rFXeL7+Q@+q~fQ*Y)`$xl60Lj=1(7?+AC=hJcZ""@RGyvUOdLQthGpLaepQv3s9dRu6T?qdth=vex;#!oY$8*TdkE?Uu7t2#vd;^mymC%9J>0dE5o$UC64)*Ei`a1133I+!RbtUOzGxC""qbWeT_V({}2o-ybQK&}WdVq#kS~^8%y{}E37av^OP$FkBD6!ce;Z$An-<at(gw5XVuF=n(%I_(IYWC9d0}!Sntrt!Q)011<5{W<8""dArFQK5b@!^q-!}nvS0UQs+5D8vk=-(&{GgNAQ?8q-QKPDH!K(v$2Eq^c^wnGxDYbX-3khYiR2QY}#j$-uTSz%X=PmDu-T{B^H?R""Kkiq|eLdJXU|jpVh*HYnk8Oe$3r?dL!E5J-61gN#NA}WLLbK(IZ)+itEO2J<FPF>pLWFte`bj<>ht_#=44GDGkpjg<f}s6E5>BFq""sw{3b8S$6%m%t^D{{!U)vVB`@I_-LD*e?UZSDSY-=Lc`F_;H*<{abP;#Xbxn9_ZU|I6xD3!BL@;&z+;R{f~Tj{s79jIya_@!9lK)""E_L91IdRHk*q~95#av3k#D+u)w`Faav0hNP49GqaB9sF>UhHHD79l}_UhQ_`m`kW;FFhwT0Ehc?BwG;A=mU41M1aQdfo3M7hGh+`""5}rM%v6k%dO)|?EAx{2m`=>6e^^0IhV@DwRO?tlUX}Dg{%03Y91kKm5xz7pN$~u+Gh%WHERhhZh8bs*U(@8YGNY(X7TK$p}1M}7M""nnEHAjmc5&SPb|Y{H=M;i%j#$%E1KFN3)VpK}@TO0CG=M|ESDG@F$#dZ}(T3fL1KW*%faX<mORK@dup;RFDtF76w9(;ef6(V;FyU""Wu_{vQx#>oRz|B!QY8>I3Ig_piE8tde@!NznZmt}mI6&dgwM|imnVB}CvISopw7}T<MB$sOYl=97oSYf4zw`A+(v=CqFC-wa<}NQ""eZ_-tx7XdVa7w99G?!ZZ<7@{?ZQdzwY}Ramw<)3Lq`&E;<Y2b+_K8`l_@z?qBRmU@Rm14KouhpQSlox{i_XKlumeJx1GdqaQ9`{_""nTqg~HZ*F{Od|gV#vQ>ay-)hPY+i^vD$Z#QxqqdN&E$)IDNDktDpZvIYdz#<NBU_cHW1ITCbO`H7JwtX3&mZKx0#vFJ=F%V$u7|;""pNwDTqThhY%e)^@sXOkU-u~W?`DI^^RVN(FW?jT~INh<TMk}}gxF_z2hVbJ)mWK)Mcr=Q8FTAM-w?-BQcoeP}iPBuYzS*Wntp$HM""+=b87QE5~~c8pdOnF|#aa+9F2kKr@8C#n;83?yzQfZjRf;C~%s%f)Kt#?PU6C$8#8k6Mi)zPI&5n=atn{W-;t-`S5VIO4bmc;Hlm""=<;QP(^Mvp&iA?dop_H8iM(!kAeV_vb6<5SCE71PxB-)$OC0w7^PS!exgq?2C4QMWRwkn+gaT=Bp`W4390|z-y_Uw}AsQbBkR%GJ""3d{Th8MDjMGvbKLpzn5suQC{k0yQS~zX>n<WsW)8@+}JxG#K+K<96>)+<JaFIckx<Jt=0U9ia#r3}6x8*6q9VIvIQE3U4win#<h`""x<$~Pb5<6mC=aF34OtCkV?N0SZLH~HFM?3xHIse-mXT$P{p$NosRDj`5-c!(2c}t-v$``xre@(ve3NV2g9C`lOh>IdI5~qDE}*k3""@A<!CatM)MOB2By)S0aE#kMh@V%iKKeSClaS~iY>I;u*gs7t~w8u#|;;lux4I3gUbjsk-eP1izoM%ZT-{xD$Zs}hsSXP+!vQQNmb""I(>|JMux5bIBc&3Qb3@Nde7+2+`qI8(v>G2cD?e1;y8!OC&jm?8tc{d*2#T%ssy0>L@FPMBV4Y>5$Z)G;jia1o2mh+oLG9ic^jG(""RhGR8xl`bNxG|C;B}Xu}?IN4ZyU{tP)1+W;V22hyDQB2ew|jj4JVR=$N0u7-Ft1x6swd~Ix*kUnOi0N6`}FZ_#<2^Mb&!g+Cx@5m""J|)on`J6Bl9H-}Ai?pzaKyjl*DJZol(#;UxvxsQZd2wS{PV12yOh4!oxTvaCVs~Lu0N;)q*)*K4MB9Hwgkljs`2WJ9bu6@@#Ru7L""lLoMYhOlrxk8!{?H_1{(J$`oA4g3_dZ{8x%L5oeEXP1j-vYM6J5p0qJ_3b;y@LqbYhr&cNb&V(&(^-V;g2BjwzZuhR+YIi@I}j{|""{{<xB>6m(ZV}zhA!6f_?mneg?f3B)3cFy<#G8t`kCWFSuTKAn*6PBKJ7xc8)#mc6?(&pjH^EA+K>acmtW>L+C1~jUQl=|Ir-1%Uh""Wx+>RA&D;+_+I!CH~Z=7+Lc0EQ`FB*JJpW9f)V51xQe>qN>aQ1GQi8lMkWEW7oo0r_lKghcZRT{U;$pW>Uz-$(r<lw<2m;pXK3n8""pD^dRI|?aG>sk}ECNY^xTf_@&Xarggx>Z-chgo){>Q^hR90&A7n&=IOuYGsk^OC%v+zt92><5%sfO50YIvDz5iz+nlVES5UJRHh3""izM1`V|P{46rvhf6ulkm%$TD*rG&IV2q{m%ANj%4XX+cn?mI)>%gnhUDdsHCM_{dLW|bxYjDz6~0EnV<{(mSalL)%Ut;IzJZ3WY}""w)rlF__5w=IhUbNEF7EzTa6T%?KCGJue2FSq#U~C0dx*oo<$}Ey9*Ql>#@p~*x7N}nc0aHnQ&4xvK^kfiNiP(5E`}x@Cf@KEWjak""SXg^3<BmP-Ia1q!On^G~A{Q?aciafc+7KR#My)@K{_*K#uqwb&y6S)^tyA~Po{XNJe=+&LBnV2#8>>C1DtGI0O5OL_k}dtz4rA*4""$dTDxPbI!7@A8uW5p>;(7z>`n%hu03VIr9&yjLDcXH;HQBVp*uY8FG!BhOUJI3FdvYo9CT?Stx~0<mh)#enX`z;gffrG8skk{PSF""+52xn=kS{*TGWFT)aScmIEoa-I8|fnn#X&jm8y(|{|tIslqI6mEQ=<go~6tT>i)fKW%_qEud}0Z-sIl_fvjS@rd&g?NdbfLc<X^t""z0&y2)o49E=s#>ExMOrtfOyIuIF$T^mj8w`DhZvFc&LaX-5`Kw+y|bqAy*YnemRGNjZi@s$FybRly?pa9;Aeu`3LIEzMDwYT>ZFe""9tjiJ!x;qmDy=J2*%4aKBnbo7`bJAFHDwWx!}#<HUwnUmvK7^u&oxHi5a{{aFtCPtgu#MUBHj?!c~!k%G+pVq>}%QYS{LuO{n`_=""ZD?h63?8uAiW;9H*)hW_+Ilm^suj&ZwnUI(!)^Dt^EmAxI&Y-yWcvTNncLLY219$TDY!q$B9YOp*-*mX0Y}8I5?^ji*GS6iuN0xl""J>8z3t)fC1%#GXf{>8;U`1^}Ld2>ex?>c8o(P%_4)ZZ(mygu^9Mbn0zo(5ddn2|0HP!>pPm*Fjc9I!tkWr8?Kdq0J<2Qi_gF^-a(""zJO|PMrVdmA&oVjBcX}(-qgxs<q*Oh*MZ7pR6OOSrqmesJbyBrqjkNK214)OoJDTWJ>DKOHp!nj<6GloG$vy@6I%TwC|i}J3?u6*""4%j<=4<h%&%(VAKX;opOh_pMI%WE59&71{itpo!<3N?Ym|H-G}x6e0scDFs+r=tzd=B^_jV_PUy5NmeRUv*EBwwl{oG=v~zws_t6""DyUJ+tanCT$ImcSqGh9^>GOS!oc&(;BLv#{V(wIZqv0Q`Ws768VsMm_#;6Moe9*jj?VhNa-Cz#q_A*)$ME_LEl`Kz1E%&okK>SsZ""27@=zQQ05i>y!@us0q&~CHeDX`0{BqF=1YYYhZv&yKePMM8)YUBL&)^A>*&8fqTjkIX&Aplo=0?_)m83zk>I}7&laW$o^i3B8Qa=""^NW${ct&LBg(ZMb(MUOt6;d9(O5vGamk8hGj$(_;Igj<50hL!Ym8_Y4fMhOyn1N@GKuZc7Q*C}$rE$D^2p<N{5;x5+!$arcMac-_""fdxl;2&HnW6QfoHId=n2gl$_afZAC4V3vK-jaEm3T}rE=2$`MpRh$XpS>+bx8`3^e>N0W%WF45u0#VA>77V(Kg5S%Ux+oOC7Zw*x""$|}OU2i=(6;1CUKOXc^+v^=w1$|UqoVBzhFtH_9|P}L85)bKd><n22PblFBM-{cuRq{sd?ts0@Cb-C~bttqfSdL$yaJYf}UuUj;Q""eE)sG8xM&G?dX=`6as3E;KVKJjTP_JU9gmzhw$&q)<BJ#5sbM52Ww#9NM=Oj5_s}LXHjt@T@_M#uw?5C5WMO>bBF>Y&4W~Bs(pn#""Jvo0MY|kfwr4Et8JqQnsMPmlM9>?*wnu$QLSI+IevQ&^ZOQNp>P=EiPiQvUOMNTSdwgT*7`t^{9<9__R8@DSxK>(itZ^AxCMUEaR""#36)Wi>?{Hh;<A?F&JTDyPxVH(FXQ86y%bUg3UFBQ?h{2S6YKb>pUWt5LmVwzZgS$;^9F!BL!<wKN2zA5=*y`8~x~tZ=9j|_S6mR""CSSD6fIoBEj7c5|{tX1npXdulU!Z!tFRH0sF@XU2p;JFjMwqAeb@tCVt<=B<TD!jNOI>>-54h9*f!+M%s~bCfEFpN}m`87&Jk@aV""NZ6=WeM+6{qq3KnhTu4hlk0e(F{|1In`S$I2ubp?*F65_WiyHCsa@w^X#%wx37ZozheVMwA;S(WnaWcfsdK$(%CLS~-)35~e8Hi$""QqTbOnTzfKs`V37jL(adsZ&6IWxzGfEp|lwY_8UqalWuF2KpV#E96)_X(cDT9znz-y<-@YF{ZrA-pjK1^sP=Bhth$OAVghd;D{5?""qWo<sAKVkc;1El<5uFOMFmh6w`HAQ&vU~ge^_t{_@*<JysQ~GeBW%6v=mdiRk6Fvl;7d3<`Vs#>6k+-Mv`@FNOwK{%vHU!Emw>{D""?`X$4N9%=TsP3_?@7MQO1t<bN`a>7et4W;G{<7*Aiocs}*oXdAHeN39mL#@u>(AxVB;tJC+OvXDX9qp7vd#Xq7r;VdQDHWG4o6i`""fQyx^Btr@#n9FU|0vYTFr~fWB;{wEpX5nAD*p<Ed)z@24B=pRw$c==u@2e|ESFW}pqmXb5w(K-b^k{YU=8Ni;`B_u4Sa+pB3`Qvs""D9!}y2?(i@$+{UY4DJrXz$wzjBN`_#%Tb<e*x2ZS)TJqJcUmsjAtWBatu_qZnU=AY#E0UAl;mIy(~Uj|!^MgU&c@%d&sCZjj?VyI""c!d#oS8Xoq{QDbA**-|*CdBz<=e&vCqAjBtVO@!*_x+bH4$KsEQh?I6MGTkssH}F{--O6IBi#$c=3m#sWlDlC_(}&xdnQ*PVfa8X""tr{tk$z=S;MT0Pa?Xc!;lDI<NxU%OI^4-E(^WD@D6sGT)BBk57K^Xq{w4BTEW2%gkc_W7Ql`2zW$t##fl<NNABF0*wpsf9nW%46t""Yu!Ks_d|kH%ELmIXBY_hXV0HusV}w*uP={?p&RE<bp=+?3Ot(^S%F_POpd*_bsfQSo?qc}Lb0E|#g_@#Wo4;BCBj2CMWUSC&JfJA""50wYrK)Zu}ul5!K_ubtiUZ2<g50u8v`WJCCmb?yE5QPlZ23p&HHs-iT7B$~Xw#%R8lxqeb>ut3h7Oq&Q9UEmO<Eq1p*+ImS6oCk9""su#Y{jII}uyA6r<7oRm?EV9l*)DcLhM}bST?!5$5z=w>17S&YRsb7(>a0<?AW1a~k++I8cuVc(X14@v){IH@$m@&NwFYozMh0(t{""+?~s6eTwX0C9{PzfkOBFhd<2i)_KnU<VEClR6q8)x)}D&L5RY1P+IKZBm?S)@GMZ}aN3JZE6|Gy1$dc4*L=sOxTtG^NZY$p7N|+O""=YHl^ARP1t7sMgsHf&1t*DFhS>`}-JLHnW>z(b=#hV{Syy~G7o-_mO)KsqeA_phT|3zbW*S57)iG^(93T3mUrGlJzSSaw+SK1|Bh""CPaU&z*|WoT;3yR$6P@#rfqWHT4yRxFVXt|>f`G|u)weN*iy6m_^pdVS2NR)yu1pu==lCteGQoi*aLA5ovmrPp;4|uMW;~=83(SU""^Bp;8qopNfuON$iu<ddhdow?gFK+s+Pc>UxLV@Acd3EHs)iv)^9&3>d!&h(Clqt^#_Q&v?$C>_>^5C0Yncg@1R^xN!fmNULYZ(TP""E){|m6vd?6?~UUJMz^#qdhUCtOH)B%#4m~8X4JUkwo@`hL$y&Z@)w85)`7uGnr)VM0NunjxzHr>F%402zb_(-C1+4_GMHQp2080r""<2hy3Kvgq{2KynY*Q=v0Yw5#^OGim^Ye@-&wb^7aTtJ5vlhH4i{{cz%b`_;>5p5VjD;Qr-c>zoHdzTfGf0g6<2VgSX-ggZ;F~<`c""T(MNO*eol|7;XeHfqf%ggpY)3l#6uGx5<@$d);;E5j_CWNl;YzFScqZ9?QTfES|;X!tKIL;;z-^eIOs$AS*{a1wX9nseXj)#3_8t""4No_$F>-?omyw5%*GKFzt6h_5)hoky>Yx}3==Z-Ch|OzxG4!UWz;Jr$kX8|1O&ArM?BN5dwDNtHO3jk~{pxy@1+G|rqrbrl=|s6`""nO9R&e%Ts?S_m1lV2J}%a`lr3Yo<tmcqmt>9}zmu1QL~<#bi@%=Ea=bC!7?KXF&ALQ&g(RIimsICP@{HzVgL0b&wl}BjO9-2qih<""xYDM`3nFkj&(RLHx6VfMEL^va(dA3b3@o$_kWN%x6}R>p>UlV0(3%oiuD)I|e-aq8Bj5PgS*i)+lfuq#4LDyXt+{W;N>4=KE+*5i""H#`x%hXK_V2mOP(to3!g<>8omNH_utG4`w_tWS?w65f~l%%I_wxXX|vB6p~<i2PlsrD$sOAY(L05?|AKq|XC6RO^gOEy5dF`amdY""BOA_EwV8td@i2E%)YC`8I_0sb;cLiE#e~Kj%O6LSw;(Y_63cyPu@6bROKfo?g2O62#~0F9VgT$aBv)a2Vd+cjuU9Kg|8{wII1ovf""c|w!jpC=w`-xbPA{I?)rJe121`F~T^Q(?ctJdp9{P!HeJM6@dss$9y_-d3Ot(!RS5u{&|UY3DWMwFyfrq<8QjUg-ISPSHltJ;%gk""j+S0+porq?G|OChu-vbm3I#$H@U6@7L9@_WJH>2lW*hmeaxIj2m78Ha0Elq##6HbE1Mg}6nDx!2`m0X~9Wy(23KUa*uCf{qezNJ=""Yp%ZI+mXOOuT!X)Q%gAD-UCZIZ;7cqcErv780eq8R_Y}Bzx03Xay^|-9d$G)BQ4ZWAB&V5Y=y;;5!uGqFluNNaTW-MVE&GuTKe>z""8<LUZ&00kH($9*opM>nLuv1T^^7|$=GcI74jiHhYz~i3Gb@49LoNUz+L!V^?SyX&iEf0kuN6lSuv6`Wx$#eBc&H1`AC3xkF#}(D@""0~=VMt69UuF)cp?3<HgTsf>DOxy02!(B-iVfeh&{D}HOLSqQxAfdnj?<8`aT7Kk7WRY5&x4_J+~O)92eY*7nz--;Y*LDQt`o)e!R""=sT(hcw}otijhfwXf~;aXT%IQxd6wf$TW%Tg07{lwXv-I#hLHSCp8gq;Y>on2kxUfT>N!(o+$rAYGVfsE>#$3CTfqHwtBIzt~xb2""ha28g<cz$!3gs+Wg;Z1Ew&Sf{4Q&|jaJT1jUDjFaE`q%`uovw%uss}AgOJhH&?)pc{k9dYLd~cqgG<$Cn@}9J&3Dl1VDb7n33Z7m""(8IyA&#<YuP*E0en#N>LBqJb0N$CKo>^2Gx9j4b4c>n4u&E@m3D!F@|KN2o@NTB){Kfx7KkHsm6E-@tVOzQpu<~co_0{Txa>pFkn""7h>eu{#W#-?{lSoE^)V0!>+eq5L<e(%|1u|or{Vf+%Jljhbd)bm6S0#dxKIgZCfg{+dGiD7mIK>Qd?a-E|Ixg*ruYkCLMH%7>cT5""UYv+j0Ly7l>LnV1dAL?;0nIH=>d|+&0D3U`u4d<5t6TF#v<!7=jw>VK{7cgzai86Epj_4^HDq2(;v0-JRK4w@{K3WPQ!MnR9|N)e""Qs-DIPVBP+0vEfkrT~d_rgEjqSyu<D^=cbkKruT2XeX0TASio9SUGj5)X~qPmB6W+Cw|NC#@rSGhNQ*}valfu9C)(FQ=f-@0=Z+A""y}fPzumC+BC#Wdw3PT(be4luB=)*fHd$zu$+Skr{qr1KD!VaNv?o0@b3`2(iFL_iS6IAu4P0^1nNa!|sdON{eL`OuAi&MB?Ieub|""%jJ8Q%?o+?+=Vq%RISUtx0Xyy;Rp2^Nd)?wfDZtDg}p5L>6vwXs8C_$I`JJIV4+4#&&4P2cNQhPnm&=+d^Do3Uzz2z2dyW>mq6GM""dm!rIi5kl#8C|*w0UJSK8ba+-F+W&{_CQZAuZ;S*Nw@RWb+%}qvyFz=Ux~9{HIQFD(~<;?RfnOTYmKDiY2nFciN5};x6q={1p;m>""tg!I7<F$`;d=#iLya#@|Tthcb*X+a8EQ4C`lg-*2=T;mv58UGw`znCF0Vyx$2zN#9Q~>3ONDrt=O99fw<{b!WkqecLH0D&5bGIk6""iaV9;ALR>Yf_3g}5|W{pFuh=9tu4BFcx@tR6QSCo8lnpzi1EtBkJKPH0u^+<jfZ@t!C85qFft4-;^EJ)I(H!rk&%OZV)sEO@_!Ky""dvCcbE}`Pdf^An`i%-JY_~?$stt<C-KAH0-E&Kh0;?;ewj{4@D=X!BGdsd0%1o6sRS(v(g*NLS9K?ubBrlpWrvr!fSMe0UU8yjWV""sQGhu5dxbx7L%Y?ZJ7NAz7|idVu0A5c|PG`1PNH<&qd=%z|&L_Tn@QR+gWA687?2|{PqytVZ&}1qZBjR6ql72#)w5S6YZJ2Y-U;y""heg9<9W&%PU$hFEKJ74llj&-PB{1+}(`*yn)CLZEvTufV7mPPjg3M;o@u3*G)jK;5gPQ!)R%hQq-CCrvG%wdU1eW#e?0@ir4f=aP"")s+r3G_9o|ysc0J$A)VM2H1``ZUHw&iLBnsxS_W^5$)^`dN@N~7)YEd*G1MgcOCW>&uiv)py}y^9_Fj_8R830LL7HAto4noWc-^_""q7CV9Qq?M-M)h1|DM64T3)EgGrf-*VZ&P~+Jeh<UR#9MJ3Bp`CL}xt3Qk3T%-RC2Sb~9xsz22(uqOMd$Ww0nxm&n9c3`Zb==f#Tn""cgflGpm$1t2LFz~EeqT5fms1v&=`}j#|)io$b;AE0~VbjNdm(emKwF`^PSK<ElVp2QW<nqaT%O(La>S4DV%}_=>W-$2^Tbujg8J@""Z(v>iQ^+&Llz~c~%wy7QK|wp-lY*MNcW|DFX=7>q@aUmX9wOyxo@XO`wfWG9Jr-vbi=h9y(3vD`3fD!cp&|NO2SKUr@W{^-`RGvc""2jjC7ZuSs!5LtyJ4>D4lXitBpTtiC7QBa9R5C(RmE0r`zy1i}Xb6KSH)Pct?&|oBI!$n!DlwWyK0v!#IyG|Wm2vliE9g;`Y(4~e9""3kdOgIqTb9nM@Kbbnl(AksL-O*=dTU<-jJG#G8E+o^1t*16|H{e<#Kx3<)-*`u9e0ep#W-Vl1sqe)>Urr+|^{|3xY>_~}zRf2H{G""<a8osI^zjaozxNtSpqC;FsRQMfA&0}uS93Gon#<W<JnuLR+S99@Q?r?89BeQOp(S5CuB;RKyL99tGU2c62oE3EGAwl#3y_&i8<Ly""CFA!jv$rtK)E9V6)Sh;e_($qul2-|>h14Px`m7bb?gZWt92OzcoFI^3j`LQu2(|)EMI#^!gUUqWIY(7{{;xS(_$3D`clFD(m@8ya""nt(&%*&wu%+injC>pO$Vm{U0BAdiR3z(+(BUL-~KRmN_&oPM4x6(|O45qqj~N=|ypHGNs*O8SU*u%TPif@12W%!@6ZWMv%d(g`Gq""(6ASR<k%-i>D?BW0L<gJ>gl?SgnRY}Syi(t`K$?eK3Kl2aAl8NBWtS_bWNh;sQ31_G^cu;iX;Ize<qh&itJ^vC2TW^+uVPx#3Fcb""T-Cga5&UeSPEEMknH1^ATWI;OO?^leLOesUf+l*w>#b4!VlNWxktkPyCfixZ>&~@xCQ3-e<Xb@`xxa(pu~uld8Y3AM@I<KvQscZG""%8eW>Tu@4*cByo769mO*J-Ri+_h=;fSRxdd!!x!wx7vA--i3;(5`ONPDPn1uGR{m^#QL;j$~nKD2FXlVf%S|V;n!D?9JiuyU5Wgw""PXr^;OUQ9$faYYR;^!{89QfMnUrl1`^(-&uZBjNMhaJ8z7{r?h&1aPk*%eZy=DR8YWgsp(^@t+69aV6x_jFqT0aLtbcK=@Q(zB@Z""balM_UU)0OtE)7@?lYVV2IU4^(NO}As<in+U8lK6=M~T#KjBmu9Wl1E!%Cr%5DQzOg9v^p$q3nnn1{cK+!voUKy4Z<eAfBNpf@Mf""GS1uKl`qw#DFR-7*`5zRUJj@aIgcV^c%i6!<AGazOwcVtV^Eu*!$Z{oh9pJPu(P_gnJezaPg8pj#yAv1*;qra7DJ^Nfo&cA@epG=""Du=QgwZmIxQJ2sjc(?~-m7!qV-sk42ti{~W_#B3OajhDNNP+?V<9Wl}^u+*2!YMs|UdJ7vp%tY8lbK}pX=Z=z5p<Jcl!SA9oo{}1""%#{_T98yv6Cas~uwX}XryhaTtBdlzFAKkT^Q-@E8@(gD<OCg7J1vz=_3Jl;%esvWcBiUqjT<W2u2JB_ehJ(*xyH@$m>Zyfq!2&XD""2%@y)D4$o7Kw`L70=%|>V|QGxQ@ggw)^LZGlt#*xG{ZT8LoZJ-Lw|}^h%WJ8+hFhKPdtK|7Sh3cW>&RKVi}{w7Hj*F9U??jL0y}7""cjT7r1sb+`%TI-HCn@vL13QPl8G?#%X009FGjq7Z)*aa3T|q`fQP;t{`_SqbL8{-UmmAs6!ApYS7$!vKX>%Kgdv02|4n<BMpZd#c""pg>I~Ul|b{vgO3*!wM=%2MokWnHIFocd>R;<CQ)~x1$aWXzc+*B9uDm-B}JWZD9}}sX$Lz<4up!66YXoqb@bnM#nPz;4pAlVQ)!w""slN3IRsG{U$=P`XlS6uz_9o^)#H4wUIw25<qN671vfikuoo^s~^y;V%B?AAg^@<<?&n372&W&^=)Whi*q&1+GU0b@#h$XT>w}c2e""2f!^ilxcytxYttC<NVIa2xw;!J;b!;6)pcuVs)+Ga!K6PM!if!GI<2~IYEfrm0I6a`7r$^P243h4J1%K7}Yt(*jZfq?PkXA(UlPb""gvRV-->Yfd{+tR<r;CNs&n+v_l#3sMwOJ9)@CmlbvjipCz5jFvFe6Y;cON>{bZ=Dr^WJPKC)@ClS$ZDqI$V|2&b|h=;cp@Us~b*^""EmhW;?~SR9Lwb@cYXxDjVbi#WWtHe)8A-f@%r=MJ=1*R827%4#g}4?<>W7oquoa>-jK2ppAvtlC-ASq}&`>7Y>aw~?OaoM=jCVhP""9tV!We?q2%L`$I66oOtqNs`sDOp(s5C^9S6#OQM8VeKGN?G7>}@|)wq+HW%vKOR8L%h8Z0<P2ROc!Ea(?WcGv=94tITnjVDHytK*""2$2^%`7n$(#M>&6iyP}@>ww*FuUcVsk4h<AGz(TGwbPvHUO(g1%^RwrxNc_O*2NGp1>R`bn~8oX9vUxW6bWS2v>hZe$<#|{tAK1*""vuSMc*Zm(fBGU`)0Rkx@Sa?dzdi>=jc}@ys7P;O*X4WiRq#>yYgBfD2nxkBq1VMX9sz-t7z<@TbQ}a2)JWATQYe|xo9d^x@_C~7M""YqLG6l7jaJh|r6!r1naj&FuQX32crFoIn-L<4@D)h;))Zkyp2yHyH}eO@Lvnve$;OB5&KL&{u;+()W0<#o(kvf$-loeHzX%%NaCM""o+o4HK5q}z(?RTk#clDL+QlqE_tL(YOn4p3;2Fb*$OG8SB7U#WEUIue4wt8iW*mOglwo~9TRxT9D+D%#+z+GU{xyNDc-OGUtNH`3""j)C<LqOa}B=Md(Wbg|0Is~$1)N;3c5_de|A24FAo&H5JcP~{Yi@hsb1FaC4&nwzglPYqvuM08Pmlzto`#8rzG>mhKTtzz*Pww7Ec""!*7Lgq!XqxZ7=(N{U@^@+bvTQ7Xlx*#j{%%{xFqk$6<Yfu!WAhDWg9S>Lm3>F~66-Y4d{IR(VzU(y@8^Skxk6|150Np6du6NjD(*""A)m?DfZoa{4%9yOy{HT2?W>jS5>c*u=BPPUXhpLZ<5S!J7n<u{U=cNd$Mc#udtUx+ps&0f#un>2MC)moesTd9*)dqcOy|`9cL%pc""=Svz=D@v9p(&1dCYu!rF<mvylxFh&~wIN@5?5JP!`qQ*TRR203UhlF4C9Ytiz#nFpLvqi0+10ul)$p*=uw(HXIKaFlde@4?tEp5P""h+W<L=u$*XNco9Aj>D+qJw{r<Nxp_#bZZSLeE}lJuRT6Ir=YJQtNXwFe~#iwwh-G#@b?OTWl|o)5OmD5PjtX1d>m`BD7~4hHb=?N""q*53S>NENM1j}elZvpqy0M;czwQlDZgi9wt7}LF&fmQQ!zb7Q51MVG>y(<r`NnKz!F=;^xmf4f4lYB!rQK3c`=Adj~)}64e%5fi4""jzm3p<f#9wja{&n$BMc_B+}GOi9zK%ipFzLw;mV3ehwcRgQ|o$t2vRX^LJ6lN-0RYoy2|D9G^SAF%^x7ZQ%YLW+duXs<siKLHmDm""<cKwJkTMc)cAk)&_{Q98*zF!qt?K6nYG=>G<*$F$2NwUoSf283l_0wM<;)4#hg}9DRT`~3Ug^c?Wa+lwEHvifrUoACxwAlKG(E24""k8Ce((|jHC$rSn%KEtb27vPes-NsaC^1#Q@m_Wo_;U5)AEchK5#UrOz8QEkyAo_o^;$;$z=}Lpl`q^Jvl?#uFi{6JXiq}OV$<s;~""QWCY^o)qWa`+4=d?OEFB)PB3Gi(#Qx9ZkQ8_Yac>Dax5s>I}K3{A5N9#~z@|-mHX~5?H!D+K~ckuM@0&J-d`+4uHR;Y2kiT6BXiR""A0D0e)`p0=0n6kTRYM#dlRz1ku!sd_rNWs;Z)U3uFK83spQ8=f+$~jl&9_j{;uuQI2bS$(%kBg|8cg!~28ND-cy`Z1V8{u(Ug$d^""dgmXHOYXSKg&r^u5N`Xu=-ISn;ls{$gn3rZ!EK@XDKy9y!P(|nuB6s|Ll1B|(#-1E_x}2!cq4e|``N@;;^ln=SwQ@jCg!`i(Xw&m""xr7OW*>U9YN3q%Wk3z<>7&nM7UN;UJ`Av>$M;v7hLiOmik(T9KO;(@R+Q5<E1K5b-`tLc{LPAgEGzWV*sY5~0v;{lGNE0}MLY*0#""+}JknJ8#wxvh;uVt2!Q8r)WZwych3QEOINoP*!h}FvY~bG#)REAf)Oq?e|2RXp~{XI^Ua2HbYVPNe<!6@NK&>t$iUz`j`bg%ssJk""^HiaTbNiJ(E&RuqvU2f{F09klX<<&vMI9Iw@1y<aFu^XB=oWS`Iy05hiHI1-7FODdCHf4<XvAm~Q>kvT5kILHI76)@!@)C~EeM#s""0odg+Gomw_W8m5_%QGRr*QsS^!Iq7fst*?#LgRbPJx{O;Xj6=JTKL;t7}El)^Gdr^`Q4;l!DA&6C#D|e_`c<GZ31vvfp$f9L4);p""jnpobqnNl9G9CI4s2N{6lrDO-)m#eID1s_A`gXORhbaxc?uV2a`~+oQ!KaeQtoZULVr70b?4M+dV1X%@ylCo6#3{2DYT#{?oza)(""#DHNXrKKhsul#tE1Vt|4>7U(>X^4Q9a`>Mu^K_PvlV@?I0~ut_5T=T8R1B(Z7P$?I)Cc0!L7h^E>$8sbXgq}%1y9q!qnJ)+r#_iS""FUVwbb)L9LGb-rDdtPB{lGn?@DpCnH*J)s8H`ZMuv14i$vol}M9`YbV=)hxuVNRUYY)$n|vk=5iLH$-}d+8%^r_(a=iM-uLppI~r""6|!$F7U%YQ!a7M|Rqv_xWa@)kqO6K})+aq$(&G02)F_bXE98Fwy$*?~0=C4!yqrlKLChh<#bNu%TmW3f^-P9wXe+tC9q=%PENRQ<""6U^U{R(z#{QT6AhW)G~+EqSTh>_MNkqM6DE4&j~<mT8g|U|1^S$1BQJ$*$;=?<u{^+1w(M`62h(WP+#raJ6Z6u2}jU@m(fYv#51L""Jeje-4>6GnLey;~Bs?QAKDyQDpX#4|$U&KIRDOy!VQX|*G3Z#fgPBJnc(6O@lNcQ5-(*Y5Ot)Sd^5g2e3(1SyOhs)B_}#CJbsrj`""&fc9Gf|_8;Ve@+guF#D^jrpWE2El7#9Y!~kFe}|rrh(3OOQ{wlTlWswX0ao0>En>FGUaa7Vv+*vio6G4`Z3G~*T>LXD*OMam5}4D""wqu{BrOw-s$e+@NbwNQ90z#{rR7oX2Wf5{kz2LsSLiFS3aZ!Fxxc4QTg+y*@(n9HAr5aYSJO-qQiO=nq2E#5${FE<=`2_muOX3OF""Fg9gyq8UINs04fQ9&~s&Eu8h}iQ}#0&4E?NXBh;xrt*3HL$4b)vozmUoky47jCg%|?!q;E$Osf4c!7OGKDoG6*v181&%R+%S~wH>""fxg{f`b%FHN7_C0is>n?a?6Ke8R*JO=FvxnZW!pRjVlsaAfzLEj8(;lY%NC@cfO_mONuN1L9xz1?CrRehT*?S$O3h1NmVwbR>@H*""ZM`|GOV6jh5mBf!B&|Wurmde~-)IIhngiR)J`cRlVl}A0^LvYY>&Cjhq-quJ3=~9@|9HDLHuB&AV$uQ^v<M-85xbvU{fHn|Op6pH""X3J!U4G{VXCgr23#aSfOg<yEjeOMn_h5Qp;9LEU`Y(xX1eXhFW(p1{Nf<4Cm$!#6G@d>srn2cQhYu}s~fs<b~#_{jP#l^CuX>vmU""31Wv8C&RU=HW1FA41D9vEGyK9Sfn)hl?LtaweFn~jd9DKetHlU;{EUTd_m>sVXn~%%|e(_^scqoyWbX(*~9116iRL%T=iv`afM(9""B*q-L3!+Ihk5H{>O%I!QBc^4ii-5!plh7@h05)T4q`w>$|Jlhz2)xMwHhBiknGcH=U1=i#jYA*AI18Ui6|lw)RDZ)2T=IpHd-1cZ""tP8;q-D2DzJK*YL6B;QzO@jE03@D6DW$u<75b$;T?GiV3fox!sClHrE9O->#rIs4iPg*)^E~RQN`n6r>*3MKH1o$TtM`F5rboP<="">|LL#LzrvLW?&c4$P{5kcHIeg?dQ{1hF$B3m=(Ttrz0+0R<F~cn!_bkIEw^Xx(>=$i=R;gcVYp5{1j2|I`0u@qnmsc8}a6;B~Njw""{uXJ#39?&}TQI&e-726xd4bd>dVahE7yUO{2Q(0N#Mns8)a<erY@P>WMP*4X%T1;GV$fjY0WQ@dZjkK@OZ{<DhTo4LYBo8ADAFJ5""Ov8kf;>R&>^te~flHsu4iU_A>pB@Eiw7ri8;1$D6ID>}%$*AWCteY{Bi5Z7#Gl%ColYI@kI_et4?{H7X)Hqdr`;w_?K4U}1$9jwY""ZI=(k9BYO5AJqIdor;z@*0nyrzQkW>ZqcNDAvEdY4(7Wq9B~7+Q#FybVrehEAG$t*X>YRuk@B{n$IJ+7qhr~SLvAR$lrI0vEFkO?""yYmE;DN-oRzQ8olwbbkvbAkzGuY+skaB3R&MCOm5PbS#~^|C_q#WZb3+~60w2fP6ZN9t@jM5dASZZKtBU8Dz`wZlKnD0smF{^LlE""LB@xj=wu;LpaWI$?~WFWf5irCh`;;$R8J=jPG5(~(>Vd+RS#=v&lusmJI%_qg??_=e+6-bWR@DtUIW|vMnnvak+RiBiOKJva-94+""MiAm3o_A%%9ZiL>3#QS(Et85cGJ2Ll#9{TrUA>Zj=U=w8dCQkc!T#pVMRV~?FBMZy&AMj1Ek0gkV)658Ide|bap8Ter7AC$SH|`;""`I(8D0+K7gK_JGlAElkJTu0}^j=7+E2CQDT6>KB3IzuWq{Q+ncnCeQ{axAmi9UxxbgWB|Vj{x_C>$w-8RW+a^`mnY>``A$NkH72M""trJ>4#NTj+;ru4W7jp4q8pAk&Fv5H<Nadc$#<QM=@fi;_V&6ooM7%nrF~UqOlWbiaG?4Iuc(1=0H686mN@t@PC$L#E21^5@%3_cD""K#j~{IClyg_Vu&58@T3xoJd78`o|i`^UlB-1zp-h+0wS7)Gt)9qvUy-u~x*xP)K^_&dfrQn{oBP0lO}xZEX6ayLRhTq5D1yGApU9""{Hg=R8;VffZ#`|*w2z2Vn>K}J2Cix?Fk8EvxV)W4El=RriV&f(mr?#=`j+r$(uzjvpzngTqnmO@<#fOE7UWcDGPq*E?y*OgdKPHa""|J!3*4QZikmKn@p640oq8i^)iPS76pA(~bX2<~&SR5#MT5^coKDTMwh;zBEqe$?S7L;|IW@t~YTxF6}^Fsr6tyYUv~o9Dmu=J)nY"";AwU~_Z%vHglgHYY`R!e6Yo%xg&&Y9FBUc(SyJ|P79(;p87kbi+KF~CW0qDgN!6k7EKf?C3apc%eGD3F5coq1&~DJa^R4U&Vu4hq""9}&aIB{`ydr-Wd;L3LIFQY*$X*5Xi!K@6*3N#$P~M0P~`=x&17B{Japx^*2Ytnn#XY2$lf8l@qOgPn&Ag_3t?8N=&7byZ_Avmbx6""5_T-MZ3~o#M^bVoep=~9hrVEa=*Op1v}62kMW;0N+6bcTlLO`hP)%^Biu90tO9JWF;Ie$mAr~NUc8?r~S3j((x6AFfaRX=q7Kd>e""kf5tLFgk`%CYYm)!I=?z+7q|y5ci&5T$;V4@QEDOZSIRbE<PpVla}C#G)Jwy?a6hXRDUEXBROf}Xg65%vS*Wsyx<TnqGW-`WC#mv""ra;ALMG*#CBLoViY`ug*6S^P_&Zlh&bjC|EpMS?>adD_lrEYT8^z6|cIR^hGRtxQEOw7?<YnPh&s>_MR+n;32F}=R|rM$35nl%~Z""BQyPrl8T?uiDn!sQp*+zixpa#ps*$*^3i)B|GqGDQ0>X`@;Sho`cX9A$upLrn6PwG&9mJ;Xxcm^!4n?Za!H-hp?N{CvN@u8Xh+8L""MMk_nNz4A#tj^3WiB8)wIHR)-w2l9#G4oD9y6DM^`3|{_109diIwnTXM@Mg%*xCq2TT}Ehr%dT<PhJ&#mXUEm^z3No=0rv{k<yD5""dcnvZS4QE5<9wZ~dls=lo7-ZSvl+fjoc#|}sQ&qNp}rXeKSWl4fM35UK~wh%rzLKvvcOI|CZOx^4K2VuMFeM@62V@++GWhIt5g(O""FF3jxN+_d6l#wp<5#C)DbpQ_3<V<Al?32?0p93dbLt_5i7_?lSb8n9Bg%xW`puqZ+^XF-I=Ue_Ln!vt74@NRs611D<q~X?MY=mu~""{nl$E`DE8@tRBnATPd93rBdrx8z}H#+xKg#nBrTPqa|;g#ljj6z!f^d?09F|OuT)6+?&MwOSQ;2X`RzI2L)5L8HDrOm=X8EJNW6s""|Aj>CdYF$k7GDQG)IOG)aDrdnW$!ytf`ut+h>4Euu{gxktyM(@hfH=Vy6bx>6BApN)q<D|j7lmy&*Hc~rv~mIj{7sNv+|G)D9c|F""O#6>LIq!S7&WWx19Wg1~5rsXZB}E-$0=-FS2QgZPq%`3>uO#&)zd*P)QE!WT@K4Es5UG@FT@D>C-5VMrtZl1*J;$>na<4lUTzVr}""m@<%va+=X4>)Yqm82zD{n)((>$NaTOI#__#8rudeqU*B}c@qFxu-&;NEi;#Y<29BahdITgGH_^lXqx<l{-APL^GMV{8Cq#w_^*-y""N_`&<`4ricZ4XWW46>kr%xPO09pt1y)O_7$N+N|~9(-<O3&it6yDf9Fy}<OjM2<j=pN2%$0<A7og5{tXR{KwvS8_bqoQG`{#fZzP""Fmmp9ptNeY8;n9?2@8bL|D-Mn`WDR4irvr)rL*S*Ejh0o^@K)C0}%+G@zAAy;z$KB`w;|VI4>Uhm=xBSZye`pOB+2$ao4Mg1$`?8""B`mq%VYx+1Czng)Q!`mtGB-Y8w|@|rb%n}q?7Z0Bh>y01yT2s71Dc(cgpSDU>W@WI;-4@2ON{Q}1vCyNKW^T;Gn?C5X2}5)N(X4n""DNoFzA{oq{))tSABs=^S(mYeTl79RQ)rNPh%IH5i!^y%;i9htW-X3!ruQ*kfDVTUh3tpTRYl<FbI^kS1CyRtLu8eO^dbz$#B7y46""^$!{5kTFt)No=+4lCGGvo~|aSghMQ+K@S8DhKLskfqXSL<n$*c)_#~K4=peI75fyuXuGk{d({XD|F{AVG3Q%L8*ITKVv!o-G8;cE""XDvT=PXKkf&+|hLy!6sqF1nkoCW4aLQWXbYo1s7yEv43t357o+_X!}D^vr~|h6>Wr62jW~aaGL%^g+4R^G{wV%4G*lBA%tq`>u0K""GgNuwY)!Pp4wBeMnyjrUsnT0eAFCYE>KuMBWn`pM@nB;8$15XaygcQ~b;k)s7UTgQ0mi0M2De5~<E;OF0{yHYfz|mht?Uv$Y2hds""nTfq5&)Akrj_V~K_DOYIMjf!IxptmW$(~nX8&{;4#<Gm-<8}j#Zss!UZV~jxF;u}riY{3~p2sh(71irazsNQvxTrGhJj%xgkqo$q""u>PIB)XNzVqK7vKKjUjHVzV>hMckv;lj_yD{mcI`Z;I$-Nw~o~MLm!woOB_M_B%IBqjMP#IQ^p*j8e~f%bjU73w0C%ds+3ZCUj!)""c2?ZmAoU|%8A5YFJ$PFi`nJQYZCu3IKB%Tb0PAz#2V>11lxRm>Q}1`iC-*~2?j221<L-012NMcik`j!c=(>taU86_Cy3%pGR!!wo""I-H9ZJJ<<-rS~-}1im-~*N9n$IOfl@-NNei*DL2v3@|hBEX-7P8&eMn@^+N|`<?x>S!wc5PMC1;3}cvu4R$SSIq5umGF?w3l%UL#""d}2ZIo9$w6nLR#Aru_#E_?Nc96Ioq#&I+z9r?_P!*v!*yAM~OXNM~GyWXbbd!SlWVg5Kego;rX4QIiCZ-|_kpVvMVQnm1Y^FM42_""X>6i2jyXe*>Lj>^m(YTMI#&rT>+d%6k&j|K_p;Z}=8?ZX^mru5Y=xxMt@u~PSY@}t@RJxHcDAW-HTp{FwD_(Y_6d!+b9uP1w9ZqY""u9z-c8d&mWKw6akL?VTh*stTF0N=B^CVh#qNCHJ#`g*si88sj3KLG^{A%>%LHNM#xK=}LISScq+h>X2Wedayk7>Dgtr!h8kH#b?%""rJ8%A5ICqaAUMh?fzO%vx}FO!nIA`=%(TM5GL|V)4*cI!aSA0PUZ9=+(VFcNpH!T_?)G645!vnUV>@~IqCCuJPKG~RXvh4-EgBAi""%p7Kh+}uN}lz~KlZd@P;1Ivh7Nyo_UfpR39@s)$WL)QQ2NQ04G9LOBOlipwLjK!cPBODi#${x;mV9(c1xiXZqV8n!^S{o;IOgGJk""Ru+;Q*3X?dKeSkD{kJPpjr-yBOg93dZ(S^PaKRx#m3Sq%v16GU<U4mnxD2Zj6sT8wiZC9b9*E4<@uGw_xZbJ)ck_q^e5M)DAXKEZ""<Jn>zef2_Zjt+eG+_{z4uLQTgHhjvl8s@|veN2sKh7yRX_k?O#cH;y6<*BZZVDLr&m=`g86qun^#^havFb{yO082`+yqTMWNkj@n""D|waTR6OR118~s7DH8C7QVhGOH~cCS%vFbVJOZ&`-{64?5ffM>+DJ-UumvN6;G@B=J8kx3^Ny01F$YKmE4=l%nWIzGFf`&yB)yXb""Nweb?wPYD0e_(ivGWivnN-0bz>3t5|WT6iYq5SB0@vELtVC<Kzr(k#yHUp2FD!%!KQLs-LWmTFoZHTX<fB*mhq_WfGPWj-=00F<2""0j!S&rxYO4vBYQl0ssI200dcD")))
\ No newline at end of file
diff --git a/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed314.log b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed314.log
new file mode 100644
index 0000000000..1257cf0aeb
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed314.log
@@ -0,0 +1,277 @@
+W0404 05:42:56.497000 3603 torch/distributed/run.py:803] 
+W0404 05:42:56.497000 3603 torch/distributed/run.py:803] *****************************************
+W0404 05:42:56.497000 3603 torch/distributed/run.py:803] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0404 05:42:56.497000 3603 torch/distributed/run.py:803] *****************************************
+Hyperparameters:
+  adam_eps: 1e-08
+  adam_wd: 0.02
+  beta1: 0.9
+  beta2: 0.95
+  compressor: brotli
+  data_dir: ./data/
+  datasets_dir: ./data/datasets/fineweb10B_sp4096
+  distributed: True
+  ema_decay: 0.997
+  embed_lr: 0.6
+  embed_wd: 0.09
+  embedding_dim: 512
+  eval_seq_len: 2048
+  eval_stride: 64
+  gptq_calibration_batches: 64
+  gptq_enabled: True
+  gptq_reserve_seconds: 10.0
+  grad_accum_steps: 1
+  grad_clip_norm: 0.3
+  head_lr: 0.008
+  is_main_process: True
+  iterations: 20000
+  ln_scale: True
+  local_rank: 0
+  logfile: logs/654954dc-8b9e-4f2b-9eed-e6ad4f070f78.txt
+  logit_softcap: 30.0
+  matrix_lr: 0.02
+  max_wallclock_seconds: 600.0
+  min_lr: 0.0
+  mlp_mult: 4.0
+  model_dim: 512
+  model_path: final_model.pt
+  muon_backend_steps: 5
+  muon_beta2: 0.95
+  muon_momentum: 0.99
+  muon_momentum_warmup_start: 0.92
+  muon_momentum_warmup_steps: 1500
+  muon_wd: 0.09
+  num_heads: 8
+  num_kv_heads: 4
+  num_layers: 11
+  parallel_start_layer: 7
+  qk_gain_init: 5.0
+  quantized_model_path: final_model.int6.ptz
+  rank: 0
+  recur_layers: 4,5
+  recur_start_step: 3000
+  rope_base: 10000.0
+  rope_dims: 16
+  rope_train_seq_len: 2048
+  run_id: 654954dc-8b9e-4f2b-9eed-e6ad4f070f78
+  scalar_lr: 0.02
+  seed: 314
+  skip_gates_enabled: True
+  sliding_window_enabled: True
+  tie_embeddings: True
+  tied_embed_init_std: 0.005
+  tied_embed_lr: 0.03
+  tokenizer_path: ./data/tokenizers/fineweb_4096_bpe.model
+  train_batch_tokens: 786432
+  train_files: ./data/datasets/fineweb10B_sp4096/fineweb_train_*.bin
+  train_log_every: 500
+  train_seq_len: 2048
+  ttt_batch_seqs: 32
+  ttt_chunk_tokens: 32768
+  ttt_enabled: True
+  ttt_epochs: 3
+  ttt_freeze_blocks: 0
+  ttt_grad_clip: 1.0
+  ttt_lr: 0.002
+  ttt_momentum: 0.9
+  val_batch_tokens: 524288
+  val_files: ./data/datasets/fineweb10B_sp4096/fineweb_val_*.bin
+  val_loss_every: 4000
+  ve_dim: 128
+  ve_enabled: True
+  ve_layers: 9,10
+  vocab_size: 4096
+  warmdown_frac: 0.667
+  warmup_steps: 20
+  world_size: 8
+  xsa_last_n: 11
+train_shards: 80
+val_tokens: 45508608
+model_params:34401372
+gptq:reserving 10s, effective=590000ms
+warmup_step: 1/20
+warmup_step: 2/20
+warmup_step: 3/20
+warmup_step: 4/20
+warmup_step: 5/20
+warmup_step: 6/20
+warmup_step: 10/20
+warmup_step: 20/20
+0/20000 val_loss: 8.3172 val_bpb: 3.6146
+1/20000 train_loss: 8.3192 train_time: 0.0m tok/s: 8357924
+2/20000 train_loss: 12.1995 train_time: 0.0m tok/s: 8306297
+3/20000 train_loss: 10.6851 train_time: 0.0m tok/s: 8223796
+4/20000 train_loss: 8.8318 train_time: 0.0m tok/s: 8190815
+5/20000 train_loss: 7.6630 train_time: 0.0m tok/s: 8172777
+500/20000 train_loss: 2.9038 train_time: 0.8m tok/s: 7970694
+1000/20000 train_loss: 2.8864 train_time: 1.6m tok/s: 7946427
+1500/20000 train_loss: 2.9106 train_time: 2.5m tok/s: 7937314
+2000/20000 train_loss: 2.6565 train_time: 3.3m tok/s: 7932891
+2500/20000 train_loss: 2.7096 train_time: 4.1m tok/s: 7929903
+3000/20000 train_loss: 2.7587 train_time: 5.0m tok/s: 7928547
+recurrence:activated at step 3000, virtual_layers=[0, 1, 2, 3, 4, 5, 4, 5, 6, 7, 8, 9, 10]
+3500/20000 train_loss: 2.6758 train_time: 6.4m tok/s: 7140710
+4000/20000 train_loss: 2.6058 train_time: 7.4m tok/s: 7101011
+4000/20000 val_loss: 2.6319 val_bpb: 1.1438
+4500/20000 train_loss: 2.5588 train_time: 8.3m tok/s: 7072297
+5000/20000 train_loss: 2.5020 train_time: 9.3m tok/s: 7047648
+5279/20000 val_loss: 2.5285 val_bpb: 1.0989
+stopping_early: wallclock_cap train_time: 590047ms step: 5279/20000
+peak memory allocated: 30164 MiB reserved: 30190 MiB
+ema:applying EMA weights
+pre-quantization post-ema val_loss:2.52622647 val_bpb:1.09786892 eval_time:2010ms
+Serialized model: 132406149 bytes
+Code size: 24671 bytes
+GPTQ:collecting Hessians from calibration data...
+GPTQ:collected 66 Hessians in 9.8s
+GPTQ quantization: 66 layers with full GPTQ, 0 fallback to clip-search
+selective_prune: unpruned=16.02MB target=16.0MB
+selective_prune: pruning 165352/9349306 lowest-error ±1 values (excess=20669B)
+Serialized model int6+brotli: 15949441 bytes
+Total submission size int6+brotli: 15974112 bytes
+final_int6_roundtrip val_loss:2.55440867 val_bpb:1.11011658 eval_time:23964ms
+final_int6_sliding_window val_loss:2.51160228 val_bpb:1.09151342 eval_time:100413ms
+ttt_sliding:start chunks=1389 chunk_tokens=32768 total_windows=711072 stride=64 ttt_lr=0.002 ttt_epochs=3 freeze_blocks=0
+ttt_sliding:params unfrozen=34401372 frozen=0
+  ttt_chunk [1/1389] bpb=1.127348 time=30.3s
+  ttt_chunk [11/1389] bpb=1.114810 time=35.4s
+  ttt_chunk [21/1389] bpb=1.079391 time=37.5s
+  ttt_chunk [31/1389] bpb=1.086754 time=39.5s
+  ttt_chunk [41/1389] bpb=1.095376 time=41.6s
+  ttt_chunk [51/1389] bpb=1.088475 time=43.7s
+  ttt_chunk [61/1389] bpb=1.084493 time=45.7s
+  ttt_chunk [71/1389] bpb=1.080428 time=47.8s
+  ttt_chunk [81/1389] bpb=1.089752 time=49.8s
+  ttt_chunk [91/1389] bpb=1.092473 time=51.9s
+  ttt_chunk [101/1389] bpb=1.090798 time=54.0s
+  ttt_chunk [111/1389] bpb=1.086302 time=56.0s
+  ttt_chunk [121/1389] bpb=1.090320 time=58.1s
+  ttt_chunk [131/1389] bpb=1.092314 time=60.1s
+  ttt_chunk [141/1389] bpb=1.095246 time=62.2s
+  ttt_chunk [151/1389] bpb=1.097732 time=64.3s
+  ttt_chunk [161/1389] bpb=1.097837 time=66.3s
+  ttt_chunk [171/1389] bpb=1.099782 time=68.4s
+  ttt_chunk [181/1389] bpb=1.099604 time=70.4s
+  ttt_chunk [191/1389] bpb=1.101394 time=72.5s
+  ttt_chunk [201/1389] bpb=1.099976 time=74.5s
+  ttt_chunk [211/1389] bpb=1.102296 time=76.6s
+  ttt_chunk [221/1389] bpb=1.104211 time=78.7s
+  ttt_chunk [231/1389] bpb=1.103940 time=80.7s
+  ttt_chunk [241/1389] bpb=1.104012 time=82.8s
+  ttt_chunk [251/1389] bpb=1.104307 time=84.9s
+  ttt_chunk [261/1389] bpb=1.105941 time=87.0s
+  ttt_chunk [271/1389] bpb=1.107934 time=89.0s
+  ttt_chunk [281/1389] bpb=1.108214 time=91.1s
+  ttt_chunk [291/1389] bpb=1.109896 time=93.2s
+  ttt_chunk [301/1389] bpb=1.108324 time=95.2s
+  ttt_chunk [311/1389] bpb=1.107919 time=97.3s
+  ttt_chunk [321/1389] bpb=1.108681 time=99.3s
+  ttt_chunk [331/1389] bpb=1.108468 time=101.4s
+  ttt_chunk [341/1389] bpb=1.108424 time=103.5s
+  ttt_chunk [351/1389] bpb=1.105785 time=105.5s
+  ttt_chunk [361/1389] bpb=1.106567 time=107.6s
+  ttt_chunk [371/1389] bpb=1.109563 time=109.6s
+  ttt_chunk [381/1389] bpb=1.106417 time=111.7s
+  ttt_chunk [391/1389] bpb=1.108109 time=113.7s
+  ttt_chunk [401/1389] bpb=1.108004 time=115.7s
+  ttt_chunk [411/1389] bpb=1.106038 time=117.8s
+  ttt_chunk [421/1389] bpb=1.103497 time=119.8s
+  ttt_chunk [431/1389] bpb=1.102606 time=121.9s
+  ttt_chunk [441/1389] bpb=1.102391 time=124.0s
+  ttt_chunk [451/1389] bpb=1.102026 time=126.0s
+  ttt_chunk [461/1389] bpb=1.100384 time=128.1s
+  ttt_chunk [471/1389] bpb=1.099923 time=130.1s
+  ttt_chunk [481/1389] bpb=1.099960 time=132.2s
+  ttt_chunk [491/1389] bpb=1.099663 time=134.2s
+  ttt_chunk [501/1389] bpb=1.099621 time=136.3s
+  ttt_chunk [511/1389] bpb=1.099851 time=138.3s
+  ttt_chunk [521/1389] bpb=1.099341 time=140.4s
+  ttt_chunk [531/1389] bpb=1.098565 time=142.5s
+  ttt_chunk [541/1389] bpb=1.098514 time=144.5s
+  ttt_chunk [551/1389] bpb=1.099201 time=146.6s
+  ttt_chunk [561/1389] bpb=1.099477 time=148.6s
+  ttt_chunk [571/1389] bpb=1.098962 time=150.7s
+  ttt_chunk [581/1389] bpb=1.099125 time=152.8s
+  ttt_chunk [591/1389] bpb=1.098695 time=154.8s
+  ttt_chunk [601/1389] bpb=1.098644 time=156.9s
+  ttt_chunk [611/1389] bpb=1.098761 time=159.0s
+  ttt_chunk [621/1389] bpb=1.098345 time=161.0s
+  ttt_chunk [631/1389] bpb=1.098143 time=163.1s
+  ttt_chunk [641/1389] bpb=1.098242 time=165.2s
+  ttt_chunk [651/1389] bpb=1.098413 time=167.2s
+  ttt_chunk [661/1389] bpb=1.098246 time=169.2s
+  ttt_chunk [671/1389] bpb=1.097271 time=171.3s
+  ttt_chunk [681/1389] bpb=1.096904 time=174.3s
+  ttt_chunk [691/1389] bpb=1.096885 time=176.4s
+  ttt_chunk [701/1389] bpb=1.097126 time=178.4s
+  ttt_chunk [711/1389] bpb=1.097580 time=180.5s
+  ttt_chunk [721/1389] bpb=1.097457 time=182.5s
+  ttt_chunk [731/1389] bpb=1.097344 time=184.6s
+  ttt_chunk [741/1389] bpb=1.097959 time=186.6s
+  ttt_chunk [751/1389] bpb=1.097916 time=188.7s
+  ttt_chunk [761/1389] bpb=1.098468 time=190.8s
+  ttt_chunk [771/1389] bpb=1.098491 time=192.8s
+  ttt_chunk [781/1389] bpb=1.098290 time=194.9s
+  ttt_chunk [791/1389] bpb=1.098107 time=196.9s
+  ttt_chunk [801/1389] bpb=1.097604 time=198.9s
+  ttt_chunk [811/1389] bpb=1.097748 time=201.0s
+  ttt_chunk [821/1389] bpb=1.098127 time=203.0s
+  ttt_chunk [831/1389] bpb=1.097955 time=205.1s
+  ttt_chunk [841/1389] bpb=1.096952 time=207.2s
+  ttt_chunk [851/1389] bpb=1.097484 time=209.2s
+  ttt_chunk [861/1389] bpb=1.097428 time=211.3s
+  ttt_chunk [871/1389] bpb=1.097301 time=213.3s
+  ttt_chunk [881/1389] bpb=1.097690 time=215.4s
+  ttt_chunk [891/1389] bpb=1.097021 time=217.4s
+  ttt_chunk [901/1389] bpb=1.096574 time=219.5s
+  ttt_chunk [911/1389] bpb=1.096044 time=221.5s
+  ttt_chunk [921/1389] bpb=1.095367 time=223.6s
+  ttt_chunk [931/1389] bpb=1.094654 time=225.6s
+  ttt_chunk [941/1389] bpb=1.094300 time=227.7s
+  ttt_chunk [951/1389] bpb=1.093788 time=229.7s
+  ttt_chunk [961/1389] bpb=1.093257 time=231.7s
+  ttt_chunk [971/1389] bpb=1.093169 time=233.8s
+  ttt_chunk [981/1389] bpb=1.092389 time=235.8s
+  ttt_chunk [991/1389] bpb=1.092472 time=237.8s
+  ttt_chunk [1001/1389] bpb=1.092522 time=239.9s
+  ttt_chunk [1011/1389] bpb=1.092527 time=241.9s
+  ttt_chunk [1021/1389] bpb=1.092118 time=243.9s
+  ttt_chunk [1031/1389] bpb=1.091741 time=246.0s
+  ttt_chunk [1041/1389] bpb=1.091469 time=248.0s
+  ttt_chunk [1051/1389] bpb=1.091907 time=250.1s
+  ttt_chunk [1061/1389] bpb=1.092532 time=252.2s
+  ttt_chunk [1071/1389] bpb=1.092534 time=254.2s
+  ttt_chunk [1081/1389] bpb=1.093267 time=256.3s
+  ttt_chunk [1091/1389] bpb=1.093350 time=258.3s
+  ttt_chunk [1101/1389] bpb=1.093063 time=260.4s
+  ttt_chunk [1111/1389] bpb=1.092545 time=262.4s
+  ttt_chunk [1121/1389] bpb=1.092939 time=264.5s
+  ttt_chunk [1131/1389] bpb=1.093828 time=266.5s
+  ttt_chunk [1141/1389] bpb=1.094179 time=268.6s
+  ttt_chunk [1151/1389] bpb=1.093981 time=270.6s
+  ttt_chunk [1161/1389] bpb=1.094321 time=272.6s
+  ttt_chunk [1171/1389] bpb=1.094494 time=274.7s
+  ttt_chunk [1181/1389] bpb=1.095050 time=276.7s
+  ttt_chunk [1191/1389] bpb=1.094972 time=278.8s
+  ttt_chunk [1201/1389] bpb=1.095399 time=280.8s
+  ttt_chunk [1211/1389] bpb=1.095523 time=282.8s
+  ttt_chunk [1221/1389] bpb=1.095434 time=284.9s
+  ttt_chunk [1231/1389] bpb=1.095637 time=286.9s
+  ttt_chunk [1241/1389] bpb=1.095774 time=288.9s
+  ttt_chunk [1251/1389] bpb=1.095904 time=291.0s
+  ttt_chunk [1261/1389] bpb=1.095304 time=293.0s
+  ttt_chunk [1271/1389] bpb=1.095093 time=295.1s
+  ttt_chunk [1281/1389] bpb=1.094801 time=297.1s
+  ttt_chunk [1291/1389] bpb=1.094560 time=299.1s
+  ttt_chunk [1301/1389] bpb=1.094508 time=301.2s
+  ttt_chunk [1311/1389] bpb=1.094365 time=303.2s
+  ttt_chunk [1321/1389] bpb=1.094307 time=305.3s
+  ttt_chunk [1331/1389] bpb=1.093594 time=307.3s
+  ttt_chunk [1341/1389] bpb=1.093242 time=309.3s
+  ttt_chunk [1351/1389] bpb=1.092585 time=311.4s
+  ttt_chunk [1361/1389] bpb=1.092305 time=313.4s
+  ttt_chunk [1371/1389] bpb=1.092083 time=315.5s
+  ttt_chunk [1381/1389] bpb=1.091971 time=317.5s
+  ttt_chunk [1389/1389] bpb=1.092046 time=333.9s
+ttt_sliding:done val_loss=2.509382 val_bpb=1.090560 elapsed=335.0s
+final_int6_ttt val_loss:2.50938234 val_bpb:1.09056017 eval_time:335375ms
diff --git a/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed42.log b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed42.log
new file mode 100644
index 0000000000..3cbe16d69a
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed42.log
@@ -0,0 +1,277 @@
+W0403 17:57:55.570000 93210 torch/distributed/run.py:803] 
+W0403 17:57:55.570000 93210 torch/distributed/run.py:803] *****************************************
+W0403 17:57:55.570000 93210 torch/distributed/run.py:803] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0403 17:57:55.570000 93210 torch/distributed/run.py:803] *****************************************
+Hyperparameters:
+  adam_eps: 1e-08
+  adam_wd: 0.02
+  beta1: 0.9
+  beta2: 0.95
+  compressor: brotli
+  data_dir: ./data/
+  datasets_dir: ./data/datasets/fineweb10B_sp4096
+  distributed: True
+  ema_decay: 0.997
+  embed_lr: 0.6
+  embed_wd: 0.09
+  embedding_dim: 512
+  eval_seq_len: 2048
+  eval_stride: 64
+  gptq_calibration_batches: 64
+  gptq_enabled: True
+  gptq_reserve_seconds: 10.0
+  grad_accum_steps: 1
+  grad_clip_norm: 0.3
+  head_lr: 0.008
+  is_main_process: True
+  iterations: 20000
+  ln_scale: True
+  local_rank: 0
+  logfile: logs/aa6a1721-157e-4313-a8fc-1457456fe274.txt
+  logit_softcap: 30.0
+  matrix_lr: 0.02
+  max_wallclock_seconds: 600.0
+  min_lr: 0.0
+  mlp_mult: 4.0
+  model_dim: 512
+  model_path: final_model.pt
+  muon_backend_steps: 5
+  muon_beta2: 0.95
+  muon_momentum: 0.99
+  muon_momentum_warmup_start: 0.92
+  muon_momentum_warmup_steps: 1500
+  muon_wd: 0.09
+  num_heads: 8
+  num_kv_heads: 4
+  num_layers: 11
+  parallel_start_layer: 7
+  qk_gain_init: 5.0
+  quantized_model_path: final_model.int6.ptz
+  rank: 0
+  recur_layers: 4,5
+  recur_start_step: 3000
+  rope_base: 10000.0
+  rope_dims: 16
+  rope_train_seq_len: 2048
+  run_id: aa6a1721-157e-4313-a8fc-1457456fe274
+  scalar_lr: 0.02
+  seed: 42
+  skip_gates_enabled: True
+  sliding_window_enabled: True
+  tie_embeddings: True
+  tied_embed_init_std: 0.005
+  tied_embed_lr: 0.03
+  tokenizer_path: ./data/tokenizers/fineweb_4096_bpe.model
+  train_batch_tokens: 786432
+  train_files: ./data/datasets/fineweb10B_sp4096/fineweb_train_*.bin
+  train_log_every: 500
+  train_seq_len: 2048
+  ttt_batch_seqs: 32
+  ttt_chunk_tokens: 32768
+  ttt_enabled: True
+  ttt_epochs: 3
+  ttt_freeze_blocks: 0
+  ttt_grad_clip: 1.0
+  ttt_lr: 0.002
+  ttt_momentum: 0.9
+  val_batch_tokens: 524288
+  val_files: ./data/datasets/fineweb10B_sp4096/fineweb_val_*.bin
+  val_loss_every: 4000
+  ve_dim: 128
+  ve_enabled: True
+  ve_layers: 9,10
+  vocab_size: 4096
+  warmdown_frac: 0.667
+  warmup_steps: 20
+  world_size: 8
+  xsa_last_n: 11
+train_shards: 80
+val_tokens: 45508608
+model_params:34401372
+gptq:reserving 10s, effective=590000ms
+warmup_step: 1/20
+warmup_step: 2/20
+warmup_step: 3/20
+warmup_step: 4/20
+warmup_step: 5/20
+warmup_step: 6/20
+warmup_step: 10/20
+warmup_step: 20/20
+0/20000 val_loss: 8.3187 val_bpb: 3.6152
+1/20000 train_loss: 8.3201 train_time: 0.0m tok/s: 8449323
+2/20000 train_loss: 12.1481 train_time: 0.0m tok/s: 8340479
+3/20000 train_loss: 10.6753 train_time: 0.0m tok/s: 8237409
+4/20000 train_loss: 8.8832 train_time: 0.0m tok/s: 8189627
+5/20000 train_loss: 7.6883 train_time: 0.0m tok/s: 8156676
+500/20000 train_loss: 2.8893 train_time: 0.8m tok/s: 7948587
+1000/20000 train_loss: 2.8857 train_time: 1.7m tok/s: 7907161
+1500/20000 train_loss: 2.9110 train_time: 2.5m tok/s: 7900939
+2000/20000 train_loss: 2.6515 train_time: 3.3m tok/s: 7903098
+2500/20000 train_loss: 2.7089 train_time: 4.1m tok/s: 7905510
+3000/20000 train_loss: 2.7570 train_time: 5.0m tok/s: 7908334
+recurrence:activated at step 3000, virtual_layers=[0, 1, 2, 3, 4, 5, 4, 5, 6, 7, 8, 9, 10]
+3500/20000 train_loss: 2.6835 train_time: 6.1m tok/s: 7503422
+4000/20000 train_loss: 2.6160 train_time: 7.1m tok/s: 7413123
+4000/20000 val_loss: 2.6401 val_bpb: 1.1473
+4500/20000 train_loss: 2.5713 train_time: 8.0m tok/s: 7344345
+5000/20000 train_loss: 2.5121 train_time: 9.0m tok/s: 7290136
+5440/20000 val_loss: 2.5261 val_bpb: 1.0978
+stopping_early: wallclock_cap train_time: 590016ms step: 5440/20000
+peak memory allocated: 30120 MiB reserved: 30154 MiB
+ema:applying EMA weights
+pre-quantization post-ema val_loss:2.52355025 val_bpb:1.09670587 eval_time:2014ms
+Serialized model: 132406149 bytes
+Code size: 24671 bytes
+GPTQ:collecting Hessians from calibration data...
+GPTQ:collected 66 Hessians in 9.8s
+GPTQ quantization: 66 layers with full GPTQ, 0 fallback to clip-search
+selective_prune: unpruned=16.00MB target=16.0MB
+selective_prune: pruning 37976/9382582 lowest-error ±1 values (excess=4747B)
+Serialized model int6+brotli: 15969398 bytes
+Total submission size int6+brotli: 15994069 bytes
+final_int6_roundtrip val_loss:2.55003948 val_bpb:1.10821778 eval_time:8326ms
+final_int6_sliding_window val_loss:2.50726240 val_bpb:1.08962735 eval_time:76348ms
+ttt_sliding:start chunks=1389 chunk_tokens=32768 total_windows=711072 stride=64 ttt_lr=0.002 ttt_epochs=3 freeze_blocks=0
+ttt_sliding:params unfrozen=34401372 frozen=0
+  ttt_chunk [1/1389] bpb=1.120397 time=30.4s
+  ttt_chunk [11/1389] bpb=1.112203 time=34.4s
+  ttt_chunk [21/1389] bpb=1.077441 time=36.4s
+  ttt_chunk [31/1389] bpb=1.084900 time=38.4s
+  ttt_chunk [41/1389] bpb=1.093201 time=40.5s
+  ttt_chunk [51/1389] bpb=1.086696 time=42.6s
+  ttt_chunk [61/1389] bpb=1.082667 time=44.6s
+  ttt_chunk [71/1389] bpb=1.078721 time=46.7s
+  ttt_chunk [81/1389] bpb=1.088149 time=48.7s
+  ttt_chunk [91/1389] bpb=1.090803 time=50.7s
+  ttt_chunk [101/1389] bpb=1.089173 time=52.8s
+  ttt_chunk [111/1389] bpb=1.084550 time=54.8s
+  ttt_chunk [121/1389] bpb=1.088518 time=56.8s
+  ttt_chunk [131/1389] bpb=1.090509 time=58.9s
+  ttt_chunk [141/1389] bpb=1.093285 time=61.0s
+  ttt_chunk [151/1389] bpb=1.095950 time=63.0s
+  ttt_chunk [161/1389] bpb=1.096100 time=65.0s
+  ttt_chunk [171/1389] bpb=1.097922 time=67.1s
+  ttt_chunk [181/1389] bpb=1.097710 time=69.1s
+  ttt_chunk [191/1389] bpb=1.099317 time=71.1s
+  ttt_chunk [201/1389] bpb=1.097938 time=73.2s
+  ttt_chunk [211/1389] bpb=1.100175 time=75.2s
+  ttt_chunk [221/1389] bpb=1.102065 time=77.3s
+  ttt_chunk [231/1389] bpb=1.101891 time=79.3s
+  ttt_chunk [241/1389] bpb=1.101911 time=81.4s
+  ttt_chunk [251/1389] bpb=1.102128 time=83.4s
+  ttt_chunk [261/1389] bpb=1.103730 time=85.5s
+  ttt_chunk [271/1389] bpb=1.105744 time=87.5s
+  ttt_chunk [281/1389] bpb=1.106086 time=89.5s
+  ttt_chunk [291/1389] bpb=1.107886 time=91.6s
+  ttt_chunk [301/1389] bpb=1.106284 time=93.6s
+  ttt_chunk [311/1389] bpb=1.105846 time=95.7s
+  ttt_chunk [321/1389] bpb=1.106627 time=97.8s
+  ttt_chunk [331/1389] bpb=1.106467 time=99.8s
+  ttt_chunk [341/1389] bpb=1.106413 time=101.8s
+  ttt_chunk [351/1389] bpb=1.103785 time=103.9s
+  ttt_chunk [361/1389] bpb=1.104596 time=105.9s
+  ttt_chunk [371/1389] bpb=1.107492 time=107.9s
+  ttt_chunk [381/1389] bpb=1.104359 time=110.0s
+  ttt_chunk [391/1389] bpb=1.106104 time=112.0s
+  ttt_chunk [401/1389] bpb=1.105954 time=114.1s
+  ttt_chunk [411/1389] bpb=1.104017 time=116.1s
+  ttt_chunk [421/1389] bpb=1.101483 time=118.1s
+  ttt_chunk [431/1389] bpb=1.100579 time=120.2s
+  ttt_chunk [441/1389] bpb=1.100400 time=122.2s
+  ttt_chunk [451/1389] bpb=1.100036 time=124.2s
+  ttt_chunk [461/1389] bpb=1.098452 time=126.3s
+  ttt_chunk [471/1389] bpb=1.097956 time=128.3s
+  ttt_chunk [481/1389] bpb=1.098036 time=130.4s
+  ttt_chunk [491/1389] bpb=1.097793 time=132.4s
+  ttt_chunk [501/1389] bpb=1.097704 time=134.4s
+  ttt_chunk [511/1389] bpb=1.097983 time=136.5s
+  ttt_chunk [521/1389] bpb=1.097488 time=138.5s
+  ttt_chunk [531/1389] bpb=1.096743 time=140.5s
+  ttt_chunk [541/1389] bpb=1.096695 time=142.6s
+  ttt_chunk [551/1389] bpb=1.097374 time=144.6s
+  ttt_chunk [561/1389] bpb=1.097649 time=146.7s
+  ttt_chunk [571/1389] bpb=1.097155 time=148.7s
+  ttt_chunk [581/1389] bpb=1.097327 time=150.8s
+  ttt_chunk [591/1389] bpb=1.096866 time=152.9s
+  ttt_chunk [601/1389] bpb=1.096831 time=155.0s
+  ttt_chunk [611/1389] bpb=1.096949 time=157.0s
+  ttt_chunk [621/1389] bpb=1.096541 time=159.0s
+  ttt_chunk [631/1389] bpb=1.096337 time=161.1s
+  ttt_chunk [641/1389] bpb=1.096417 time=163.2s
+  ttt_chunk [651/1389] bpb=1.096553 time=165.2s
+  ttt_chunk [661/1389] bpb=1.096397 time=167.3s
+  ttt_chunk [671/1389] bpb=1.095441 time=169.4s
+  ttt_chunk [681/1389] bpb=1.095053 time=171.4s
+  ttt_chunk [691/1389] bpb=1.095035 time=173.5s
+  ttt_chunk [701/1389] bpb=1.095257 time=175.5s
+  ttt_chunk [711/1389] bpb=1.095734 time=177.6s
+  ttt_chunk [721/1389] bpb=1.095651 time=179.7s
+  ttt_chunk [731/1389] bpb=1.095513 time=181.7s
+  ttt_chunk [741/1389] bpb=1.096076 time=183.8s
+  ttt_chunk [751/1389] bpb=1.096059 time=185.9s
+  ttt_chunk [761/1389] bpb=1.096608 time=188.0s
+  ttt_chunk [771/1389] bpb=1.096650 time=190.0s
+  ttt_chunk [781/1389] bpb=1.096446 time=192.0s
+  ttt_chunk [791/1389] bpb=1.096270 time=194.1s
+  ttt_chunk [801/1389] bpb=1.095773 time=196.2s
+  ttt_chunk [811/1389] bpb=1.095935 time=198.3s
+  ttt_chunk [821/1389] bpb=1.096315 time=200.3s
+  ttt_chunk [831/1389] bpb=1.096149 time=202.4s
+  ttt_chunk [841/1389] bpb=1.095167 time=204.5s
+  ttt_chunk [851/1389] bpb=1.095702 time=206.6s
+  ttt_chunk [861/1389] bpb=1.095613 time=208.7s
+  ttt_chunk [871/1389] bpb=1.095484 time=210.7s
+  ttt_chunk [881/1389] bpb=1.095906 time=212.8s
+  ttt_chunk [891/1389] bpb=1.095254 time=214.8s
+  ttt_chunk [901/1389] bpb=1.094833 time=216.9s
+  ttt_chunk [911/1389] bpb=1.094287 time=218.9s
+  ttt_chunk [921/1389] bpb=1.093635 time=221.0s
+  ttt_chunk [931/1389] bpb=1.092947 time=223.1s
+  ttt_chunk [941/1389] bpb=1.092573 time=225.1s
+  ttt_chunk [951/1389] bpb=1.092056 time=227.2s
+  ttt_chunk [961/1389] bpb=1.091527 time=229.3s
+  ttt_chunk [971/1389] bpb=1.091421 time=231.4s
+  ttt_chunk [981/1389] bpb=1.090655 time=233.4s
+  ttt_chunk [991/1389] bpb=1.090727 time=235.5s
+  ttt_chunk [1001/1389] bpb=1.090772 time=237.6s
+  ttt_chunk [1011/1389] bpb=1.090754 time=239.6s
+  ttt_chunk [1021/1389] bpb=1.090373 time=241.7s
+  ttt_chunk [1031/1389] bpb=1.090024 time=243.7s
+  ttt_chunk [1041/1389] bpb=1.089758 time=245.8s
+  ttt_chunk [1051/1389] bpb=1.090203 time=247.9s
+  ttt_chunk [1061/1389] bpb=1.090827 time=250.0s
+  ttt_chunk [1071/1389] bpb=1.090846 time=252.0s
+  ttt_chunk [1081/1389] bpb=1.091549 time=254.1s
+  ttt_chunk [1091/1389] bpb=1.091627 time=256.1s
+  ttt_chunk [1101/1389] bpb=1.091333 time=258.2s
+  ttt_chunk [1111/1389] bpb=1.090798 time=260.3s
+  ttt_chunk [1121/1389] bpb=1.091172 time=262.3s
+  ttt_chunk [1131/1389] bpb=1.092077 time=264.4s
+  ttt_chunk [1141/1389] bpb=1.092409 time=266.4s
+  ttt_chunk [1151/1389] bpb=1.092227 time=268.5s
+  ttt_chunk [1161/1389] bpb=1.092574 time=270.6s
+  ttt_chunk [1171/1389] bpb=1.092767 time=272.7s
+  ttt_chunk [1181/1389] bpb=1.093336 time=274.8s
+  ttt_chunk [1191/1389] bpb=1.093238 time=276.8s
+  ttt_chunk [1201/1389] bpb=1.093667 time=278.8s
+  ttt_chunk [1211/1389] bpb=1.093805 time=280.9s
+  ttt_chunk [1221/1389] bpb=1.093728 time=283.0s
+  ttt_chunk [1231/1389] bpb=1.093916 time=285.0s
+  ttt_chunk [1241/1389] bpb=1.094063 time=287.1s
+  ttt_chunk [1251/1389] bpb=1.094182 time=289.2s
+  ttt_chunk [1261/1389] bpb=1.093586 time=291.3s
+  ttt_chunk [1271/1389] bpb=1.093374 time=293.3s
+  ttt_chunk [1281/1389] bpb=1.093082 time=295.4s
+  ttt_chunk [1291/1389] bpb=1.092835 time=297.5s
+  ttt_chunk [1301/1389] bpb=1.092807 time=299.5s
+  ttt_chunk [1311/1389] bpb=1.092669 time=301.6s
+  ttt_chunk [1321/1389] bpb=1.092617 time=303.6s
+  ttt_chunk [1331/1389] bpb=1.091912 time=305.7s
+  ttt_chunk [1341/1389] bpb=1.091562 time=307.8s
+  ttt_chunk [1351/1389] bpb=1.090845 time=309.8s
+  ttt_chunk [1361/1389] bpb=1.090573 time=311.9s
+  ttt_chunk [1371/1389] bpb=1.090367 time=314.0s
+  ttt_chunk [1381/1389] bpb=1.090258 time=316.0s
+  ttt_chunk [1389/1389] bpb=1.090328 time=333.4s
+ttt_sliding:done val_loss=2.505656 val_bpb=1.088941 elapsed=333.4s
+final_int6_ttt val_loss:2.50565571 val_bpb:1.08894061 eval_time:333786ms
diff --git a/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed999.log b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed999.log
new file mode 100644
index 0000000000..54a6a22ef1
--- /dev/null
+++ b/records/track_10min_16mb/2026-04-04_SP4096_DepthRecurrence_ParallelResid_MuonEqR_TTT/train_seed999.log
@@ -0,0 +1,277 @@
+W0404 06:09:14.987000 64767 torch/distributed/run.py:803] 
+W0404 06:09:14.987000 64767 torch/distributed/run.py:803] *****************************************
+W0404 06:09:14.987000 64767 torch/distributed/run.py:803] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0404 06:09:14.987000 64767 torch/distributed/run.py:803] *****************************************
+Hyperparameters:
+  adam_eps: 1e-08
+  adam_wd: 0.02
+  beta1: 0.9
+  beta2: 0.95
+  compressor: brotli
+  data_dir: ./data/
+  datasets_dir: ./data/datasets/fineweb10B_sp4096
+  distributed: True
+  ema_decay: 0.997
+  embed_lr: 0.6
+  embed_wd: 0.09
+  embedding_dim: 512
+  eval_seq_len: 2048
+  eval_stride: 64
+  gptq_calibration_batches: 64
+  gptq_enabled: True
+  gptq_reserve_seconds: 10.0
+  grad_accum_steps: 1
+  grad_clip_norm: 0.3
+  head_lr: 0.008
+  is_main_process: True
+  iterations: 20000
+  ln_scale: True
+  local_rank: 0
+  logfile: logs/0f19e1ba-d1ba-4d8c-a54a-00d5627524bd.txt
+  logit_softcap: 30.0
+  matrix_lr: 0.02
+  max_wallclock_seconds: 600.0
+  min_lr: 0.0
+  mlp_mult: 4.0
+  model_dim: 512
+  model_path: final_model.pt
+  muon_backend_steps: 5
+  muon_beta2: 0.95
+  muon_momentum: 0.99
+  muon_momentum_warmup_start: 0.92
+  muon_momentum_warmup_steps: 1500
+  muon_wd: 0.09
+  num_heads: 8
+  num_kv_heads: 4
+  num_layers: 11
+  parallel_start_layer: 7
+  qk_gain_init: 5.0
+  quantized_model_path: final_model.int6.ptz
+  rank: 0
+  recur_layers: 4,5
+  recur_start_step: 3000
+  rope_base: 10000.0
+  rope_dims: 16
+  rope_train_seq_len: 2048
+  run_id: 0f19e1ba-d1ba-4d8c-a54a-00d5627524bd
+  scalar_lr: 0.02
+  seed: 999
+  skip_gates_enabled: True
+  sliding_window_enabled: True
+  tie_embeddings: True
+  tied_embed_init_std: 0.005
+  tied_embed_lr: 0.03
+  tokenizer_path: ./data/tokenizers/fineweb_4096_bpe.model
+  train_batch_tokens: 786432
+  train_files: ./data/datasets/fineweb10B_sp4096/fineweb_train_*.bin
+  train_log_every: 500
+  train_seq_len: 2048
+  ttt_batch_seqs: 32
+  ttt_chunk_tokens: 32768
+  ttt_enabled: True
+  ttt_epochs: 3
+  ttt_freeze_blocks: 0
+  ttt_grad_clip: 1.0
+  ttt_lr: 0.002
+  ttt_momentum: 0.9
+  val_batch_tokens: 524288
+  val_files: ./data/datasets/fineweb10B_sp4096/fineweb_val_*.bin
+  val_loss_every: 4000
+  ve_dim: 128
+  ve_enabled: True
+  ve_layers: 9,10
+  vocab_size: 4096
+  warmdown_frac: 0.667
+  warmup_steps: 20
+  world_size: 8
+  xsa_last_n: 11
+train_shards: 80
+val_tokens: 45508608
+model_params:34401372
+gptq:reserving 10s, effective=590000ms
+warmup_step: 1/20
+warmup_step: 2/20
+warmup_step: 3/20
+warmup_step: 4/20
+warmup_step: 5/20
+warmup_step: 6/20
+warmup_step: 10/20
+warmup_step: 20/20
+0/20000 val_loss: 8.3152 val_bpb: 3.6137
+1/20000 train_loss: 8.3175 train_time: 0.0m tok/s: 8452811
+2/20000 train_loss: 12.1471 train_time: 0.0m tok/s: 8354536
+3/20000 train_loss: 10.6844 train_time: 0.0m tok/s: 8261627
+4/20000 train_loss: 8.8563 train_time: 0.0m tok/s: 8209588
+5/20000 train_loss: 7.6729 train_time: 0.0m tok/s: 8185726
+500/20000 train_loss: 2.8939 train_time: 0.8m tok/s: 7958648
+1000/20000 train_loss: 2.8888 train_time: 1.7m tok/s: 7942072
+1500/20000 train_loss: 2.9149 train_time: 2.5m tok/s: 7936863
+2000/20000 train_loss: 2.6571 train_time: 3.3m tok/s: 7934521
+2500/20000 train_loss: 2.7098 train_time: 4.1m tok/s: 7932839
+3000/20000 train_loss: 2.7611 train_time: 5.0m tok/s: 7931512
+recurrence:activated at step 3000, virtual_layers=[0, 1, 2, 3, 4, 5, 4, 5, 6, 7, 8, 9, 10]
+3500/20000 train_loss: 2.6875 train_time: 6.1m tok/s: 7525862
+4000/20000 train_loss: 2.6175 train_time: 7.1m tok/s: 7432761
+4000/20000 val_loss: 2.6423 val_bpb: 1.1483
+4500/20000 train_loss: 2.5752 train_time: 8.0m tok/s: 7362265
+5000/20000 train_loss: 2.5163 train_time: 9.0m tok/s: 7306598
+5452/20000 val_loss: 2.5271 val_bpb: 1.0982
+stopping_early: wallclock_cap train_time: 590112ms step: 5452/20000
+peak memory allocated: 30120 MiB reserved: 30154 MiB
+ema:applying EMA weights
+pre-quantization post-ema val_loss:2.52466886 val_bpb:1.09719200 eval_time:2008ms
+Serialized model: 132406149 bytes
+Code size: 24671 bytes
+GPTQ:collecting Hessians from calibration data...
+GPTQ:collected 66 Hessians in 9.8s
+GPTQ quantization: 66 layers with full GPTQ, 0 fallback to clip-search
+selective_prune: unpruned=16.00MB target=16.0MB
+selective_prune: pruning 38760/9370762 lowest-error ±1 values (excess=4845B)
+Serialized model int6+brotli: 15971330 bytes
+Total submission size int6+brotli: 15996001 bytes
+final_int6_roundtrip val_loss:2.55124117 val_bpb:1.10874002 eval_time:8124ms
+final_int6_sliding_window val_loss:2.50835068 val_bpb:1.09010031 eval_time:76268ms
+ttt_sliding:start chunks=1389 chunk_tokens=32768 total_windows=711072 stride=64 ttt_lr=0.002 ttt_epochs=3 freeze_blocks=0
+ttt_sliding:params unfrozen=34401372 frozen=0
+  ttt_chunk [1/1389] bpb=1.125346 time=4.2s
+  ttt_chunk [11/1389] bpb=1.114484 time=8.8s
+  ttt_chunk [21/1389] bpb=1.077517 time=10.9s
+  ttt_chunk [31/1389] bpb=1.084636 time=12.9s
+  ttt_chunk [41/1389] bpb=1.093020 time=14.9s
+  ttt_chunk [51/1389] bpb=1.086370 time=17.0s
+  ttt_chunk [61/1389] bpb=1.082429 time=19.0s
+  ttt_chunk [71/1389] bpb=1.078291 time=21.1s
+  ttt_chunk [81/1389] bpb=1.088034 time=23.1s
+  ttt_chunk [91/1389] bpb=1.090571 time=25.2s
+  ttt_chunk [101/1389] bpb=1.088779 time=27.2s
+  ttt_chunk [111/1389] bpb=1.084326 time=29.3s
+  ttt_chunk [121/1389] bpb=1.088238 time=31.3s
+  ttt_chunk [131/1389] bpb=1.090319 time=33.4s
+  ttt_chunk [141/1389] bpb=1.093219 time=35.4s
+  ttt_chunk [151/1389] bpb=1.095735 time=37.4s
+  ttt_chunk [161/1389] bpb=1.095831 time=39.5s
+  ttt_chunk [171/1389] bpb=1.097722 time=41.5s
+  ttt_chunk [181/1389] bpb=1.097562 time=43.6s
+  ttt_chunk [191/1389] bpb=1.099349 time=45.7s
+  ttt_chunk [201/1389] bpb=1.097927 time=47.7s
+  ttt_chunk [211/1389] bpb=1.100217 time=49.8s
+  ttt_chunk [221/1389] bpb=1.102277 time=51.8s
+  ttt_chunk [231/1389] bpb=1.102021 time=53.9s
+  ttt_chunk [241/1389] bpb=1.102105 time=56.0s
+  ttt_chunk [251/1389] bpb=1.102379 time=58.0s
+  ttt_chunk [261/1389] bpb=1.103988 time=60.0s
+  ttt_chunk [271/1389] bpb=1.106042 time=62.1s
+  ttt_chunk [281/1389] bpb=1.106314 time=64.2s
+  ttt_chunk [291/1389] bpb=1.107961 time=66.3s
+  ttt_chunk [301/1389] bpb=1.106377 time=68.3s
+  ttt_chunk [311/1389] bpb=1.105994 time=70.4s
+  ttt_chunk [321/1389] bpb=1.106756 time=72.5s
+  ttt_chunk [331/1389] bpb=1.106564 time=74.5s
+  ttt_chunk [341/1389] bpb=1.106575 time=76.6s
+  ttt_chunk [351/1389] bpb=1.103995 time=78.7s
+  ttt_chunk [361/1389] bpb=1.104791 time=80.7s
+  ttt_chunk [371/1389] bpb=1.107815 time=82.8s
+  ttt_chunk [381/1389] bpb=1.104693 time=84.9s
+  ttt_chunk [391/1389] bpb=1.106431 time=86.9s
+  ttt_chunk [401/1389] bpb=1.106291 time=89.0s
+  ttt_chunk [411/1389] bpb=1.104343 time=91.1s
+  ttt_chunk [421/1389] bpb=1.101846 time=93.1s
+  ttt_chunk [431/1389] bpb=1.100983 time=95.2s
+  ttt_chunk [441/1389] bpb=1.100816 time=97.3s
+  ttt_chunk [451/1389] bpb=1.100489 time=99.3s
+  ttt_chunk [461/1389] bpb=1.098853 time=101.4s
+  ttt_chunk [471/1389] bpb=1.098333 time=103.5s
+  ttt_chunk [481/1389] bpb=1.098438 time=105.5s
+  ttt_chunk [491/1389] bpb=1.098115 time=107.6s
+  ttt_chunk [501/1389] bpb=1.098045 time=109.7s
+  ttt_chunk [511/1389] bpb=1.098263 time=111.8s
+  ttt_chunk [521/1389] bpb=1.097766 time=113.9s
+  ttt_chunk [531/1389] bpb=1.096998 time=116.0s
+  ttt_chunk [541/1389] bpb=1.096947 time=118.1s
+  ttt_chunk [551/1389] bpb=1.097646 time=120.1s
+  ttt_chunk [561/1389] bpb=1.097874 time=122.2s
+  ttt_chunk [571/1389] bpb=1.097383 time=124.3s
+  ttt_chunk [581/1389] bpb=1.097584 time=126.3s
+  ttt_chunk [591/1389] bpb=1.097145 time=128.4s
+  ttt_chunk [601/1389] bpb=1.097115 time=130.5s
+  ttt_chunk [611/1389] bpb=1.097253 time=132.6s
+  ttt_chunk [621/1389] bpb=1.096852 time=134.7s
+  ttt_chunk [631/1389] bpb=1.096666 time=136.8s
+  ttt_chunk [641/1389] bpb=1.096732 time=138.9s
+  ttt_chunk [651/1389] bpb=1.096898 time=140.9s
+  ttt_chunk [661/1389] bpb=1.096712 time=143.0s
+  ttt_chunk [671/1389] bpb=1.095761 time=145.1s
+  ttt_chunk [681/1389] bpb=1.095365 time=147.2s
+  ttt_chunk [691/1389] bpb=1.095372 time=149.2s
+  ttt_chunk [701/1389] bpb=1.095603 time=151.3s
+  ttt_chunk [711/1389] bpb=1.096082 time=153.3s
+  ttt_chunk [721/1389] bpb=1.095999 time=155.4s
+  ttt_chunk [731/1389] bpb=1.095901 time=157.5s
+  ttt_chunk [741/1389] bpb=1.096498 time=159.5s
+  ttt_chunk [751/1389] bpb=1.096464 time=161.6s
+  ttt_chunk [761/1389] bpb=1.097021 time=163.7s
+  ttt_chunk [771/1389] bpb=1.097069 time=165.7s
+  ttt_chunk [781/1389] bpb=1.096870 time=167.8s
+  ttt_chunk [791/1389] bpb=1.096700 time=169.9s
+  ttt_chunk [801/1389] bpb=1.096205 time=171.9s
+  ttt_chunk [811/1389] bpb=1.096355 time=174.0s
+  ttt_chunk [821/1389] bpb=1.096730 time=176.1s
+  ttt_chunk [831/1389] bpb=1.096561 time=178.1s
+  ttt_chunk [841/1389] bpb=1.095559 time=180.2s
+  ttt_chunk [851/1389] bpb=1.096082 time=182.2s
+  ttt_chunk [861/1389] bpb=1.096025 time=184.3s
+  ttt_chunk [871/1389] bpb=1.095928 time=186.4s
+  ttt_chunk [881/1389] bpb=1.096352 time=188.4s
+  ttt_chunk [891/1389] bpb=1.095678 time=190.5s
+  ttt_chunk [901/1389] bpb=1.095241 time=192.6s
+  ttt_chunk [911/1389] bpb=1.094724 time=194.7s
+  ttt_chunk [921/1389] bpb=1.094051 time=196.7s
+  ttt_chunk [931/1389] bpb=1.093375 time=198.8s
+  ttt_chunk [941/1389] bpb=1.093010 time=200.8s
+  ttt_chunk [951/1389] bpb=1.092503 time=202.9s
+  ttt_chunk [961/1389] bpb=1.091986 time=205.0s
+  ttt_chunk [971/1389] bpb=1.091882 time=207.1s
+  ttt_chunk [981/1389] bpb=1.091146 time=209.2s
+  ttt_chunk [991/1389] bpb=1.091192 time=211.2s
+  ttt_chunk [1001/1389] bpb=1.091213 time=213.3s
+  ttt_chunk [1011/1389] bpb=1.091213 time=215.4s
+  ttt_chunk [1021/1389] bpb=1.090822 time=217.5s
+  ttt_chunk [1031/1389] bpb=1.090439 time=219.5s
+  ttt_chunk [1041/1389] bpb=1.090158 time=221.6s
+  ttt_chunk [1051/1389] bpb=1.090549 time=223.7s
+  ttt_chunk [1061/1389] bpb=1.091187 time=225.7s
+  ttt_chunk [1071/1389] bpb=1.091188 time=227.8s
+  ttt_chunk [1081/1389] bpb=1.091899 time=229.8s
+  ttt_chunk [1091/1389] bpb=1.091968 time=231.8s
+  ttt_chunk [1101/1389] bpb=1.091675 time=233.9s
+  ttt_chunk [1111/1389] bpb=1.091164 time=236.0s
+  ttt_chunk [1121/1389] bpb=1.091521 time=238.0s
+  ttt_chunk [1131/1389] bpb=1.092438 time=240.1s
+  ttt_chunk [1141/1389] bpb=1.092760 time=242.1s
+  ttt_chunk [1151/1389] bpb=1.092558 time=244.2s
+  ttt_chunk [1161/1389] bpb=1.092905 time=246.3s
+  ttt_chunk [1171/1389] bpb=1.093085 time=248.3s
+  ttt_chunk [1181/1389] bpb=1.093638 time=250.4s
+  ttt_chunk [1191/1389] bpb=1.093556 time=252.4s
+  ttt_chunk [1201/1389] bpb=1.093976 time=254.5s
+  ttt_chunk [1211/1389] bpb=1.094114 time=256.5s
+  ttt_chunk [1221/1389] bpb=1.094036 time=258.6s
+  ttt_chunk [1231/1389] bpb=1.094223 time=260.7s
+  ttt_chunk [1241/1389] bpb=1.094375 time=262.8s
+  ttt_chunk [1251/1389] bpb=1.094501 time=264.8s
+  ttt_chunk [1261/1389] bpb=1.093921 time=266.8s
+  ttt_chunk [1271/1389] bpb=1.093705 time=268.9s
+  ttt_chunk [1281/1389] bpb=1.093424 time=271.0s
+  ttt_chunk [1291/1389] bpb=1.093179 time=273.0s
+  ttt_chunk [1301/1389] bpb=1.093135 time=275.1s
+  ttt_chunk [1311/1389] bpb=1.093004 time=277.2s
+  ttt_chunk [1321/1389] bpb=1.092945 time=279.2s
+  ttt_chunk [1331/1389] bpb=1.092232 time=281.3s
+  ttt_chunk [1341/1389] bpb=1.091885 time=283.4s
+  ttt_chunk [1351/1389] bpb=1.091200 time=285.5s
+  ttt_chunk [1361/1389] bpb=1.090937 time=287.5s
+  ttt_chunk [1371/1389] bpb=1.090724 time=289.6s
+  ttt_chunk [1381/1389] bpb=1.090618 time=291.7s
+  ttt_chunk [1389/1389] bpb=1.090681 time=295.1s
+ttt_sliding:done val_loss=2.506726 val_bpb=1.089406 elapsed=295.2s
+final_int6_ttt val_loss:2.50672616 val_bpb:1.08940582 eval_time:295540ms