memory: snapshot — Worldtree U10 done, U11a prepped, U8 wrapper for infra-hermes in progress; Blender extensions live; Bonsai spike closed; 3 entries archived

archival-memory.md

@@ -278,6 +278,14 @@ follow-ups in `docs/pfi/headscale-mesh-plan.md` § CUTOVER EXECUTED. Headlines:
 - Dormancy = disabled+retained (flip UDM object back to enabled=true to restore); NO auto
   failover wired. Bonus: exit nodes → free multi-location egress proxy (parked).
 
+- `[2026-09-14]` **fv-ml1 rebalance: cyberprev→`sec` (mog-sec retired), NEW gen-small A3B seat, all sec/gen/char at native 262K in-band, coder reclaimed, seat catalog + bench shipped.** cyberprev = hotdogs cyber-SFT (name-repaired past a tripled-prefix unsloth export bug, house NVFP4 quant); gen-small = llmfan46 Qwen3.6-35B-A3B Heretic (already on disk), MTP 69.6%. Serial depth-tested all seats clean (0 OOM); warm tok/s 62.7-337.3. Commits 1418edb→dfa91a8. → `persistent-memory.d/2026-09-14-fv-seat-rebalance-gen-small.md`
+  _Archived 2026-09-29._
+
+- `[2026-09-14]` fv-ml1 all-night seat reorg — MTP k=3 on gen-large (+52%@conc1), gen consolidated onto flash-next (27B dense retired, 38 GB freed), char-rp restored to MeroMero-v2-31B, Sentinel-R3 served + dflash cutover (beat MTP 2.40 vs 2.18). ✅ gen-large RESOLVED 2026-09-14 — orcarouter serving: PLE bf16→FP8 convert + `ple_embedding_dtype` + `layer_types` rename; NO source build needed. → `persistent-memory.d/2026-09-14-fv-seat-reorg-and-orca-blocker.md`
+  _Archived 2026-09-29._
+
+- `[2026-09-13]` ⭐⭐ **Qwen3.8-Flash-Next serving on ONE card with its 51B n-gram table in host RAM — the first seat whose weights do not fit its GPU.** `stacks/flash-next-seat/`, fv-ml1 GPU 2 `:8022`, plus a `gen-large` LiteLLM alias. Measured: 74.36 GiB weights resident, 14.00 GiB KV = 560,654 tokens at the full 262,144 context, 67 GiB host RSS, 75.5/212.3/387.8 tok/s at conc 1/4/8 (⚠ n=1). ⭐ The offload is vLLM **#54371 (UVA, merged 2026-09-09)** which **supersedes the paused #53899** — it has no worker process, so #53899's whole bug family (TP=1 deadlock #53960, `pidfd_getfd`/ptrace gate, stale-output-under-graphs) is designed out; in `v0.29.1rc0`, **not** `v0.29.0`. ⚠ **`text_config.ple_embedding_dtype` is the load-or-fail discriminator** for any community build. ⚠⚠ **`--kv-cache-memory` makes vLLM SKIP MEMORY PROFILING and ignore `--gpu-memory-utilization`** — 16 GiB nearly OOM'd on a 155K prefill with no visible failure; 14 GiB is the measured-safe value and vLLM's own "17.46 GiB to fully utilize" is 3.5 GiB too high. ⚠ MTP is off **pending measurement here, not written off** — the recipe's number is cross-harness and tested k=3 only, while the head is ONE layer run autoregressively, so k=1 is unpublished and may win (`services/flash-next-mtp-bench/`, one `off_A` rep banked before the outage). ⚠ A container once ran `(healthy)` with `PORTS=[]` — verify `docker port`, not the healthcheck. → `persistent-memory.d/2026-09-13-flash-next-seat-and-fv-outage.md`
+  _Archived 2026-09-29._
 
 ## Exit nodes (2026-09-06, operator-requested)
 All three routers advertise+serve exit nodes (approved). Clients pick location: