sm121/launcher

Alexander Korolev b984f59269 night plan 2026-10-22/23: reproduction of the long-prefill host hang, the three envelopes to measure (KV-per-token, prefill scratch, spec depth), and the guardrails		vor 12 Stunden
..
2026-09-05.md	c957beb710 eager A/B done (159 vs 158, rumor not confirmed); aiden 4x 163 = campaign record; night closed	vor 1 Monat
2026-09-07.md	7d850a136d roce A/B: rocenante +31-90% decode on r27 4x dcp4 (160/176 vs 155), restore prod	vor 1 Monat
2026-09-08.md	f4abb53cbf recipe [env] refs: values can resolve from recipe fields ({ from = "port" })	vor 1 Monat
2026-09-10.md	37e4be4a9f r31 campaign: 4x A/B win (159/176, +10% decode), 2x below baseline, eugr recipe ready	vor 1 Monat
2026-09-11.md	7b9d588ac8 dsv41: b12x A/B — 167/176 (95, Excellent), +9 points and +18-33% speed	vor 1 Monat
2026-09-12.md	2bde637398 r36 @512K + Engram NVMe: 154/176, 3,774 pp / 822 ms TTFT (2.3x prod prefill, decode -3%) — candidate lane; + forum 381350#521 b12x loader finding (verified: 320B/18B from the checkpoint config)	vor 1 Monat
2026-09-13.md	2f717631ff report: stale cross-check of run ledger vs live probe (orange + runlock hint); new 'sm121 runlock list\|close' offline ledger command	vor 1 Monat
2026-09-18.md	498137d8b2 history: 09-18 end-state note (power off policy)	vor 1 Monat
2026-09-22.md	f1e87ec828 history: night 09-21/22 shift log (r39 verified, r38 4x repro, control baseline)	vor 1 Monat
2026-09-25.md	2aa9023c2f bench: MiMo-V2.6-Pro-RL 8x first run — 165/176 (94/100), perf pp ~1.7K t/s	vor 3 Wochen
2026-09-26.md	c31f0f107f shift notes: qwen pinned-image verdict — bench 94, live agent clean, dgx3 off	vor 3 Wochen
2026-09-27.md	4ace7f4d06 bench roce-quality: 168/184 (91/100), resp 79, decode +10-15% — RoCEnante is quality-neutral	vor 3 Wochen
2026-09-28.md	e3f7dbc229 mimo26-vllm: sparkring-contract bench 170/184 (92/100) + c2 long-context collapse logged for morning	vor 3 Wochen
2026-09-29.md	c9f74f0cb6 sparkring variants: 2x direct-pair (2nodes-dac, tp2 qualified shape + PAIR_PATHS, f1/GID3 verified) and 8x Pro-MOPD (TP8 contract, arch/processor/dflash verified); document lanes+bench-artifact case (specs/tech-architecture/sparkring-mimo-lanes.md); mark RL recipes retired; AGENTS: depth-suite c1+c2 + warm-prefix flag + sparkring block; shift log evening section	vor 3 Wochen
2026-10-02.md	571a5499f1 history: 2026-10-02 night — kindling 4x e2e (spec-config doubling, triple-quote form, upstream 27cf839 re-vendor, GB10 UMA gate coin flip fixed via GMU 0.86 + last-moment drop_caches mod), serve-boot-triage skill, headless dgx6/7/8, upstream sweep (MiMo/GLM/DFlash2 pins current, eugr nightly-20261001 new)	vor 3 Wochen
2026-10-03.md	1421e95250 campaign 2026-10-03: MiMo 2x/Pro 8x benched (pluto/wired), kindling 2x+TP6, karmic-kraken image, depth-hang forensics	vor 2 Wochen
2026-10-14.md	6c4469ae4f shift logs: 10-14 (gr8/DGPP/control benches, template fix, full-cluster power-off), 10-15 (dead nightly base → v0.31.0 re-pin, sm121-native rebuild)	vor 1 Woche
2026-10-15.md	06a4103738 shift log: first-class gr8 + DGPP engine drivers (Phases 1-3)	vor 1 Woche
2026-10-16-night-plan.md	7e0fefcbb2 night-shift plan 2026-10-16: DGPP w2 finish, Ling-3.1-flash + Step-5-Preview tests, forum/repo mining, end-of-shift 8x power-off	vor 1 Woche
2026-10-16.md	909162aef1 upstream-watch + shift log 10-16: live model probes — Ling fp8 feasible on GB10, Step-5 SM121 DSA blocker; entrypoint=[vllm,serve] gap noted	vor 1 Woche
2026-10-17-night-plan.md	28ed284c75 Night-plan Task1: Ling fp4 38/54 done; NAS cold-read collapsed to 1.3MB/s (external, not our cluster) = blocker; converter fixes (copy_file_range NFS hang, 1-worker, NAS is the only bottleneck); resume command	vor 6 Tagen
2026-10-20-night-plan.md	7a8c147767 night shift 2026-10-20: FLARE 1017 tested and rejected (2x slower decode, crashes on agent prompts); Beam-1.0 NVFP4 lane added at 90/100 (fleet best); four-lane comparison recorded	vor 2 Tagen
2026-10-20.md	7a8c147767 night shift 2026-10-20: FLARE 1017 tested and rejected (2x slower decode, crashes on agent prompts); Beam-1.0 NVFP4 lane added at 90/100 (fleet best); four-lane comparison recorded	vor 2 Tagen
2026-10-21-night-plan.md	73b8ee0800 night plan 2026-10-21: TF 1.0.12 TP4 (upstream moved from our pinned 0.6.0), RedHat GLM-5.3-MXFP4xFP8 (415GB/4 nodes), local-inference-lab GitHub, daily HF deltas	vor 1 Tag
2026-10-21.md	06722fefd9 night shift 2026-10-21: long-context results (131K/327K needle, both correct), depth sweep, 384K context	vor 2 Tagen
2026-10-22-night-plan.md	b984f59269 night plan 2026-10-22/23: reproduction of the long-prefill host hang, the three envelopes to measure (KV-per-token, prefill scratch, spec depth), and the guardrails	vor 12 Stunden
2026-10-22.md	87e885d967 night report 2026-10-21/22: dashboard fixes 1.3.0-1.3.3, recipe-relative hooks, TF 1.0.12 blocked (MissingEntry) with full diagnosis, GLM-5.3 full MXFP4xFP8 blocked at 4-node worker init, GLM-5.3-Flash 2x repro 95/100	vor 1 Tag