coulomb/freedom-intelligence
4
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
freedom-intelligence/workplans/FI-WP-0003-seed-open-weight-reserve.md
tegwick 77d4ffe86e Finish FI-WP-0002 and FI-WP-0003: daily rhythm and R-spine reserve. 
Enable fi-daily-research-brief, prove fi_brief_status idempotence with the
2026-07-28 brief, approve R/S catalog entries, collect and verify embeds plus
Qwen3-8B and R1-Distill-14B on VAULT, and document residuals (HF-gated Llama,
deferred S giants, railiance ConfigMap apply).
2026-07-28 01:30:55 +02:00

3.1 KiB
Raw Permalink Blame History

id type title domain repo status owner topic_slug priority created updated depends_on state_hub_workstream_id
FI-WP-0003 workplan Seed open-weight reserve from baseline recommendations agents freedom-intelligence finished grok freedom-intelligence medium 2026-07-24 2026-07-28
FI-WP-0001
7513a853-24bf-4826-9ed9-3e8ed8ab0fa5

FI-WP-0003 — Seed open-weight reserve from baseline recommendations

Goal

Turn baseline + strategic reserve policy into a real, policy-compliant reserve: approve high-priority catalog candidates, download to the pinned bulk store, verify checksums, and leave lower-priority items as candidates for later briefs.

Storage pin: D:\vault\coulomb\freedom-intelligence\ / /mnt/d/vault/coulomb/freedom-intelligence/

Soft quota 850 GiB. Tool: scripts/collect_model.py (weights-only, MANIFEST).

Tasks

T01 — Seed catalog YAML

id: FI-WP-0003-T01
status: done
priority: high
state_hub_task_id: "5d2a493b-a189-4eae-8a40-18ce8e0ab17d"

T02 — Operator approve R spine + S priority order

id: FI-WP-0003-T02
status: done
priority: high
state_hub_task_id: "909263c7-7148-45fa-98bc-5d2bafacf507"

2026-07-28 approvals (approved_by: bernd):

Tier Models
R Qwen3-8B, Llama-3.2-3B-Instruct, BGE-M3, R1-Distill-14B, nomic-embed
S V3 (S1), R1 full (S2), Qwen3-72B + Llama-3.3-70B (S4) — collect deferred

T03 — Download R + strategic S + verify

id: FI-WP-0003-T03
status: done
priority: high
state_hub_task_id: "222969d0-f3af-46fb-a303-f76800c80ff9"

Verified on VAULT (weights + MANIFEST sha256):

  • nomic-embed-text-v1.5 (~0.5 GiB)
  • BGE-M3 (~2.1 GiB)
  • Qwen3-8B (~15.3 GiB)
  • R1-Distill-Qwen-14B (~27.5 GiB)

Residual:

  • Llama-3.2-3B-Instruct — deferred (HF gate; need HF_TOKEN + Meta license)
  • S giants — deferred per out-of-scope / capacity (prefer official compressed later)

T04 — W-tier and remaining S decision pass

id: FI-WP-0003-T04
status: done
priority: medium
state_hub_task_id: "33daa2cf-f9fd-430d-a4e4-58996c900ca6"

W remain candidate: Qwen3-14B, R1-Distill-32B, bge-reranker-v2-m3.
S approved with collection_deferred history events.

T05 — Reserve index snapshot

id: FI-WP-0003-T05
status: done
priority: low
state_hub_task_id: "9974b98a-e1b3-4b7a-b2c2-0fd2cba65f97"

inventory/RESERVE-STATUS.md written.

Acceptance (workplan-level)

  • Catalog seeded + profile/SWOT
  • Storage pin complete
  • R + S approved (S deferred on disk)
  • R spine verified on media (embeds + Qwen3-8B + R1-Distill-14B; Llama gated residual)
  • RESERVE-STATUS snapshot

Residual handoffs

Residual Action
Llama-3.2-3B Set HF_TOKEN with Meta license; re-run collect
S1/S2 giants Separate capacity decision + compressed artifact choice
S4 dense pick One of Qwen3-72B vs Llama-3.3-70B under quota

Out of scope

  • Hosting inference endpoints
  • Full DeepSeek-V3/R1 671B pull without separate capacity decision
  • Dataset reserve