DGR-049: Keep V4 attention and auxiliary state shard-local #33

New Issue

popov · 2026-07-16T20:36:05Z

popov commented

2026-07-16 20:36:05 +00:00

DGR-049: Keep V4 attention and auxiliary state shard-local

Status / triage: specification only; ready-for-agent; passes: false
Execution mode: AFK
Milestone: M3
Dependencies: DGR-046, DGR-047
Blocks (derived): DGR-051
Labels: area:deepseek-v4, area:attention, type:runtime, priority:p0, ready-for-agent
Evidence class: real-model
Hardware: optional
Model: deepseek-v4-flash
Upstream: yes

Objective / description

Fresh Ralph session: read .scratch/distributed-gguf-runtime/RALPH-CONTEXT.md, source issue .scratch/distributed-gguf-runtime/issues/049-keep-v4-attention-and-auxiliary-state-shard-local.md, and evidence READMEs for dependencies (DGR-046, DGR-047) before changing code. Inspect live source/tests rather than trusting legacy pass states. Objective: Bind V4 CSA/HCA/SWA/indexer/compressor and KV state to the owning shard session/epoch rather than transferring caches over the WAN.

Acceptance criteria

Map CSA, HCA, SWA, indexer, compressor, and KV ownership from the pinned upstream graph to each owned layer range.
Key all such caches/state by route session and epoch with bounded lifecycle, memory accounting, stale-epoch rejection, release, and eviction.
Prove cross-stage messages carry only the typed architecture boundary and cache expectations, never per-layer cache contents.
On route change or state loss, use explicit cache miss and re-prefill/restart; do not migrate V4 caches across workers.
Applicable shared quality gates in prd.json pass, and the evidence handoff records exact commands/results, changed files, limitations, and dependency handoff.

Shared quality gates

Targeted deterministic tests pass; Python changes also pass python -m compileall packages tests.
git diff --check passes.
Default tests are model-download-free, API-credit-free, and GPU-free.
Evidence README records exact changed files, commands/results, limitations, and dependency handoff; no fabricated evidence or inherited completion credit.

Evidence handoff

Write and verify .scratch/distributed-gguf-runtime/evidence/DGR-049/README.md. Until every criterion and applicable gate has real evidence, this story remains passes: false. Legacy evidence is provenance only, not completion credit.

# DGR-049: Keep V4 attention and auxiliary state shard-local - **Status / triage:** specification only; `ready-for-agent`; `passes: false` - **Execution mode:** `AFK` - **Milestone:** `M3` - **Dependencies:** `DGR-046`, `DGR-047` - **Blocks (derived):** `DGR-051` - **Labels:** `area:deepseek-v4`, `area:attention`, `type:runtime`, `priority:p0`, `ready-for-agent` - **Evidence class:** `real-model` - **Hardware:** `optional` - **Model:** `deepseek-v4-flash` - **Upstream:** `yes` ## Objective / description Fresh Ralph session: read `.scratch/distributed-gguf-runtime/RALPH-CONTEXT.md`, source issue `.scratch/distributed-gguf-runtime/issues/049-keep-v4-attention-and-auxiliary-state-shard-local.md`, and evidence READMEs for dependencies (DGR-046, DGR-047) before changing code. Inspect live source/tests rather than trusting legacy pass states. Objective: Bind V4 CSA/HCA/SWA/indexer/compressor and KV state to the owning shard session/epoch rather than transferring caches over the WAN. ## Acceptance criteria - [ ] Map CSA, HCA, SWA, indexer, compressor, and KV ownership from the pinned upstream graph to each owned layer range. - [ ] Key all such caches/state by route session and epoch with bounded lifecycle, memory accounting, stale-epoch rejection, release, and eviction. - [ ] Prove cross-stage messages carry only the typed architecture boundary and cache expectations, never per-layer cache contents. - [ ] On route change or state loss, use explicit cache miss and re-prefill/restart; do not migrate V4 caches across workers. - [ ] Applicable shared quality gates in `prd.json` pass, and the evidence handoff records exact commands/results, changed files, limitations, and dependency handoff. ## Shared quality gates - Targeted deterministic tests pass; Python changes also pass `python -m compileall packages tests`. - `git diff --check` passes. - Default tests are model-download-free, API-credit-free, and GPU-free. - Evidence README records exact changed files, commands/results, limitations, and dependency handoff; no fabricated evidence or inherited completion credit. ## Evidence handoff Write and verify `.scratch/distributed-gguf-runtime/evidence/DGR-049/README.md`. Until every criterion and applicable gate has real evidence, this story remains `passes: false`. Legacy evidence is provenance only, not completion credit.

popov added this to the M3 milestone 2026-07-16 20:36:05 +00:00

popov added the area:attention type:runtime area:deepseek-v4 ready-for-agent priority:p0 status:blocked labels 2026-07-16 20:36:05 +00:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: popov/neuron-tai#33