fix(desktop): stop pinning local models in memory and cap audio buffers
Some checks failed
ci / 정본·보안·린트·타입·테스트 (push) Successful in 56s
ci / .NET API 서버 테스트 (push) Successful in 10s
deploy-site / deploy (push) Successful in 38s
ci / 모바일 린트·타입·Jest (push) Successful in 44s
ci / 워크스페이스 빌드 검증 (push) Successful in 54s
ci / Supabase Edge Functions + Cloudflare Worker (push) Successful in 22s
portable-unsigned / portable-windows (push) Failing after 35s
release / release-windows (push) Failing after 13s
Some checks failed
ci / 정본·보안·린트·타입·테스트 (push) Successful in 56s
ci / .NET API 서버 테스트 (push) Successful in 10s
deploy-site / deploy (push) Successful in 38s
ci / 모바일 린트·타입·Jest (push) Successful in 44s
ci / 워크스페이스 빌드 검증 (push) Successful in 54s
ci / Supabase Edge Functions + Cloudflare Worker (push) Successful in 22s
portable-unsigned / portable-windows (push) Failing after 35s
release / release-windows (push) Failing after 13s
The PC-wide slowdown was not a leak but model residency: boot warmup loaded gemma4:e4b (~11.6 GB RAM incl. 5.9 GB pinned host buffer, no mmap on Windows CUDA) and kept it for 30m, next to the large-v3 sidecar. Measured on the HEAD build: free RAM 7.9 GB -> 32 MB within 20 s; after this change 10.8 GB with no Ollama model loaded until typing starts. - single LOCAL_LLM_KEEP_ALIVE (10m) for every local request; no boot warmup - unload this app's resident Ollama models (keep_alive 0) on quit - sidecar POST /unload for the aux slot; captions release it on stop - partial dictation copies only the tail window, not the whole recording - cap the voice-conversation listen buffer at 5 minutes - mic test: generation guard so re-entry cannot orphan lease/timers - startRealtime: single-flight and removeChannel to stop duplicate bindings - meeting WAV save writes header and PCM without a concat copy Docs: AI-01 policy note, GAP-PERF-01 with remaining work.
This commit is contained in:
parent
a7c0d6fb21
commit
d63854db67
19 changed files with 268 additions and 45 deletions
|
|
@ -37,7 +37,7 @@ Status quick-reference: `[x]` done+verified · `[~]` partial/unverified · `[ ]`
|
|||
|
||||
| ID | Feature | D | W | M | B | Anchors / notes |
|
||||
|---|---|---|---|---|---|---|
|
||||
| AI-01 | Local LLM (Ollama) | [~] | [-] | [ ] | [-] | Desktop bundled Ollama. 2026-09-22 guard contract: no boot warmup; suggestion `keep_alive: 2m`; request-owned cancellation; generate/stream 2048 tokens / 120 s and chat 512 / 60 s; `done` frame required with incomplete-stream cleanup; deduplicated Ollama spawn/poll plus lifecycle disposal. GAP-LLM-03 has targeted evidence (6 test files / 69 passed / 0 failed; changed code/tests ESLint and diff check exit 0). Raw Ollama confirms cold timeout cleanup, a 553 ms `num_predict=1`/`keep_alive='2m'` response, and after the about-119.9 s expiry a 19:48:59 +09:00 `/api/ps` HTTP 200 / 45.8 ms `{models:[]}` observation with no intervening generate/unload/kill/retry; `/api/version` was HTTP 200 / 7.3 ms / `0.32.13`. This is raw API expiry/unload evidence, not app-restart or GUI/runtime proof (GAP-LLM-04). |
|
||||
| AI-01 | Local LLM (Ollama) | [~] | [-] | [ ] | [-] | Desktop bundled Ollama. 2026-09-22 guard contract: no boot warmup; suggestion `keep_alive: 2m`; request-owned cancellation; generate/stream 2048 tokens / 120 s and chat 512 / 60 s; `done` frame required with incomplete-stream cleanup; deduplicated Ollama spawn/poll plus lifecycle disposal. GAP-LLM-03 has targeted evidence (6 test files / 69 passed / 0 failed; changed code/tests ESLint and diff check exit 0). Raw Ollama confirms cold timeout cleanup, a 553 ms `num_predict=1`/`keep_alive='2m'` response, and after the about-119.9 s expiry a 19:48:59 +09:00 `/api/ps` HTTP 200 / 45.8 ms `{models:[]}` observation with no intervening generate/unload/kill/retry; `/api/version` was HTTP 200 / 7.3 ms / `0.32.13`. This is raw API expiry/unload evidence, not app-restart or GUI/runtime proof (GAP-LLM-04). **2026-10-11 residency policy (supersedes 2m / later 30m):** single `LOCAL_LLM_KEEP_ALIVE='10m'` for every local request (`services/llm/model-residency.ts`), no boot warmup (first keystroke warms via `SuggestionService._keepWarm`), keep-warm refresh only when <4 min left, and `keep_alive: 0` unload of this app's resident models on quit (`LocalLLMService.unloadResidentModels`). Reason: gemma4:e4b holds ~11.6 GB RAM (5.9 GB pinned host buffer, no mmap on Windows CUDA) — boot warmup took free RAM 7.9 GB→32 MB within 20 s (GAP-PERF-01). |
|
||||
| AI-02 | Cloud LLM (Claude/OpenAI) | [x] | [x] | [x] | [x] | Desktop `PremiumLLMService`; web/mobile via `llm-proxy`; .NET `LlmProxyService` |
|
||||
| AI-03 | Auto Polish (cleanup/filler removal) | [x] | [~] | [~] | [x] | Desktop built-in; web/mobile via commands. Desktop Auto Polish is the plain `refine` action (`llm-prompts.ts:14`), not a custom instruction, so it was **not** affected by the 2026-09-21 instruction-prompt fix (AI-05); regression cases now pin `refine`/`summarize`/`grammar`/`expand` (`VoiceModeService.test.ts:435`, `llm-prompts.test.ts:153`) |
|
||||
| AI-04 | Translate / summarize / rephrase | [~] | [x] | [x] | [x] | Built-in instructions. **Desktop has two paths and only one of them worked.** The plain-action path (Settings → `defaultLLMAction`, `SettingsModal.tsx:653`) reads `BASE_SYSTEM_PROMPTS` directly and was always correct. The built-in *instruction* presets (`CustomInstructionService.ts:26/35/44/53/62`) ran through the custom-instruction path and inserted the instruction's own wording instead of the result — see AI-05. Fixed in `9c2b4d4` (2026-09-21), **not verified in a running app** (`11` GAP-LLM-02). Translate still always targets English: `AppConfig` has no target-language key and neither `language` (UI locale) nor `sttLanguage` (source language) can stand in (`llm-prompts.ts:37-52`, `11` GAP-LLM-01) |
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue