till/managarten
1
0
Fork 0
mirror of https://github.com/Memo-2023/mana-monorepo.git synced 2026-05-20 19:26:41 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

feat: add Ollama memory optimization, LLM metrics, and chat streaming

Browse source 
Three improvements to the unified LLM infrastructure:

1. Ollama memory optimization (scripts/mac-mini/configure-ollama.sh):
   - OLLAMA_KEEP_ALIVE=5m → models unload after 5min idle (saves 3-16GB RAM)
   - OLLAMA_NUM_PARALLEL=1 → predictable memory usage
   - OLLAMA_MAX_LOADED_MODELS=1 → max 1 model in RAM at a time

2. Request-level metrics in @manacore/shared-llm:
   - LlmRequestMetrics interface (model, latency, tokens, fallback detection)
   - LlmMetricsCollector class with summary stats (for health endpoints)
   - Optional onMetrics callback in LlmModuleOptions
   - Automatic metrics emission in chatMessages() (success + error)

3. Chat streaming (token-by-token SSE):
   - Backend: POST /chat/completions/stream SSE endpoint
   - OllamaService.createStreamingCompletion() via llm.chatStreamMessages()
   - ChatService.createStreamingCompletion() with upfront credit consumption
   - Web: chatApi.createStreamingCompletion() SSE consumer
   - Chat store: sendMessage() now streams tokens into assistant message
   - UI updates reactively as each token arrives

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Till JS 2026-03-24 09:41:33 +01:00
parent ecda4535d8
commit 56ffcbac39
13 changed files with 462 additions and 29 deletions
Download patch file Download diff file Expand all files Collapse all files

5
docs/MAC_MINI_SERVER.md
View file

@ -594,8 +594,13 @@ Systemeinstellungen → Datenschutz & Sicherheit → Voller Festplattenzugriff
**LaunchAgent:** `~/Library/LaunchAgents/homebrew.mxcl.ollama.plist`
Optimierungen bereits aktiviert:
- `OLLAMA_KEEP_ALIVE=5m` - Modelle nach 5min Inaktivität aus RAM entladen (spart 3-16 GB)
- `OLLAMA_FLASH_ATTENTION=1` - Schnellere Attention-Berechnung
- `OLLAMA_KV_CACHE_TYPE=q8_0` - Effizienterer KV-Cache
- `OLLAMA_NUM_PARALLEL=1` - Max 1 paralleler Request (vorhersagbarer RAM)
- `OLLAMA_MAX_LOADED_MODELS=1` - Max 1 Modell gleichzeitig im RAM
Setup-Script: `./scripts/mac-mini/configure-ollama.sh`
### Speicherort