fizzlepoof/truenas-stacks
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

workflow 08: v1.1 fixes + docs update

Browse Source 
- 08-class-recording-rag-v1.1.json: Whisper URL → 10.5.1.16:8786,
  model → faster-distil-whisper-small.en, auth headers → env vars,
  inputDataFieldName fix, Paperless response format text
- PLAUSIBLEDENABILITY.md: remove Whisper from GPU list (moved to N.O.M.A.D.)
- n8n-workflows/README.md: fix class recordings path to /mnt/tank,
  correct workflow 08 trigger (webhook, not folder watcher),
  add curl upload example and docker-compose setup instructions
This commit is contained in:
Paccoco
2026-05-09 19:07:27 -05:00
parent 2ee549d843
commit 31edbf1970
3 changed files with 567 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
docs/servers/PLAUSIBLEDENABILITY.md
View File

@@ -31,9 +31,10 @@ Primary Docker host running all production workloads.
RTX 2080 Ti used by:
- Plex (hardware transcoding)
- immich-ml (CUDA face recognition)
- Whisper (CUDA speech-to-text, port 8786)
- Ollama light-tier inference (qwen2.5:14b and smaller)
> **Note:** Whisper moved to N.O.M.A.D. (10.5.1.16:8786) as CPU inference — avoids VRAM conflict with Ollama. See [NOMAD.md](NOMAD.md).
Recommended models for 11GB VRAM:
| Model | VRAM | Best for |
|-------|------|---------|