Hacker-Software/ARIA-AGENT
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 236 Wiki Activity

fix(xtts): GPU-Pinning an 8GB+12GB-Realitaet anpassen

Browse Source 
Die zwei 3060 sind ungleich (GPU0=8GB, GPU1=12GB), nicht 12+12. Groesstes Modell (STT, spaeter Voxtral-STT-3B ~9GB) muss auf die 12GB-Karte: whisper->GPU1. TTS(F5)+LLM auf die 8GB-Karte: ->GPU0.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
duffyduck
2026-08-15 01:53:59 +02:00
co-authored by Claude Opus 4.8
parent 37aaa90239
commit 75daadf72d
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
xtts/docker-compose.yml
+3 -3
View File
@@ -30,7 +30,7 @@ services:
reservations:
devices:
- driver: nvidia
device_ids: ["1"] # TTS-Rolle → GPU 1 (spaeter: Voxtral-TTS)
device_ids: ["0"] # TTS → GPU 0 (8 GB; F5 ist klein)
capabilities: [gpu]
volumes:
- ./voices:/voices # WAV + TXT Referenz
@@ -68,7 +68,7 @@ services:
reservations:
devices:
- driver: nvidia
device_ids: ["0"] # STT-Rolle → GPU 0 (spaeter: Voxtral-STT-3B)
device_ids: ["1"] # STT/groesstes Modell → GPU 1 (12 GB; spaeter Voxtral-STT-3B ~9 GB)
capabilities: [gpu]
environment:
- RVS_HOST=${RVS_HOST}
@@ -108,7 +108,7 @@ services:
reservations:
devices:
- driver: nvidia
device_ids: ["0"] # LLM → GPU 0 (teilt sich mit dem kleinen STT)
device_ids: ["0"] # LLM → GPU 0 (8 GB, teilt sich mit TTS)
capabilities: [gpu]
volumes:
- ./models:/models # HF-Download-Cache (persistent)