DevOps/software-infrastructure
2
0
Fork 0
Code Issues 16 Pull Requests Actions Packages Projects Releases Wiki Activity

LocalAI: unload idle models after 15 minutes

Browse Source 
A resident model holds the passively cooled Tesla T4 in P0 at about
30 W, which keeps the poweredge-t640 fans near 4,800 RPM. Unloading
idle models lets the card return to P8 at about 15 W.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
claude
2026-11-08 14:28:19 -05:00
co-authored by Claude Opus 5.5
parent 39f22f176a
commit 41faafd072
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
localai.yaml
+2 -2
View File
@@ -8,8 +8,8 @@
vars: vars:
localai_disable_webui: "false" localai_disable_webui: "false"
localai_debug: "true" localai_debug: "true"
localai_watchdog_idle: "false" localai_watchdog_idle: "true"
localai_watchdog_idle_timeout: "1m" localai_watchdog_idle_timeout: "15m"
localai_watchdog_busy: "false" localai_watchdog_busy: "false"
localai_p2p: "false" localai_p2p: "false"
localai_federated: "false" localai_federated: "false"