search

psugihara / FreeChat

llama.cpp based AI chat app for macOS
https://www.freechat.run
MIT License
425 stars 37 forks source link

55 improve reliability when switching between models #56

Closed psugihara closed 8 months ago

psugihara commented 8 months ago

@shavit I used ServerHealth but couldn't get the score to work correctly with the singleton, not sure why. This seems to work pretty well though.

Going to merge and keep testing on the TestFlight.

fixes #55

vercel[bot] commented 8 months ago

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name Status Preview Comments Updated (UTC)
free-chat ✅ Ready (Inspect) Visit Preview 💬 Add feedback Feb 3, 2024 4:16pm
shavit commented 8 months ago

The idea was to use the server health indicator in views, although I integrated it only in the settings view. Initially I added weighted score of last pings, to revive the server faster if it comes back online, but decided to simplify it. So if it takes few seconds to score up, that's the reason.

psugihara commented 8 months ago

The idea was to use the server health indicator in views, although I integrated it only in the settings view. Initially I added weighted score of last pings, to revive the server faster if it comes back online, but decided to simplify it. So if it takes few seconds to score up, that's the reason.

ok cool, i think the way I'm using for this case is OK then.