bug: macos network installer not install llama-cpp engines by default

janhq / cortex.cpp

Local AI API Platform

https://cortex.so

Apache License 2.0

2.09k stars 119 forks source link

bug: macos network installer not install llama-cpp engines by default #1616

Closed hiento09 closed 1 week ago

hiento09 commented 1 week ago

Context

From nightly version v1.0.0-191, cortex-install on macos not install llama-cpp engine by default

Goal

Fix the bug
Add e2e testing for installer

hiento09 commented 1 week ago

Thanks @vansangpfiev for finding out this bug

gabrielle-ong commented 1 week ago

Hi @hiento09, I tried on v223 mac network installer

seems that llama-cpp engine is not installed, am I missing something?

Steps: uninstalled engines, uninstalled cortex-nightly, reinstalled cortex-nightly

gabrielle-ong commented 1 week ago

Thanks @hiento09! Sorry I'm still not getting it on install on the v227 network installer

to confirm, llama-cpp should be listed in engines list upon install right?
it does successfully pull llama-cpp on cortex run

hiento09 commented 1 week ago

Hi @gabrielle-ong , can you help test version 226? Version 226 was built from my branch for testing and has not been merged into the dev branch. Version 227 is the nightly version from yesterday and does not include my fix PR.

gabrielle-ong commented 1 week ago

Adding the logs for future reference - thanks @hiento09 for investigating. tested v226 -

Works on VM (test-macos-13-1)
Doesnt work on my local mac - /var/log/install.log attached cortex-installer.log

gabrielle-ong commented 1 week ago

tested on v228 for branch feat/cicd-docker, works on my local now. Thanks @hiento09!