Skip to content

llamacpp provider never show llama.cpp server started message and the marquee never stop even though the server successfully started #82

Description

@affhp

Continue of #74

After this setup, I restart nvim, and perform the same steps shown in the README gif (Count to five), but it doesn't seems to respond back.
Image

This is probably caused by this line:

if out and out:find('HTTP server listening') then

It looks like the plugin is grepping the 'HTTP server listening' log message from llama-server to determine if the server is started.

However in recent llama-server, this log message 'HTTP server listening' was replaced by something like 'server is listening on', so that model.nvim can never determine llama-server is started or not, so the function following this step in the llamacpp provider can not be successfully executed, making the marquee never stop.

However I forked this plugin and simply changed this line to the correct log message but with no luck. I have no idea why. I have basically no lua knowledge.

Hope someone can help!

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions