Completion instead of Completions about continue HOT 7 CLOSED

Yes it works! Thanks a lot!

from continue.

Thanks for finding this! I'll fix right away and include it in the version I'm releasing tomorrow morning

from continue.

@xaranex1 I took a quick look into this - it seems the problem is actually our documentation...

The LlamaCpp class is made for this server, so has a different endpoint name from OpenAI (this is not a typo)

This isn't completely obvious from our documentation, so I'll make some improvements there. Could you share which server you are trying to use and maybe I can help get that setup?

from continue.

I use vast.ai with https://github.com/continuedev/ggml-server-example @sestinj

from continue.

And i get error HTTP ERROR 404 on completion because endpoints in llmama cpp python are completions if i go to /docs its okay

from continue.

@xaranex1 Could you try using the GGML class and let me know if this works? This is the one that we intended to be used with llama-cpp-python, though of course this naming is a bit backward

from continue.

Fantastic! I'll close the issue, but let me know if you have any other issues : ) Can ask in Discord if that's ever easier

I'll also rename the GGML class probably to avoid this confusion

from continue.