Yuge Zhang
Yuge Zhang
You didn't provide your provider and apiUrl. Neither did you compare your config with the config I provided. So I can hardly help you. I think it's probably a silly...
I checked the docs of https://models.github.ai/inference and figured that they don't support response API yet. I'll post a hot-fix to revert to chat completion API quickly.
The hot-fix is merged. If everything went well, it will appear in the nightly build tonight. aihubmix is a Chinese website which I used most often. They provide features similar...
I notice the gpt-4o in this snapshot. So you configured to use deepseek-chat, but here it used gpt-4o?
Just tried deepseek-chat on aihubmix. It does not support response API either. Chat API works fine. I recommend to try the nightly build https://poml-vscode-nightly.scottyugochang.workers.dev/ The problem seems quite annoying if...
I changed the title so that whoever encounters a similar problem could find the solution.
Yep. Still a response API error. Try the nightly build please: https://poml-vscode-nightly.scottyugochang.workers.dev/
POML is a pure-local library that doesn't expose any service to any network. So I think the outdated dependencies are mostly "you can't hurt yourself if you didn't mean it"...
duplicate of #80
Duplicate of #76