Daniel Nowak

Left this on the stackoverflow as well since I'm not sure which is more likely to capture proper attention. Chris, just happened to check back here. Thanks for your awesome...

Duplicate of https://github.com/portainer/portainer/issues/12319

This may have introduced a regression by falling back to "gpt-3.5-turbo-1106" model (if one not given in the UI). The regression would be that the backend no longer respects the...

+1 for this feature. I switched from Ollama (which has it) to text-generation-webui in the hopes that I could serve _watt-ai/watt-tool-8B_ from Hugging Face (which is near the top of...

Sure, not built from source - pasting relevant distro-provided packages: ``` > ii bluez 5.66-1+rpt1+deb12u1 arm64 Bluetooth tools and daemons > ii bluez-alsa-utils 4.0.0-2 arm64 Bluetooth Audio ALSA Backend (utils)...

It would also be helpful to know how to set debug logging to true, without having to mount a config.toml into a docker container > but all these vars are...

By the way, DEBUG_RUNTIME didn't seem to work, but RUNTIME_DEBUG did. That one isn't mentioned on the doc page at all

Unfortunately I'm still having an issue after a fresh pull and _make build_. The UI is telling me that I have the model I want "ollama/deepseek-coder:instruct" , and I have...

the FE already has the model I want (see screenshot below). And you can see my config.toml is correct too (futher below). Yet still it keeps printing this error ...

> It might make more sense for a condensation strategy to work over tokens instead of Event objects Agreed - is this being considered? Conversations on a local LLM (Devstral)...