Matthieu Borgognon
Matthieu Borgognon
Not sure it is, or maybe I make something wrong, but it always fails to set architect as new model Here my process: 1. Update local server and strat it...
It does not fix the problem (no visible effect)
It looks better, at least I have other problems now, so I will need to solve them first before completely validate this path.
No logs anywhere... Maybe I did something wrong 
Definitely nice catch, but still a bit confusing for me for such a big model as Gemma3, what could I do here because it support MM interaction as far as...
Thank you very much for pointing me out there, I appreciate your help. Unfortunately I have different error but still always the same blank page result at the end, here...
Thanks, the "Could not parse response." was apparently fixed by #1027 But your point to the EOF problem give me no clue. If I extrapolate, I can imagine it for...
Thanks @pirate but it seems that this is not solving our problem so far I understand here: ## Update package ``` uv pip install git+https://github.com/browser-use/browser-use Resolved 59 packages in 1.06s...
@maranello-o Sounds not, because the LLM Provider is still using OpenAI how to support Ollama models to have a full local implementation. Not only the transcribe_provider
Sorry folks I have a lot on the plate actually. I planned to review mine until Sunday, maybe @kadiryildiz283 will be quicker who knows. Best,