Josh Leverette

@lashahub Nice. I ended up moving all of this machine learning stuff over to a different computer running Ubuntu.

What's especially weird/annoying here is that this could be fixed (basically) just by removing a single if statement... https://github.com/menloresearch/jan/blob/c74e5280e75512751711da8f99e24ca7da5470f4/web-app/src/services/providers.ts#L174 Providing an API key breaks because of CORS, since Ollama doesn't...

This is a complete patch to add Ollama support to jan: ```patch diff --git a/web-app/src/containers/DropdownModelProvider.tsx b/web-app/src/containers/DropdownModelProvider.tsx index 14a53c12..3b12ccc5 100644 --- a/web-app/src/containers/DropdownModelProvider.tsx +++ b/web-app/src/containers/DropdownModelProvider.tsx @@ -151,8 +151,12 @@ const DropdownModelProvider =...

@ed0c I’m not associated with Jan in any way, so I can’t tell you anything other than that you can build the software yourself with those changes, or just wait...

@louis-menlo this still does not work. Did anyone test the fix in #6224 against olllama?

I run `ollama` on a separate (more powerful) machine. If I open Jan's developer console, this is the output I see when refreshing the list of models: ```json { "status":...

@louis-menlo Ok, setting `OLLAMA_ORIGINS=*` seems to have resolved the issue. Thanks!

Anecdotally, I observed that the previous PR (and presumably this PR too) essentially fixed tool calling for qwen3-coder. Although when trying to use it with `codex`, `qwen3-coder` absolutely refuses to...

> which for the 10M context length is going to be giant I don't think they said anything about fitting onto a single GPU with 10M context. Very, _very_ few...

New release of Phi-3.5-vision-instruct today: https://huggingface.co/microsoft/Phi-3.5-vision-instruct (As well as a [16x3.8B MoE](https://huggingface.co/microsoft/Phi-3.5-MoE-instruct) and an updated version of the basic [Phi-3.5-mini](https://huggingface.co/microsoft/Phi-3.5-mini-instruct))