Chris Van Pelt (CVP)

💯 plus one. Good to know I'm not the only one excited about this.

It's super simple to add models that serve with the same API as OpenAI. The Gemini series uses a google specific api protocol so we'd have to pull in that...

Hmm, I just tried ollama in Codespaces and it's working for me, I see this in the settings dialog: It sounds like what you're running into is the requirement for...

@hirenchauhan2 and @BeeTwenty those errors look like potentially a ollama compatibility issue. Can you verify you're running a fairly recent version of Ollama or share what version you have? You...

Hey guys, I just updated the README [with instructions](https://github.com/wandb/openui?tab=readme-ov-file#docker-compose) for running via Docker Compose. That might be easiest all be it slow.

@YangZhiBoGreenHand if you're not running the backend on the default port you'll need to rebuild the container in "hosted" mode. From the frontend directory run: ``` npm install npm run...

Should be possible, I'll look into adding that option.

@MMoneer I just [pushed a fix ](https://github.com/wandb/openui/commit/9aeff3503fe46c907ca8a6db6dfa0e4819cfe903) that should resolve this, not sure what's going on exactly but can you try pulling and seeing if you still get the error?

Hey @alfi4000 I'm going to make the codespace setup just work with Ollama and provide more instructions. I'll link to it and close this ticket once it's done hopefully later...

Hey guys, just pushed the changes, should work now. [Instructions for GH codespaces](https://github.com/wandb/openui?tab=readme-ov-file#ollama) here.