Issue using continue with several ollama models

[lencic]

I'm trying to use multiple ollama models for different tasks. But when continue need to switch to the second model, the first model turns off. How can I make multiple models run at the same time instead of shutting down if a different model starts to be used? I have enough memory to use two models at the same time

[bdougie]

@lencic can you provide more specifics to assist in triaging. Which models and which modes are you using?

[lencic]

I'm using qwen2.5-coder:1.5b for autocomplete and qwen2.5-coder:14b for other tasks. I have 16gb vram and If these models are run manually through ollama, they work perfectly at the same time.

[uinstinct]

Do you mean this happens when selecting models from the chat panel as well (meaning clicking on the dropdown and then selecting one)?

[lencic]

No, I don't change models from the chat panel, that's not why the problem happens