Image + Text support

[kghandour]

A user should have the ability to select a local photo, and insert a text prompt, and ask the multimodal model about it.

• [x] Add support for Ollama multimodal input
• [x] Insert file import option
• [x] Show the photo imported to the user above the text prompt field
• [x] Show the image and the text prompt in the chat bubble of the user.

[shayneoneill]

This would be spectacular. bakllava is a great little model.