Bin-Huang / chatbox

User-friendly Desktop Client App for AI Models/LLMs (GPT, Claude, Gemini, Ollama...)
https://chatboxai.app
GNU General Public License v3.0
20.64k stars 2.09k forks source link

[BUG] ollama llava:7b is a vision model, but chatbox said it's not. #1446

Closed nne998 closed 5 days ago

nne998 commented 1 month ago

Bug Description

ollama llava:7b is a vision model ( https://ollama.com/library/llava ), but chatbox said it's not.

how to reproduce chat with an image

JPZ4-5 commented 1 month ago

It occurred when testing llava:34B on Windows, too.

llava:34B works perfectly well on open-webui when sending pictures.

chllei commented 1 month ago

refer to #1390, please consider image support in ollama mode

ldwnt commented 1 month ago

so is this a bug or due to biz choice

nne998 commented 5 days ago

fixed by v1.4.0