This is [Qwen/Qwen3-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct) converted to [OpenVINO](https://github.com/openvinotoolkit/openvino) with INT4 weight quantization.

This is a preview model. It is not yet fully supported by OpenVINO and Optimum Intel and there may be issues.
See https://gist.github.com/helena-intel/5d773ba0a57c895ed9a1060fe7805227 for an inference script, and the exact dependencies to install.