Multi Lora on NPU

### Describe the documentation issue

I have been trying to first auto-opt my awq model to vitis which doesnt work. And second of all I have created an onnx model for npu and hybrid but cannot connect with adapters, saying inconsistent size. I would appreciate any help.

### Page / URL

_No response_

No page, file, test, or entry point is identified. Start by clarifying the intended AWQ-to-Vitis workflow and the ONNX NPU or hybrid adapter setup, including the inconsistent-size error. Done would require a specific reproducible workflow and documentation location, neither of which the report currently provides.

Written by the indexing model from the issue text.