Comment by cyanydeez

I assume it was put out to get inference engines ready to do qwen4 models. For AMD halo, the Halgogen engine is the fastest i dound so fare