![]() ![]() ![]() The reason why MacOS ARM64 build is not available is that it needs ARM Python Dev Libs which I have no ideas on how to For other platforms, you may compile it on your own. ![]() Linux 圆4, Windows 圆4 and MacOS 圆4 releases are available now. Of course, the auto tilesize and prepadding settings are implemented, so don't worry about them. As a result, functions other than the core upscaling and denoising such as multi-thread loading and saving are not available. This project only wrapped the original Waifu2x class. Runs fast on Intel / AMD / Nvidia with Vulkan API. Waifu2x-ncnn-vulkan is nihui's ncnn implementation of waifu2x converter. A Python FFI of nihui/waifu2x-ncnn-vulkan achieved with SWIG. ![]()
0 Comments
Leave a Reply. |