On Android, the non-CPU backends of llama.cpp have limited usability. As an alternative, other inference engines like MNN could be integrated in the future. Achieving this will likely require extensive validation and application architecture refactoring.
https://github.com/alibaba/MNN
On Android, the non-CPU backends of llama.cpp have limited usability. As an alternative, other inference engines like MNN could be integrated in the future. Achieving this will likely require extensive validation and application architecture refactoring.
https://github.com/alibaba/MNN