Lightweight inference library for ONNX files, written in C++. It can run Stable Diffusion XL 1.0 on a RPI Zero 2 (or in 298MB of RAM) but also Mistral 7B on desktops and servers. ARM, x86, WASM, RISC-V supported. Accelerated by XNNPACK.
Rereading that sentence, I decided to avoid any possibility of misunderstanding and I reformulated the entire text (also adding a more accurate technical context).
If something is 0.5x slower, then it is 2x faster. I presume, and the performance data seems to confirm, that it was intended to say "1.5-2x slower".