#machine-learning #back-end #models #inference #blazingly #diffusion #quantization

yanked diffuse_rs_backend

Blazingly fast inference of diffusion models

4 releases

0.1.3 Jan 7, 2025
0.1.2 Jan 7, 2025
0.1.1 Jan 7, 2025
0.1.0 Jan 7, 2025

#14 in #diffusion


Used in 2 crates (via diffuse_rs_core)

MIT license

3MB
73K SLoC

Rust 51K SLoC // 0.0% comments Metal Shading Language 15K SLoC // 0.1% comments CUDA 6.5K SLoC // 0.0% comments Python 25 SLoC // 0.3% comments

diffuse_rs_backend

Backend for quantization in diffuse-rs.

Dependencies

~29–43MB
~755K SLoC