summaryrefslogtreecommitdiff
path: root/candle-transformers/src/models/mmdit/mod.rs
blob: ce4872e0b2887d8051e695a542569a2fbd271056 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
//! Mix of Multi-scale Dilated and Traditional Convolutions
//!
//! Mix of Multi-scale Dilated and Traditional Convolutions (MMDiT) is an architecture
//! introduced for Stable Diffusion 3, with the MMDiT-X variant used in Stable Diffusion 3.5.
//!
//! - [Research Paper](https://arxiv.org/abs/2403.03206)
//! - ComfyUI [reference implementation](https://github.com/comfyanonymous/ComfyUI/blob/78e133d0415784924cd2674e2ee48f3eeca8a2aa/comfy/ldm/modules/diffusionmodules/mmdit.py)
//! - Stability-AI [MMDiT-X implementation](https://github.com/Stability-AI/sd3.5/blob/4e484e05308d83fb77ae6f680028e6c313f9da54/mmditx.py)

pub mod blocks;
pub mod embedding;
pub mod model;
pub mod projections;