Stars
You can use Diffusion for Language too!!
A unified inference and post-training framework for accelerated video generation.
[ICML2025, NeurIPS2025 Spotlight] Sparse VideoGen 1 & 2: Accelerating Video Diffusion Transformers with Sparse Attention
Mastering Diverse Domains through World Models
State-of-the-art Image & Video CLIP, Multimodal Large Language Models, and More!
SkyReels-V2: Infinite-length Film Generative model
Inference-time scaling of diffusion-based image and video generation models.
A FlashAttention implementation for JAX with support for efficient document mask computation and context parallelism.
JAX bindings for the flash-attention2 kernels
A collection of resources and papers on Diffusion Models
paper list, tutorial, and nano code snippet for Diffusion Large Language Models.
Microbenchmarking hyperparameter tuning for JAX functions.
Minimal yet performant LLM examples in pure JAX
Pytorch implementation of Transfusion, "Predict the Next Token and Diffuse Images with One Multi-Modal Model", from MetaAI
USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference
xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism
Lets make video diffusion practical!
DeerFlow is a community-driven Deep Research framework, combining language models with tools like web search, crawling, and Python execution, while contributing back to the open-source community.
Efficient Triton Kernels for LLM Training
Tile primitives for speedy kernels