thu-ml / SpargeAttnLinks
SpargeAttention: A training-free sparse attention that can accelerate any model inference.
β695Updated 2 weeks ago
Alternatives and similar repositories for SpargeAttn
Users that are interested in SpargeAttn are comparing it to the libraries listed below
Sorting:
- πA curated list of Awesome Diffusion Inference Papers with Codes: Sampling, Cache, Quantization, Parallelism, etc.πβ380Updated last week
- Model Compression Toolbox for Large Language Models and Diffusion Modelsβ607Updated 2 weeks ago
- A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Trainingβ493Updated this week
- https://wavespeed.ai/ Context parallel attention that accelerates DiT model inference with dynamic cachingβ365Updated last month
- [ICML2025] Sparse VideoGen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsityβ417Updated this week
- β173Updated 7 months ago
- A sparse attention kernel supporting mix sparse patternsβ281Updated 6 months ago
- [CVPR 2024 Highlight] DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Modelsβ703Updated 8 months ago
- [ICCV2025] From Reusing to Forecasting: Accelerating Diffusion Models with TaylorSeersβ269Updated 2 weeks ago
- Radial Attention Official Implementationβ485Updated 3 weeks ago
- π Collection of awesome generation acceleration resources.β322Updated last month
- [ICLR2025] Accelerating Diffusion Transformers with Token-wise Feature Cachingβ173Updated 5 months ago
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inferenceβ550Updated last month
- Quantized Attention achieves speedup of 2-5x and 3-11x compared to FlashAttention and xformers, without lossing end-to-end metrics acrossβ¦β2,296Updated 3 weeks ago
- [ICLR'25] ViDiT-Q: Efficient and Accurate Quantization of Diffusion Transformers for Image and Video Generationβ115Updated 5 months ago
- π€An Unified Cache Acceleration Toolbox for DiTs: Qwen-Image, Qwen-Image-Edit, FLUX.1, Wan2.1/2.2, etc.β229Updated this week
- A parallelism VAE avoids OOM for high resolution image generationβ75Updated 3 weeks ago
- β408Updated 2 weeks ago
- [ICML 2025] XAttention: Block Sparse Attention with Antidiagonal Scoringβ222Updated last month
- Official implementation of "Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding"β386Updated this week
- VeOmni: Scaling any Modality Model Training to any Accelerators with PyTorch native Training Frameworkβ975Updated this week
- An open-source implementation of Regional Adaptive Sampling (RAS), a novel diffusion model sampling strategy that introduces regional varβ¦β138Updated 2 months ago
- xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelismβ2,235Updated last month
- π³ Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"β833Updated 5 months ago
- High performance inference engine for diffusion modelsβ79Updated 3 weeks ago
- Combining Teacache with xDiT to Accelerate Visual Generation Modelsβ31Updated 4 months ago
- Accelerate Video Diffusion Inference via Sketching-Rendering Cooperationβ16Updated 2 months ago
- [ICCV 2023] Q-Diffusion: Quantizing Diffusion Models.β354Updated last year
- [ICLR 2025] DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Headsβ487Updated 6 months ago
- [ICLR 2025] COAT: Compressing Optimizer States and Activation for Memory-Efficient FP8 Trainingβ231Updated 3 weeks ago