Alternatives and similar repositories for draft-attention

Users that are interested in draft-attention are comparing it to the libraries listed below

• thu-nics / DiTFastAttn
  ☆190Updated last year
• Peyton-Chen / Sparse-vDiT
  The official implementation of "Sparse-vDiT: Unleashing the Power of Sparse Attention to Accelerate Video Diffusion Transformers" (arXiv …
  ☆50Updated 8 months ago
• thu-ml / SLA
  SLA: Beyond Sparsity in Diffusion Transformers via Fine-Tunable Sparse–Linear Attention
  ☆264Updated 3 weeks ago
• KlingTeam / VMoBA
  Official implementation of paper "VMoBA: Mixture-of-Block Attention for Video Diffusion Models"
  ☆62Updated 7 months ago
• prathebaselva / FORA
  FORA introduces simple yet effective caching mechanism in Diffusion Transformer Architecture for faster inference sampling.
  ☆52Updated last year
• AdaCache-DiT / AdaCache
  Code for our ICCV 2025 paper "Adaptive Caching for Faster Video Generation with Diffusion Transformers"
  ☆165Updated last year
• horseee / learning-to-cache
  [NeurIPS 2024] Learning-to-Cache: Accelerating Diffusion Transformer via Layer Caching
  ☆116Updated last year
• TZW1998 / ParaTAA-Diffusion
  This is the official repo for the paper "Accelerating Parallel Sampling of Diffusion Models" Tang et al. ICML 2024 https://openreview.net…
  ☆16Updated last year
• xdit-project / DistVAE
  A parallelism VAE avoids OOM for high resolution image generation
  ☆85Updated 6 months ago
• thu-nics / ViDiT-Q
  [ICLR'25] ViDiT-Q: Efficient and Accurate Quantization of Diffusion Transformers for Image and Video Generation
  ☆149Updated 10 months ago
• Tele-AI / TeleTron
  To pioneer training long-context multi-modal transformer models
  ☆72Updated 6 months ago
• svg-project / Sparse-VideoGen
  [ICML2025, NeurIPS2025 Spotlight] Sparse VideoGen 1 & 2: Accelerating Video Diffusion Transformers with Sparse Attention
  ☆627Updated last week
• SingleZombie / LLSA
  Official implementation of Log-linear Sparse Attention (LLSA).
  ☆56Updated last week
• microsoft / RAS
  An open-source implementation of Regional Adaptive Sampling (RAS), a novel diffusion model sampling strategy that introduces regional var…
  ☆150Updated 7 months ago
• thu-nics / MixDQ
  [ECCV24] MixDQ: Memory-Efficient Few-Step Text-to-Image Diffusion Models with Metric-Decoupled Mixed Precision Quantization
  ☆49Updated last year
• ModelTC / TFMQ-DM
  [CVPR 2024 Highlight & TPAMI 2025] This is the official PyTorch implementation of "TFMQ-DM: Temporal Feature Maintenance Quantization for…
  ☆108Updated 4 months ago
• FelixMessi / Awesome-Efficient-dLLMs
  📚 A curated list of Awesome Efficient dLLMs Papers with Codes
  ☆94Updated last week
• Tencent-Hunyuan / flex-block-attn
  flex-block-attn: an efficient block sparse attention computation library
  ☆108Updated last month
• Bujiazi / DiCache
  [ICLR 2026] Official implementation of DiCache: Let Diffusion Model Determine Its Own Cache
  ☆55Updated 2 weeks ago
• mit-han-lab / patch_conv
  Patch convolution to avoid large GPU memory usage of Conv2D
  ☆95Updated last year
• NoakLiu / FastCache-xDiT
  FastCache: Fast Caching for Diffusion Transformer Through Learnable Linear Approximation [Efficient ML Model]
  ☆46Updated 2 weeks ago
• Shenyi-Z / TaylorSeer
  [ICCV2025] From Reusing to Forecasting: Accelerating Diffusion Models with TaylorSeers
  ☆365Updated 6 months ago
• mit-han-lab / Block-Sparse-Attention
  A sparse attention kernel supporting mix sparse patterns
  ☆453Updated 3 weeks ago
• Shenyi-Z / ToCa
  [ICLR2025] Accelerating Diffusion Transformers with Token-wise Feature Caching
  ☆209Updated 10 months ago
• mit-han-lab / lpd
  [ICLR 2026 Oral] Locality-aware Parallel Decoding for Efficient Autoregressive Image Generation
  ☆81Updated this week
• sandyresearch / chipmunk
  🎬 3.7× faster video generation E2E 🖼️ 1.6× faster image generation E2E ⚡ ColumnSparseAttn 9.3× vs FlashAttn‑3 💨 ColumnSparseGEMM 2.5× …
  ☆101Updated 5 months ago
• Juanerx / Q-DiT
  [CVPR 2025] Q-DiT: Accurate Post-Training Quantization for Diffusion Transformers
  ☆74Updated last year
• ThisisBillhe / NAR
  [ICCV 2025] The official implementation of "Neighboring Autoregressive Modeling for Efficient Visual Generation"
  ☆58Updated 10 months ago
• ThisisBillhe / ZipAR
  [ICML 2025] This is the official PyTorch implementation of "ZipAR: Accelerating Auto-regressive Image Generation through Spatial Locality…
  ☆53Updated 10 months ago
• ziplab / BLADE
  This is the official PyTorch implementation of "BLADE: Block-Sparse Attention Meets Step Distillation for Efficient Video Generation."
  ☆40Updated 4 months ago