A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Training
☆946Sep 23, 2026Updated this week
Alternatives and similar repositories for MagiAttention
Users that are interested in MagiAttention are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- A plug-and-play compiler that delivers free-lunch optimizations for both inference and training.☆333Updated this week
- Accelerating MoE with IO and Tile-aware Optimizations☆769Aug 29, 2026Updated 3 weeks ago
- Distributed Compiler and Optimized Parallel Kernels☆1,552Updated this week
- MAGI-1: Autoregressive Video Generation at Scale☆3,789Jun 17, 2026Updated 3 months ago
- VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo☆2,224Updated this week
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,364Aug 28, 2025Updated last year
- ☆255Nov 19, 2025Updated 10 months ago
- [ICML2025, NeurIPS2025 Spotlight] Sparse VideoGen 1 & 2: Accelerating Video Diffusion Transformers with Sparse Attention☆708Jul 4, 2026Updated 2 months ago
- A unified inference and post-training framework for accelerated video generation.☆4,496Updated this week
- A Quirky Assortment of CuTe Kernels☆1,152Updated this week
- [ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.☆1,226Feb 25, 2026Updated 6 months ago
- Ring attention implementation with flash attention☆1,059Sep 10, 2025Updated last year
- 🚀 Efficient implementations for emerging model architectures☆5,786Updated this week
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference☆693Sep 16, 2026Updated last week
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on H…☆3,550Updated this week
- xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism☆2,727Updated this week
- Helpful tools and examples for working with flex-attention☆1,252Updated this week
- FlashInfer: Kernel Library for LLM Serving☆6,495Updated this week
- 🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"☆1,021Feb 5, 2026Updated 7 months ago
- [ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-t…☆3,909Jan 17, 2026Updated 8 months ago
- ☆37Aug 7, 2025Updated last year
- Tile primitives for speedy kernels☆3,725Sep 12, 2026Updated last week
- [NeurIPS 2025] Radial Attention: O(nlogn) Sparse Attention with Energy Decay for Long Video Generation☆608Nov 11, 2025Updated 10 months ago
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- flex-block-attn: an efficient block sparse attention computation library☆132Dec 26, 2025Updated 8 months ago
- An efficient implementation of the NSA (Native Sparse Attention) kernel☆135Jun 24, 2025Updated last year
- A sparse attention kernel supporting mix sparse patterns☆557Jan 18, 2026Updated 8 months ago
- DeepSeek-V3.2-Exp DSA Warmup Lightning Indexer training operator based on tilelang☆52Nov 19, 2025Updated 10 months ago
- CUDA kernels for linear attention variants, written in CuTe DSL and CUTLASS C++.☆551Updated this week
- Perplexity GPU Kernels☆608Nov 7, 2025Updated 10 months ago
- high-performance linear attention kernel library built on TileLang☆706Updated this week
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆204Feb 11, 2026Updated 7 months ago
- Kernel Library for Large Headdim Attention (64~1024, BF16/FP8/FP4), 1.5x~15x↑ vs PyTorch SDPA.☆334Updated this week
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels☆7,464Updated this week
- A Survey of Efficient Attention Methods: Hardware-efficient, Sparse, Compact, and Linear Attention☆309Dec 1, 2025Updated 9 months ago
- Allow torch tensor memory to be released and resumed later☆276Updated this week
- Efficient triton implementation of Native Sparse Attention.☆283May 23, 2025Updated last year
- Kernels, of the mega variety :)☆833May 26, 2026Updated 3 months ago
- A PyTorch-native inference engine with cache, parallelism, quantization and cpu offload for DiTs.☆1,285Updated this week
- Distributed MoE in a Single Kernel [NeurIPS '25]☆291May 5, 2026Updated 4 months ago