fla-org / native-sparse-attentionLinks
🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"
☆952Updated 9 months ago
Alternatives and similar repositories for native-sparse-attention
Users that are interested in native-sparse-attention are comparing it to the libraries listed below
Sorting:
- Implementation of the sparse attention pattern proposed by the Deepseek team in their "Native Sparse Attention" paper☆791Updated 5 months ago
- Ring attention implementation with flash attention☆963Updated 4 months ago
- Muon is Scalable for LLM Training☆1,397Updated 5 months ago
- TransMLA: Multi-Head Latent Attention Is All You Need (NeurIPS 2025 Spotlight)☆423Updated 3 months ago
- Official implementation of "Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding"☆773Updated last month
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference☆626Updated this week
- Efficient triton implementation of Native Sparse Attention.☆258Updated 7 months ago
- Helpful tools and examples for working with flex-attention☆1,108Updated this week
- [ICLR 2025] Official PyTorch Implementation of Gated Delta Networks: Improving Mamba2 with Delta Rule☆421Updated 4 months ago
- VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo☆1,524Updated this week
- A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Training☆607Updated this week
- Parallel Scaling Law for Language Model — Beyond Parameter and Inference Time Scaling☆467Updated 8 months ago
- 🔥 A minimal training framework for scaling FLA models☆335Updated 2 months ago
- ☆578Updated 3 months ago
- ☆444Updated 5 months ago
- Efficient LLM Inference over Long Sequences☆393Updated 6 months ago
- [ICLR 2025] DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads☆519Updated 11 months ago
- Miles is an enterprise-facing reinforcement learning framework for large-scale MoE post-training and production workloads, forked from an…☆714Updated this week
- Training library for Megatron-based models with bi-directional Hugging Face conversion capability☆363Updated this week
- Scalable toolkit for efficient model reinforcement☆1,227Updated this week
- Accelerating MoE with IO and Tile-aware Optimizations☆542Updated this week
- Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs☆198Updated last month
- MoBA: Mixture of Block Attention for Long-Context LLMs☆2,031Updated 9 months ago
- ☆208Updated 2 months ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆626Updated this week
- SeerAttention: Learning Intrinsic Sparse Attention in Your LLMs☆184Updated 3 months ago
- ☆816Updated 7 months ago
- ☆949Updated 2 months ago
- LLM KV cache compression made easy☆749Updated last month
- [NeurIPS'24 Spotlight, ICLR'25, ICML'25] To speed up Long-context LLMs' inference, approximate and dynamic sparse calculate the attention…☆1,171Updated 3 months ago