XunhaoLai/native-sparse-attention-triton
Efficient triton implementation of Native Sparse Attention.
This project provides an optimized way to train and use large language models (LLMs) more efficiently. It takes in sequence data for your LLM and outputs the results of an attention mechanism that is much faster than traditional methods. This is for researchers and engineers working on developing or deploying LLMs who need to process long sequences of text or other data quickly.
269 stars. No commits in the last 6 months.
Use this if you are developing or fine-tuning large language models and need to accelerate attention computations for both training and inference, especially with long input sequences.
Not ideal if you are a casual user of existing LLMs and do not need to implement or optimize the underlying attention mechanisms.
Stars
269
Forks
19
Language
Python
License
Apache-2.0
Category
Last pushed
May 23, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/XunhaoLai/native-sparse-attention-triton"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
fla-org/flash-linear-attention
🚀 Efficient implementations of state-of-the-art linear attention models
thu-ml/SageAttention
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x...
thu-ml/SpargeAttn
[ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.
fla-org/flame
🔥 A minimal training framework for scaling FLA models
foundation-model-stack/fms-fsdp
🚀 Efficiently (pre)training foundation models with native PyTorch features, including FSDP for...