MaxLSB/flash-attn2
FlashAttention for sliding window attention in Triton (fwd + bwd pass)
This project helps machine learning engineers accelerate the core 'attention' mechanism in large language models. It takes your model's attention computations and processes them much faster on NVIDIA GPUs. The result is significantly quicker training and inference for models that use sliding window, global, or causal attention, making your LLM workflows more efficient.
No commits in the last 6 months.
Use this if you are developing or training large language models and need to speed up the attention computation on NVIDIA GPUs, especially for models employing sliding window attention.
Not ideal if you are not working with large language models, do not have access to NVIDIA GPUs, or require features like dropout support or other attention mechanisms.
Stars
11
Forks
1
Language
Python
License
MIT
Category
Last pushed
Jun 25, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/MaxLSB/flash-attn2"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
triton-inference-server/server
The Triton Inference Server provides an optimized cloud and edge inferencing solution.
gpu-mode/Triton-Puzzles
Puzzles for learning Triton
hailo-ai/hailo_model_zoo
The Hailo Model Zoo includes pre-trained models and a full building and evaluation environment
open-mmlab/mmdeploy
OpenMMLab Model Deployment Framework
hyperai/tvm-cn
TVM Documentation in Chinese Simplified / TVM 中文文档