Pinned Loading

1. Tencent-Hunyuan/Simple-Attention-Sparsification Tencent-Hunyuan/Simple-Attention-Sparsification Public

   Code for our resarch paper "SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking"

   Python 60 1

2. sglang sglang Public

   A fork of official SGLang that supports block-sparse attention. The backend and two algorithms (SeerAttention-R and MoBA) are implemented.

   Python 10

3. BiFormer BiFormer Public

   [CVPR 2023] Official code release of our paper "BiFormer: Vision Transformer with Bi-Level Routing Attention"

   Python 584 42

4. vllm-ra vllm-ra Public

   [ACL 2024] RelayAttention for Efficient Large Language Model Serving with Long System Prompts

   Python 40 6

5. GLMix GLMix Public

   [NeurIPS 2024] official code release for our paper "Revisiting the Integration of Convolution and Attention for Vision Backbone".

   Python 43 3

6. docker-cuda-codeserver docker-cuda-codeserver Public

   Forked from works-on-my-machine/pytorch-code-server

   Docker code server with CUDA development environment.

   Shell 3