66RING

tiny-flash-attention. flash attention tutorial written in python, triton, cuda, cutlass

dotfiles. My dotfiles

CritiPrefill. Code repo for "CritiPrefill: A Segment-wise Criticality-based Approach for Prefilling Acceleration in LLMs".

learn-vgpu-the-hard-way. qemu, cuda, virtio and kernel driver etc, none of which I understand, I just in awe.

single-gpu-passthrough. Shell

ring-attention-pytorch. tiny ring attention implement for learning purpose

LLMTest_NeedleInAHaystack-Local. run Needle In A Haystack with local LLM. check the makefile

LongShortTokenDecoding. Long short token decoding speed up 4x for long context LLM. A hundred lines of core code. Open source for learning.

bookmarks-nvim. Interactive mark and jump action in neovim.

.config. my .config file

Notes. my note things

.vim. my vim config file

pytorch-cuda-binding-tutorial. Tutorial for building a custom CUDA and C function for torch

scripts. some scripts

ringemacs. Emacs Lisp

Counting-Stars-Local. Counting-Stars scripts for evaluating local llm.

66RING.github.io. https://66ring.github.io/

ContinuousBatching. A demo about continuous batching, which is simple than you think.

the-chaos. Rust

llama-playground. play with llama

qCUDA. qCUDA: GPGPU Virtualization at a New API Remoting Method with Para-virtualization

nvim-lua-guide-zh. https://github.com/nanotee/nvim-lua-guide chinese version

TH-CPL. 清华大学计算机学科推荐学术会议和期刊列表