66RING

tiny-flash-attention

flash attention tutorial written in python, triton, cuda, cutlass

dotfiles

My dotfiles

LongShortTokenDecoding

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

ring-attention-pytorch

tiny ring attention implement for learning purpose

scripts

some scripts

66RING.github.io

https://66ring.github.io/

Counting-Stars-Local

Counting-Stars scripts for evaluating local llm.

Notes

my note things

single-gpu-passthrough

pytorch-cuda-binding-tutorial

Tutorial for building a custom CUDA and C function for torch

the-chaos

15445-bootcamp

A basic introduction to coding in modern C++.

66RING

academicpages.github.io

Github Pages template for academic personal websites, forked from mmistakes/minimal-mistakes

bufferline.nvim

A snazzy bufferline for Neovim

clash-verge

A Clash GUI based on tauri. Supports Windows, macOS and Linux.

ContinuousBatching

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

flash-attention

Fast and memory-efficient exact attention

LightSeq

Official repository for LightSeq: Sequence Level Parallelism for Distributed Training of Long Context Transformers

llama-playground

play with llama

LLMTest_NeedleInAHaystack-Local

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

minitorch

The full minitorch student suite.

paperDB

PoSE

Positional Skip-wise Training for Efficient Context Window Extension of LLMs to Extremely Length

RULER

This repo contains the source code for RULER: What’s the Real Context Size of Your Long-Context Language Models?

st

my st

ThunderKittens

Tile primitives for speedy kernels

vattention

Dynamic Memory Management for Serving LLMs without PagedAttention

wallpapers

zephyr-nvim

Customized nvimdev/zephyr-nvim