Zilin Zhu

ring-flash-attention. Ring attention implementation with flash attention

NP_ML. A tool library of classical machine learning algorithms with only numpy.

whisper-openvino. openvino version of openai/whisper

faster-nougat. Implementation of nougat that focuses on processing pdf locally.

pdf-with-its-own-md5. A PDF template that contains its own MD5!

monkey. A C++ version monkey language interpreter. From Write An Interpreter In Go

flash-attention-with-sink. Python

ncnn-swift. An example on using ncnn with Swift.

es. A JavaScript interpreter from scratch, supporting ES5 syntax.

chatgpt-desktop. Desktop version of ChatGPT, support manually set cookie

seastar-cn. Seastar教程（文档）中文翻译。A Chinese translation for Seastar tutorial and relevant documents.

pytorch-malloc. An external memory allocator example for PyTorch.

google-translate-desktop. Google Translate Desktop built with Electron

vllm-group. Python

NeZha. Organizing ssh servers in one shell.

simple-pandas. A much simpler pandas!!!

wandb-discord-bot. A discord bot for monitoring wandb project and runs.

SwiftPEG. A PEG parser generator written in swift 5.3.

aqt-pytorch. Python

li. another mini text editor with 500 loc and simple interfaces for short cut.

ecdh-psi. A Go Implementation of ECDH-PSI

cumem_allocator. Python

PIXEL.css. the pixel art code from the fun and famous NES.css

pytorch-extension-gcc. How to use gcc instead of setup.py to build an pytorch extension.

pytorch-reloadable-pg. PyTorch Reloadable Process Group

OpenRLHF. An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & Mixtral)

blog. my blog~

llama. Inference code for LLaMA models

on-device_recommendation_tflite. Swift

zhuzilin.

dafny-exercises. some formal verification exercises using dafny.

torchrec_mapper. C++

nvcc_daxpy_example. C++

sglang. SGLang is a fast serving framework for large language models and vision language models.

scattermoe. Triton-based implementation of Sparse Mixture of Experts.

unilm. Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

bfc. A really fast non-JIT brainfuck interpreter.

triton. Development repository for the Triton language and compiler

electron-fc. A electron based famicom(NES) emulator

flash-attention. Fast and memory-efficient exact attention

garbled_circuit. A python implementation of Yao's GC.

torch_memory_saver. Allow torch tensor memory to be released and resumed later

autodiff. pytorch like automatic differentiation library in numpy