LeoXinhaoLee

big_vision

Official codebase used to develop Vision Transformer, SigLIP, MLP-Mixer, LiT and more.

maxtext

A simple, performant and scalable Jax LLM!

AMS

Code released for our ECCV 2022 paper "Interpretable Open-Set Domain Adaptation via Angular Margin Separation".

flash-attention

Fast and memory-efficient exact attention

Imbalanced-Source-free-Domain-Adaptation

Code released for our ACM MM 2021 paper "Imbalanced Source-free Domain Adaptation".

LeoXinhaoLee.github.io

ThunderKittens

Tile primitives for speedy kernels

TK_Study

transformers

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

mttt

LeoXinhaoLee/AMS

Code released for our ECCV 2022 paper "Interpretable Open-Set Domain Adaptation via Angular Margin Separation".

LeoXinhaoLee/Imbalanced-Source-free-Domain-Adaptation

Code released for our ACM MM 2021 paper "Imbalanced Source-free Domain Adaptation".

LeoXinhaoLee/flash-attention

Fast and memory-efficient exact attention

LeoXinhaoLee/LeoXinhaoLee.github.io

LeoXinhaoLee/ThunderKittens

Tile primitives for speedy kernels

LeoXinhaoLee/TK_Study

LeoXinhaoLee/transformers

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

LeoXinhaoLee/vllm

A high-throughput and memory-efficient inference and serving engine for LLMs