Name: Xiangrui Yu
Type: User
Company: Data Science and Analytic Thrust, Information Hub, HKUST(GZ)
Bio: I am a first year mphil in HKUST(GZ). My research interest based on accelerating LLM using GPUs.
Before that, I got my CS bachelor's degree from CUP.
Location: Guangzhou
Xiangrui Yu's Projects
ACM程序设计竞赛、Codeforces比赛、各种训练赛
acwing周赛的一些题目
study of Ampere' Sparse tensor core Matmul
This is my GPU course final project in MICS600J. The main content is my attempt to handwrite the attention process.
Graph Sampling for GNN, using GPU. Build and use alias table for random search, especially.
Distributed-SpMV, c/mpi/openmp, this work was accepted by IEEE/ACM CCGrid'23.
Fine-tuning Llama-2-7B for Text classification. Datasets: imdb , framework: deepspeed.
MICS600J - GPU Architectures and Programming, Homework 1
Something need to be noted
一个基于SpringBoot框架的人机对战平台
我总结的ACM模版,实时更新~
TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT-LLM also contains components to create Python and C++ runtimes that execute those TensorRT engines.
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.