Welcome to Shizhuo's Profile

About Me

Hi, I'm Shizhuo Li, a senior undergraduate at Carnegie Mellon University, majoring in computer science. I have a solid foundation in programming, software development, and machine learning. My passion lies in deep learning, natural language processing, and research involving large language models, particularly in generative AI and its practical applications. You can view my resume here.

Education

Carnegie Mellon University

Degree: Bachelor of Science in Computer Science

Field of Study: Computer Science, Machine Learning

Location: Pittsburgh / PA

Expected Year of Graduation: Spring 2025

Contact Information

Email: shizhuol [at] andrew [dot] cmu [dot] edu

Location: Pittsburgh / PA

Project Experience

RESEARCH: SUIGPT: LLM Code Generation for Novel Blockchain Languages

Carnegie Mellon University - Learn Lab, Pittsburgh, PA   (May 2024 ~ Now)

  • Developed and explored methods in enhancing LLM code generation performance of Sui Move Programming Language
  • Performed Data Cleaning of a Sui Move code dataset using Python Pandas, and built a vector dataset with optimal embedding model
  • Assisted designing architecture of applying Retrieval Augmented Generation in enhancing SuiGPT's performance
  • Tested different retrieval methods such as FAISS, ElasticSearch, and Pinecone for applied RAG to code generation process
  • Assisted with prompt engineering, fine-tuning and in-context learning for SuiGPT, and constructed APIs for seamless LLM pipeline integration.

    • TEAM PROJECT: Retrieval-Augmented Multi-hop Code Generation with CodeLlama and Unlimiformer (Click for Project Report)

    Carnegie Mellon University, Pittsburgh, PA   (March 2024 ~ April 2024)

    TEAM PROJECT: End-to-End NLP System Building-An InfoBox for LTI (Click for Project Report)

    Carnegie Mellon University, Pittsburgh, PA   (February 2024 ~ March 2024)

  • Design and build an end-to-end NLP system LTI-Chat capable to answer question related to Carnegie Mellon University and its Language Technology Institute questions in team of 3
  • Collected and processed relevant to CMU and LTI for vector database construction by scripting, processed the data for retrieval use.
  • Conducted experiment on multiple language models for system optimization, applied few-shots prompting and basic finetuning.
  • Achieved good performance (0.84 overall recall and 0.73 precision) in answering on a large set of self-annotated test questions.
    •

    TEAM PROJECT: A Review of Generative Music Models (Click for Project Report)

    Carnegie Mellon University, Pittsburgh, PA   (March 2024 ~ April 2024)

  • Conducted extensive research on existing generative music models including MusicLM, MusicGen, and Neural Network models by reviewing papers.
  • Conducted experiments on testing selected generative music models' performance including music quality, consistency of content in generation, and speed of generation.
  • Summarized the experiment and review in a well-formated report

    • TEAM PROJECT: LSP Based Distributed Bitcoin Miner

    Carnegie Mellon University, Pittsburgh, PA   (September 2023 ~ October 2023)

  • Designed and implemented novel network protocol: Live-Sequence Protocol (LSP) which combines advantages of UDP and TCP for purpose of distributed system inner communication using Golang
  • Implemented a Client-Server concurrent interaction model with LSP with failure handling / tolerance and lost / duplicated packet handling
  • Design and implemented a scalable distributed bitcoin miner with LSP s client-server model, and achieved expected performance.
  • Improved the system's performance by refining scheduling algorithms and failure tolerance of the system.

    • INDIVIDUAL COURSE PROJECT - Malloc Lab:

    Carnegie Mellon University, Pittsburgh, PA   (March 2023)

  • Implemented functions malloc, calloc, and realloc in C with newly designed data structures and algorithms such as explicit list and seglist
  • Tested the performance including utility and speed performance of malloc of different data structures
  • Refined the the implementation with techniques such as mini memory blocks for malloc performance boosting

    • Courses

    Coureses Taken/Taking

    15-440 Distributed Systems Fall 2023
    15-213 Introduction to Computer Systems Spring 2023
    15-317 Constructive Logic Spring 2023
    15-210 Parallel and Sequential Data Structures and Algorithms Fall 2022
    15-251 Great Ideas in Theoretical Computer Science Fall 2022
    15-150 Principles of Functional Programming Summer 2022
    15-122 Principles of Imperative Computation Spring 2022
    10-701 Introduction to Machine Learning (PHD) Spring 2024
    11-711 Natural Language Processing Spring 2024
    10-735 Responsible AI Spring 2024
    16-385 Computer Vision Fall 2023

    Teaching Assistant Courses

    21-241 Matrices and Linear Transformation Spring 2023
    21-120 Differential and Integral Calculus Fall 2022

    Skills

    Programming                

    Python

    C

    Golang

    Java

    JavaScript

    HTML

    CSS

    Standard ML

    Sui Move

    Software / Systems             

    Data Structures and Algorithms

    Distributed Systems

    Software / System Development

    Git

    Node.js

    Machine Learning Related                    

    Deep Learning

    Pytorch Framework

    Natural Language Processing

    LangChain

    Large Language Models

    Computer Vision

    Platforms: AWS Sagemaker, Colab