Skip to content
View yanghui1-arch's full-sized avatar
  • Shanghai, China

Block or report yanghui1-arch

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
yanghui1-arch/README.md

简体中文

Hi, I'm yanghui1-arch 👋

Computer Science · Open Source Contributor · LLM Systems & AI Applications

About me

  • 🎓 Third-year M.S. student in Computer Science at Shanghai University
  • 🏫 Undergraduate study in Computer Science at Nanchang University
  • 🔭 Interested in LLM inference, AI agents, observability, and full-stack systems
  • 🌱 I enjoy turning real-world bugs into focused, well-tested open-source contributions

Open-source contributions

Contributions around model execution, CUDA graph correctness, speculative decoding, structured outputs, and model compatibility.

PR Status / Role Contribution
#51113 Merged · Co-author Keep Mamba align prefill chunks block-aligned past last_cache_position
#37873 Merged · Author Fix RoBERTa position_id accumulation in CUDA graph padding

Contributions to agent integrations, ACP reliability, chat behavior, and UI robustness.

PR Status / Role Contribution
#804 Merged · Author Integrate nanobot
#550 Merged · Author Prevent long content from overflowing its block

Four merged fixes improving speech-provider compatibility and voice selection in the web experience.

PR Status / Role Contribution
#593 Merged · Author Show voice search results correctly
#579 Merged · Author Fix voice testing with OpenAI and CosyVoice v2
#576 Merged · Author Add CosyVoice SSML support
#568 Merged · Author Fix CosyVoice v1 testing in the speech module

My project

An open-source LLM observability platform for tracing OpenAI, Claude, Gemini, and OpenAI-compatible models. It combines a Python SDK, a web dashboard, and Java/Cloudflare backends to help developers inspect trace trees, understand latency and cost, and improve agent systems.


GitHub

Popular repositories Loading

  1. mwin mwin Public

    mwin: Track OpenAI, Claude, Gemini and OpenAI-compatible models then give solutions to improve your agent system. Mwin's goal is to make agent system more valuable and effortlessly improve agent sy…

    TypeScript 4 2

  2. petmate petmate Public

    TypeScript 3

  3. VsDB VsDB Public

    A low-memory database visualization GUI

    C++ 1

  4. cpython cpython Public

    Forked from python/cpython

    The Python programming language

    Python

  5. electron electron Public

    Forked from electron/electron

    :electron: Build cross-platform desktop apps with JavaScript, HTML, and CSS

    C++

  6. superagent superagent Public

    A super agent do all of things you can image.

    Python