sciwork icon
sciwork conference
December 14th
National Taiwan University

Open Source Projects Participating in the Conference

Explore a diverse range of open source projects and join the vibrant community of developers, contributors, and enthusiasts. Here are the amazing projects participating in our conference:


Projects Overview

Modmesh

  • Languages: Python, C++
  • Focus Area: PDE Solving and Visualization
  • Contact:
    • Yung-Yu Chen (Discord: @yyc#7718)
    • Chun-Hsu Lai (Discord: @Chun-Hsu#6296)
  • Repository: GitHub

Description:

Modmesh seamlessly integrates C++ and Python using pybind11, leveraging the strengths of both languages for efficient PDE solving. It features Qt and Python-based visualization for computational results, offering enhanced understanding of PDE solutions. Modmesh supports mesh visualization in the Gmsh format and is compatible across Windows, Linux, and macOS. Recent updates include improved UI/UX for better user experience.

Discussion topic:

  • Development environment, array library, time-series analysis, metal GPU code, Qt/Qt3D

Python Official Documents Translation (zh_TW)

  • Languages: Python, Translation
  • Focus Area: Localization of Python Documentation
  • Contact:
    • Matt Wang (Discord: matt.wang#3443)
    • Josix Wang (Discord: josix#3800)
  • Repository: GitHub

Description:

LLMs are highly skilled in translation. In the translation project workflow, which includes translation, terminology selection and definition, proofreading, and more, there are additional considerations such as formatting of the translated text, consistency in terminology usage, localization of terms, and the fluency of the translation.

We should consider what role LLMs should play in these processes, which tasks are best suited for LLMs to handle, and how adjustments to interactions with LLMs can enhance outcomes. Since this is still in the design phase, we welcome everyone to join the discussion!

Discussion topic:

  • How should we apply LLM for translation in the era of AI?

sciwork Portal

  • Technologies: Pelican, TailwindCSS, Netlify
  • Focus Area: Website Development and Maintenance
  • Contacts:
    • Chester (Discord: @chester)
    • Wuxian (Discord: @5x9527)
  • Repository: GitHub

Description:

sciwork Portal maintains the official website, sciwork.dev, and related event pages, including conf.sciwork.dev. Built with Pelican and TailwindCSS, it offers promotional pages for meetups and sprint events. Our team is dedicated to enhancing user experience through better information presentation and visuals.

Discussion topic:

  • Volunteer System Frontend Deployment (Wuxian)
  • Volunteer System API Testing with Google Workspace API (Chester)

uTensor

  • Language: C++
  • Focus Area: Lightweight Machine Learning Inference Framework
  • Contact:
    • Dboy (Discord: @dboyliao#1295)
  • Repository: GitHub

Description:

uTensor simplifies model deployment by converting TensorFlow-trained models into efficient C++ files, enabling easy integration with optimized libraries like CMSIS-NN by ARM. It provides defaults for tensors, operators, and memory allocation, ensuring flexibility for embedding engineers.

Discussion topic:

  • Edge AI: Challenges and Tools
  • Lab: Let’s build a offline memory allocation optimizer!

Cytnx

  • Languages: C++, Python
  • Focus Area: Tensor Networks for Physics Simulations
  • Contacts:
    • Lin Chang Teng (Discord: @jeffry1829#0568)
    • Ke Hsu (Discord: @kjjjjjjjjj1104#5128)
  • Repository: GitHub

Description:

Cytnx is a tensor network library designed for classical and quantum physics simulations. It provides interfaces resembling popular libraries like NumPy, SciPy, and PyTorch, enabling users to define symmetries and contract large tensor networks optimally. Future developments include GPU backends and new physics applications.

Discussion topic:

  • Cytnx-core: Core components of Cytnx

AItlas

  • Focus Area: Knowledge Graphs for AI and NLU
  • Contact:
    • PeterWolf (Discord: @PeterWolf#1422)
  • Repository: GitHub

Description:

AItlas combines AI with knowledge graphs to represent entity relations across multiple dimensions, aligning with recent RAG concepts in NLU. This project explores expanding the utility of knowledge graphs in AI to enhance semantic computation and more.

Discussion topic:

  • The End of End-to-End
  • Divide and Conquer: A Renaissance Architecture of Strong AI

ReGLOSS hiodoshiao Offline Tea Gathering

  • Time: 13:00 - 17:00

Description:

To commemorate various milestones achieved by hiodoshiao, a member of hololive DEV_IS's ReGLOSS, over the past year, including reaching 500,000 subscribers, her 1st anniversary, her 3D debut, and the release of the group’s first album, we are organizing a small offline tea gathering. This event aims to bring fans together to look back on the growth and key moments of hiodoshiao over the past year, sharing insights and heartfelt memories. We warmly welcome all fans who love hiodoshiao to join this offline gathering!


These projects showcase the power of open source collaboration. Whether you're a developer, designer, or enthusiast, there's a place for you to contribute. Dive in and be part of the innovation!