DeepSeek-Powered HapHiC Knowledge Base (Beta) #107
Pinned
zengxiaofei
announced in
Announcements
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
🚀 Introducing the HapHiC Knowledge Base!
We’re excited to share our initial attempt at building an AI-powered knowledge base for the HapHiC project! This beta version is built on Tencent’s ima platform using RAG (Retrieval-Augmented Generation) technology. For now, you can try asking DeepSeek R1 questions about HapHiC in English or Chinese, and we’d love to hear your feedback!
What’s included in this knowledge base?
✅ Full research papers and technical documentation
✅ Key code snippets for quick reference
✅ A curated collection of GitHub issues addressing common questions and solutions
Important notes:
🔹 WeChat registration is required (availability outside China may be limited).
🔹 As this is a beta version, AI-generated answers may occasionally be imperfect. Please always cross-check critical details for accuracy.
🔹 For English responses, end your query with “Answer me in English” to ensure replies in English.
🔹 Still unsure? Visit the HapHiC GitHub Issues for community support.
Try it now:
Scan the QR code below using WeChat:

Let’s work together to make genome assembly and scaffolding smarter! 🧬💻
Beta Was this translation helpful? Give feedback.
All reactions