Henry Du
AI & Data: Custom ML Solutions & Analysis
- $1k+
- Earned
- 2x
- Hired
- 5.00
- Rating
- 4
- Followers
SDS Knowledge Base — RAG + LoRA Fine-Tuning
Built a local question-answering system over 497,000 chemical safety data sheets: two parsers unify raw corpora (Kaggle SDS files + PubChem XML) into one validated schema, and a hybrid retrieval pipeline SQLite FTS5 for exact identifiers, vector embeddings for semantics, fused with reciprocal rank fusion and re-ranked by a cross-encoder answers queries with cited
sources and honest "no data" responses instead of guesses. Fine-tuned Qwen 2.5 7B with LoRA on 4,000+ Q&A pairs generated from the corpus, exported the adapter to GGUF.
2
2
73
A Rust speed-reading app that runs inside the terminal, showing words one at a time at a fixed screen position (RSVP method) so readers can sustain much higher speeds than normal reading. Supports PDF, EPUB, and clipboard text with adjustable WPM, ghost words, progress indicators,
and themes shipped as a single ~8MB self-contained binary.
0
25
P&ID Drawing Analyzer Turn Paper Plant Diagrams into a Smart, Searchable Database. Converts scanned piping & instrumentation diagrams (P&IDs) into a connected digital model of your entire plant every pipe, valve, and instrument identified automatically and linked across all drawings. No manual data entry, no cloud dependency.
1
35
cells (@henrythdu/cells) is a TypeScript CLI that partitions a codebase into context-bounded cells so an LLM can work on any part of it without blowing its context window.
1
24