5 projects
paperkiln
Paste an arXiv ID, get a model you can chat with — a readable C++ LLM training stack with provenance-carrying paper extraction
paperkiln-fetch
Drop an arXiv id, get the paper's transformer architecture back - every value with the evidence snippet it was read from, emitted as a Hugging Face config or normalized JSON.
tinyllama-cpp
Python bindings for the tinyllama.cpp inference engine
chunking-pandas
A package for experimenting with different data chunking strategies
chunking-experiment
A package for experimenting with different data chunking strategies
