Things I’ve had a hand in building.
Open-source software and open-science research, built in the open with brilliant people.
ViBench 2.0: A Benchmark for Long-Horizon Software Building
Evaluating AI agents on long-horizon software building: 17 full-stack apps, 241 feature requests, 544 test plans.
arxiv.papers
Agent 4: Teamwork and Collaboration for Vibe-Coding
Teamwork and collaboration patterns for AI agents that vibe-code.
dl-acm.papers
CalmeRys-78B
Cream of the crop of my LLM finetunes.
calmerys-78b.tech
Polyglot or Not?
Measuring multilingual encyclopedic knowledge in foundation models.
emnlp-main-2023.papers
sft-demos
Lightweight demos for finetuning LLMs. Powered by transformers and pytorch.
sft-demos.tech
Species distribution modeling
A Python tutorial and autoML package for geospatial modeling.
pysdms.tech
Awesome ChatGPT prompts
Text clustering: HDBSCAN is probably all you need.
hdbscan.tech