Lightning Talks Day 2
summary
Short, concrete demos and tools dominate tonight: an open source wrapper called lfs spec that plugs data versioning into FS spec reads and writes and adds post‑execution hooks; Recon for NER dataset debugging that surfaces label disparities, per‑example F1 trouble spots, and fixes on CONLL2003; a data‑cleaning and display tool Buckaroo that replaces pandas’ view with downsampling, histograms and low‑code cleaning plus code export; an AutoML package that produces leaderboards and docs and a PlotAI that generates Python plotting code from prompts, and a Prefect workflow framework.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 5:19Debugging named entity recognition
- 11:10Automating ML pipelines
- 14:31EGS in Python
- 19:31Data analysis with Buckaroo
- 23:31Workflow automation with Prefect
