datasets▌

A curated index of scientific datasets, models, and demos for AI/ML research in various scientific domains.

Molecular ML with diverse featurizers and pre-built datasets for property prediction and drug discovery.

Create, query, and manage datasets on Hugging Face Hub with SQL-based transformation and streaming updates. \n \n Initialize new dataset repositories with template-based schemas (chat, classification, QA, completion, tabular) and custom system prompts \n Query any Hugging Face dataset using DuckDB SQL via the hf:// protocol, including filtering, aggregations, joins, and regex operations \n Stream rows efficiently without downloading entire datasets, with JSON validation and batch processing for