
Hugging Face, a popular platform for AI datasets and models, has recently seen the release of several new datasets. These include a large open AI preference dataset called OpenHermesPreferences 1M, a public textual human feedback dataset, and an analysis of spaces dataset created from 20K code files. The datasets aim to provide valuable resources for the AI community and showcase the power of open science and collaboration.

When @argilla_io & @huggingface embraces we get a great new public textual human feedback dataset https://t.co/fBBPlNIs4L
So much fun to build this dataset together with our friends @huggingface Open science, software and collaboration are unstoppable 🚀 https://t.co/2RH4KBu4CK
Introducing OpenHermesPreferences 1M! 🦋 We just released the largest open AI preference dataset on Hugging Face! 🤯 Together with @argilla_io, we extended the OpenHermes (@Teknium1) dataset into a pair-wise comparison dataset for RLHF and DPO. 🧠 Dataset: 📦 Size: ~1 million AI… https://t.co/O5iucdR62Z