pile EleutherAI
The Pile is a 825 GiB diverse, open source language modelling data set that consists of 22 smaller, high-quality datasets combined together.
- Type
- dataset
- License
- other
- Language
- en
- Downloads
- 2,877
- Likes
- 502
- Access
- public
- Files
- 0
The Pile is a 825 GiB diverse, open source language modelling data set that consists of 22 smaller, high-quality datasets combined together.