File size: 992 Bytes
ea7f228 d4d7d56 ea7f228 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
---
datasets:
- bs-la/xP3ru
license: bigscience-bloom-rail-1.0
---
# Model Summary
[bloom-7b1](https://huggingface.co/bigscience/bloom-7b1) finetuned on Russian multitask data. Hence the same as [bloomz-7b1](https://huggingface.co/bigscience/bloomz-7b1), but with **only** Russian finetuning data. 4b stands for 4 billion finetuning tokens (same as bloomz-7b1).
# Citation
```
BLOOM+1 - TODO
```
```bibtex
@misc{muennighoff2022crosslingual,
title={Crosslingual Generalization through Multitask Finetuning},
author={Niklas Muennighoff and Thomas Wang and Lintang Sutawika and Adam Roberts and Stella Biderman and Teven Le Scao and M Saiful Bari and Sheng Shen and Zheng-Xin Yong and Hailey Schoelkopf and Xiangru Tang and Dragomir Radev and Alham Fikri Aji and Khalid Almubarak and Samuel Albanie and Zaid Alyafeai and Albert Webson and Edward Raff and Colin Raffel},
year={2022},
eprint={2211.01786},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
``` |