README.md

---
language:
- ru
tags:
- generation
- question answering
- instruction tuning
license: cc-by-nc-4.0
---

###  Model Description

This HF repository contains base LLMs instruction tuned (SFT) with full-parameter fine-tuning and then used to study whether monolingual or multilingual instruction tuning is more favourable.
* [GitHub](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main)
* [Paper](https://arxiv.org/abs/2309.08958)

#### Instruction tuning details
* Base model: [bloom-3b](https://huggingface.co/bloom-3b)
* Instruction tuning language: Russian
* Training method: full-parameter fine-tuning.
* Best checkpoint: best cross-entropy on a validation set, trained for 3 epochs.
* Dataset: machine-translated from [yahma/alpaca-cleaned](https://huggingface.co/datasets/yahma/alpaca-cleaned). You can download our data [HERE](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main/training-data).

#### Usage
The model checkpoint should be loaded using `transformers` library.

Please refer to our Github repository [HERE](https://github.com/hplt-project/monolingual-multilingual-instruction-tuning/tree/main/fpft) for inference and training instructions.

#### Citation
```
@inproceedings{chen-etal-2024-monolingual,
  title="Monolingual or multilingual instruction tuning: Which makes a better {Alpaca}",
  author="Pinzhen Chen and Shaoxiong Ji and Nikolay Bogoychev and Andrey Kutuzov and Barry Haddow and Kenneth Heafield",
  year="2024",
  booktitle = "Findings of the Association for Computational Linguistics: EACL 2024",
}
```