Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2412.13922 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929637773279232 |
|---|---|
| author | Corral, Ander Sarasua, Ixak Saralegi, Xabier |
| author_facet | Corral, Ander Sarasua, Ixak Saralegi, Xabier |
| contents | Large language models (LLMs) are typically optimized for resource-rich languages like English, exacerbating the gap between high-resource and underrepresented languages. This work presents a detailed analysis of strategies for developing a model capable of following instructions in a low-resource language, specifically Basque, by focusing on three key stages: pre-training, instruction tuning, and alignment with human preferences. Our findings demonstrate that continual pre-training with a high-quality Basque corpus of around 600 million words improves natural language understanding (NLU) of the foundational model by over 12 points. Moreover, instruction tuning and human preference alignment using automatically translated datasets proved highly effective, resulting in a 24-point improvement in instruction-following performance. The resulting models, Llama-eus-8B and Llama-eus-8B-instruct, establish a new state-of-the-art for Basque in the sub-10B parameter category. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_13922 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Pipeline Analysis for Developing Instruct LLMs in Low-Resource Languages: A Case Study on Basque Corral, Ander Sarasua, Ixak Saralegi, Xabier Computation and Language Artificial Intelligence Machine Learning Large language models (LLMs) are typically optimized for resource-rich languages like English, exacerbating the gap between high-resource and underrepresented languages. This work presents a detailed analysis of strategies for developing a model capable of following instructions in a low-resource language, specifically Basque, by focusing on three key stages: pre-training, instruction tuning, and alignment with human preferences. Our findings demonstrate that continual pre-training with a high-quality Basque corpus of around 600 million words improves natural language understanding (NLU) of the foundational model by over 12 points. Moreover, instruction tuning and human preference alignment using automatically translated datasets proved highly effective, resulting in a 24-point improvement in instruction-following performance. The resulting models, Llama-eus-8B and Llama-eus-8B-instruct, establish a new state-of-the-art for Basque in the sub-10B parameter category. |
| title | Pipeline Analysis for Developing Instruct LLMs in Low-Resource Languages: A Case Study on Basque |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2412.13922 |