Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2412.10244 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910743937417216 |
|---|---|
| author | Nag, Arijit Chakrabarti, Soumen Mukherjee, Animesh Ganguly, Niloy |
| author_facet | Nag, Arijit Chakrabarti, Soumen Mukherjee, Animesh Ganguly, Niloy |
| contents | Open-source Large Language models (OsLLMs) propel the democratization of natural language research by giving the flexibility to augment or update model parameters for performance improvement. Nevertheless, like proprietary LLMs, Os-LLMs offer poorer performance on low-resource languages (LRLs) than high-resource languages (HRLs), owing to smaller amounts of training data and underrepresented vocabulary. On the other hand, continual pre-training (CPT) with large amounts of language-specific data is a costly proposition in terms of data acquisition and computational resources. Our goal is to drastically reduce CPT cost. To that end, we first develop a new algorithm to select a subset of texts from a larger corpus. We show the effectiveness of our technique using very little CPT data. In search of further improvement, we design a new algorithm to select tokens to include in the LLM vocabulary. We experiment with the recent Llama-3 model and nine Indian languages with diverse scripts and extent of resource availability. For evaluation, we use IndicGenBench, a generation task benchmark dataset for Indic languages. We experiment with various CPT corpora and augmented vocabulary size and offer insights across language families. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_10244 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Efficient Continual Pre-training of LLMs for Low-resource Languages Nag, Arijit Chakrabarti, Soumen Mukherjee, Animesh Ganguly, Niloy Computation and Language Machine Learning Open-source Large Language models (OsLLMs) propel the democratization of natural language research by giving the flexibility to augment or update model parameters for performance improvement. Nevertheless, like proprietary LLMs, Os-LLMs offer poorer performance on low-resource languages (LRLs) than high-resource languages (HRLs), owing to smaller amounts of training data and underrepresented vocabulary. On the other hand, continual pre-training (CPT) with large amounts of language-specific data is a costly proposition in terms of data acquisition and computational resources. Our goal is to drastically reduce CPT cost. To that end, we first develop a new algorithm to select a subset of texts from a larger corpus. We show the effectiveness of our technique using very little CPT data. In search of further improvement, we design a new algorithm to select tokens to include in the LLM vocabulary. We experiment with the recent Llama-3 model and nine Indian languages with diverse scripts and extent of resource availability. For evaluation, we use IndicGenBench, a generation task benchmark dataset for Indic languages. We experiment with various CPT corpora and augmented vocabulary size and offer insights across language families. |
| title | Efficient Continual Pre-training of LLMs for Low-resource Languages |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2412.10244 |