Bilingual Adaptation of Monolingual Foundation Models

• URL: http://arxiv.org/abs/2407.12869v2
• Date: Thu, 25 Jul 2024 22:51:39 GMT
• Title: Bilingual Adaptation of Monolingual Foundation Models
• Authors: Gurpreet Gosal, Yishi Xu, Gokul Ramakrishnan, Rituraj Joshi, Avraham Sheinin, Zhiming, Chen, Biswajit Mishra, Natalia Vassilieva, Joel Hestness, Neha Sengupta, Sunil Kumar Sahu, Bokang Jia, Onkar Pandit, Satheesh Katipomu, Samta Kamboj, Samujjwal Ghosh, Rahul Pal, Parvez Mullah, Soundar Doraiswamy, Mohamed El Karim Chami, Preslav Nakov,
• Abstract summary: We present an efficient method for adapting a monolingual Large Language Model (LLM) to another language. Our two-stage approach begins with expanding the vocabulary and training only the embeddings matrix. By continually pre-training on a mix of Arabic and English corpora, the model retains its proficiency in English while acquiring capabilities in Arabic.
• Score: 48.859227944759986
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present an efficient method for adapting a monolingual Large Language Model (LLM) to another language, addressing challenges of catastrophic forgetting and tokenizer limitations. We focus this study on adapting Llama 2 to Arabic. Our two-stage approach begins with expanding the vocabulary and training only the embeddings matrix, followed by full model continual pre-training on a bilingual corpus. By continually pre-training on a mix of Arabic and English corpora, the model retains its proficiency in English while acquiring capabilities in Arabic. Our approach results in significant improvements in Arabic and slight enhancements in English, demonstrating cost-effective cross-lingual transfer. We perform ablations on embedding initialization techniques, data mix ratios, and learning rates and release a detailed training recipe. To demonstrate generalizability of this approach we also adapted Llama 3 8B to Arabic and Llama 2 13B to Hindi.

Related papers

• ALLaM: Large Language Models for Arabic and English [9.881560166505452]
  We present ALLaM: Arabic Large Language Model, a series of large language models to support the ecosystem of Arabic Language Technologies (ALT) Our autoregressive decoder-only architecture models demonstrate how second-language acquisition via vocabulary expansion and pretraining can steer a model towards a new language (Arabic) without any catastrophic forgetting in the original language (English) We show that extensive alignment with human preferences can significantly enhance the performance of a language model compared to models of a larger scale with lower quality alignment.
  arXiv  Detail & Related papers  (2024-07-22T05:35:17Z)
• AlcLaM: Arabic Dialectal Language Model [2.8477895544986955]
  We construct an Arabic dialectal corpus comprising 3.4M sentences gathered from social media platforms. We utilize this corpus to expand the vocabulary and retrain a BERT-based model from scratch. Named AlcLaM, our model was trained using only 13 GB of text, which represents a fraction of the data used by existing models.
  arXiv  Detail & Related papers  (2024-07-18T02:13:50Z)
• No Train but Gain: Language Arithmetic for training-free Language Adapters enhancement [59.37775534633868]
  We introduce a novel method called language arithmetic, which enables training-free post-processing. The effectiveness of the proposed solution is demonstrated on three downstream tasks in a MAD-X-based set of cross-lingual schemes.
  arXiv  Detail & Related papers  (2024-04-24T08:52:40Z)
• Parameter and Data Efficient Continual Pre-training for Robustness to Dialectal Variance in Arabic [9.004920233490642]
  We show that multilingual-BERT (mBERT) incrementally pretrained on Arabic monolingual data takes less training time and yields comparable accuracy when compared to our custom monolingual Arabic model. We then explore two continual pre-training methods-- (1) using small amounts of dialectical data for continual finetuning and (2) parallel Arabic to English data and a Translation Language Modeling loss function.
  arXiv  Detail & Related papers  (2022-11-08T02:51:57Z)
• Language Contamination Explains the Cross-lingual Capabilities of English Pretrained Models [79.38278330678965]
  We find that common English pretraining corpora contain significant amounts of non-English text. This leads to hundreds of millions of foreign language tokens in large-scale datasets. We then demonstrate that even these small percentages of non-English data facilitate cross-lingual transfer for models trained on them.
  arXiv  Detail & Related papers  (2022-04-17T23:56:54Z)
• Supporting Undotted Arabic with Pre-trained Language Models [0.0]
  We study the effect of applying pre-trained Arabic language models on "undotted" Arabic texts. We suggest several ways of supporting undotted texts with pre-trained models, without additional training, and measure their performance on two Arabic natural-language-processing tasks.
  arXiv  Detail & Related papers  (2021-11-18T16:47:56Z)
• Improving the Lexical Ability of Pretrained Language Models for Unsupervised Neural Machine Translation [127.81351683335143]
  Cross-lingual pretraining requires models to align the lexical- and high-level representations of the two languages. Previous research has shown that this is because the representations are not sufficiently aligned. In this paper, we enhance the bilingual masked language model pretraining with lexical-level information by using type-level cross-lingual subword embeddings.
  arXiv  Detail & Related papers  (2021-03-18T21:17:58Z)
• Cross-lingual Machine Reading Comprehension with Language Branch Knowledge Distillation [105.41167108465085]
  Cross-lingual Machine Reading (CLMRC) remains a challenging problem due to the lack of large-scale datasets in low-source languages. We propose a novel augmentation approach named Language Branch Machine Reading (LBMRC) LBMRC trains multiple machine reading comprehension (MRC) models proficient in individual language. We devise a multilingual distillation approach to amalgamate knowledge from multiple language branch models to a single model for all target languages.
  arXiv  Detail & Related papers  (2020-10-27T13:12:17Z)
• Mixed-Lingual Pre-training for Cross-lingual Summarization [54.4823498438831]
  Cross-lingual Summarization aims at producing a summary in the target language for an article in the source language. We propose a solution based on mixed-lingual pre-training that leverages both cross-lingual tasks like translation and monolingual tasks like masked language models. Our model achieves an improvement of 2.82 (English to Chinese) and 1.15 (Chinese to English) ROUGE-1 scores over state-of-the-art results.
  arXiv  Detail & Related papers  (2020-10-18T00:21:53Z)
• Can Multilingual Language Models Transfer to an Unseen Dialect? A Case Study on North African Arabizi [2.76240219662896]
  We study the ability of multilingual language models to process an unseen dialect. We take user generated North-African Arabic as our case study. We show in zero-shot and unsupervised adaptation scenarios that multilingual language models are able to transfer to such an unseen dialect.
  arXiv  Detail & Related papers  (2020-05-01T11:29:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.