FuLG: 150B Romanian Corpus for Language Model Pretraining

• URL: http://arxiv.org/abs/2407.13657v1
• Date: Thu, 18 Jul 2024 16:32:48 GMT
• Title: FuLG: 150B Romanian Corpus for Language Model Pretraining
• Authors: Vlad-Andrei Bădoiu, Mihai-Valentin Dumitru, Alexandru M. Gherghescu, Alexandru Agache, Costin Raiciu,
• Abstract summary: We introduce FuLG, a hundred-fifty-billion-token Romanian corpus extracted from CommonCrawl. We present our methodology for filtering FuLG and compare it via ablation studies against existing Romanian corpora.
• Score: 76.09455151754062
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Research in the field of language models is rapidly evolving, with many open models being released to the public. Openly available pretraining corpora usually focus on only a handful of languages, with many others either missing completely or extremely underrepresented. In this report, we introduce FuLG, a hundred-fifty-billion-token Romanian corpus extracted from CommonCrawl. We present our methodology for filtering FuLG and compare it via ablation studies against existing Romanian corpora.

Related papers

• LLMic: Romanian Foundation Language Model [76.09455151754062]
  We present LLMic, a foundation language model designed specifically for the Romanian Language. We show that fine-tuning LLMic for language translation after the initial pretraining phase outperforms existing solutions in English-to-Romanian translation tasks.
  arXiv  Detail & Related papers  (2025-01-13T22:14:45Z)
• GlotCC: An Open Broad-Coverage CommonCrawl Corpus and Pipeline for Minority Languages [53.56700754408902]
  GlotCC is a clean, document-level, 2TB general domain corpus derived from CommonCrawl. We make GlotCC and the system used to generate it available to the research community.
  arXiv  Detail & Related papers  (2024-10-31T11:14:12Z)
• Language Models on a Diet: Cost-Efficient Development of Encoders for Closely-Related Languages via Additional Pretraining [4.38070902806635]
  We set up a benchmark for languages Croatian, Serbian, Bosnian and Montenegrin. We show that comparable performance to dedicated from-scratch models can be obtained by additionally pretraining available multilingual models. We also show that neighboring languages, in our case Slovenian, can be included in the additional pretraining with little to no loss in the performance of the final model.
  arXiv  Detail & Related papers  (2024-04-08T11:55:44Z)
• Poro 34B and the Blessing of Multilinguality [3.270981284471548]
  Poro 34B is a 34 billion parameter model trained for 1 trillion tokens of Finnish, English, and programming languages. We show that a multilingual training approach can produce a model that substantially advances over the capabilities of existing models for Finnish.
  arXiv  Detail & Related papers  (2024-04-02T11:34:12Z)
• FinGPT: Large Generative Models for a Small Language [48.46240937758779]
  We create large language models (LLMs) for Finnish, a language spoken by less than 0.1% of the world population. We train seven monolingual models from scratch (186M to 13B parameters) dubbed FinGPT. We continue the pretraining of the multilingual BLOOM model on a mix of its original training data and Finnish, resulting in a 176 billion parameter model we call BLUUMI.
  arXiv  Detail & Related papers  (2023-11-03T08:05:04Z)
• Skywork: A More Open Bilingual Foundation Model [55.927396986873816]
  We present Skywork-13B, a family of large language models (LLMs) trained on a corpus of over 3.2 trillion tokens drawn from both English and Chinese texts. We show that our model not only excels on popular benchmarks, but also achieves emphstate of the art performance in Chinese language modeling on diverse domains.
  arXiv  Detail & Related papers  (2023-10-30T08:31:47Z)
• Zero- and Few-Shot Prompting with LLMs: A Comparative Study with Fine-tuned Models for Bangla Sentiment Analysis [6.471458199049549]
  In this study, we present a sizeable manually annotated dataset encompassing 33,606 Bangla news tweets and Facebook comments. We also investigate zero- and few-shot in-context learning with several language models, including Flan-T5, GPT-4, and Bloomz. Our findings suggest that monolingual transformer-based models consistently outperform other models, even in zero and few-shot scenarios.
  arXiv  Detail & Related papers  (2023-08-21T15:19:10Z)
• BLOOM: A 176B-Parameter Open-Access Multilingual Language Model [264.96498474333697]
  Large language models (LLMs) have been shown to be able to perform new tasks based on a few demonstrations or natural language instructions. We present BLOOM, a 176B- parameter open-access language model designed and built thanks to a collaboration of hundreds of researchers. BLOOM is a decoder-only Transformer language model that was trained on the ROOTS corpus, a dataset comprising hundreds of sources in 46 natural and 13 programming languages.
  arXiv  Detail & Related papers  (2022-11-09T18:48:09Z)
• The birth of Romanian BERT [1.377045689881944]
  This paper introduces Romanian BERT, the first purely Romanian transformer-based language model, pretrained on a large text corpus. We discuss corpus composition and cleaning, the model training process, as well as an extensive evaluation of the model on various Romanian datasets.
  arXiv  Detail & Related papers  (2020-09-18T09:30:48Z)
• Unsupervised Cross-lingual Representation Learning for Speech Recognition [63.85924123692923]
  XLSR learns cross-lingual speech representations by pretraining a single model from the raw waveform of speech in multiple languages. We build on wav2vec 2.0 which is trained by solving a contrastive task over masked latent speech representations. Experiments show that cross-lingual pretraining significantly outperforms monolingual pretraining.
  arXiv  Detail & Related papers  (2020-06-24T18:25:05Z)

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.