paper-with-me

홈 › Papers

DeBERTinha: A Multistep Approach to Adapt DebertaV3 XSmall for Brazilian Portuguese Natural Language Processing Task

2023-09-28 · Israel Campiotti, Matheus Rodrigues, Yuri Albuquerque, Rafael Azevedo, Alyson Andrade

This paper presents an approach for adapting the DebertaV3 XSmall model pre-trained in English for Brazilian Portuguese natural language processing (NLP) tasks. A key aspect of the methodology involves a multistep training process to ensure the model is effectively tuned for the Portuguese language. Initial datasets from Carolina and BrWac are preprocessed to address issues like emojis, HTML tags, and encodings. A Portuguese-specific vocabulary of 50,000 tokens is created using SentencePiece. Rather than training from scratch, the weights of the pre-trained English model are used to initialize most of the network, with random embeddings, recognizing the expensive cost of training from scratch. The model is fine-tuned using the replaced token detection task in the same format of DebertaV3 training. The adapted model, called DeBERTinha, demonstrates effectiveness on downstream tasks like named entity recognition, sentiment analysis, and determining sentence relatedness, outperforming BERTimbau-Large in two tasks despite having only 40M parameters.

📄 PDF Abstract BibTeX arXiv:2309.16844

Code (0)

등록된 구현이 없습니다.

Tasks

named-entity-recognitionNamed Entity RecognitionSentenceSentiment Analysis

Methods 이 논문이 사용한 방법론

BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
SentencePiece 설명 없음

Similar Papers 제목 키워드 기반

Solution of DeBERTaV3 on CommonsenseQA

2022-04-30 · Letian Peng, Zuchao Li, Hai Zhao

We report the performance of DeBERTaV3 on CommonsenseQA in this report. We simply formalize the answer selection as a text classification for DeBERTaV3. The strong natural language inference ability of DeBERTaV3 helps it…

Answer SelectionNatural Language Inferencetext-classificationText Classification

ModernBERT or DeBERTaV3? Examining Architecture and Data Influence on Transformer Encoder Models Performance

2025-04-11 · Wissam Antoun, Benoît Sagot, Djamé Seddah

Pretrained transformer-encoder models like DeBERTaV3 and ModernBERT introduce architectural advancements aimed at improving efficiency and performance. Although the authors of ModernBERT report improved performance over …

DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing

2021-11-18 · Pengcheng He, Jianfeng Gao, Weizhu Chen

This paper presents a new pre-trained language model, DeBERTaV3, which improves the original DeBERTa model by replacing mask language modeling (MLM) with replaced token detection (RTD), a more sample-efficient pre-traini…

Language ModelingLanguage ModellingNatural Language InferenceNatural Language Understanding+2

Image captioning for Brazilian Portuguese using GRIT model

2024-02-07 · Rafael Silva de Alencar, William Alberto Cruz Castañeda, Marcellus Amadeus

This work presents the early development of a model of image captioning for the Brazilian Portuguese language. We used the GRIT (Grid - and Region-based Image captioning Transformer) model to accomplish this work. GRIT i…

Image Captioningmodel

Pyramidal Hidden Markov Model For Multivariate Time Series Forecasting

2023-10-22 · YeXin Huang

The Hidden Markov Model (HMM) can predict the future value of a time series based on its current and previous values, making it a powerful algorithm for handling various types of time series. Numerous studies have explor…

Multivariate Time Series ForecastingTime SeriesTime Series Forecasting