paper-with-me

홈 › Papers

AVocaDo: Strategy for Adapting Vocabulary to Downstream Domain

2021-10-26 · EMNLP 2021 11 · Jimin Hong, Taehee Kim, Hyesu Lim, Jaegul Choo

During the fine-tuning phase of transfer learning, the pretrained vocabulary remains unchanged, while model parameters are updated. The vocabulary generated based on the pretrained data is suboptimal for downstream data when domain discrepancy exists. We propose to consider the vocabulary as an optimizable parameter, allowing us to update the vocabulary by expanding it with domain-specific vocabulary based on a tokenization statistic. Furthermore, we preserve the embeddings of the added words from overfitting to downstream data by utilizing knowledge learned from a pretrained language model with a regularization term. Our method achieved consistent performance improvements on diverse domains (i.e., biomedical, computer science, news, and reviews).

📄 PDF Abstract BibTeX arXiv:2110.13434

Code (1)

Jimin9401/avocado 공식 구현 pytorch

Tasks

Language ModelingLanguage ModellingTransfer Learning

Similar Papers 제목 키워드 기반

Domain Adaptation for Large-Vocabulary Object Detectors

2024-01-13 · Kai Jiang, Jiaxing Huang, Weiying Xie, Jie Lei 외

Large-vocabulary object detectors (LVDs) aim to detect objects of many categories, which learn super objectness features and can locate objects accurately while applied to various downstream data. However, LVDs often str…

Domain AdaptationKnowledge GraphsObjectObject Localization

Evaluating Pretraining Strategies for Clinical BERT Models

2022-06-01 · LREC 2022 6 · Anastasios Lamproudis, Aron Henriksson, Hercules Dalianis

Research suggests that using generic language models in specialized domains may be sub-optimal due to significant domain differences. As a result, various strategies for developing domain-specific language models have be…

Language ModelingLanguage Modelling

Graph Neural Networks for Adapting Off-the-shelf General Domain Language Models to Low-Resource Specialised Domains

2022-07-01 · NAACL (DLG4NLP) 2022 7 · Merieme Bouhandi, Emmanuel Morin, Thierry Hamon

Language models encode linguistic proprieties and are used as input for more specific models. Using their word representations as-is for specialised and low-resource domains might be less efficient. Methods of adapting t…

Avocado Price Prediction Using a Hybrid Deep Learning Model: TCN-MLP-Attention Architecture

2025-05-15 · Linwei Zhang, LuFeng, Ruijia Liang

With the growing demand for healthy foods, agricultural product price forecasting has become increasingly important. Hass avocados, as a high-value crop, exhibit complex price fluctuations influenced by factors such as s…

ImputationTime Series Forecasting

AVoCaDO: An Audiovisual Video Captioner Driven by Temporal Orchestration

2025-10-12 · Xinlong Chen, Yue Ding, Weihong Lin, Jingyun Hua 외 arxiv

Audiovisual video captioning aims to generate semantically rich descriptions with temporal alignment between visual and auditory events, thereby benefiting both video understanding and generation. In this paper, we prese…

Video Captioning