paper-with-me

홈 › Papers

Minimizing PLM-Based Few-Shot Intent Detectors

2024-07-13 · Haode Zhang, Albert Y. S. Lam, Xiao-Ming Wu

Recent research has demonstrated the feasibility of training efficient intent detectors based on pre-trained language model~(PLM) with limited labeled data. However, deploying these detectors in resource-constrained environments such as mobile devices poses challenges due to their large sizes. In this work, we aim to address this issue by exploring techniques to minimize the size of PLM-based intent detectors trained with few-shot data. Specifically, we utilize large language models (LLMs) for data augmentation, employ a cutting-edge model compression method for knowledge distillation, and devise a vocabulary pruning mechanism called V-Prune. Through these approaches, we successfully achieve a compression ratio of 21 in model memory usage, including both Transformer and the vocabulary, while maintaining almost identical performance levels on four real-world benchmarks.

📄 PDF Abstract BibTeX arXiv:2407.09943

Code (1)

hdzhang-code/smallID 공식 구현 jax

Tasks

Data AugmentationKnowledge DistillationLanguage ModelingLanguage ModellingModel Compression

Methods 이 논문이 사용한 방법론

Attention 설명 없음
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…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
Adam 설명 없음
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Multi-Head Attention 설명 없음

Similar Papers 제목 키워드 기반

Efficient Intent Detection with Dual Sentence Encoders

2020-03-10 · WS 2020 7 · Iñigo Casanueva, Tadas Temčinas, Daniela Gerz, Matthew Henderson 외

Building conversational systems in new domains and with added functionality requires resource-efficient models that work under low-data regimes (i.e., in few-shot setups). Motivated by these requirements, we introduce in…

CPUIntent DetectionSentence

Multilingual and Cross-Lingual Intent Detection from Spoken Data

2021-04-17 · EMNLP 2021 11 · Daniela Gerz, Pei-Hao Su, Razvan Kusztos, Avishek Mondal 외

We present a systematic study on multilingual and cross-lingual intent detection from spoken data. The study leverages a new resource put forth in this work, termed MInDS-14, a first training and evaluation resource for …

Few-Shot LearningIntent DetectionMachine TranslationSentence+3

Navigating Hallucinations for Reasoning of Unintentional Activities

2024-02-29 · Shresth Grover, Vibhav Vineet, Yogesh S Rawat

In this work we present a novel task of understanding unintentional human activities in videos. We formalize this problem as a reasoning task under zero-shot scenario, where given a video of an unintentional activity we …

HallucinationNavigate

Bridging the Gap Between Object Detection and User Intent via Query-Modulation

2021-06-18 · Marco Fornoni, Chaochao Yan, Liangchen Luo, Kimberly Wilber 외

When interacting with objects through cameras, or pictures, users often have a specific intent. For example, they may want to perform a visual search. With most object detection models relying on image pixels as their so…

Objectobject-detectionObject DetectionReferring Expression

IntRec: Intent-based Retrieval with Contrastive Refinement

2026-02-19 · Pourya Shamsolmoali, Masoumeh Zareapoor, Eric Granger, Yue Lu arxiv

Retrieving user-specified objects from complex scenes remains a challenging task, especially when queries are ambiguous or involve multiple similar objects. Existing open-vocabulary detectors operate in a one-shot manner…