Source codes for the paper "You Truly Understand What I Need: Intellectual and Friendly Dialogue Agents grounding Knowledge and Persona", accepted at EMNLP 2022 Findings.
We propose an effective QAGen framework that enhances diversity and quality in the QA pair generation. Our framework consists of a QFS-based answer generator, iterative QA generator, and relevancy-aware ranker.
PEEP-Talk is an educational platform with a deep learning-based persona conversation system and a feedback function for correcting English grammar. In addition, unlike the existing persona conversation system, a Context Detector (CD) module that can automatically determine the flow of conversation and change the conversation topic in real time can be applied to give the user a feeling of talking to a real person.
고려대학교 NLP&AI 연구실의 LLM 구름 코드
입력 문장에 현재 언어에 대한 특수 심볼을 추가해서 두 언어 대상 개체명 인식의 성능을 끌어올린 모델
의존 구문 정보를 활용해 주어진 두 문장 사이의 의미적 관계(수반/상반/중립)를 추론하는 모델
This paper presents a novel cost aggregation network, called Volumetric Aggregation with Transformers (VAT), for few-shot segmentation. The use of transformers can benefit correlation map aggregation through self-attention over a global receptive field. However, the tokenization of a correlation map for transformer processing can be detrimental, because the discontinuity at token boundaries reduces the local context available near the token edges and decreases inductive bias. To address this pro
Text-To-Text Transfer Transformer for Korean languages
knowledge triple 생성을 위한 관계 추출
AMR과 상식 graph를 이용한 상식 질의응답