일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | ||
6 | 7 | 8 | 9 | 10 | 11 | 12 |
13 | 14 | 15 | 16 | 17 | 18 | 19 |
20 | 21 | 22 | 23 | 24 | 25 | 26 |
27 | 28 | 29 | 30 |
- Python
- 딥러닝기초
- fasttext text classification 한글
- 데이터 합성
- 2020정보처리기사필기
- 정보처리기사전공자
- 프로그래머스
- Zero-shot transfer learning with synthesized data for multi-domain dialogue state tracking
- 다이나믹 프로그래밍
- SUMBT:Slot-Utterance Matching for Universal and Scalable Belief Tracking
- Leveraging Slot Descriptions for Zero-Shot Cross-Domain Dialogue State Tracking
- 자연어처리 논문 리뷰
- dialogue state tracking
- 정보처리기사 수제비
- 정보처리기사 책 추천
- Few Shot Dialogue State Tracking using Meta-learning
- 정보처리기사전공자합격후기
- 백준
- 파이썬을 파이썬답게
- classification text
- nlp논문리뷰
- DST zeroshot learning
- DST fewshot learning
- How Much Knowledge Can You Pack Into the Parameters of a Language Model?
- 모두의딥러닝
- 검색엔진
- MySQL
- til
- From Machine Reading Comprehension to Dialogue State Tracking: Bridging the Gap
- few shot dst
- Today
- Total
🌲자라나는청년
[NLP논문리뷰]How Much Knowledge Can You Pack Into the Parameters of a Language Model? 본문
[NLP논문리뷰]How Much Knowledge Can You Pack Into the Parameters of a Language Model?
JihyunLee 2021. 9. 6. 16:34제목 : HowMuchKnowledge Can You Pack Into the Parameters of a Language Model?
저자 : Adam Roberts, Colin Raffel, Noam Shazeer
발행년도 : 2020
paper : https://arxiv.org/abs/2002.08910
code : https://github.com/google-research/google-research/tree/master/t5_closed_book_qa
Review
QA모델에 대한 논문 중 google에서 출간한 논문.
기존의 QA task는 passage를 보여주고 이에 대한 답을 찾아가는 방식이었다. 그러나 이 논문은 passage를 보여주지 않고 질문 부터 한다.

즉 다양한 task에 대해 학습한 T5에게 "루즈벨트 대통령이 태어난 해는?" 이라는 질문을 한다면, T5는 위키피디아의 지식 없이도 1882이라고 답을 하는지 실험해 본 논문이다.
논문에서는 Passage를 보여주고 Passage안에서 답을 찾도록 한 방식을 Open book test, passage를 보여주지 않고 답을 찾도록 한 방식을 Closed book test라고 비유했다!
Experimetns
실험에는 Natural Question, WebQuestions, TriviaQA 데이터를 사용하였고, passage는 사용하지 않고, 질문-답만 사용해서 학습시켰다.
Result

여기서 T5.1.1 은 unlabeled data only로 학습된 T5를 의미한다. 놀랍게도, passage를 보여주지 않았음에도 SOTA의 성능을 달성한 Task들이 있었으며, 다른 데이터 셋에 대한 성능도 SOTA에서 크게 뒤쳐지지 않았다.