Skip to content

intermediate_source/scaled_dot_product_attention_tutorial.py 번역 진행 #747

New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Closed
ganghe74 opened this issue Aug 26, 2023 · 0 comments
Closed

Comments

@ganghe74
Copy link
Contributor

문서 URL

(대략적인) 예상 완료 일정

  • 예상 완료 일정: 1주일 (~23.09.02 토)

관련 이슈

ganghe74 added a commit to ganghe74/tutorials-kr that referenced this issue Sep 5, 2023
1. Scaled Dot Product Attention는 번역하지 않는게 좋다고 판단했습니다.
2. 트랜스포머는 처음에 음차와 영어 원문(Transformers)를 병행 표기했습니다.
   이후 Transformer 용어가 등장하면 "트랜스포머"로 음차 표기할 예정입니다.
3. query, key, value 는 각각 쿼리, 키, 값으로 번역했습니다.
  번역 사례 2가지를 참고했습니다.
  ref 1) "딥 러닝을 이용한 자연어 처리 입문" (https://wikidocs.net/31379)
  ref 2) "트랜스포머를 활용한 자연어 처리" 루이스 턴스톨, 레안드로 폰
         베라, 토마스 울프 저/박해선 역
4. fused 는 적당한 번역을 생각해내지 못해서 음차와 영어 원문을 병행
   표기하는 것으로 결정했습니다.

issue) PyTorchKorea#747
ganghe74 added a commit to ganghe74/tutorials-kr that referenced this issue Sep 5, 2023
ganghe74 added a commit to ganghe74/tutorials-kr that referenced this issue Sep 5, 2023
Dispatcher는 번역할 마땅한 단어를 찾지 못해서,
    영문 표기를 그대로 사용했습니다.
Helpful arguments mapper 또한 영문 표기를 유지했습니다.

issue) PyTorchKorea#747
ganghe74 added a commit to ganghe74/tutorials-kr that referenced this issue Sep 5, 2023
ganghe74 added a commit to ganghe74/tutorials-kr that referenced this issue Sep 5, 2023
마지막 전 문단 (line 308~312) 번역에 도움이 필요합니다.
이해를 돕기 위해 약간의 의역이 필요할 것 같습니다.

issue) PyTorchKorea#747
ganghe74 added a commit to ganghe74/tutorials-kr that referenced this issue Sep 5, 2023
# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant