Skip to content
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

intermediate_source/speech_recognition_pipeline_tutorial.py 번역 #582

Merged
merged 2 commits into from
Sep 9, 2022
Merged

intermediate_source/speech_recognition_pipeline_tutorial.py 번역 #582

merged 2 commits into from
Sep 9, 2022

Conversation

terri1102
Copy link
Contributor

@terri1102 terri1102 commented Aug 29, 2022

라이선스 동의

변경해주시는 내용에 BSD 3항 라이선스가 적용됨을 동의해주셔야 합니다.

더 자세한 내용은 기여하기 문서를 참고해주세요.

동의하시면 아래 [ ][x]로 만들어주세요.

  • 기여하기 문서를 확인하였으며, 본 PR 내용에 BSD 3항 라이선스가 적용됨에 동의합니다.

관련 이슈 번호

이 Pull Request와 관련있는 이슈 번호를 적어주세요.

이슈 또는 PR 번호 앞에 #을 붙이시면 제목을 바로 확인하실 수 있습니다. (예. #999 )

PR 종류

이 PR에 해당되는 종류 앞의 [ ][x]로 변경해주세요.

  • 오탈자를 수정하거나 번역을 개선하는 기여
  • 번역되지 않은 튜토리얼을 번역하는 기여
  • 공식 튜토리얼 내용을 반영하는 기여
  • 위 종류에 포함되지 않는 기여

PR 설명

이 PR로 무엇이 달라지는지 대략적으로 알려주세요.
intermediate_source/speech_recognition_pipeline_tutorial.py를 번역하였습니다.

line16에 Heading이 깨지는데 원본 문서도 깨져있어서 그대로 두었습니다.
혹시 원인을 아시는 분은 코멘트해주세요!

Copy link
Contributor

@bub3690 bub3690 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

looks good to me!
고생하셨습니다

Copy link
Contributor

@updaun updaun left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

내용이 많습니다~! 수고하셨습니다!
lgtm!!

@hyoyoung hyoyoung added the 컨트리뷰톤 오픈소스 컨트리뷰톤 관련 이슈/PR label Sep 4, 2022
Copy link
Contributor

@hjvision96 hjvision96 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

잘 번역하신 것 같습니다. 고생하셨습니다!

# The bundle object provides the interface to instantiate model and other
# information. Sampling rate and the class labels are found as follow.
# 아래의 코드에서 번들 객체(object)는 모델을 생성하고(instantiate)하고 다른 정보를 얻기 위한
# 인터페이스를 제공합니다. 이를 이용해 샘플링 레이트와 클래스 라벨을 확인하겠습니다.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

오타인 것 같습니다. 다음과 같이 수정하면 어떨까요?
생성하고(instantiate)하고 -> 생성(instantiate)하고

Copy link
Contributor

@YunjinJo YunjinJo left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

번역 되게 잘하셨네요! 큰 문제는 못찾았고 일부 사소한 문제들만 리뷰 남겼습니다.

# Wav2Vec2 (and HuBERT) models are trained in self-supervised manner. They
# are firstly trained with audio only for representation learning, then
# fine-tuned for a specific task with additional labels.
# Wav2Vec2와 HuBERT 모델은 자기 지도 방식으로 학습된 모델입니다. 두 모델은 먼저
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

원문과 같이 (그리고 HuBERT) 라고 쓰는건 어떤가요?

#

with torch.inference_mode():
features, _ = model.extract_features(waveform)


######################################################################
# The returned features is a list of tensors. Each tensor is the output of
# a transformer layer.
# 반환되는 특성은 tensor의 리스트이고 각 tensor는 transformer 레이어의 출력입니다.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

리스트 -> 배열
음차하지 않아도 될 것 같습니다.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

보통 배열은 'array'를 가리켜서 리스트로 음차했는데, 여기서는 리스트 자료형을 의미하지 않아서 '배열'로 써도 될 것 같긴 하네요.

# labels. They are bundled together and available under
# :py:func:`torchaudio.pipelines` module.
# Torchaudio는 사전 학습된 모델 가중치와 기대 샘플링 레이트나 클래스 라벨과 같은
# 관련 정보를 쉽게 제공합니다. 이런 정보들은
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

관련 정보를 쉽게 제공합니다 -> 관련 정보를 제공합니다
쉽게 제공한다는 말이 어색하기도 하고, 의미 전달에 큰 영향이 없어 번역하지 않아도 될 것 같습니다.

# There are two types of Wav2Vec2 pre-trained weights available in
# torchaudio. The ones fine-tuned for ASR task, and the ones not
# fine-tuned.
# torchaudio에는 사전 학습된 Wav2Vec2 모델 가중치가 두 종류 있습니다.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

torchaudio에는 -> Torchaudio에는

# tutorial, we also show how to perform feature extraction here.
# Wav2Vec2 모델은 ASR 태스크를 위해 미세 조정되어 특성 추출과 분류를
# 한 번에 수행할 수 있습니다.
# 하지만 이번 튜토리얼에서는 특성 추출을 하는 방법도 설명하겠습니다.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

하지만 이번 튜토리얼에서는 -> 하지만 자세한 설명을 위해
for the sake of 의 뜻을 살려서 약간의 의역을 하였습니다. 튜토리얼 단어를 넣은 번역문은 생각이 안나네요.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

'하지만 자세한 설명을 위해 이번 튜토리얼에서는'으로 쓰겠습니다.

Copy link
Member

@hyoyoung hyoyoung left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

긴 글 번역하느라 수고하셨습니다.

몇가지 제안을 드립니다.

# Model can be constructed as following. This process will automatically
# fetch the pre-trained weights and load it into the model.
# 모델은 아래와 같이 생성할 수 있습니다. 이 과정을 통해 사전 학습된 모델의 가중치가
# 다운로드되고 모델에 불러와집니다.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

fetch를 다운로드로 번역해도 의미가 통하겠지만

이 과정을 통해 사전 학습된 모델의 가중치를 받아서 모델에 넣어줍니다.

정도로 바꾸는건 어떨까요?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

'이 과정을 통해 사전 학습된 모델의 가중치를 가져와서 모델에 넣어줍니다.'로 수정하였습니다.

#
# We start by defining greedy decoding algorithm.
# 그리디 디코딩 알고리즘을 정의하겠습니다.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

greedy는 보통 탐욕적인, 욕심쟁이 등으로 번역하는 듯 합니다

# a model and getting the emission is as short as two lines.
# 이번 튜토리얼에서 음향 특성 추출과 음성 인식을 위해서
# :py:mod:`torchaudio.pipelines` 를 어떻게 사용하는지 알아보았습니다.
# 모델을 만들고 emission을 얻는 모든 과정은 아래의 2줄 만으로도 가능합니다.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

emission은 산출물로 번역해도 뜻이 통하지 않을까요?

@terri1102 terri1102 requested review from hjvision96, hyoyoung and YunjinJo and removed request for hjvision96, hyoyoung and YunjinJo September 9, 2022 13:22
Copy link
Member

@hyoyoung hyoyoung left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

good

@hyoyoung hyoyoung merged commit 8f37c36 into PyTorchKorea:master Sep 9, 2022
@terri1102 terri1102 deleted the translation branch September 11, 2022 05:17
# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
컨트리뷰톤 오픈소스 컨트리뷰톤 관련 이슈/PR
Projects
None yet
Development

Successfully merging this pull request may close these issues.

6 participants