Skip to content
New issue

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

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

Already on GitHub? Sign in to your account

librispeech_decode : split save_results() -> save_asr_output() + save_wer_results() #1712

Merged
merged 1 commit into from
Aug 13, 2024

Conversation

KarelVesely84
Copy link
Contributor

@KarelVesely84 KarelVesely84 commented Aug 8, 2024

THE IDEA

  • the idea is to support --skip-scoring argument passed to a decoding script
  • created for Transducer decoding (non-streaming, streaming) and CTC decoding (non-streaming)

OTHER CHANGES

  • also added --label for extra label in streaming_decode.py
  • and also added set_caching_enabled(True)
    • this has no effect on librispeech, but it leads to faster runtime on DBs with long recordings (assuming librispeech/zipformer scripts are the example scripts for other setups)
  • replaced some '-' with '_' and vice versa, to get more readable filenames with results

TESTING

  • still need to run the code to make sure it works properly (it is meld-ed from other recipe)
  • at this point, interested in getting early feedback from you...

Thank you.

@KarelVesely84 KarelVesely84 changed the title split save_results() -> save_asr_output() + save_wer_results() decode : split save_results() -> save_asr_output() + save_wer_results() Aug 8, 2024
- the idea is to support `--skip-scoring` argument passed to a decoding
  script
- created for Transducer decoding (non-streaming, streaming)
- it can be done also for CTC decoding... (not yet)

- also added `--label` for extra label in `streaming_decode.py`
- and also added `set_caching_enabled(True)`, which has no effect on
  librispeech, but it leads to faster runtime on DBs with long
  recordings (assuming `librispeech/zipformer` scripts are the
  example scripts for other setups)

- still need to run the code to make sure it works properly
- at this point, interested in getting early feedback from you...

Thank you.
@KarelVesely84 KarelVesely84 changed the title decode : split save_results() -> save_asr_output() + save_wer_results() librispeech decode : split save_results() -> save_asr_output() + save_wer_results() Aug 12, 2024
@KarelVesely84 KarelVesely84 changed the title librispeech decode : split save_results() -> save_asr_output() + save_wer_results() librispeech_decode : split save_results() -> save_asr_output() + save_wer_results() Aug 12, 2024
@KarelVesely84
Copy link
Contributor Author

Hi @csukuangfj ,
would it be ok to merge somethig like this PR ?
(adding --skip-scoring to the example decode scripts, it's handy for processing evaluation data)

What should be the next step ?

Best regards,
Karel

@csukuangfj
Copy link
Collaborator

Looks good to me. Thank you for your contribution!

@csukuangfj csukuangfj merged commit 1730fce into k2-fsa:master Aug 13, 2024
275 of 278 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants