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

add additonal condition for prev msdd cfgs #9182

Open
wants to merge 15 commits into
base: r2.0.0rc0
Choose a base branch
from

Conversation

nithinraok
Copy link
Collaborator

What does this PR do ?

TitaNet model was updated but some previous configurations were not.
This PR provides support for prevevious msdd configuration versions.

Collection: SD

Changelog

  • Catch the _target_ code in newer titanet versions, which is not available in previous versions.

GitHub Actions CI

The Jenkins CI system has been replaced by GitHub Actions self-hosted runners.

The GitHub Actions CI will run automatically when the "Run CICD" label is added to the PR.
To re-run CI remove and add the label again.
To run CI on an untrusted fork, a NeMo user with write access must first click "Approve and run".

Before your PR is "Ready for review"

Pre checks:

  • Make sure you read and followed Contributor guidelines
  • Did you write any new necessary tests?
  • Did you add or update any necessary documentation?
  • Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
    • Reviewer: Does the PR have correct import guards for all optional libraries?

PR Type:

  • New Feature
  • Bugfix
  • Documentation

Signed-off-by: Nithin Rao Koluguri <nithinraok>
@github-actions github-actions bot added the ASR label May 13, 2024
Signed-off-by: nithinraok <nithinraok@users.noreply.github.com>
tango4j and others added 5 commits May 14, 2024 14:38
Signed-off-by: Taejin Park <tango4j@gmail.com>
Signed-off-by: tango4j <tango4j@users.noreply.github.com>
Signed-off-by: Taejin Park <tango4j@gmail.com>
@tango4j
Copy link
Collaborator

tango4j commented May 14, 2024

  • Changed Speaker embedding extraction to avoid torch.cat in every for loop step for speed up.
    This does not make a lot of difference in general, but if CPU is busy with other tasks, this could make some difference in processing time.

  • Now, device=null and if torch.cuda.is_cuda_available() == True, then MSDD is running on GPU. This means diarization default setting will now run on GPU if cuda is available.

tango4j
tango4j previously approved these changes May 14, 2024
titu1994
titu1994 previously approved these changes May 15, 2024
Signed-off-by: Taejin Park <tango4j@gmail.com>
@tango4j tango4j dismissed stale reviews from titu1994 and themself via bdb8ca3 May 22, 2024 20:55
tango4j and others added 2 commits May 22, 2024 20:56
if isinstance(audio_file, list):
if len(audio_file) == 0:
raise ValueError(f"Empty audio file list: {audio_file}")
audio_file_name = sorted(audio_file)[0]

Check warning

Code scanning / CodeQL

Variable defined multiple times Warning

This assignment to 'audio_file_name' is unnecessary as it is
redefined
before this value is used.
raise ValueError(f"Empty audio file list: {audio_file}")
audio_file_name = sorted(audio_file)[0]
else:
audio_file_name = audio_file

Check warning

Code scanning / CodeQL

Variable defined multiple times Warning

This assignment to 'audio_file_name' is unnecessary as it is
redefined
before this value is used.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

4 participants