-
-
Notifications
You must be signed in to change notification settings - Fork 615
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
WandBLogger and TensorboardLogger have different APIs for logging audio #2791
Comments
@turian thanks for reporting. I think you do not need such workaround as we propagate automatically attributes to ignite/ignite/contrib/handlers/wandb_logger.py Lines 137 to 138 in a36e62d
So, I think the following should work (haven't explicitly checked) logger.log({"validation": [wandb.Audio(x.t(), caption="mixture", sample_rate=44100)]})
logger.log({"validation": [wandb.Audio(x.t(), caption="vocals", sample_rate=44100)]})
I agree that in case of TensorboardLogger, user can access |
Hey @vfdev-5 I want to contribute to this issue, can you please give me a general idea what should I do to make the APIs consistent |
@guptaaryan16 as it is described in #2791 (comment) we can add - logger.writer.add_audio('mixture', x.t(), engine.state.epoch)
+ logger.add_audio('mixture', x.t(), engine.state.epoch) Is it more clear now ? |
thanks for the guidance , I will try to complete it as fast as possible |
🚀 Feature
The following code doesn't work:
This is how you would typically add audio with a tensorboard logger:
The workaround (similar to discussed in Lightning-AI/pytorch-lightning#7028) would be to use the underlying _wandb object:
Is there a proposal for a standardized media logging API?
The text was updated successfully, but these errors were encountered: