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
Dead code? #21
Comments
Hi @martingasser ! Good find. Feel free to PR a change or I will when I have time to update the code. Regarding releasing the code for training, we intend to push it sometime during the summer. Thanks for your patience. |
Hi @drubinstein https://github.com/spotify/basic-pitch/blob/drubinstein/wip-training/basic_pitch/train.py
|
@drubinstein For example, do I need |
Hey @eunkoh, I'm planning on working on the training code in the near future, so i'll try to answer your questions to the best of my ability. I think it's worth pointing out that the code in that branch is still 'wip' as labeled, so we're aware on our side that it doesn't work in it's current form. As for your question about Please let me know if i'm misunderstanding your question and clarify further if so! |
Hey @drubinstein @rabitt, Was in intended to be the reduced version, with the three skipped layers that are still present in the code? Or was it just a happy accident that it happened to work very well without them? As the model structure diagram in the paper features the original Conv-BatchNorm-ReLU that's been skipped. If so would you consider retraining with the original intended model structure and see if the model accuracy improves? |
Hi,
first of all, congratulations to this work!
During examining the model, I noticed that
x_contours
is not used and overwritten afterbasic-pitch/basic_pitch/models.py
Lines 207 to 216 in fc5c319
I also wanted to ask if you plan on releasing the script used for training the model as well.
Thanks,
Martin
The text was updated successfully, but these errors were encountered: