-
Notifications
You must be signed in to change notification settings - Fork 90
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
Can't run the deep learning based source separation model #230
Comments
Have you trained a model? In the code snipped you provided:
it is assumed that you have a trained model at the path specified by the variable |
Hi Ethan:
Whether the downloaded model is pre-trained ? Or I have to train it by myself? Thx! |
Your problem is here: Unfortunately, we don't have any pre-trained models available currently. If you want to use one with nussl, you will have to train your own. Alternatively, there are pre-trained models available from asteroid. We don't currently have the bandwidth to provide a nice suite of available models. Sorry and best of luck! |
BEFORE POSTING A BUG REPORT Please look through existing issues (both open and closed) to see if it's already been reported or fixed!
Describe the bug
When I am using the deep mask estimation machine learning model to try the source separation, I get the following error at this step:
separator = nussl.separation.deep.DeepMaskEstimation(
audio_signal, mask_type='soft', model_path=model_path)
ERROR:
Error(s) in loading state_dict for SeparationModel:
Missing key(s) in state_dict: "normalization.batch_norm.weight", "normalization.batch_norm.bias", "normalization.batch_norm.running_mean", "normalization.batch_norm.running_var", "recurrent_stack.rnn.weight_ih_l0", "recurrent_stack.rnn.weight_hh_l0", "recurrent_stack.rnn.bias_ih_l0", "recurrent_stack.rnn.bias_hh_l0", "recurrent_stack.rnn.weight_ih_l0_reverse", "recurrent_stack.rnn.weight_hh_l0_reverse", "recurrent_stack.rnn.bias_ih_l0_reverse", "recurrent_stack.rnn.bias_hh_l0_reverse", "recurrent_stack.rnn.weight_ih_l1", "recurrent_stack.rnn.weight_hh_l1", "recurrent_stack.rnn.bias_ih_l1", "recurrent_stack.rnn.bias_hh_l1", "recurrent_stack.rnn.weight_ih_l1_reverse", "recurrent_stack.rnn.weight_hh_l1_reverse", "recurrent_stack.rnn.bias_ih_l1_reverse", "recurrent_stack.rnn.bias_hh_l1_reverse", "recurrent_stack.rnn.weight_ih_l2", "recurrent_stack.rnn.weight_hh_l2", "recurrent_stack.rnn.bias_ih_l2", "recurrent_stack.rnn.bias_hh_l2", "recurrent_stack.rnn.weight_ih_l2_reverse", "recurrent_stack.rnn.weight_hh_l2_reverse", "recurrent_stack.rnn.bias_ih_l2_reverse", "recurrent_stack.rnn.bias_hh_l2_reverse", "recurrent_stack.rnn.weight_ih_l3", "recurrent_stack.rnn.weight_hh_l3", "recurrent_stack.rnn.bias_ih_l3", "recurrent_stack.rnn.bias_hh_l3", "recurrent_stack.rnn.weight_ih_l3_reverse", "recurrent_stack.rnn.weight_hh_l3_reverse", "recurrent_stack.rnn.bias_ih_l3_reverse", "recurrent_stack.rnn.bias_hh_l3_reverse", "mask.linear.weight", "mask.linear.bias".
Steps To Reproduce
Expected behavior
By looking at the tutorial, it should give the result of separated audio.
What did happen
A clear and concise description of what did happen.
Audio output
If applicable, please link to audio examples that you uploaded to help us diagnose the issue.
Screenshots
If applicable, add screenshots to help explain your problem.
Software versions*
Additional context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered: