-
Notifications
You must be signed in to change notification settings - Fork 1
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
Could you provide the weights of the models for extracting features from scratch? #1
Comments
Hi @AndreJJXu, I finetuned the feature extraction models during my research, therefore you see To extract eg the features for UCF by yourself, you would have to run |
When I want to load weight from the files downloaded from "https://github.com/XinhaoMei/WavCaps", specifically for the "WavCaps/retrieval/pretrained_models/audio_encoders/HTSAT_BERT_zero_shot.pt", I always get the error { |
Hi, did you use the right conda environment? Also, you have to adjust the model path in the scripts where the features are created. For UCF it would be
|
@AndreJJXu I was facing a similar problem, I used https://github.com/XinhaoMei/WavCaps/blob/master/retrieval/work.yaml and created a new env. That solved the issue for me |
In your section "Extracting Features from Scratch", I find that you have leveraged the pre-trained models fine-tuned by yourself. Since I want to run the whole structure of your work, can you provide these weights or provide more details about "Extracting Features from Scratch"? Thanks!
The text was updated successfully, but these errors were encountered: