The documentation mentions using model.save_torch_dict and CNN.from_torch_dict to save and load model weights between opensoundscape versions, but there's no tutorial. I think it would be helpful to have a tutorial showing this.
e.g. something like
# save the model
model.save_torch_dict...
# load the model
CNN.from_torch_dict...
# recreate the preprocessor, make sure the settings are the same as in the old version!
hop_size = 256
window_length = 512
preprocessor = ....
The documentation mentions using
model.save_torch_dict
andCNN.from_torch_dict
to save and load model weights between opensoundscape versions, but there's no tutorial. I think it would be helpful to have a tutorial showing this.e.g. something like