This folder contains the implemented models and the testing results appeared the report's Parameter Matching section. The training is performed on Google Colab so the file directory might need changing if you want retrain the model. Content in Model_Param and Res needs to be downloaded from https://drive.google.com/drive/folders/181wBma-MXyyFgpl6MOZew1NWl-8Zv_iK?usp=sharing
/Res contains the preprocessed DIVA dataset. The raw dataset(:https://nubo.ircam.fr/index.php/s/nL3NQomqxced6eJ) Including the txt of 32 parameters.
/Result contains loss data in numpy array and example audio in the error normalisation section of parameter matching in the report
/old_models contains earlier atttempt when learning pytorch
/figures contains some of the plots and images presented in the report
/Model_Param contains pytorch model checkpoints