Since GitHub does not allow to upload big files, to run this script, you need to download a supplementary file available in this link [https://drive.google.com/drive/folders/176qhDmYUDzQg5yrv-IPzkpKSCR3Jv083]. Therefore, to correctly run the python program, you need to download the repository folder, download the "cassette-th-data.pck" in the link above, and copy the .pck file in the repository folder.
The data and model are inspired by the DeepSleepNet (https://github.com/akaraspt/deepsleepnet) architecture. All of the information can be read in the .pdf file in the repository, corresponding to my report on the model.
Enjoy :)