This release is offering a faster and more memory efficient code than the previously released version
Key Changes:
- Moving paddings and mini-batches related tensors creation to the EHR_dataloader
- Creating the mini-batches list once before running the epochs
- Adding RETAIN to the models list