The code to enable training is now implemented but not tested extensively.
Like, the functions run and training does not crash. But it appears that testing this and running this has to be on HPC. My poor MacBook can't run training with the same number of batches and audio length as the implementation we ripped off.
The code to enable training is now implemented but not tested extensively. Like, the functions run and training does not crash. But it appears that testing this and running this has to be on HPC. My poor MacBook can't run training with the same number of batches and audio length as the implementation we ripped off.