Clean speech data is required

[LiamHorton]

Hi:

I am trying to use the dns_challenge recipie. I have passed stage 0 but when it goes into stage 1 and runs the noisyspeech_synthesizer_singleprocess.py from the cloned DNS-Challenge repo, I get the following error:

File "noisyspeech_synthesizer_singleprocess.py", line 289, in main_body
assert False, ('Clean speech data is required')
AssertionError: Clean speech data is required

Does anyone know how to fix this? I am trying to download the dataset, so I will not have the clean speech before this.

Thanks,
Liam