Error while doing 'Getting started' steps

[SatenHarutyunyan]

Hi. Thanks for the work.
I tried to install this repo and I followed the steps in the readme.
I downloaded the dataset succesfully and now have speech and transcripts folders.
On the step 6. compute_features.py I get this error

return open_fn(path, mode)
FileNotFoundError: [Errno 2] No such file or directory: './data/manifests/recordings_dev.jsonl'

Because I have only these files in that folder - recordings_train.jsonl and supervisions_train.jsonl.
What had gone wrong?

[SatenHarutyunyan]

I changed the line in compute_features.py from

SPLITS = ['train', 'dev', 'test']

to

SPLITS = ['train']

Then I start getting

FileNotFoundError: [Errno 2] No such file or directory: './data/icsi/data_dfs/train_df.csv'

[LasseWolter]

I'm sorry, I've been busy with work and am travelling now.
I'll try to have a look at it over the weekend.

Out of curiosity, are you working on a similar project?

[LasseWolter]

I just cloned a fresh version of the repo and followed the steps.
Works for me.
I have recordings_dev.jsonl in my data/manifests folder and the cutsets for each split (dev, test and train) are generated successfully, see screenshot below:

I'd guess that something went wrong during audio download or manifest creation.
Did you only install the pip packages mentioned in the setup steps, i.e. you ran pip install -r requirements in a separate virtual environment?
Just want to make sure it's not some issue with a different version of lhotse or something like that.

Could you also confirm that you have all 75 meetings downloaded in your data/icsi/speech folder? You can do so by running ls data/icsi/speech | wc -l which should give you 75. (note this command is a unix command and will not work on powershell on windows).

Looking forward to hearing back from you and hope we can get this to work :)