Albert Zeyer

Why open a separate new PR then? You could have just updated #60. Please do that in the future. But anyway, leave it like this now.

> > Why open a separate new PR then? You could have just updated #60. Please do that in the future. But anyway, leave it like this now. > >...

> > Btw, I see that you also added the dataset there. Please separate this ... > > Yes. It should have been put in the main config. So can...

> > But making (and dumping) the alignment is independent from the label topology? > > Yes, if you try to change the chunking to make up for the topology....

This PR still has stuff about the dummy dataset. Can you remove this here?

This PR still looks very much work-in-progress. Can you mark it as draft, until you consider it ready? Also, can you comment what the state is now?

If this PR is work-in-progress, please do not mark it as "ready" but as "draft" instead. Also remove the "WIP" from the title. And once you consider it ready, mark...

> For FixedPath training, different datasets, are to be handled differently. For example for Switchboard, `seq_order_seq_lens_file` has to be provided whereas LibriSpeech has `seq_tags`. No, those mean different things. In...

See my previous comment: Also remove the "WIP" from the title.

[Question on StackOverflow](https://stackoverflow.com/questions/72588572/can-i-configure-github-to-block-large-files), without a real answer yet, though. However, after some short research, I think sth like this could work: Maybe via the branch protection rules. I saw some...