Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

moreNj_2_splitScp #3419

Closed
wants to merge 1 commit into from
Closed

moreNj_2_splitScp #3419

wants to merge 1 commit into from

Conversation

wanglong001
Copy link

using utils/split_scp.pl, --NJ must be little than num of speakers. it's not good for me to train.

./utils/split_scp.2.pl: Refusing to split data because number of speakers 2 is less than the number of output .scp files 4

and i want to fix it.

  1. Previously in speaker units, now in different classes of utt units.
  2. Different classes are more evenly distributed in each split

@danpovey
Copy link
Contributor

This would break certain things, e.g. in SAT training, FMLLR. You can split with --per-utt if this becomes a constraint, but it may require script changes.

@danpovey danpovey closed this Jun 26, 2019
@wanglong001 wanglong001 deleted the moreNj_2_splitScp branch June 26, 2019 17:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants