usnistgov / ccu_validation_scoring

Other
5 stars 0 forks source link

LDC2022E18-V6.0 segments.tab file doesn't pass validation #11

Open mary-octavia opened 1 year ago

mary-octavia commented 1 year ago

Hello,

When I try to validate the reference directory for LDC release E18, I get this error:

CCU_scoring validate-ref -ref ldc2022e18_v6/ ERROR:VALIDATION:Invalid file ldc2022e18_v6/docs/segments.tab: ERROR:VALIDATION:Start 457.9 is equal to/higher than end 457.9 in audio/video ldc2022e18_v6/docs/segments.tab

Raj gets the same validation error as well.

mary-octavia commented 1 year ago

from segments.tab M010005OM M010005OM_0005 457.9 457.9

jfiscus commented 1 year ago

Since it happens from time to time during data production that audio/video segments have zero seconds duration, we are going to drop those segments during scoring. We are also going to change the validator to issue a warning rather than failing.