Datasets:

danielrosehill
/

Proofed-ASR-Cleanup-Training-Data

Tasks:

Automatic Speech Recognition

Languages:

Size:

Tags:

inverse-text-normalization

post-processing

License:

Dataset card Files Files and versions

Proofed-ASR-Cleanup-Training-Data / data /train /metadata.jsonl

Commit History

Oct 25 2 days ago

12:52
Transcripts within the allowed characters: no ; quotes dashes & + or accents, which tokenize to <unk> (ASR-Training scripts/normalize_transcripts.py --chars-only)

danielrosehill

Verified signature
e886d71

Oct 24 3 days ago

12:07
First snapshot: 600 takes, ~75 min, 9 tasks (part 3)

danielrosehill

Verified signature
2c51919