3 lines
149 B
Markdown
3 lines
149 B
Markdown
|
|
# seriatim
|
||
|
|
|
||
|
|
Seriatim merges per-speaker whisperx transcripts into a single output transcript that preserves speaker identity and chronological order.
|