Open Root-FTW opened 9 months ago
Maybe --chunk_size
is what you need.
https://github.com/m-bain/whisperX/blob/f8cc46c6f7fa3b8509bc6aa04cdf4a62a702bb42/whisperx/transcribe.py#L44
Maybe
--chunk_size
is what you need.
Wouldn't this cause worse transcription results, since the context is lost with smaller chunks of audio? Seems like it would make more sense to split it after transcription.
How can I make my subtitles 1 line long and no more than 5 words?
This is the command I use on my windows with Nvidia GTX 1080 graphics card and the result is multiple lines of subtitles and many words resulting in a super long and long paragraph to read.
whisperx video.mp4 --model large-v3 --highlight_words False --align_model WAV2VEC2_ASR_LARGE_LV60K_960H --compute_type int8 --max_line_width 42 --max_line_count 2 --output_format srt