Make words last untill new once are said?

[juliuswms]

Is there a way to make the words last untill new ones come in? I want to avoid having no subtitles on the screen. I would like to make the last said words to last untill new onces are said. I saw there is suppress_silence and suppress_ts_tokens but I think I'm using it wrong. Any help is much appreciated :)

def whisper_transcribe(tts_title_path, tts_description_path, srt_output_path):
    model = stable_whisper.load_model('base.en')
    result = model.transcribe(tts_description_path, regroup=True, suppress_silence=False, suppress_ts_tokens=True)
    (
        result
        .split_by_duration(0.6)
    )
    result.to_srt_vtt(srt_output_path, strip=True, word_level=False, segment_level=True)

(This is the code I am using right now)

[jianfch]

suppress_ts_tokens has no direct effect on the word timestamps.
To make words last until new ones, just set the end of the each word to the start of the next word.
With something like this:

all_words = result.all_words()
for i, word in enumerate(all_words[:-1], 1):
    word.end = all_words[i].start