Unable to generate SRT format

SETUP: Docker :latest-cpu

Hello, what i wanna achive is the same result that i obtain with che classic whisper cli using this command: whisper voice.mp3 --model medium --language Italian --word_timestamps True --max_words_per_line 1 -f 'srt'

thats gives me as outpur an SRT file with 1 word for timestamp. Beside of this I'm not sure if speaches supports --max_words_per_line param, i'm also unable to get an srt file as output.

I tried different curl sintax but each time i just obtain an json output:

curl "http://local.ip/v1/audio/transcriptions" -F "[email protected]" -F "model=guillaumekln/faster-whisper-medium" -F "--language=Italian" -F "--response-format=srt" -F "--word-timestamps=true" -F "--max_words_per_line=1"

I'm pasting just this as an example but i tried everything that i could find in the docs.