Does this project support other models of whisper base tiny

[MiningIrving]

I want to run whisper on my low hash rate terminal device and speed it up, but does the project only speed up for the large model

[riffov]

Changue this code line:

from faster_whisper import WhisperModel
model_size = "large-v2"

to

from faster_whisper import WhisperModel
model_size = "tiny"

Try switching to the whisper model that best suits your needs, being the following:

• tiny: a model of 39 million parameters that can do speech recognition and translation in English and multilingual1.
• base: a model of 74 million parameters that can do speech recognition and translation in English and multilingual1.
• small: a model of 244 million parameters that can do speech recognition and translation in English and multilingual1.
• medium: a model of 769 million parameters that can do speech recognition and translation in English and multilingual1.
• large: a model of 1550 million parameters that can only do speech recognition and translation in multilingual1.
• large-v2: a model of 1550 million parameters that can only do speech recognition and translation in multilingual, and that has better accuracy and robustness than the large model.

[Vaibhavs10]

Hi @MiningIrving - you can pass any other Whisper model via --model-name parameter mentioned here

(closing this issue now, since this is mostly now fixed).