Where does faster-whisper downloads the model ?

[shrangideqode]

When initialising the whisper model
model = WhisperModel("distil-large-v3", device="cuda", compute_type='float16')
What path does it download the model ?
How can i download the model beforehand so that while initialising it does not download and only load the model.
I want to download the model in Docker image so that i start my application without waiting for it to downlaod.

[MahmoudAshraf97]

it's downloaded from huggingface to the huggingface cache dir

[hmf]

If you are in Linux that would be in (user is your user name):

/home/user/.cache/huggingface/hub/

For a list of model names see:

faster-whisper/faster_whisper/utils.py

Line 12 in 97a4785

_MODELS = {

HTHs