Is it possible to load a model 100% offline using EncoderClassifier.from_hparams?

[silvioprog]

Hi.

I have an environment with zero internet access, so any model must be loaded offline from a sdcard.

So, is it possible to load for example the model lang-id-voxlingua107-ecapa but 100% offline? Something like this:

from speechbrain.inference.classifiers import EncoderClassifier

language_id = EncoderClassifier.from_hparams(
  source="/path/with/the/model/lang-id-voxlingua107-ecapa"
)

TIA for any help!

[silvioprog]

I have just solved using this:

from speechbrain.inference.classifiers import EncoderClassifier

hparams_file = '/path/with/the/model/lang-id-voxlingua107-ecapa/hyperparams.yaml'
saved_model_dir = '/path/with/the/model/lang-id-voxlingua107-ecapa'
classifier = EncoderClassifier.from_hparams(source=hparams_file, savedir=saved_model_dir)

predictions = classifier.classify_file("~/audio.wav")
print(predictions[3])