loading my trained model // Error: AssertionError: Cannot stride by more than max generation duration.

#76
by deleted - opened
deleted
edited May 3

Screenshot_3.jpg

I further trained the model on my data (facebook/musicgen-small), but when loading the model I received the following error with this code:
import audiocraft.models
model = audiocraft.models.MusicGen.get_pretrained('/mnt/my_audio_lm/')

path '/mnt/my_audio_lm/' - mine is kept there state_dict.bit and compression_state_dict (link to facebook/musicgen-small)

What causes this error?

Sign up or log in to comment