RuntimeError: shape '[80, 1, 251]' is invalid for input of size 10040

[vovasilenko]

I tried using the following code:

import torch
from transformers import AutoModel
sincnet = AutoModel.from_pretrained("D4ve-R/sincnet")
# Forward pass
y = sincnet(torch.randn(1, 1, 16000))

I got the following error:

        band_pass = torch.cat([band_pass_left,band_pass_center,band_pass_right],dim=1)
        band_pass = band_pass / (2*band[:,None])
---> return band_pass.view(self._out_channels, 1, self._kernel_size)

RuntimeError: shape '[80, 1, 251]' is invalid for input of size 10040

[Toto-Africa]

I am experiencing the same error. Replying so that @D4ve-R hopefully sees these messages and could help us out.

@Nenteru : were you able to resolve this bug yourself?