failed models

[dallaszkorben]

1. when I run the whisper with 'tiny' model, I got python error:

ValueError: Expected parameter logits (Tensor of shape (5, 51865)) of distribution Categorical(logits: torch.Size([5, 51865])) to satisfy the constraint IndependentConstraint(Real(), 1), but found invalid values:
tensor([[nan, nan, nan,  ..., nan, nan, nan],
        [nan, nan, nan,  ..., nan, nan, nan],
        [nan, nan, nan,  ..., nan, nan, nan],
        [nan, nan, nan,  ..., nan, nan, nan],
        [nan, nan, nan,  ..., nan, nan, nan]], device='cuda:0')

2. When I run the whisper with 'medium' model, the command just got killed. No error message only 'Killed' message
3. It works well only with 'base' and 'small' models

In all case I use the same wav file.

Can you tell me what I did wrong or missed?