SWAV Loss function

Hello,

Congratulation for this great work. I have a question regarding the loss function from SWAV.

```

def cross_entropy_loss(q, p): 
   return torch.mean(torch.sum(q * F.log_softmax(p, dim=1), dim=1))
```

That ``F.log_softmax`` will return a negative number. Should it be ``-F.log_softmax(p, dim=1)``?

Thank you.