e2f66c0de2600308044ec3ab7731dae9017378fa,allennlp/modules/seq2seq_encoders/bidirectional_language_model_transformer.py,,subsequent_mask,#Any#Any#,40

Before Change

    Mask out subsequent positions.
    attn_shape = (1, size, size)
    mask = np.triu(np.ones(attn_shape), k=1).astype("uint8")
    mask = (torch.from_numpy(mask) == 0)
    return mask.to(device)


class PositionalEncoding(torch.nn.Module):

After Change

def subsequent_mask(size: int, device: str = "cpu") -> torch.Tensor:
    Mask out subsequent positions.
    mask = torch.tril(torch.ones(size, size, device=device, dtype=torch.int32)).unsqueeze(0)
    return mask


class PositionalEncoding(torch.nn.Module):

Instances