Self-Attention Code

1. ScaledDotProductAttention

class ScaledDotProductAttention(nn.Module):
    
    def __init__(self, dim: int):
        super(ScaledDotProductAttention, self).__init__()
        self.sqrt_dim = np.sqrt(dim)

    def forward(self, query: Tensor, key: Tensor, value: Tensor) -> Tuple[Tensor, Tensor]:
        score = torch.bmm(query, key.transpose(1, 2)) / self.sqrt_dim
        attn = F.softmax(score, -1)
        context = torch.bmm(attn, value)
        return context, attn

2. Multi-Head Self-Attention

[Reference] https://github.com/hyunwoongko/transformer/blob/master/models/layers/multi_head_attention.py

class MultiHeadAttention(nn.Module):

    def __init__(self, d_model, n_head):
        super(MultiHeadAttention, self).__init__()
        self.n_head = n_head
        self.attention = ScaleDotProductAttention()
        self.w_q = nn.Linear(d_model, d_model)
        self.w_k = nn.Linear(d_model, d_model)
        self.w_v = nn.Linear(d_model, d_model)
        self.w_concat = nn.Linear(d_model, d_model)

    def forward(self, q, k, v, mask=None):
        # 1. dot product with weight matrices
        q, k, v = self.w_q(q), self.w_k(k), self.w_v(v)

        # 2. split tensor by number of heads
        q, k, v = self.split(q), self.split(k), self.split(v)

        # 3. do scale dot product to compute similarity
        out, attention = self.attention(q, k, v, mask=mask)

        # 4. concat and pass to linear layer
        out = self.concat(out)
        out = self.w_concat(out)

        # 5. visualize attention map
        # TODO : we should implement visualization

        return out

    def split(self, tensor):
        """
        split tensor by number of head
        :param tensor: [batch_size, length, d_model]
        :return: [batch_size, head, length, d_tensor]
        """
        batch_size, length, d_model = tensor.size()

        d_tensor = d_model // self.n_head
        tensor = tensor.view(batch_size, length, self.n_head, d_tensor).transpose(1, 2)
        # it is similar with group convolution (split by number of heads)

        return tensor

    def concat(self, tensor):
        """
        inverse function of self.split(tensor : torch.Tensor)
        :param tensor: [batch_size, head, length, d_tensor]
        :return: [batch_size, length, d_model]
        """
        batch_size, head, length, d_tensor = tensor.size()
        d_model = head * d_tensor

        tensor = tensor.transpose(1, 2).contiguous().view(batch_size, length, d_model)
        return 

---

Author by 김주영
Edit by 김주영