Source code for monai.networks.blocks.transformerblock

# Copyright (c) MONAI Consortium
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#     http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import annotations

import torch.nn as nn

from monai.networks.blocks.mlp import MLPBlock
from monai.networks.blocks.selfattention import SABlock


class TransformerBlock(nn.Module):
    """
    A transformer block, based on: "Dosovitskiy et al.,
    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale <https://arxiv.org/abs/2010.11929>"
    """

    def __init__(
        self,
        hidden_size: int,
        mlp_dim: int,
        num_heads: int,
        dropout_rate: float = 0.0,
        qkv_bias: bool = False,
        save_attn: bool = False,
    ) -> None:
        """
        Args:
            hidden_size (int): dimension of hidden layer.
            mlp_dim (int): dimension of feedforward layer.
            num_heads (int): number of attention heads.
            dropout_rate (float, optional): faction of the input units to drop. Defaults to 0.0.
            qkv_bias (bool, optional): apply bias term for the qkv linear layer. Defaults to False.
            save_attn (bool, optional): to make accessible the attention matrix. Defaults to False.

        """

        super().__init__()

        if not (0 <= dropout_rate <= 1):
            raise ValueError("dropout_rate should be between 0 and 1.")

        if hidden_size % num_heads != 0:
            raise ValueError("hidden_size should be divisible by num_heads.")

        self.mlp = MLPBlock(hidden_size, mlp_dim, dropout_rate)
        self.norm1 = nn.LayerNorm(hidden_size)
        self.attn = SABlock(hidden_size, num_heads, dropout_rate, qkv_bias, save_attn)
        self.norm2 = nn.LayerNorm(hidden_size)

    def forward(self, x):
        x = x + self.attn(self.norm1(x))
        x = x + self.mlp(self.norm2(x))
        return x