ImageBind/imagebind/models/transformer.py

#!/usr/bin/env python3
# Portions Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.

# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.

# Code modified from
# https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py ;
# https://github.com/facebookresearch/deit/blob/main/models.py
# and https://github.com/facebookresearch/vissl/blob/main/vissl/models/trunks/vision_transformer.py


from functools import partial
from inspect import signature
from typing import Callable, List, Optional, Sequence

import torch
import torch.nn as nn
import torch.utils.checkpoint as checkpoint
from timm.layers import DropPath, trunc_normal_


class Attention(nn.Module):
    def __init__(
        self,
        dim,
        num_heads=8,
        qkv_bias=False,
        qk_scale=None,
        attn_drop=0.0,
        proj_drop=0.0,
    ):
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        # NOTE scale factor was wrong in my original version,
        # can set manually to be compat with prev weights
        self.scale = qk_scale or head_dim**-0.5

        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

    def forward(self, x):
        B, N, C = x.shape
        qkv = (
            self.qkv(x)
            .reshape(B, N, 3, self.num_heads, C // self.num_heads)
            .permute(2, 0, 3, 1, 4)
        )
        q, k, v = (
            qkv[0],
            qkv[1],
            qkv[2],
        )  # make torchscript happy (cannot use tensor as tuple)

        attn = (q @ k.transpose(-2, -1)) * self.scale
        attn = attn.softmax(dim=-1)
        attn = self.attn_drop(attn)

        x = (attn @ v).transpose(1, 2).reshape(B, N, C)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x


class Mlp(nn.Module):
    def __init__(
        self,
        in_features,
        hidden_features=None,
        out_features=None,
        act_layer=nn.GELU,
        drop=0.0,
    ):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.drop = nn.Dropout(drop)

    def forward(self, x):
        x = self.fc1(x)
        x = self.act(x)
        x = self.drop(x)
        x = self.fc2(x)
        x = self.drop(x)
        return x


class MultiheadAttention(nn.MultiheadAttention):
    """
    Same as ``nn.MultiheadAttention`` for ImageBind trunks, with an optional capture path
    so attention weights can be read for visualization without duplicating the forward pass.

    Call ``set_attention_capture(True)`` before ``forward``; averaged weights are stored in
    ``last_attn_weights`` (shape ``(B, L, L)``) when capture is enabled.
    """

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self._capture_attention = False
        self.last_attn_weights: Optional[torch.Tensor] = None

    def set_attention_capture(self, enabled: bool) -> None:
        self._capture_attention = bool(enabled)
        if not enabled:
            self.last_attn_weights = None

    def forward(self, x: torch.Tensor, attn_mask: torch.Tensor):
        if not self._capture_attention:
            return super().forward(x, x, x, need_weights=False, attn_mask=attn_mask)[0]
        kwargs = dict(
            key_padding_mask=None,
            need_weights=True,
            attn_mask=attn_mask,
            average_attn_weights=True,
        )
        if "is_causal" in signature(super().forward).parameters:
            kwargs["is_causal"] = False
        out, attn_weights = super().forward(x, x, x, **kwargs)
        self.last_attn_weights = attn_weights
        return out


class ViTAttention(Attention):
    def forward(self, x: torch.Tensor, attn_mask: torch.Tensor):
        assert attn_mask is None
        return super().forward(x)


class BlockWithMasking(nn.Module):
    def __init__(
        self,
        dim: int,
        attn_target: Callable,
        mlp_ratio: int = 4,
        act_layer: Callable = nn.GELU,
        norm_layer: Callable = nn.LayerNorm,
        ffn_dropout_rate: float = 0.0,
        drop_path: float = 0.0,
        layer_scale_type: Optional[str] = None,
        layer_scale_init_value: float = 1e-4,
    ):
        super().__init__()

        assert not isinstance(
            attn_target, nn.Module
        ), "attn_target should be a Callable. Otherwise attn_target is shared across blocks!"
        self.attn = attn_target()
        if drop_path > 0.0:
            self.drop_path = DropPath(drop_path)
        else:
            self.drop_path = nn.Identity()
        self.norm_1 = norm_layer(dim)
        mlp_hidden_dim = int(mlp_ratio * dim)
        self.mlp = Mlp(
            in_features=dim,
            hidden_features=mlp_hidden_dim,
            act_layer=act_layer,
            drop=ffn_dropout_rate,
        )
        self.norm_2 = norm_layer(dim)
        self.layer_scale_type = layer_scale_type
        if self.layer_scale_type is not None:
            assert self.layer_scale_type in [
                "per_channel",
                "scalar",
            ], f"Found Layer scale type {self.layer_scale_type}"
            if self.layer_scale_type == "per_channel":
                # one gamma value per channel
                gamma_shape = [1, 1, dim]
            elif self.layer_scale_type == "scalar":
                # single gamma value for all channels
                gamma_shape = [1, 1, 1]
            # two gammas: for each part of the fwd in the encoder
            self.layer_scale_gamma1 = nn.Parameter(
                torch.ones(size=gamma_shape) * layer_scale_init_value,
                requires_grad=True,
            )
            self.layer_scale_gamma2 = nn.Parameter(
                torch.ones(size=gamma_shape) * layer_scale_init_value,
                requires_grad=True,
            )

    def forward(self, x: torch.Tensor, attn_mask: torch.Tensor):
        if self.layer_scale_type is None:
            x = x + self.drop_path(self.attn(self.norm_1(x), attn_mask))
            x = x + self.drop_path(self.mlp(self.norm_2(x)))
        else:
            x = (
                x
                + self.drop_path(self.attn(self.norm_1(x), attn_mask))
                * self.layer_scale_gamma1
            )
            x = x + self.drop_path(self.mlp(self.norm_2(x))) * self.layer_scale_gamma2
        return x


_LAYER_NORM = partial(nn.LayerNorm, eps=1e-6)


class SimpleTransformer(nn.Module):
    def __init__(
        self,
        attn_target: Callable,
        embed_dim: int,
        num_blocks: int,
        block: Callable = BlockWithMasking,
        pre_transformer_layer: Optional[Callable] = None,
        post_transformer_layer: Optional[Callable] = None,
        drop_path_rate: float = 0.0,
        drop_path_type: str = "progressive",
        norm_layer: Callable = _LAYER_NORM,
        mlp_ratio: int = 4,
        ffn_dropout_rate: float = 0.0,
        layer_scale_type: Optional[str] = None,  # from cait; possible values are None, "per_channel", "scalar"
        layer_scale_init_value: float = 1e-4,  # from cait; float
        weight_init_style: str = "jax",  # possible values jax or pytorch
    ):
        """
        Simple Transformer with the following features
        1. Supports masked attention
        2. Supports DropPath
        3. Supports LayerScale
        4. Supports Dropout in Attention and FFN
        5. Makes few assumptions about the input except that it is a Tensor
        """
        super().__init__()
        self.pre_transformer_layer = pre_transformer_layer
        if drop_path_type == "progressive":
            dpr = [x.item() for x in torch.linspace(0, drop_path_rate, num_blocks)]
        elif drop_path_type == "uniform":
            dpr = [drop_path_rate for i in range(num_blocks)]
        else:
            raise ValueError(f"Unknown drop_path_type: {drop_path_type}")

        self.blocks = nn.Sequential(
            *[
                block(
                    dim=embed_dim,
                    attn_target=attn_target,
                    mlp_ratio=mlp_ratio,
                    ffn_dropout_rate=ffn_dropout_rate,
                    drop_path=dpr[i],
                    norm_layer=norm_layer,
                    layer_scale_type=layer_scale_type,
                    layer_scale_init_value=layer_scale_init_value,
                )
                for i in range(num_blocks)
            ]
        )
        self.post_transformer_layer = post_transformer_layer
        self.weight_init_style = weight_init_style
        self.apply(self._init_weights)

    def _init_weights(self, m):
        if isinstance(m, nn.Linear):
            if self.weight_init_style == "jax":
                # Based on MAE and official Jax ViT implementation
                torch.nn.init.xavier_uniform_(m.weight)
            elif self.weight_init_style == "pytorch":
                # PyTorch ViT uses trunc_normal_
                trunc_normal_(m.weight, std=0.02)

            if m.bias is not None:
                nn.init.constant_(m.bias, 0)
        elif isinstance(m, (nn.LayerNorm)):
            nn.init.constant_(m.bias, 0)
            nn.init.constant_(m.weight, 1.0)

    def set_attention_capture(
        self,
        enabled: bool,
        block_indices: Optional[Sequence[int]] = None,
    ) -> None:
        """
        Enable or disable attention weight capture on ``MultiheadAttention`` blocks.

        When ``enabled`` is True and ``block_indices`` is None, only the **last** block
        records attention (typical for CLS–patch maps). Pass an explicit index sequence
        (e.g. ``range(num_blocks)``) to capture specific layers.
        """
        n = len(self.blocks)
        for i, blk in enumerate(self.blocks):
            if not isinstance(blk.attn, MultiheadAttention):
                continue
            if not enabled:
                blk.attn.set_attention_capture(False)
                continue
            if block_indices is None:
                capture = i == n - 1
            else:
                capture = i in block_indices
            blk.attn.set_attention_capture(capture)

    def forward(
        self,
        tokens: torch.Tensor,
        attn_mask: torch.Tensor = None,
        use_checkpoint: bool = False,
        checkpoint_every_n: int = 1,
        checkpoint_blk_ids: Optional[List[int]] = None,
    ):
        """
        Inputs
        - tokens: data of shape N x L x D (or L x N x D depending on the attention implementation)
        - attn: mask of shape L x L

        Output
        - x: data of shape N x L x D (or L x N x D depending on the attention implementation)
        """
        if self.pre_transformer_layer:
            tokens = self.pre_transformer_layer(tokens)
        if use_checkpoint and checkpoint_blk_ids is None:
            checkpoint_blk_ids = [
                blk_id
                for blk_id in range(len(self.blocks))
                if blk_id % checkpoint_every_n == 0
            ]
        if checkpoint_blk_ids:
            checkpoint_blk_ids = set(checkpoint_blk_ids)
        for blk_id, blk in enumerate(self.blocks):
            if use_checkpoint and blk_id in checkpoint_blk_ids:
                tokens = checkpoint.checkpoint(
                    blk, tokens, attn_mask, use_reentrant=False
                )
            else:
                tokens = blk(tokens, attn_mask=attn_mask)
        if self.post_transformer_layer:
            tokens = self.post_transformer_layer(tokens)
        return tokens