src/module/unet/unet_2d_blocks.py

# Adapted from https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_blocks.py

import torch
from torch import nn

from .attention_2d import Transformer2DModel
from .resnet_2d import Downsample2D, ResnetBlock2D, Upsample2D


def get_down_block(
    down_block_type,
    num_layers,
    in_channels,
    out_channels,
    temb_channels,
    add_downsample,
    resnet_eps,
    resnet_act_fn,
    attn_num_head_channels,
    resnet_groups=None,
    cross_attention_dim=None,
    downsample_padding=None,
    dual_cross_attention=False,
    use_linear_projection=False,
    only_cross_attention=False,
    upcast_attention=False,
    resnet_time_scale_shift="default",
    use_sc_attn=False,
    use_st_attn=False,
    layer_id = 0,
):
    down_block_type = down_block_type[7:] if down_block_type.startswith("UNetRes") else down_block_type
    if down_block_type == "DownBlock2D":
        return DownBlock2D(
            num_layers=num_layers,
            in_channels=in_channels,
            out_channels=out_channels,
            temb_channels=temb_channels,
            add_downsample=add_downsample,
            resnet_eps=resnet_eps,
            resnet_act_fn=resnet_act_fn,
            resnet_groups=resnet_groups,
            downsample_padding=downsample_padding,
            resnet_time_scale_shift=resnet_time_scale_shift,
        )
    elif down_block_type == "CrossAttnDownBlock2D":
        if cross_attention_dim is None:
            raise ValueError("cross_attention_dim must be specified for CrossAttnDownBlock2D")
        return CrossAttnDownBlock2D(
            num_layers=num_layers,
            in_channels=in_channels,
            out_channels=out_channels,
            temb_channels=temb_channels,
            add_downsample=add_downsample,
            resnet_eps=resnet_eps,
            resnet_act_fn=resnet_act_fn,
            resnet_groups=resnet_groups,
            downsample_padding=downsample_padding,
            cross_attention_dim=cross_attention_dim,
            attn_num_head_channels=attn_num_head_channels,
            dual_cross_attention=dual_cross_attention,
            use_linear_projection=use_linear_projection,
            only_cross_attention=only_cross_attention,
            upcast_attention=upcast_attention,
            resnet_time_scale_shift=resnet_time_scale_shift,
            use_sc_attn=use_sc_attn,
            use_st_attn=use_st_attn,
            layer_id=layer_id
        )
    raise ValueError(f"{down_block_type} does not exist.")


def get_up_block(
    up_block_type,
    num_layers,
    in_channels,
    out_channels,
    prev_output_channel,
    temb_channels,
    add_upsample,
    resnet_eps,
    resnet_act_fn,
    attn_num_head_channels,
    resnet_groups=None,
    cross_attention_dim=None,
    dual_cross_attention=False,
    use_linear_projection=False,
    only_cross_attention=False,
    upcast_attention=False,
    resnet_time_scale_shift="default",
    use_sc_attn=False,
    use_st_attn=False,
    layer_id=0
):
    up_block_type = up_block_type[7:] if up_block_type.startswith("UNetRes") else up_block_type
    if up_block_type == "UpBlock2D":
        return UpBlock2D(
            num_layers=num_layers,
            in_channels=in_channels,
            out_channels=out_channels,
            prev_output_channel=prev_output_channel,
            temb_channels=temb_channels,
            add_upsample=add_upsample,
            resnet_eps=resnet_eps,
            resnet_act_fn=resnet_act_fn,
            resnet_groups=resnet_groups,
            resnet_time_scale_shift=resnet_time_scale_shift,
        )
    elif up_block_type == "CrossAttnUpBlock2D":
        if cross_attention_dim is None:
            raise ValueError("cross_attention_dim must be specified for CrossAttnUpBlock2D")
        return CrossAttnUpBlock2D(
            num_layers=num_layers,
            in_channels=in_channels,
            out_channels=out_channels,
            prev_output_channel=prev_output_channel,
            temb_channels=temb_channels,
            add_upsample=add_upsample,
            resnet_eps=resnet_eps,
            resnet_act_fn=resnet_act_fn,
            resnet_groups=resnet_groups,
            cross_attention_dim=cross_attention_dim,
            attn_num_head_channels=attn_num_head_channels,
            dual_cross_attention=dual_cross_attention,
            use_linear_projection=use_linear_projection,
            only_cross_attention=only_cross_attention,
            upcast_attention=upcast_attention,
            resnet_time_scale_shift=resnet_time_scale_shift,
            use_sc_attn=use_sc_attn,
            use_st_attn=use_st_attn,
            layer_id=layer_id,
        )
    raise ValueError(f"{up_block_type} does not exist.")


class UNetMidBlock2DCrossAttn(nn.Module):
    def __init__(
        self,
        in_channels: int,
        temb_channels: int,
        dropout: float = 0.0,
        num_layers: int = 1,
        resnet_eps: float = 1e-6,
        resnet_time_scale_shift: str = "default",
        resnet_act_fn: str = "swish",
        resnet_groups: int = 32,
        resnet_pre_norm: bool = True,
        attn_num_head_channels=1,
        output_scale_factor=1.0,
        cross_attention_dim=1280,
        dual_cross_attention=False,
        use_linear_projection=False,
        upcast_attention=False,
        use_sc_attn=False,
        use_st_attn=False,
    ):
        super().__init__()

        self.has_cross_attention = True
        self.attn_num_head_channels = attn_num_head_channels
        resnet_groups = resnet_groups if resnet_groups is not None else min(in_channels // 4, 32)

        # there is always at least one resnet
        resnets = [
            ResnetBlock2D(
                in_channels=in_channels,
                out_channels=in_channels,
                temb_channels=temb_channels,
                eps=resnet_eps,
                groups=resnet_groups,
                dropout=dropout,
                time_embedding_norm=resnet_time_scale_shift,
                non_linearity=resnet_act_fn,
                output_scale_factor=output_scale_factor,
                pre_norm=resnet_pre_norm,
            )
        ]
        attentions = []

        for _ in range(num_layers):
            if dual_cross_attention:
                raise NotImplementedError
            attentions.append(
                Transformer2DModel(
                    attn_num_head_channels,
                    in_channels // attn_num_head_channels,
                    in_channels=in_channels,
                    num_layers=1,
                    cross_attention_dim=cross_attention_dim,
                    norm_num_groups=resnet_groups,
                    use_linear_projection=use_linear_projection,
                    upcast_attention=upcast_attention,
                    use_sc_attn=use_sc_attn,
                    use_st_attn=True if (use_st_attn and _ == 0) else False,
                    updown='mid'
                )
            )
            resnets.append(
                ResnetBlock2D(
                    in_channels=in_channels,
                    out_channels=in_channels,
                    temb_channels=temb_channels,
                    eps=resnet_eps,
                    groups=resnet_groups,
                    dropout=dropout,
                    time_embedding_norm=resnet_time_scale_shift,
                    non_linearity=resnet_act_fn,
                    output_scale_factor=output_scale_factor,
                    pre_norm=resnet_pre_norm,
                )
            )

        self.attentions = nn.ModuleList(attentions)
        self.resnets = nn.ModuleList(resnets)
        self.layer_id=0

    def forward(
        self, 
        hidden_states, 
        temb=None, 
        encoder_hidden_states=None, 
        attention_mask=None,
        encoder_attention_mask=None,
        iter_cur=0, 
        save_kv=True,
        mode='drag',
        mask=None
    ):
        hidden_states = self.resnets[0](hidden_states, temb)
        for attn, resnet in zip(self.attentions, self.resnets[1:]):
            hidden_states = attn(
                hidden_states, 
                encoder_hidden_states=encoder_hidden_states,
                encoder_attention_mask=encoder_attention_mask,
                iter_cur=iter_cur, 
                save_kv=save_kv, 
                mode=mode,
                mask=mask
            ).sample
            hidden_states = resnet(hidden_states, temb)
        return hidden_states


class CrossAttnDownBlock2D(nn.Module):
    def __init__(
        self,
        in_channels: int,
        out_channels: int,
        temb_channels: int,
        dropout: float = 0.0,
        num_layers: int = 1,
        resnet_eps: float = 1e-6,
        resnet_time_scale_shift: str = "default",
        resnet_act_fn: str = "swish",
        resnet_groups: int = 32,
        resnet_pre_norm: bool = True,
        attn_num_head_channels=1,
        cross_attention_dim=1280,
        output_scale_factor=1.0,
        downsample_padding=1,
        add_downsample=True,
        dual_cross_attention=False,
        use_linear_projection=False,
        only_cross_attention=False,
        upcast_attention=False,
        use_sc_attn=False,
        use_st_attn=False,
        layer_id=0
    ):
        super().__init__()
        resnets = []
        attentions = []

        self.has_cross_attention = True
        self.attn_num_head_channels = attn_num_head_channels

        for i in range(num_layers):
            in_channels = in_channels if i == 0 else out_channels
            resnets.append(
                ResnetBlock2D(
                    in_channels=in_channels,
                    out_channels=out_channels,
                    temb_channels=temb_channels,
                    eps=resnet_eps,
                    groups=resnet_groups,
                    dropout=dropout,
                    time_embedding_norm=resnet_time_scale_shift,
                    non_linearity=resnet_act_fn,
                    output_scale_factor=output_scale_factor,
                    pre_norm=resnet_pre_norm,
                )
            )
            if dual_cross_attention:
                raise NotImplementedError
            attentions.append(
                Transformer2DModel(
                    attn_num_head_channels,
                    out_channels // attn_num_head_channels,
                    in_channels=out_channels,
                    num_layers=1,
                    cross_attention_dim=cross_attention_dim,
                    norm_num_groups=resnet_groups,
                    use_linear_projection=use_linear_projection,
                    only_cross_attention=only_cross_attention,
                    upcast_attention=upcast_attention,
                    use_sc_attn=use_sc_attn,
                    use_st_attn=True if (use_st_attn and i == 0) else False,
                    updown = 'down',
                    layer_id=layer_id
                )
            )
        self.attentions = nn.ModuleList(attentions)
        self.resnets = nn.ModuleList(resnets)

        if add_downsample:
            self.downsamplers = nn.ModuleList(
                [
                    Downsample2D(
                        out_channels, use_conv=True, out_channels=out_channels, padding=downsample_padding, name="op"
                    )
                ]
            )
        else:
            self.downsamplers = None

        self.gradient_checkpointing = False

    def forward(
        self, 
        hidden_states, 
        temb=None, 
        encoder_hidden_states=None, 
        attention_mask=None,
        encoder_attention_mask=None,
        iter_cur=0, 
        save_kv=True, 
        mode='drag',
        mask=None
    ): # FIXMIE: attn mask
        output_states = ()

        for resnet, attn in zip(self.resnets, self.attentions):
            if self.training and self.gradient_checkpointing:

                def create_custom_forward(module, return_dict=None):
                    def custom_forward(*inputs):
                        if return_dict is not None:
                            return module(*inputs, return_dict=return_dict, iter_cur=iter_cur)
                        else:
                            return module(*inputs)

                    return custom_forward

                hidden_states = torch.utils.checkpoint.checkpoint(create_custom_forward(resnet), hidden_states, temb)
                hidden_states = torch.utils.checkpoint.checkpoint(
                    create_custom_forward(attn, return_dict=False, iter_cur=iter_cur),
                    hidden_states,
                    encoder_hidden_states,
                    encoder_attention_mask,
                )[0]
            else:
                hidden_states = resnet(hidden_states, temb)
                hidden_states = attn(
                    hidden_states, 
                    encoder_hidden_states=encoder_hidden_states,
                    encoder_attention_mask=encoder_attention_mask,
                    iter_cur=iter_cur, 
                    save_kv=save_kv, 
                    mode=mode,
                    mask=mask
                ).sample

            output_states += (hidden_states,)

        if self.downsamplers is not None:
            for downsampler in self.downsamplers:
                hidden_states = downsampler(hidden_states)

            output_states += (hidden_states,) 

        return hidden_states, output_states


class DownBlock2D(nn.Module):
    def __init__(
        self,
        in_channels: int,
        out_channels: int,
        temb_channels: int,
        dropout: float = 0.0,
        num_layers: int = 1,
        resnet_eps: float = 1e-6,
        resnet_time_scale_shift: str = "default",
        resnet_act_fn: str = "swish",
        resnet_groups: int = 32,
        resnet_pre_norm: bool = True,
        output_scale_factor=1.0,
        add_downsample=True,
        downsample_padding=1,
    ):
        super().__init__()
        resnets = []

        for i in range(num_layers):
            in_channels = in_channels if i == 0 else out_channels
            resnets.append(
                ResnetBlock2D(
                    in_channels=in_channels,
                    out_channels=out_channels,
                    temb_channels=temb_channels,
                    eps=resnet_eps,
                    groups=resnet_groups,
                    dropout=dropout,
                    time_embedding_norm=resnet_time_scale_shift,
                    non_linearity=resnet_act_fn,
                    output_scale_factor=output_scale_factor,
                    pre_norm=resnet_pre_norm,
                )
            )

        self.resnets = nn.ModuleList(resnets)

        if add_downsample:
            self.downsamplers = nn.ModuleList(
                [
                    Downsample2D(
                        out_channels, use_conv=True, out_channels=out_channels, padding=downsample_padding, name="op"
                    )
                ]
            )
        else:
            self.downsamplers = None

        self.gradient_checkpointing = False

    def forward(self, hidden_states, temb=None):
        output_states = ()

        for resnet in self.resnets:
            if self.training and self.gradient_checkpointing:

                def create_custom_forward(module):
                    def custom_forward(*inputs):
                        return module(*inputs)

                    return custom_forward

                hidden_states = torch.utils.checkpoint.checkpoint(create_custom_forward(resnet), hidden_states, temb)
            else:
                hidden_states = resnet(hidden_states, temb)

            output_states += (hidden_states,)

        if self.downsamplers is not None:
            for downsampler in self.downsamplers:
                hidden_states = downsampler(hidden_states)

            output_states += (hidden_states,)

        return hidden_states, output_states

from torch.fft import fftn, fftshift, ifftn, ifftshift

def fourier_filter(x_in: torch.Tensor, threshold: int, scale: int) -> torch.Tensor:
    """Fourier filter as introduced in FreeU (https://arxiv.org/abs/2309.11497).

    This version of the method comes from here:
    https://github.com/huggingface/diffusers/pull/5164#issuecomment-1732638706
    """
    x = x_in[:,:,0]
    B, C, H, W = x.shape

    # Non-power of 2 images must be float32
    if (W & (W - 1)) != 0 or (H & (H - 1)) != 0:
        x = x.to(dtype=torch.float32)

    # FFT
    x_freq = fftn(x, dim=(-2, -1))
    x_freq = fftshift(x_freq, dim=(-2, -1))

    B, C, H, W = x_freq.shape
    mask = torch.ones((B, C, H, W), device=x.device)

    crow, ccol = H // 2, W // 2
    mask[..., crow - threshold : crow + threshold, ccol - threshold : ccol + threshold] = scale
    x_freq = x_freq * mask

    # IFFT
    x_freq = ifftshift(x_freq, dim=(-2, -1))
    x_filtered = ifftn(x_freq, dim=(-2, -1)).real
    x_filtered = x_filtered.unsqueeze(2)

    return x_filtered.to(dtype=x_in.dtype)


def apply_freeu(
    resolution_idx: int, hidden_states: torch.Tensor, res_hidden_states: torch.Tensor, **freeu_kwargs
):
    """Applies the FreeU mechanism as introduced in https:
    //arxiv.org/abs/2309.11497. Adapted from the official code repository: https://github.com/ChenyangSi/FreeU.

    Args:
        resolution_idx (`int`): Integer denoting the UNet block where FreeU is being applied.
        hidden_states (`torch.Tensor`): Inputs to the underlying block.
        res_hidden_states (`torch.Tensor`): Features from the skip block corresponding to the underlying block.
        s1 (`float`): Scaling factor for stage 1 to attenuate the contributions of the skip features.
        s2 (`float`): Scaling factor for stage 2 to attenuate the contributions of the skip features.
        b1 (`float`): Scaling factor for stage 1 to amplify the contributions of backbone features.
        b2 (`float`): Scaling factor for stage 2 to amplify the contributions of backbone features.
    """
    if resolution_idx == 0:
        num_half_channels = hidden_states.shape[1] // 2
        hidden_states[:, :num_half_channels] = hidden_states[:, :num_half_channels] * freeu_kwargs["b1"]
        res_hidden_states = fourier_filter(res_hidden_states, threshold=1, scale=freeu_kwargs["s1"])
    if resolution_idx == 1:
        num_half_channels = hidden_states.shape[1] // 2
        hidden_states[:, :num_half_channels] = hidden_states[:, :num_half_channels] * freeu_kwargs["b2"]
        res_hidden_states = fourier_filter(res_hidden_states, threshold=1, scale=freeu_kwargs["s2"])

    return hidden_states, res_hidden_states

class CrossAttnUpBlock2D(nn.Module):
    def __init__(
        self,
        in_channels: int,
        out_channels: int,
        prev_output_channel: int,
        temb_channels: int,
        dropout: float = 0.0,
        num_layers: int = 1,
        resnet_eps: float = 1e-6,
        resnet_time_scale_shift: str = "default",
        resnet_act_fn: str = "swish",
        resnet_groups: int = 32,
        resnet_pre_norm: bool = True,
        attn_num_head_channels=1,
        cross_attention_dim=1280,
        output_scale_factor=1.0,
        add_upsample=True,
        dual_cross_attention=False,
        use_linear_projection=False,
        only_cross_attention=False,
        upcast_attention=False,
        use_sc_attn=False,
        use_st_attn=False,
        layer_id=0
    ):
        super().__init__()
        resnets = []
        attentions = []

        self.has_cross_attention = True
        self.attn_num_head_channels = attn_num_head_channels

        for i in range(num_layers):
            res_skip_channels = in_channels if (i == num_layers - 1) else out_channels
            resnet_in_channels = prev_output_channel if i == 0 else out_channels

            resnets.append(
                ResnetBlock2D(
                    in_channels=resnet_in_channels + res_skip_channels,
                    out_channels=out_channels,
                    temb_channels=temb_channels,
                    eps=resnet_eps,
                    groups=resnet_groups,
                    dropout=dropout,
                    time_embedding_norm=resnet_time_scale_shift,
                    non_linearity=resnet_act_fn,
                    output_scale_factor=output_scale_factor,
                    pre_norm=resnet_pre_norm,
                )
            )
            if dual_cross_attention:
                raise NotImplementedError
            attentions.append(
                Transformer2DModel(
                    attn_num_head_channels,
                    out_channels // attn_num_head_channels,
                    in_channels=out_channels,
                    num_layers=1,
                    cross_attention_dim=cross_attention_dim,
                    norm_num_groups=resnet_groups,
                    use_linear_projection=use_linear_projection,
                    only_cross_attention=only_cross_attention,
                    upcast_attention=upcast_attention,
                    use_sc_attn=use_sc_attn,
                    use_st_attn=True if (use_st_attn and i == 0) else False,
                    updown = 'up',
                    layer_id=layer_id
                )
            )

        self.attentions = nn.ModuleList(attentions)
        self.resnets = nn.ModuleList(resnets)

        if add_upsample:
            self.upsamplers = nn.ModuleList([Upsample2D(out_channels, use_conv=True, out_channels=out_channels)])
        else:
            self.upsamplers = None

        self.gradient_checkpointing = False

    def forward(
        self,
        hidden_states,
        res_hidden_states_tuple,
        temb=None,
        encoder_hidden_states=None,
        upsample_size=None,
        attention_mask=None,
        encoder_attention_mask=None,
        iter_cur=0,
        save_kv=True, 
        mode='drag',
        mask=None
    ):
        for resnet, attn in zip(self.resnets, self.attentions):
            # pop res hidden states
            res_hidden_states = res_hidden_states_tuple[-1]
            res_hidden_states_tuple = res_hidden_states_tuple[:-1]

            # # --------------- FreeU code -----------------------
            # if save_kv == False:
            #     self.b1 = 1.1
            #     self.b2 = 1.2
            #     self.s1 = 0.6
            #     self.s2 = 0.4
            #     # Only operate on the first two stages
            #     hidden_states, res_hidden_states = apply_freeu(
            #             self.resolution_idx,
            #             hidden_states,
            #             res_hidden_states,
            #             s1=self.s1,
            #             s2=self.s2,
            #             b1=self.b1,
            #             b2=self.b2,
            #         )
            # # ---------------------------------------------------------

            hidden_states = torch.cat([hidden_states, res_hidden_states], dim=1)

            if self.training and self.gradient_checkpointing:
                def create_custom_forward(module, return_dict=None, iter_cur=iter_cur):
                    def custom_forward(*inputs):
                        if return_dict is not None:
                            return module(*inputs, return_dict=return_dict, iter_cur=iter_cur)
                        else:
                            return module(*inputs)

                    return custom_forward

                hidden_states = torch.utils.checkpoint.checkpoint(create_custom_forward(resnet), hidden_states, temb)
                hidden_states = torch.utils.checkpoint.checkpoint(
                    create_custom_forward(attn, return_dict=False, iter_cur=iter_cur),
                    hidden_states,
                    encoder_hidden_states,
                    encoder_attention_mask,
                )[0]
            else:
                hidden_states = resnet(hidden_states, temb)
                hidden_states = attn(
                    hidden_states, 
                    encoder_hidden_states=encoder_hidden_states,
                    encoder_attention_mask=encoder_attention_mask,
                    iter_cur=iter_cur, 
                    save_kv=save_kv, 
                    mode=mode,
                    mask=mask
                ).sample

        if self.upsamplers is not None:
            for upsampler in self.upsamplers:
                hidden_states = upsampler(hidden_states, upsample_size)

        return hidden_states


class UpBlock2D(nn.Module):
    def __init__(
        self,
        in_channels: int,
        prev_output_channel: int,
        out_channels: int,
        temb_channels: int,
        dropout: float = 0.0,
        num_layers: int = 1,
        resnet_eps: float = 1e-6,
        resnet_time_scale_shift: str = "default",
        resnet_act_fn: str = "swish",
        resnet_groups: int = 32,
        resnet_pre_norm: bool = True,
        output_scale_factor=1.0,
        add_upsample=True,
    ):
        super().__init__()
        resnets = []

        for i in range(num_layers):
            res_skip_channels = in_channels if (i == num_layers - 1) else out_channels
            resnet_in_channels = prev_output_channel if i == 0 else out_channels

            resnets.append(
                ResnetBlock2D(
                    in_channels=resnet_in_channels + res_skip_channels,
                    out_channels=out_channels,
                    temb_channels=temb_channels,
                    eps=resnet_eps,
                    groups=resnet_groups,
                    dropout=dropout,
                    time_embedding_norm=resnet_time_scale_shift,
                    non_linearity=resnet_act_fn,
                    output_scale_factor=output_scale_factor,
                    pre_norm=resnet_pre_norm,
                )
            )

        self.resnets = nn.ModuleList(resnets)

        if add_upsample:
            self.upsamplers = nn.ModuleList([Upsample2D(out_channels, use_conv=True, out_channels=out_channels)])
        else:
            self.upsamplers = None

        self.gradient_checkpointing = False

    def forward(self, hidden_states, res_hidden_states_tuple, temb=None, upsample_size=None):
        for resnet in self.resnets:
            # pop res hidden states
            res_hidden_states = res_hidden_states_tuple[-1]
            res_hidden_states_tuple = res_hidden_states_tuple[:-1]
            hidden_states = torch.cat([hidden_states, res_hidden_states], dim=1)

            if self.training and self.gradient_checkpointing:

                def create_custom_forward(module):
                    def custom_forward(*inputs):
                        return module(*inputs)

                    return custom_forward

                hidden_states = torch.utils.checkpoint.checkpoint(create_custom_forward(resnet), hidden_states, temb)
            else:
                hidden_states = resnet(hidden_states, temb)

        if self.upsamplers is not None:
            for upsampler in self.upsamplers:
                hidden_states = upsampler(hidden_states, upsample_size)

        return hidden_states