animatelcm/models/unet.py

# Adapted from https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/unet_2d_condition.py

from dataclasses import dataclass
from typing import List, Optional, Tuple, Union

import os
import json

import torch
import torch.nn as nn
import torch.utils.checkpoint

from diffusers.configuration_utils import ConfigMixin, register_to_config
from diffusers.modeling_utils import ModelMixin
from diffusers.utils import BaseOutput, logging
from animatelcm.models.embeddings import TimestepEmbedding, Timesteps
from .unet_blocks import (
    CrossAttnDownBlock3D,
    CrossAttnUpBlock3D,
    DownBlock3D,
    UNetMidBlock3DCrossAttn,
    UpBlock3D,
    get_down_block,
    get_up_block,
)
from .resnet import InflatedConv3d, InflatedGroupNorm
# from .adapter import Adapter, PixelAdapter # Not ready
from einops import repeat


logger = logging.get_logger(__name__)  # pylint: disable=invalid-name


@dataclass
class UNet3DConditionOutput(BaseOutput):
    sample: torch.FloatTensor


class UNet3DConditionModel(ModelMixin, ConfigMixin):
    _supports_gradient_checkpointing = True

    @register_to_config
    def __init__(
        self,
        sample_size: Optional[int] = None,
        in_channels: int = 4,
        out_channels: int = 4,
        center_input_sample: bool = False,
        flip_sin_to_cos: bool = True,
        freq_shift: int = 0,
        down_block_types: Tuple[str] = (
            "CrossAttnDownBlock3D",
            "CrossAttnDownBlock3D",
            "CrossAttnDownBlock3D",
            "DownBlock3D",
        ),
        mid_block_type: str = "UNetMidBlock3DCrossAttn",
        up_block_types: Tuple[str] = (
            "UpBlock3D",
            "CrossAttnUpBlock3D",
            "CrossAttnUpBlock3D",
            "CrossAttnUpBlock3D"
        ),
        only_cross_attention: Union[bool, Tuple[bool]] = False,
        block_out_channels: Tuple[int] = (320, 640, 1280, 1280),
        layers_per_block: int = 2,
        downsample_padding: int = 1,
        mid_block_scale_factor: float = 1,
        act_fn: str = "silu",
        norm_num_groups: int = 32,
        norm_eps: float = 1e-5,
        cross_attention_dim: int = 1280,
        attention_head_dim: Union[int, Tuple[int]] = 8,
        dual_cross_attention: bool = False,
        use_linear_projection: bool = False,
        class_embed_type: Optional[str] = None,
        num_class_embeds: Optional[int] = None,
        upcast_attention: bool = False,
        resnet_time_scale_shift: str = "default",

        use_inflated_groupnorm=False,

        # Additional
        use_motion_module=False,
        use_motion_resnet=False,
        motion_module_resolutions=(1, 2, 4, 8),
        motion_module_mid_block=False,
        motion_module_decoder_only=False,
        motion_module_type=None,
        motion_module_kwargs={},
        unet_use_cross_frame_attention=None,
        unet_use_temporal_attention=None,
        time_cond_proj_dim=None, # not ready
        use_img_encoder=False,
        use_pixel_encoder=False,
    ):
        super().__init__()

        self.sample_size = sample_size
        time_embed_dim = block_out_channels[0] * 4

        self.img_encoder = None if use_img_encoder else None # not ready
        self.pixel_encoder = None if use_pixel_encoder else None # not ready
        
        
        # input
        self.conv_in = InflatedConv3d(
            in_channels, block_out_channels[0], kernel_size=3, padding=(1, 1))

        # time
        self.time_proj = Timesteps(
            block_out_channels[0], flip_sin_to_cos, freq_shift)
        timestep_input_dim = block_out_channels[0]

        self.time_embedding = TimestepEmbedding(
            timestep_input_dim, time_embed_dim, time_cond_proj_dim=time_cond_proj_dim)

        # class embedding
        if class_embed_type is None and num_class_embeds is not None:
            self.class_embedding = nn.Embedding(
                num_class_embeds, time_embed_dim)
        elif class_embed_type == "timestep":
            self.class_embedding = TimestepEmbedding(
                timestep_input_dim, time_embed_dim)
        elif class_embed_type == "identity":
            self.class_embedding = nn.Identity(time_embed_dim, time_embed_dim)
        else:
            self.class_embedding = None

        self.down_blocks = nn.ModuleList([])
        self.mid_block = None
        self.up_blocks = nn.ModuleList([])

        if isinstance(only_cross_attention, bool):
            only_cross_attention = [
                only_cross_attention] * len(down_block_types)

        if isinstance(attention_head_dim, int):
            attention_head_dim = (attention_head_dim,) * len(down_block_types)

        # down
        output_channel = block_out_channels[0]
        for i, down_block_type in enumerate(down_block_types):
            res = 2 ** i
            input_channel = output_channel
            output_channel = block_out_channels[i]
            is_final_block = i == len(block_out_channels) - 1

            down_block = get_down_block(
                down_block_type,
                num_layers=layers_per_block,
                in_channels=input_channel,
                out_channels=output_channel,
                temb_channels=time_embed_dim,
                add_downsample=not is_final_block,
                resnet_eps=norm_eps,
                resnet_act_fn=act_fn,
                resnet_groups=norm_num_groups,
                cross_attention_dim=cross_attention_dim,
                attn_num_head_channels=attention_head_dim[i],
                downsample_padding=downsample_padding,
                dual_cross_attention=dual_cross_attention,
                use_linear_projection=use_linear_projection,
                only_cross_attention=only_cross_attention[i],
                upcast_attention=upcast_attention,
                resnet_time_scale_shift=resnet_time_scale_shift,

                unet_use_cross_frame_attention=unet_use_cross_frame_attention,
                unet_use_temporal_attention=unet_use_temporal_attention,
                use_inflated_groupnorm=use_inflated_groupnorm,

                use_motion_module=use_motion_module and (
                    res in motion_module_resolutions) and (not motion_module_decoder_only),
                use_motion_resnet=use_motion_resnet and (
                    res in motion_module_resolutions) and (not motion_module_decoder_only),
                motion_module_type=motion_module_type,
                motion_module_kwargs=motion_module_kwargs,
            )
            self.down_blocks.append(down_block)

        # mid
        if mid_block_type == "UNetMidBlock3DCrossAttn":
            self.mid_block = UNetMidBlock3DCrossAttn(
                in_channels=block_out_channels[-1],
                temb_channels=time_embed_dim,
                resnet_eps=norm_eps,
                resnet_act_fn=act_fn,
                output_scale_factor=mid_block_scale_factor,
                resnet_time_scale_shift=resnet_time_scale_shift,
                cross_attention_dim=cross_attention_dim,
                attn_num_head_channels=attention_head_dim[-1],
                resnet_groups=norm_num_groups,
                dual_cross_attention=dual_cross_attention,
                use_linear_projection=use_linear_projection,
                upcast_attention=upcast_attention,

                unet_use_cross_frame_attention=unet_use_cross_frame_attention,
                unet_use_temporal_attention=unet_use_temporal_attention,
                use_inflated_groupnorm=use_inflated_groupnorm,

                use_motion_module=use_motion_module and motion_module_mid_block,
                use_motion_resnet=use_motion_resnet and motion_module_mid_block,

                motion_module_type=motion_module_type,
                motion_module_kwargs=motion_module_kwargs,
            )
        else:
            raise ValueError(f"unknown mid_block_type : {mid_block_type}")

        # count how many layers upsample the videos
        self.num_upsamplers = 0

        # up
        reversed_block_out_channels = list(reversed(block_out_channels))
        reversed_attention_head_dim = list(reversed(attention_head_dim))
        only_cross_attention = list(reversed(only_cross_attention))
        output_channel = reversed_block_out_channels[0]
        for i, up_block_type in enumerate(up_block_types):
            res = 2 ** (3 - i)
            is_final_block = i == len(block_out_channels) - 1

            prev_output_channel = output_channel
            output_channel = reversed_block_out_channels[i]
            input_channel = reversed_block_out_channels[min(
                i + 1, len(block_out_channels) - 1)]

            # add upsample block for all BUT final layer
            if not is_final_block:
                add_upsample = True
                self.num_upsamplers += 1
            else:
                add_upsample = False

            up_block = get_up_block(
                up_block_type,
                num_layers=layers_per_block + 1,
                in_channels=input_channel,
                out_channels=output_channel,
                prev_output_channel=prev_output_channel,
                temb_channels=time_embed_dim,
                add_upsample=add_upsample,
                resnet_eps=norm_eps,
                resnet_act_fn=act_fn,
                resnet_groups=norm_num_groups,
                cross_attention_dim=cross_attention_dim,
                attn_num_head_channels=reversed_attention_head_dim[i],
                dual_cross_attention=dual_cross_attention,
                use_linear_projection=use_linear_projection,
                only_cross_attention=only_cross_attention[i],
                upcast_attention=upcast_attention,
                resnet_time_scale_shift=resnet_time_scale_shift,

                unet_use_cross_frame_attention=unet_use_cross_frame_attention,
                unet_use_temporal_attention=unet_use_temporal_attention,
                use_inflated_groupnorm=use_inflated_groupnorm,

                use_motion_module=use_motion_module and (
                    res in motion_module_resolutions),
                use_motion_resnet=use_motion_resnet and (
                    res in motion_module_resolutions),

                motion_module_type=motion_module_type,
                motion_module_kwargs=motion_module_kwargs,
            )
            self.up_blocks.append(up_block)
            prev_output_channel = output_channel

        # out
        if use_inflated_groupnorm:
            self.conv_norm_out = InflatedGroupNorm(
                num_channels=block_out_channels[0], num_groups=norm_num_groups, eps=norm_eps)
        else:
            self.conv_norm_out = nn.GroupNorm(
                num_channels=block_out_channels[0], num_groups=norm_num_groups, eps=norm_eps)
        self.conv_act = nn.SiLU()
        self.conv_out = InflatedConv3d(
            block_out_channels[0], out_channels, kernel_size=3, padding=1)

    def set_attention_slice(self, slice_size):
        r"""
        Enable sliced attention computation.

        When this option is enabled, the attention module will split the input tensor in slices, to compute attention
        in several steps. This is useful to save some memory in exchange for a small speed decrease.

        Args:
            slice_size (`str` or `int` or `list(int)`, *optional*, defaults to `"auto"`):
                When `"auto"`, halves the input to the attention heads, so attention will be computed in two steps. If
                `"max"`, maxium amount of memory will be saved by running only one slice at a time. If a number is
                provided, uses as many slices as `attention_head_dim // slice_size`. In this case, `attention_head_dim`
                must be a multiple of `slice_size`.
        """
        sliceable_head_dims = []

        def fn_recursive_retrieve_slicable_dims(module: torch.nn.Module):
            if hasattr(module, "set_attention_slice"):
                sliceable_head_dims.append(module.sliceable_head_dim)

            for child in module.children():
                fn_recursive_retrieve_slicable_dims(child)

        # retrieve number of attention layers
        for module in self.children():
            fn_recursive_retrieve_slicable_dims(module)

        num_slicable_layers = len(sliceable_head_dims)

        if slice_size == "auto":
            # half the attention head size is usually a good trade-off between
            # speed and memory
            slice_size = [dim // 2 for dim in sliceable_head_dims]
        elif slice_size == "max":
            # make smallest slice possible
            slice_size = num_slicable_layers * [1]

        slice_size = num_slicable_layers * \
            [slice_size] if not isinstance(slice_size, list) else slice_size

        if len(slice_size) != len(sliceable_head_dims):
            raise ValueError(
                f"You have provided {len(slice_size)}, but {self.config} has {len(sliceable_head_dims)} different"
                f" attention layers. Make sure to match `len(slice_size)` to be {len(sliceable_head_dims)}."
            )

        for i in range(len(slice_size)):
            size = slice_size[i]
            dim = sliceable_head_dims[i]
            if size is not None and size > dim:
                raise ValueError(
                    f"size {size} has to be smaller or equal to {dim}.")

        # Recursively walk through all the children.
        # Any children which exposes the set_attention_slice method
        # gets the message
        def fn_recursive_set_attention_slice(module: torch.nn.Module, slice_size: List[int]):
            if hasattr(module, "set_attention_slice"):
                module.set_attention_slice(slice_size.pop())

            for child in module.children():
                fn_recursive_set_attention_slice(child, slice_size)

        reversed_slice_size = list(reversed(slice_size))
        for module in self.children():
            fn_recursive_set_attention_slice(module, reversed_slice_size)

    def _set_gradient_checkpointing(self, module, value=False):
        if isinstance(module, (CrossAttnDownBlock3D, DownBlock3D, CrossAttnUpBlock3D, UpBlock3D)):
            module.gradient_checkpointing = value

    def forward(
        self,
        sample: torch.FloatTensor,
        timestep: Union[torch.Tensor, float, int],
        encoder_hidden_states: torch.Tensor,
        img_latent: torch.FloatTensor = None,
        control: torch.FloatTensor = None,
        time_cond: torch.FloatTensor = None,  # not ready
        class_labels: Optional[torch.Tensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        return_dict: bool = True,
    ) -> Union[UNet3DConditionOutput, Tuple]:
        r"""
        Args:
            sample (`torch.FloatTensor`): (batch, channel, height, width) noisy inputs tensor
            timestep (`torch.FloatTensor` or `float` or `int`): (batch) timesteps
            encoder_hidden_states (`torch.FloatTensor`): (batch, sequence_length, feature_dim) encoder hidden states
            return_dict (`bool`, *optional*, defaults to `True`):
                Whether or not to return a [`models.unet_2d_condition.UNet2DConditionOutput`] instead of a plain tuple.

        Returns:
            [`~models.unet_2d_condition.UNet2DConditionOutput`] or `tuple`:
            [`~models.unet_2d_condition.UNet2DConditionOutput`] if `return_dict` is True, otherwise a `tuple`. When
            returning a tuple, the first element is the sample tensor.
        """

        if img_latent is not None and self.img_encoder is not None:
            f = sample.shape[2]
            img_latent = repeat(img_latent, "b c  h w  -> b c f h w",
                                f=f) if img_latent.ndim == 4 else img_latent
            img_features = self.img_encoder(img_latent)
        else:
            img_features = None

        if control is not None and self.pixel_encoder is not None:
            ctrl_features = self.pixel_encoder(control)
        else:
            # assert 0
            ctrl_features = None

        # By default samples have to be AT least a multiple of the overall upsampling factor.
        # The overall upsampling factor is equal to 2 ** (# num of upsampling layears).
        # However, the upsampling interpolation output size can be forced to fit any upsampling size
        # on the fly if necessary.
        default_overall_up_factor = 2**self.num_upsamplers

        # upsample size should be forwarded when sample is not a multiple of `default_overall_up_factor`
        forward_upsample_size = False
        upsample_size = None

        if any(s % default_overall_up_factor != 0 for s in sample.shape[-2:]):
            logger.info(
                "Forward upsample size to force interpolation output size.")
            forward_upsample_size = True

        # prepare attention_mask
        if attention_mask is not None:
            attention_mask = (1 - attention_mask.to(sample.dtype)) * -10000.0
            attention_mask = attention_mask.unsqueeze(1)

        # center input if necessary
        if self.config.center_input_sample:
            sample = 2 * sample - 1.0

        # time
        timesteps = timestep
        if not torch.is_tensor(timesteps):
            # This would be a good case for the `match` statement (Python 3.10+)
            is_mps = sample.device.type == "mps"
            if isinstance(timestep, float):
                dtype = torch.float32 if is_mps else torch.float64
            else:
                dtype = torch.int32 if is_mps else torch.int64
            timesteps = torch.tensor(
                [timesteps], dtype=dtype, device=sample.device)
        elif len(timesteps.shape) == 0:
            timesteps = timesteps[None].to(sample.device)

        # broadcast to batch dimension in a way that's compatible with ONNX/Core ML
        timesteps = timesteps.expand(sample.shape[0])

        t_emb = self.time_proj(timesteps)

        # timesteps does not contain any weights and will always return f32 tensors
        # but time_embedding might actually be running in fp16. so we need to cast here.
        # there might be better ways to encapsulate this.
        t_emb = t_emb.to(dtype=self.dtype)

        emb = self.time_embedding(t_emb)

        if self.class_embedding is not None:
            if class_labels is None:
                raise ValueError(
                    "class_labels should be provided when num_class_embeds > 0")

            if self.config.class_embed_type == "timestep":
                class_labels = self.time_proj(class_labels)

            class_emb = self.class_embedding(class_labels).to(dtype=self.dtype)
            emb = emb + class_emb

        # pre-process
        sample = self.conv_in(sample)

        # down

        down_block_res_samples = (sample,)

        img_feature_idx = 0

        for downsample_block in self.down_blocks:

            added_feature = img_features[img_feature_idx] if img_features is not None else torch.tensor(
                0.).to(sample.device, sample.dtype)
            added_feature = added_feature + \
                ctrl_features[img_feature_idx] if ctrl_features is not None else added_feature
            added_feature = None if added_feature.abs().mean() == 0 else added_feature
            if hasattr(downsample_block, "has_cross_attention") and downsample_block.has_cross_attention:
                sample, res_samples = downsample_block(
                    hidden_states=sample,
                    temb=emb,
                    encoder_hidden_states=encoder_hidden_states,
                    attention_mask=attention_mask,
                    img_feature=added_feature
                )
            else:
                sample, res_samples = downsample_block(
                    hidden_states=sample, temb=emb, encoder_hidden_states=encoder_hidden_states, img_feature=added_feature)

            down_block_res_samples += res_samples
            img_feature_idx += 1
        # mid
        sample = self.mid_block(
            sample, emb, encoder_hidden_states=encoder_hidden_states, attention_mask=attention_mask
        )

        # up
        for i, upsample_block in enumerate(self.up_blocks):
            is_final_block = i == len(self.up_blocks) - 1

            res_samples = down_block_res_samples[-len(upsample_block.resnets):]
            down_block_res_samples = down_block_res_samples[: -len(
                upsample_block.resnets)]

            # if we have not reached the final block and need to forward the
            # upsample size, we do it here
            if not is_final_block and forward_upsample_size:
                upsample_size = down_block_res_samples[-1].shape[2:]

            if hasattr(upsample_block, "has_cross_attention") and upsample_block.has_cross_attention:
                sample = upsample_block(
                    hidden_states=sample,
                    temb=emb,
                    res_hidden_states_tuple=res_samples,
                    encoder_hidden_states=encoder_hidden_states,
                    upsample_size=upsample_size,
                    attention_mask=attention_mask,
                )
            else:
                sample = upsample_block(
                    hidden_states=sample, temb=emb, res_hidden_states_tuple=res_samples, upsample_size=upsample_size, encoder_hidden_states=encoder_hidden_states,
                )

        # post-process
        sample = self.conv_norm_out(sample)
        sample = self.conv_act(sample)
        sample = self.conv_out(sample)

        if not return_dict:
            return (sample,)

        return UNet3DConditionOutput(sample=sample)

    @classmethod
    def from_pretrained_2d(cls, pretrained_model_path, subfolder=None, unet_additional_kwargs=None):
        if subfolder is not None:
            pretrained_model_path = os.path.join(
                pretrained_model_path, subfolder)
        print(
            f"loaded temporal unet's pretrained weights from {pretrained_model_path} ...")

        config_file = os.path.join(pretrained_model_path, 'config.json')
        if not os.path.isfile(config_file):
            raise RuntimeError(f"{config_file} does not exist")
        with open(config_file, "r") as f:
            config = json.load(f)
        config["_class_name"] = cls.__name__
        config["down_block_types"] = [
            "CrossAttnDownBlock3D",
            "CrossAttnDownBlock3D",
            "CrossAttnDownBlock3D",
            "DownBlock3D"
        ]
        config["up_block_types"] = [
            "UpBlock3D",
            "CrossAttnUpBlock3D",
            "CrossAttnUpBlock3D",
            "CrossAttnUpBlock3D"
        ]

        from diffusers.utils import WEIGHTS_NAME
        model = cls.from_config(config, **unet_additional_kwargs)
        model_file = os.path.join(pretrained_model_path, WEIGHTS_NAME)
        if not os.path.isfile(model_file):
            raise RuntimeError(f"{model_file} does not exist")
        state_dict = torch.load(model_file, map_location="cpu")
        if "state_dict" in state_dict.keys():
            state_dict = state_dict["state_dict"]
            state_dict = {k.replace("module.", ""): v for k,
                          v in state_dict.items()}
        m, u = model.load_state_dict(state_dict, strict=False)
        print("###load unet weights")
        print(f"### missing keys: {len(m)}; \n### unexpected keys: {len(u)};")

        params = [p.numel() if "motion" in n else 0 for n,
                  p in model.named_parameters()]
        print(f"### Temporal Module Parameters: {sum(params) / 1e6} M")

        return model