_utils/attn_utils.py

import abc

import cv2
import numpy as np
import torch
from IPython.display import display
from PIL import Image
from typing import Union, Tuple, List
from einops import rearrange, repeat
import math
from torch import nn, einsum
from inspect import isfunction
from diffusers.utils import logging
try:
    from diffusers.models.unet_2d_condition import UNet2DConditionOutput
except:
    from diffusers.models.unets.unet_2d_condition import UNet2DConditionOutput

try:
    from diffusers.models.cross_attention import CrossAttention
except:
    from diffusers.models.attention_processor import Attention as CrossAttention

MAX_NUM_WORDS = 77
LOW_RESOURCE = False 

class CountingCrossAttnProcessor1:

    def __init__(self, attnstore, place_in_unet):
        super().__init__()
        self.attnstore = attnstore
        self.place_in_unet = place_in_unet

    def __call__(self, attn_layer: CrossAttention, hidden_states, encoder_hidden_states=None, attention_mask=None):
        batch_size, sequence_length, dim = hidden_states.shape
        h = attn_layer.heads
        q = attn_layer.to_q(hidden_states)
        is_cross = encoder_hidden_states is not None
        context = encoder_hidden_states if is_cross else hidden_states
        k = attn_layer.to_k(context)
        v = attn_layer.to_v(context)
        # q = attn_layer.reshape_heads_to_batch_dim(q)
        # k = attn_layer.reshape_heads_to_batch_dim(k)
        # v = attn_layer.reshape_heads_to_batch_dim(v)
        # q = attn_layer.head_to_batch_dim(q)
        # k = attn_layer.head_to_batch_dim(k)
        # v = attn_layer.head_to_batch_dim(v)
        q = self.head_to_batch_dim(q, h)
        k = self.head_to_batch_dim(k, h)
        v = self.head_to_batch_dim(v, h)

        sim = torch.einsum("b i d, b j d -> b i j", q, k) * attn_layer.scale

        if attention_mask is not None:
            attention_mask = attention_mask.reshape(batch_size, -1)
            max_neg_value = -torch.finfo(sim.dtype).max
            attention_mask = attention_mask[:, None, :].repeat(h, 1, 1)
            sim.masked_fill_(~attention_mask, max_neg_value)

        # attention, what we cannot get enough of
        attn_ = sim.softmax(dim=-1).clone()
        # softmax = nn.Softmax(dim=-1)
        # attn_ = softmax(sim)
        self.attnstore(attn_, is_cross, self.place_in_unet)
        out = torch.einsum("b i j, b j d -> b i d", attn_, v)
        # out = attn_layer.batch_to_head_dim(out)
        out = self.batch_to_head_dim(out, h)

        if type(attn_layer.to_out) is torch.nn.modules.container.ModuleList:
            to_out = attn_layer.to_out[0]
        else:
            to_out = attn_layer.to_out

        out = to_out(out)
        return out
    
    def batch_to_head_dim(self, tensor, head_size):
        # head_size = self.heads
        batch_size, seq_len, dim = tensor.shape
        tensor = tensor.reshape(batch_size // head_size, head_size, seq_len, dim)
        tensor = tensor.permute(0, 2, 1, 3).reshape(batch_size // head_size, seq_len, dim * head_size)
        return tensor

    def head_to_batch_dim(self, tensor, head_size, out_dim=3):
        # head_size = self.heads
        batch_size, seq_len, dim = tensor.shape
        tensor = tensor.reshape(batch_size, seq_len, head_size, dim // head_size)
        tensor = tensor.permute(0, 2, 1, 3)

        if out_dim == 3:
            tensor = tensor.reshape(batch_size * head_size, seq_len, dim // head_size)

        return tensor


def register_attention_control(model, controller):

    attn_procs = {}
    cross_att_count = 0
    for name in model.unet.attn_processors.keys():
        cross_attention_dim = None if name.endswith("attn1.processor") else model.unet.config.cross_attention_dim
        if name.startswith("mid_block"):
            hidden_size = model.unet.config.block_out_channels[-1]
            place_in_unet = "mid"
        elif name.startswith("up_blocks"):
            block_id = int(name[len("up_blocks.")])
            hidden_size = list(reversed(model.unet.config.block_out_channels))[block_id]
            place_in_unet = "up"
        elif name.startswith("down_blocks"):
            block_id = int(name[len("down_blocks.")])
            hidden_size = model.unet.config.block_out_channels[block_id]
            place_in_unet = "down"
        else:
            continue

        cross_att_count += 1
        # attn_procs[name] = AttendExciteCrossAttnProcessor(
        #     attnstore=controller, place_in_unet=place_in_unet
        # )
        attn_procs[name] = CountingCrossAttnProcessor1(
            attnstore=controller, place_in_unet=place_in_unet
        )

    model.unet.set_attn_processor(attn_procs)
    controller.num_att_layers = cross_att_count

def register_hier_output(model):
    self = model.unet
    from ldm.modules.diffusionmodules.util import checkpoint, timestep_embedding
    logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
    def forward(sample, timestep=None, encoder_hidden_states=None, class_labels=None, timestep_cond=None, 
                attention_mask=None, cross_attention_kwargs=None, added_cond_kwargs=None, down_block_additional_residuals=None,
                mid_block_additional_residual=None, encoder_attention_mask=None, return_dict=True):

        out_list = []

        
        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

        if attention_mask is not None:
            # assume that mask is expressed as:
            #   (1 = keep,      0 = discard)
            # convert mask into a bias that can be added to attention scores:
            #       (keep = +0,     discard = -10000.0)
            attention_mask = (1 - attention_mask.to(sample.dtype)) * -10000.0
            attention_mask = attention_mask.unsqueeze(1)

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

        if self.config.center_input_sample:
            sample = 2 * sample - 1.0

        timesteps = timestep
        if not torch.is_tensor(timesteps):
            # TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
            # 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)

        t_emb = t_emb.to(dtype=sample.dtype)

        emb = self.time_embedding(t_emb, timestep_cond)
        aug_emb = None

        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)

                # `Timesteps` does not contain any weights and will always return f32 tensors
                # there might be better ways to encapsulate this.
                class_labels = class_labels.to(dtype=sample.dtype)

            class_emb = self.class_embedding(class_labels).to(dtype=sample.dtype)

            if self.config.class_embeddings_concat:
                emb = torch.cat([emb, class_emb], dim=-1)
            else:
                emb = emb + class_emb

        if self.config.addition_embed_type == "text":
            aug_emb = self.add_embedding(encoder_hidden_states)
        elif self.config.addition_embed_type == "text_image":
            # Kandinsky 2.1 - style
            if "image_embeds" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `addition_embed_type` set to 'text_image' which requires the keyword argument `image_embeds` to be passed in `added_cond_kwargs`"
                )

            image_embs = added_cond_kwargs.get("image_embeds")
            text_embs = added_cond_kwargs.get("text_embeds", encoder_hidden_states)
            aug_emb = self.add_embedding(text_embs, image_embs)
        elif self.config.addition_embed_type == "text_time":
            # SDXL - style
            if "text_embeds" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `addition_embed_type` set to 'text_time' which requires the keyword argument `text_embeds` to be passed in `added_cond_kwargs`"
                )
            text_embeds = added_cond_kwargs.get("text_embeds")
            if "time_ids" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `addition_embed_type` set to 'text_time' which requires the keyword argument `time_ids` to be passed in `added_cond_kwargs`"
                )
            time_ids = added_cond_kwargs.get("time_ids")
            time_embeds = self.add_time_proj(time_ids.flatten())
            time_embeds = time_embeds.reshape((text_embeds.shape[0], -1))

            add_embeds = torch.concat([text_embeds, time_embeds], dim=-1)
            add_embeds = add_embeds.to(emb.dtype)
            aug_emb = self.add_embedding(add_embeds)
        elif self.config.addition_embed_type == "image":
            # Kandinsky 2.2 - style
            if "image_embeds" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `addition_embed_type` set to 'image' which requires the keyword argument `image_embeds` to be passed in `added_cond_kwargs`"
                )
            image_embs = added_cond_kwargs.get("image_embeds")
            aug_emb = self.add_embedding(image_embs)
        elif self.config.addition_embed_type == "image_hint":
            # Kandinsky 2.2 - style
            if "image_embeds" not in added_cond_kwargs or "hint" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `addition_embed_type` set to 'image_hint' which requires the keyword arguments `image_embeds` and `hint` to be passed in `added_cond_kwargs`"
                )
            image_embs = added_cond_kwargs.get("image_embeds")
            hint = added_cond_kwargs.get("hint")
            aug_emb, hint = self.add_embedding(image_embs, hint)
            sample = torch.cat([sample, hint], dim=1)

        emb = emb + aug_emb if aug_emb is not None else emb

        if self.time_embed_act is not None:
            emb = self.time_embed_act(emb)

        if self.encoder_hid_proj is not None and self.config.encoder_hid_dim_type == "text_proj":
            encoder_hidden_states = self.encoder_hid_proj(encoder_hidden_states)
        elif self.encoder_hid_proj is not None and self.config.encoder_hid_dim_type == "text_image_proj":
            # Kadinsky 2.1 - style
            if "image_embeds" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `encoder_hid_dim_type` set to 'text_image_proj' which requires the keyword argument `image_embeds` to be passed in  `added_conditions`"
                )

            image_embeds = added_cond_kwargs.get("image_embeds")
            encoder_hidden_states = self.encoder_hid_proj(encoder_hidden_states, image_embeds)
        elif self.encoder_hid_proj is not None and self.config.encoder_hid_dim_type == "image_proj":
            # Kandinsky 2.2 - style
            if "image_embeds" not in added_cond_kwargs:
                raise ValueError(
                    f"{self.__class__} has the config param `encoder_hid_dim_type` set to 'image_proj' which requires the keyword argument `image_embeds` to be passed in  `added_conditions`"
                )
            image_embeds = added_cond_kwargs.get("image_embeds")
            encoder_hidden_states = self.encoder_hid_proj(image_embeds)
        # 2. pre-process
        sample = self.conv_in(sample) # 1, 320, 64, 64

        # 2.5 GLIGEN position net
        if cross_attention_kwargs is not None and cross_attention_kwargs.get("gligen", None) is not None:
            cross_attention_kwargs = cross_attention_kwargs.copy()
            gligen_args = cross_attention_kwargs.pop("gligen")
            cross_attention_kwargs["gligen"] = {"objs": self.position_net(**gligen_args)}

        # 3. down
        lora_scale = cross_attention_kwargs.get("scale", 1.0) if cross_attention_kwargs is not None else 1.0

        is_controlnet = mid_block_additional_residual is not None and down_block_additional_residuals is not None
        is_adapter = mid_block_additional_residual is None and down_block_additional_residuals is not None

        down_block_res_samples = (sample,)

        for downsample_block in self.down_blocks:
            if hasattr(downsample_block, "has_cross_attention") and downsample_block.has_cross_attention:
                # For t2i-adapter CrossAttnDownBlock2D
                additional_residuals = {}
                if is_adapter and len(down_block_additional_residuals) > 0:
                    additional_residuals["additional_residuals"] = down_block_additional_residuals.pop(0)

                sample, res_samples = downsample_block(
                    hidden_states=sample,
                    temb=emb,
                    encoder_hidden_states=encoder_hidden_states,
                    attention_mask=attention_mask,
                    cross_attention_kwargs=cross_attention_kwargs,
                    encoder_attention_mask=encoder_attention_mask,
                    **additional_residuals,
                )
            else:
                sample, res_samples = downsample_block(hidden_states=sample, temb=emb, scale=lora_scale)

                if is_adapter and len(down_block_additional_residuals) > 0:
                    sample += down_block_additional_residuals.pop(0)

            down_block_res_samples += res_samples

        if is_controlnet:
            new_down_block_res_samples = ()

            for down_block_res_sample, down_block_additional_residual in zip(
                down_block_res_samples, down_block_additional_residuals
            ):
                down_block_res_sample = down_block_res_sample + down_block_additional_residual
                new_down_block_res_samples = new_down_block_res_samples + (down_block_res_sample,)

            down_block_res_samples = new_down_block_res_samples

        # 4. mid
        if self.mid_block is not None:
            sample = self.mid_block(
                sample,
                emb,
                encoder_hidden_states=encoder_hidden_states,
                attention_mask=attention_mask,
                cross_attention_kwargs=cross_attention_kwargs,
                encoder_attention_mask=encoder_attention_mask,
            )
            # To support T2I-Adapter-XL
            if (
                is_adapter
                and len(down_block_additional_residuals) > 0
                and sample.shape == down_block_additional_residuals[0].shape
            ):
                sample += down_block_additional_residuals.pop(0)

        if is_controlnet:
            sample = sample + mid_block_additional_residual

        # 5. 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,
                    cross_attention_kwargs=cross_attention_kwargs,
                    upsample_size=upsample_size,
                    attention_mask=attention_mask,
                    encoder_attention_mask=encoder_attention_mask,
                )
            else:
                sample = upsample_block(
                    hidden_states=sample,
                    temb=emb,
                    res_hidden_states_tuple=res_samples,
                    upsample_size=upsample_size,
                    scale=lora_scale,
                )

            # if i in [1, 4, 7]:
            out_list.append(sample)

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

        if not return_dict:
            return (sample,)

        return UNet2DConditionOutput(sample=sample), out_list
    
    self.forward = forward


class AttentionControl(abc.ABC):

    def step_callback(self, x_t):
        return x_t

    def between_steps(self):
        return

    @property
    def num_uncond_att_layers(self):
        return 0

    @abc.abstractmethod
    def forward(self, attn, is_cross: bool, place_in_unet: str):
        raise NotImplementedError

    def __call__(self, attn, is_cross: bool, place_in_unet: str):
        if self.cur_att_layer >= self.num_uncond_att_layers:
            # self.forward(attn, is_cross, place_in_unet)
            if LOW_RESOURCE:
                attn = self.forward(attn, is_cross, place_in_unet)
            else:
                h = attn.shape[0]
                attn[h // 2:] = self.forward(attn[h // 2:], is_cross, place_in_unet)
        self.cur_att_layer += 1
        if self.cur_att_layer == self.num_att_layers + self.num_uncond_att_layers:
            self.cur_att_layer = 0
            self.cur_step += 1
            self.between_steps()
        return attn

    def reset(self):
        self.cur_step = 0
        self.cur_att_layer = 0

    def __init__(self):
        self.cur_step = 0
        self.num_att_layers = -1
        self.cur_att_layer = 0


class EmptyControl(AttentionControl):

    def forward(self, attn, is_cross: bool, place_in_unet: str):
        return attn


class AttentionStore(AttentionControl):

    @staticmethod
    def get_empty_store():
        return {"down_cross": [], "mid_cross": [], "up_cross": [],
                "down_self": [], "mid_self": [], "up_self": []}

    def forward(self, attn, is_cross: bool, place_in_unet: str):
        key = f"{place_in_unet}_{'cross' if is_cross else 'self'}"
        if attn.shape[1] <= self.max_size ** 2:  # avoid memory overhead
            self.step_store[key].append(attn)
        return attn

    def between_steps(self):
        self.attention_store = self.step_store
        if self.save_global_store:
            with torch.no_grad():
                if len(self.global_store) == 0:
                    self.global_store = self.step_store
                else:
                    for key in self.global_store:
                        for i in range(len(self.global_store[key])):
                            self.global_store[key][i] += self.step_store[key][i].detach()
        self.step_store = self.get_empty_store()
        self.step_store = self.get_empty_store()

    def get_average_attention(self):
        average_attention = self.attention_store
        return average_attention

    def get_average_global_attention(self):
        average_attention = {key: [item / self.cur_step for item in self.global_store[key]] for key in
                             self.attention_store}
        return average_attention

    def reset(self):
        super(AttentionStore, self).reset()
        self.step_store = self.get_empty_store()
        self.attention_store = {}
        self.global_store = {}

    def __init__(self, max_size=32, save_global_store=False):
        '''
        Initialize an empty AttentionStore
        :param step_index: used to visualize only a specific step in the diffusion process
        '''
        super(AttentionStore, self).__init__()
        self.save_global_store = save_global_store
        self.max_size = max_size
        self.step_store = self.get_empty_store()
        self.attention_store = {}
        self.global_store = {}
        self.curr_step_index = 0

def aggregate_attention(prompts, attention_store: AttentionStore, res: int, from_where: List[str], is_cross: bool, select: int):
    out = []
    attention_maps = attention_store.get_average_attention()
    num_pixels = res ** 2
    for location in from_where:
        for item in attention_maps[f"{location}_{'cross' if is_cross else 'self'}"]:
            if item.shape[1] == num_pixels:
                cross_maps = item.reshape(len(prompts), -1, res, res, item.shape[-1])[select]
                out.append(cross_maps)
    out = torch.cat(out, dim=0)
    out = out.sum(0) / out.shape[0]
    return out


def show_cross_attention(tokenizer, prompts, attention_store: AttentionStore, res: int, from_where: List[str], select: int = 0):
    tokens = tokenizer.encode(prompts[select])
    decoder = tokenizer.decode
    attention_maps = aggregate_attention(attention_store, res, from_where, True, select)
    images = []
    for i in range(len(tokens)):
        image = attention_maps[:, :, i]
        image = 255 * image / image.max()
        image = image.unsqueeze(-1).expand(*image.shape, 3)
        image = image.numpy().astype(np.uint8)
        image = np.array(Image.fromarray(image).resize((256, 256)))
        image = text_under_image(image, decoder(int(tokens[i])))
        images.append(image)
    view_images(np.stack(images, axis=0))
    

def show_self_attention_comp(attention_store: AttentionStore, res: int, from_where: List[str],
                        max_com=10, select: int = 0):
    attention_maps = aggregate_attention(attention_store, res, from_where, False, select).numpy().reshape((res ** 2, res ** 2))
    u, s, vh = np.linalg.svd(attention_maps - np.mean(attention_maps, axis=1, keepdims=True))
    images = []
    for i in range(max_com):
        image = vh[i].reshape(res, res)
        image = image - image.min()
        image = 255 * image / image.max()
        image = np.repeat(np.expand_dims(image, axis=2), 3, axis=2).astype(np.uint8)
        image = Image.fromarray(image).resize((256, 256))
        image = np.array(image)
        images.append(image)
    view_images(np.concatenate(images, axis=1))

def text_under_image(image: np.ndarray, text: str, text_color: Tuple[int, int, int] = (0, 0, 0)):
    h, w, c = image.shape
    offset = int(h * .2)
    img = np.ones((h + offset, w, c), dtype=np.uint8) * 255
    font = cv2.FONT_HERSHEY_SIMPLEX
    # font = ImageFont.truetype("/usr/share/fonts/truetype/noto/NotoMono-Regular.ttf", font_size)
    img[:h] = image
    textsize = cv2.getTextSize(text, font, 1, 2)[0]
    text_x, text_y = (w - textsize[0]) // 2, h + offset - textsize[1] // 2
    cv2.putText(img, text, (text_x, text_y ), font, 1, text_color, 2)
    return img


def view_images(images, num_rows=1, offset_ratio=0.02):
    if type(images) is list:
        num_empty = len(images) % num_rows
    elif images.ndim == 4:
        num_empty = images.shape[0] % num_rows
    else:
        images = [images]
        num_empty = 0

    empty_images = np.ones(images[0].shape, dtype=np.uint8) * 255
    images = [image.astype(np.uint8) for image in images] + [empty_images] * num_empty
    num_items = len(images)

    h, w, c = images[0].shape
    offset = int(h * offset_ratio)
    num_cols = num_items // num_rows
    image_ = np.ones((h * num_rows + offset * (num_rows - 1),
                      w * num_cols + offset * (num_cols - 1), 3), dtype=np.uint8) * 255
    for i in range(num_rows):
        for j in range(num_cols):
            image_[i * (h + offset): i * (h + offset) + h:, j * (w + offset): j * (w + offset) + w] = images[
                i * num_cols + j]

    pil_img = Image.fromarray(image_)
    display(pil_img)

def self_cross_attn(self_attn, cross_attn):
    res = self_attn.shape[0]
    assert res == cross_attn.shape[0]
    # cross attn [res, res] -> [res*res]
    cross_attn_ = cross_attn.reshape([res*res])
    # self_attn [res, res, res*res]
    self_cross_attn = cross_attn_ * self_attn
    self_cross_attn = self_cross_attn.mean(-1).unsqueeze(0).unsqueeze(0)
    return self_cross_attn