Spaces:
Running
on
Zero
Running
on
Zero
# Copyright 2023 The HuggingFace Team. All rights reserved. | |
# | |
# 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. | |
import inspect | |
from dataclasses import dataclass | |
from typing import Callable, Dict, List, Optional, Union | |
import cv2, os, sys | |
import numpy as np | |
import PIL.Image | |
import torch | |
import torch.nn as nn | |
from transformers import CLIPImageProcessor, CLIPVisionModelWithProjection | |
from diffusers.image_processor import VaeImageProcessor | |
from diffusers.schedulers import EulerDiscreteScheduler | |
from diffusers.models import AutoencoderKLTemporalDecoder | |
from diffusers.utils import BaseOutput, logging | |
from diffusers.utils.torch_utils import randn_tensor | |
from diffusers import DiffusionPipeline | |
# Import files from the local folder | |
root_path = os.path.abspath('.') | |
sys.path.append(root_path) | |
from utils.img_utils import tensor2np | |
from svd.temporal_controlnet import ControlNetModel | |
from svd.unet_spatio_temporal_condition import UNetSpatioTemporalConditionModel | |
logger = logging.get_logger(__name__) # pylint: disable=invalid-name | |
def _append_dims(x, target_dims): | |
"""Appends dimensions to the end of a tensor until it has target_dims dimensions.""" | |
dims_to_append = target_dims - x.ndim | |
if dims_to_append < 0: | |
raise ValueError(f"input has {x.ndim} dims but target_dims is {target_dims}, which is less") | |
return x[(...,) + (None,) * dims_to_append] | |
def tensor2vid(video: torch.Tensor, processor, output_type="np"): | |
# Based on: | |
# https://github.com/modelscope/modelscope/blob/1509fdb973e5871f37148a4b5e5964cafd43e64d/modelscope/pipelines/multi_modal/text_to_video_synthesis_pipeline.py#L78 | |
batch_size, channels, num_frames, height, width = video.shape | |
outputs = [] | |
for batch_idx in range(batch_size): | |
batch_vid = video[batch_idx].permute(1, 0, 2, 3) | |
batch_output = processor.postprocess(batch_vid, output_type) | |
outputs.append(batch_output) | |
return outputs | |
class StableVideoDiffusionPipelineOutput(BaseOutput): | |
r""" | |
Output class for zero-shot text-to-video pipeline. | |
Args: | |
frames (`[List[PIL.Image.Image]`, `np.ndarray`]): | |
List of denoised PIL images of length `batch_size` or NumPy array of shape `(batch_size, height, width, | |
num_channels)`. | |
""" | |
frames: Union[List[PIL.Image.Image], np.ndarray] | |
class StableVideoDiffusionControlNetPipeline(DiffusionPipeline): | |
r""" | |
Pipeline to generate video from an input image using Stable Video Diffusion. | |
This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods | |
implemented for all pipelines (downloading, saving, running on a particular device, etc.). | |
Args: | |
vae ([`AutoencoderKL`]): | |
Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations. | |
image_encoder ([`~transformers.CLIPVisionModelWithProjection`]): | |
Frozen CLIP image-encoder ([laion/CLIP-ViT-H-14-laion2B-s32B-b79K](https://huggingface.co/laion/CLIP-ViT-H-14-laion2B-s32B-b79K)). | |
unet ([`UNetSpatioTemporalConditionModel`]): | |
A `UNetSpatioTemporalConditionModel` to denoise the encoded image latents. | |
scheduler ([`EulerDiscreteScheduler`]): | |
A scheduler to be used in combination with `unet` to denoise the encoded image latents. | |
feature_extractor ([`~transformers.CLIPImageProcessor`]): | |
A `CLIPImageProcessor` to extract features from generated images. | |
""" | |
model_cpu_offload_seq = "image_encoder->unet->vae" | |
_callback_tensor_inputs = ["latents"] | |
def __init__( | |
self, | |
vae: AutoencoderKLTemporalDecoder, | |
image_encoder: CLIPVisionModelWithProjection, | |
unet: UNetSpatioTemporalConditionModel, | |
scheduler: EulerDiscreteScheduler, | |
feature_extractor: CLIPImageProcessor, | |
): | |
super().__init__() | |
# TODO: multi-controlnet consideration | |
self.register_modules( | |
vae = vae, | |
image_encoder = image_encoder, | |
unet = unet, | |
scheduler = scheduler, | |
feature_extractor = feature_extractor, | |
) | |
self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) | |
self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True) # The vae_scale_factor is for image dimension, not for image size | |
self.control_image_processor = VaeImageProcessor( | |
vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False | |
) | |
def encode_clip(self, image, prompt, use_text, text_encoder, device, num_videos_per_prompt, do_classifier_free_guidance, use_instructpix2pix): | |
dtype = next(self.image_encoder.parameters()).dtype | |
if not isinstance(image, torch.Tensor): | |
image = self.image_processor.pil_to_numpy(image) # Map [0, 255] to [0, 1] range | |
image = self.image_processor.numpy_to_pt(image) | |
# We normalize the image before resizing to match with the original implementation. | |
# Then, we unnormalize it after resizing. | |
image = image * 2.0 - 1.0 # [-1, 1] range | |
image = _resize_with_antialiasing(image, (224, 224)) # Resize to square image | |
image = (image + 1.0) / 2.0 # [0, 1] range | |
# Normalize the image with for CLIP input | |
image = self.feature_extractor( | |
images=image, | |
do_normalize=True, | |
do_center_crop=False, | |
do_resize=False, | |
do_rescale=False, | |
return_tensors="pt", | |
).pixel_values # The value range is a little deviated now, and I got [-1.76, 2.15] for one sample | |
image = image.to(device=device, dtype=dtype) | |
image_embeddings = self.image_encoder(image).image_embeds | |
image_embeddings = image_embeddings.unsqueeze(1) | |
# duplicate image embeddings for each generation per prompt, using mps friendly method | |
bs_embed, seq_len, _ = image_embeddings.shape | |
image_embeddings = image_embeddings.repeat(1, num_videos_per_prompt, 1) | |
encoder_hidden_states = image_embeddings.view(bs_embed * num_videos_per_prompt, seq_len, -1) | |
# Prepare for the text embeddings if needed | |
if use_text: | |
text_embeddings = text_encoder(prompt)[0] | |
# Concat two embeddings together on dim 1 | |
encoder_hidden_states = torch.cat((text_embeddings, encoder_hidden_states), dim=1) | |
# Layer norm on the last dim TODO: 这里order小改了一下顺序,变成先encoder hidden states了 | |
layer_norm = nn.LayerNorm((78, 1024)).to(device=device, dtype=dtype) | |
encoder_hidden_states = layer_norm(encoder_hidden_states) | |
if do_classifier_free_guidance: | |
negative_encoder_hidden_states = torch.zeros_like(encoder_hidden_states) | |
# For classifier free guidance, we need to do two forward passes. | |
# Here we concatenate the unconditional and text embeddings into a single batch | |
# to avoid doing two forward passes | |
if use_instructpix2pix: | |
encoder_hidden_states = torch.cat([encoder_hidden_states, negative_encoder_hidden_states, negative_encoder_hidden_states]) | |
else: | |
encoder_hidden_states = torch.cat([negative_encoder_hidden_states, encoder_hidden_states]) | |
return encoder_hidden_states | |
def _encode_vae_image( | |
self, | |
image: torch.Tensor, | |
device, | |
num_videos_per_prompt, | |
do_classifier_free_guidance, | |
use_instructpix2pix, | |
): | |
image = image.to(device=device) | |
image_latents = self.vae.encode(image).latent_dist.mode() | |
if do_classifier_free_guidance: | |
negative_image_latents = torch.zeros_like(image_latents) | |
# For classifier free guidance, we need to do two forward passes. | |
# Here we concatenate the unconditional and text embeddings into a single batch | |
# to avoid doing two forward passes | |
if use_instructpix2pix: | |
image_latents = torch.cat([image_latents, image_latents, negative_image_latents]) | |
else: | |
image_latents = torch.cat([negative_image_latents, image_latents]) | |
# duplicate image_latents for each generation per prompt, using mps friendly method | |
image_latents = image_latents.repeat(num_videos_per_prompt, 1, 1, 1) | |
return image_latents | |
def _get_add_time_ids( | |
self, | |
fps, | |
motion_bucket_id, | |
noise_aug_strength, | |
dtype, | |
batch_size, | |
num_videos_per_prompt, | |
do_classifier_free_guidance, | |
guess_mode, | |
use_instructpix2pix, | |
): | |
# Define the default values from SVD | |
add_time_ids = [fps, motion_bucket_id, noise_aug_strength] | |
# Sanity Check | |
passed_add_embed_dim = self.unet.config.addition_time_embed_dim * len(add_time_ids) | |
expected_add_embed_dim = self.unet.add_embedding.linear_1.in_features | |
if expected_add_embed_dim != passed_add_embed_dim: | |
raise ValueError( | |
f"Model expects an added time embedding vector of length {expected_add_embed_dim}, but a vector of {passed_add_embed_dim} was created. The model has an incorrect config. Please check `unet.config.time_embedding_type` and `text_encoder_2.config.projection_dim`." | |
) | |
add_time_ids = torch.tensor([add_time_ids], dtype=dtype) | |
add_time_ids = add_time_ids.repeat(batch_size * num_videos_per_prompt, 1) | |
if do_classifier_free_guidance: | |
if use_instructpix2pix: | |
add_time_ids = torch.cat([add_time_ids, add_time_ids, add_time_ids]) | |
else: | |
add_time_ids = torch.cat([add_time_ids, add_time_ids]) | |
# Return the info | |
return add_time_ids | |
def decode_latents(self, latents, num_frames, decode_chunk_size=14): | |
# [batch, frames, channels, height, width] -> [batch*frames, channels, height, width] | |
latents = latents.flatten(0, 1) | |
latents = 1 / self.vae.config.scaling_factor * latents | |
accepts_num_frames = "num_frames" in set(inspect.signature(self.vae.forward).parameters.keys()) | |
# decode decode_chunk_size frames at a time to avoid OOM | |
frames = [] | |
for i in range(0, latents.shape[0], decode_chunk_size): | |
num_frames_in = latents[i : i + decode_chunk_size].shape[0] | |
decode_kwargs = {} | |
if accepts_num_frames: | |
# we only pass num_frames_in if it's expected | |
decode_kwargs["num_frames"] = num_frames_in | |
frame = self.vae.decode(latents[i : i + decode_chunk_size], **decode_kwargs).sample | |
frames.append(frame) | |
frames = torch.cat(frames, dim=0) | |
# [batch*frames, channels, height, width] -> [batch, channels, frames, height, width] | |
frames = frames.reshape(-1, num_frames, *frames.shape[1:]).permute(0, 2, 1, 3, 4) | |
# we always cast to float32 as this does not cause significant overhead and is compatible with bfloat16 | |
frames = frames.float() | |
return frames | |
def check_inputs(self, | |
image, | |
height, | |
width, | |
): | |
if ( | |
not isinstance(image, torch.Tensor) | |
and not isinstance(image, PIL.Image.Image) | |
and not isinstance(image, list) | |
): | |
raise ValueError( | |
"`image` has to be of type `torch.FloatTensor` or `PIL.Image.Image` or `List[PIL.Image.Image]` but is" | |
f" {type(image)}" | |
) | |
if height % 8 != 0 or width % 8 != 0: | |
raise ValueError(f"`height` and `width` have to be divisible by 8 but are {height} and {width}.") | |
# TODO: I didn't test input for controlnet_conditioning_scale, control_guidance_start, and control_guidance_end | |
def prepare_latents( | |
self, | |
batch_size, | |
num_frames, | |
num_channels_latents, | |
height, | |
width, | |
dtype, | |
device, | |
generator, | |
latents=None, | |
): | |
shape = ( | |
batch_size, | |
num_frames, | |
num_channels_latents // 2, | |
height // self.vae_scale_factor, | |
width // self.vae_scale_factor, | |
) | |
if isinstance(generator, list) and len(generator) != batch_size: | |
raise ValueError( | |
f"You have passed a list of generators of length {len(generator)}, but requested an effective batch" | |
f" size of {batch_size}. Make sure the batch size matches the length of the generators." | |
) | |
if latents is None: | |
latents = randn_tensor(shape, generator=generator, device=device, dtype=dtype) | |
else: | |
latents = latents.to(device) | |
# scale the initial noise by the standard deviation required by the scheduler | |
latents = latents * self.scheduler.init_noise_sigma | |
return latents | |
def guidance_scale(self): | |
return self._guidance_scale | |
# We don't directly have do_classifier_free_guidance function, we judge simply by max_guidance | |
def num_timesteps(self): | |
return self._num_timesteps | |
def prepare_condition_image( | |
self, | |
condition_img, | |
width, | |
height, | |
batch_size, | |
num_videos_per_prompt, | |
device, | |
dtype, | |
do_classifier_free_guidance=False, | |
guess_mode=False, | |
): | |
# The input of condition_img is already in the range [0, 1] | |
condition_img = torch.from_numpy(condition_img) # hwc -> chw | |
condition_img = condition_img.to(torch.float16).to(self._execution_device) # Set this in default | |
# CFG will be done in main function, not here now | |
return condition_img # [0, 1] range && Torch data type | |
def __call__( | |
self, | |
image: Union[PIL.Image.Image, List[PIL.Image.Image], torch.FloatTensor], | |
condition_img: np, | |
controlnet: ControlNetModel, | |
prompt = None, | |
use_text: bool = False, | |
text_encoder = None, | |
height: int = 576, | |
width: int = 1024, | |
num_frames: Optional[int] = None, | |
num_inference_steps: int = 25, | |
min_guidance_scale: float = 1.0, | |
max_guidance_scale: float = 3.0, | |
fps: int = 7, | |
motion_bucket_id: int = 127, | |
# controlnet_image_index: Optional[int] = [0], | |
# coordinate_values = None, | |
noise_aug_strength: int = 0.02, | |
decode_chunk_size: Optional[int] = None, | |
num_videos_per_prompt: Optional[int] = 1, | |
generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None, | |
latents: Optional[torch.FloatTensor] = None, | |
output_type: Optional[str] = "pil", | |
callback_on_step_end: Optional[Callable[[int, int, Dict], None]] = None, | |
callback_on_step_end_tensor_inputs: List[str] = ["latents"], | |
return_dict: bool = True, | |
controlnet_conditioning_scale: Union[float, List[float]] = 1.0, | |
use_instructpix2pix: bool = False, | |
control_guidance_start: Union[float, List[float]] = 0.0, | |
control_guidance_end: Union[float, List[float]] = 1.0, | |
inner_conditioning_scale: float = 1.0, | |
guess_mode: bool = True, | |
image_guidance_scale: float = 7.5, | |
): | |
r""" | |
The call function to the pipeline for generation. | |
Args: | |
image (`PIL.Image.Image` or `List[PIL.Image.Image]` or `torch.FloatTensor`): | |
Image or images to guide image generation. If you provide a tensor, it needs to be compatible with | |
[`CLIPImageProcessor`](https://huggingface.co/lambdalabs/sd-image-variations-diffusers/blob/main/feature_extractor/preprocessor_config.json). | |
height (`int`, *optional*, defaults to `self.unet.config.sample_size * self.vae_scale_factor`): | |
The height in pixels of the generated image. | |
width (`int`, *optional*, defaults to `self.unet.config.sample_size * self.vae_scale_factor`): | |
The width in pixels of the generated image. | |
num_frames (`int`, *optional*): | |
The number of video frames to generate. Defaults to 14 for `stable-video-diffusion-img2vid` and to 25 for `stable-video-diffusion-img2vid-xt` | |
num_inference_steps (`int`, *optional*, defaults to 25): | |
The number of denoising steps. More denoising steps usually lead to a higher quality image at the | |
expense of slower inference. This parameter is modulated by `strength`. | |
min_guidance_scale (`float`, *optional*, defaults to 1.0): | |
The minimum guidance scale. Used for the classifier free guidance with first frame. | |
max_guidance_scale (`float`, *optional*, defaults to 3.0): | |
The maximum guidance scale. Used for the classifier free guidance with last frame. | |
fps (`int`, *optional*, defaults to 7): | |
Frames per second. The rate at which the generated images shall be exported to a video after generation. | |
Note that Stable Diffusion Video's UNet was micro-conditioned on fps-1 during training. | |
motion_bucket_id (`int`, *optional*, defaults to 127): | |
The motion bucket ID. Used as conditioning for the generation. The higher the number the more motion will be in the video. | |
noise_aug_strength (`int`, *optional*, defaults to 0.02): | |
The amount of noise added to the init image, the higher it is the less the video will look like the init image. Increase it for more motion. | |
decode_chunk_size (`int`, *optional*): | |
The number of frames to decode at a time. The higher the chunk size, the higher the temporal consistency | |
between frames, but also the higher the memory consumption. By default, the decoder will decode all frames at once | |
for maximal quality. Reduce `decode_chunk_size` to reduce memory usage. | |
num_videos_per_prompt (`int`, *optional*, defaults to 1): | |
The number of images to generate per prompt. | |
generator (`torch.Generator` or `List[torch.Generator]`, *optional*): | |
A [`torch.Generator`](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make | |
generation deterministic. | |
latents (`torch.FloatTensor`, *optional*): | |
Pre-generated noisy latents sampled from a Gaussian distribution, to be used as inputs for image | |
generation. Can be used to tweak the same generation with different prompts. If not provided, a latents | |
tensor is generated by sampling using the supplied random `generator`. | |
output_type (`str`, *optional*, defaults to `"pil"`): | |
The output format of the generated image. Choose between `PIL.Image` or `np.array`. | |
callback_on_step_end (`Callable`, *optional*): | |
A function that calls at the end of each denoising steps during the inference. The function is called | |
with the following arguments: `callback_on_step_end(self: DiffusionPipeline, step: int, timestep: int, | |
callback_kwargs: Dict)`. `callback_kwargs` will include a list of all tensors as specified by | |
`callback_on_step_end_tensor_inputs`. | |
callback_on_step_end_tensor_inputs (`List`, *optional*): | |
The list of tensor inputs for the `callback_on_step_end` function. The tensors specified in the list | |
will be passed as `callback_kwargs` argument. You will only be able to include variables listed in the | |
`._callback_tensor_inputs` attribute of your pipeline class. | |
return_dict (`bool`, *optional*, defaults to `True`): | |
Whether or not to return a [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] instead of a | |
plain tuple. | |
Returns: | |
[`~pipelines.stable_diffusion.StableVideoDiffusionPipelineOutput`] or `tuple`: | |
If `return_dict` is `True`, [`~pipelines.stable_diffusion.StableVideoDiffusionPipelineOutput`] is returned, | |
otherwise a `tuple` is returned where the first element is a list of list with the generated frames. | |
Examples: | |
```py | |
from diffusers import StableVideoDiffusionPipeline | |
from diffusers.utils import load_image, export_to_video | |
pipe = StableVideoDiffusionPipeline.from_pretrained("stabilityai/stable-video-diffusion-img2vid-xt", torch_dtype=torch.float16, variant="fp16") | |
pipe.to("cuda") | |
image = load_image("https://lh3.googleusercontent.com/y-iFOHfLTwkuQSUegpwDdgKmOjRSTvPxat63dQLB25xkTs4lhIbRUFeNBWZzYf370g=s1200") | |
image = image.resize((1024, 576)) | |
frames = pipe(image, num_frames=25, decode_chunk_size=8).frames[0] | |
export_to_video(frames, "generated.mp4", fps=7) | |
``` | |
""" | |
# align format for control guidance | |
mult = 1 | |
control_guidance_start, control_guidance_end = ( | |
mult * [control_guidance_start], | |
mult * [control_guidance_end], | |
) | |
# 0. Default height and width to unet | |
height = height or self.unet.config.sample_size * self.vae_scale_factor | |
width = width or self.unet.config.sample_size * self.vae_scale_factor # It seems that self.unet.config.sample_size * self.vae_scale_factor is a default image size input setting | |
num_frames = num_frames if num_frames is not None else self.unet.config.num_frames | |
decode_chunk_size = decode_chunk_size if decode_chunk_size is not None else num_frames | |
# 1. Check inputs. Raise error if not correct | |
self.check_inputs(image, height, width) | |
# 2. Define call parameters | |
if isinstance(image, PIL.Image.Image): | |
batch_size = 1 | |
elif isinstance(image, list): | |
batch_size = len(image) | |
else: | |
batch_size = image.shape[0] | |
device = self._execution_device | |
# here `guidance_scale` is defined analog to the guidance weight `w` of equation (2) | |
# of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1` | |
# corresponds to doing no classifier free guidance. | |
do_classifier_free_guidance = max_guidance_scale > 1.0 | |
if do_classifier_free_guidance: | |
print("We will use CFG!!!") | |
# 3. Encode input image | |
encoder_hidden_states = self.encode_clip(image, prompt, use_text, text_encoder, device, num_videos_per_prompt, do_classifier_free_guidance, use_instructpix2pix) | |
# NOTE: Stable Diffusion Video was conditioned on fps - 1, which | |
# is why it is reduced here. | |
# See: https://github.com/Stability-AI/generative-models/blob/ed0997173f98eaf8f4edf7ba5fe8f15c6b877fd3/scripts/sampling/simple_video_sample.py#L188 | |
fps = fps - 1 | |
# 4. Encode input image using VAE | |
image = self.image_processor.preprocess(image, height=height, width=width) # [0, 255] to [-1, 1] | |
noise = randn_tensor(image.shape, generator=generator, device=image.device, dtype=image.dtype) | |
image = image + noise_aug_strength * noise | |
needs_upcasting = self.vae.dtype == torch.float16 and self.vae.config.force_upcast | |
if needs_upcasting: | |
self.vae.to(dtype=torch.float32) | |
image_latents = self._encode_vae_image(image, device, num_videos_per_prompt, do_classifier_free_guidance, use_instructpix2pix) | |
image_latents = image_latents.to(encoder_hidden_states.dtype) | |
# cast back to fp16 if needed | |
if needs_upcasting: | |
self.vae.to(dtype=torch.float16) | |
# Repeat the image latents for each frame so we can concatenate them with the noise | |
# image_latents [batch, channels, height, width] ->[batch, num_frames, channels, height, width] | |
image_latents = image_latents.unsqueeze(1).repeat(1, num_frames, 1, 1, 1) | |
# 4.5 Prepare control image (Will need to consider multiControlNet) | |
condition_img = self.prepare_condition_image( | |
condition_img = condition_img, | |
width = width, | |
height = height, | |
batch_size = batch_size * num_videos_per_prompt, | |
num_videos_per_prompt = num_videos_per_prompt, | |
device = device, | |
dtype = controlnet.dtype, | |
do_classifier_free_guidance = do_classifier_free_guidance, | |
guess_mode = guess_mode, | |
) # [0, 255] to [0, 1] range | |
# 5. Get Added Time IDs | |
added_time_ids = self._get_add_time_ids( | |
fps, | |
motion_bucket_id, | |
noise_aug_strength, | |
# coordinate_values, | |
encoder_hidden_states.dtype, | |
batch_size, | |
num_videos_per_prompt, | |
do_classifier_free_guidance, | |
guess_mode = guess_mode, | |
use_instructpix2pix = use_instructpix2pix, | |
) | |
added_time_ids = added_time_ids.to(device) | |
# 4. Prepare timesteps | |
self.scheduler.set_timesteps(num_inference_steps, device=device) | |
timesteps = self.scheduler.timesteps | |
# 5. Prepare latent variables | |
num_channels_latents = self.unet.config.in_channels | |
latents = self.prepare_latents( | |
batch_size * num_videos_per_prompt, | |
num_frames, | |
num_channels_latents, | |
height, | |
width, | |
encoder_hidden_states.dtype, | |
device, | |
generator, | |
latents, | |
) # Nosiy latents across all frames needed | |
# 7. Prepare guidance scale | |
guidance_scale = torch.linspace(min_guidance_scale, max_guidance_scale, num_frames).unsqueeze(0) | |
guidance_scale = guidance_scale.to(device, latents.dtype) | |
guidance_scale = guidance_scale.repeat(batch_size * num_videos_per_prompt, 1) | |
guidance_scale = _append_dims(guidance_scale, latents.ndim) | |
self._guidance_scale = guidance_scale | |
# 7.5 Create tensor stating which controlnets to keep | |
controlnet_keep = [] | |
for i in range(len(timesteps)): | |
keeps = [ | |
1.0 - float(i / len(timesteps) < s or (i + 1) / len(timesteps) > e) | |
for s, e in zip(control_guidance_start, control_guidance_end) | |
] | |
controlnet_keep.append(keeps[0] if isinstance(controlnet, ControlNetModel) else keeps) | |
# 8. Denoising loop | |
num_warmup_steps = len(timesteps) - num_inference_steps * self.scheduler.order | |
self._num_timesteps = len(timesteps) | |
with self.progress_bar(total=num_inference_steps) as progress_bar: | |
for i, t in enumerate(timesteps): | |
# expand the latents if we are doing classifier free guidance | |
if use_instructpix2pix: | |
latent_model_input = torch.cat([latents] * 3) if do_classifier_free_guidance else latents | |
else: | |
latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents | |
latent_model_input = self.scheduler.scale_model_input(latent_model_input, t) # I think that this is where sequential generation takes influence | |
# Concatenate image_latents over channels dimension for video diffusion purposes | |
latent_model_input = torch.cat([latent_model_input, image_latents], dim=2) # image_latents is fixed and latent_model_input will be based on latents which is updated frequently | |
# ControlNet Scale | |
if isinstance(controlnet_keep[i], list): | |
cond_scale = [c * s for c, s in zip(controlnet_conditioning_scale, controlnet_keep[i])] | |
else: | |
controlnet_cond_scale = controlnet_conditioning_scale | |
if isinstance(controlnet_cond_scale, list): | |
controlnet_cond_scale = controlnet_cond_scale[0] | |
cond_scale = controlnet_cond_scale * controlnet_keep[i] | |
# assert condition_img.shape[1] >= len(controlnet_image_index) | |
# VAE encode | |
controlnet_cond = self.vae.encode(condition_img).latent_dist.mode() | |
if do_classifier_free_guidance: | |
if use_instructpix2pix: | |
controlnet_cond = torch.cat([controlnet_cond, controlnet_cond, controlnet_cond]) | |
# controlnet_conditioning_mask = torch.cat([controlnet_conditioning_mask, controlnet_conditioning_mask, controlnet_conditioning_mask]) | |
else: | |
controlnet_cond = torch.cat([controlnet_cond, controlnet_cond]) | |
# controlnet_conditioning_mask = torch.cat([controlnet_conditioning_mask, controlnet_conditioning_mask]) | |
down_block_res_samples, mid_block_res_sample = controlnet( | |
sample = latent_model_input, | |
timestep = t, | |
encoder_hidden_states = encoder_hidden_states, | |
added_time_ids = added_time_ids, | |
controlnet_cond = controlnet_cond, | |
return_dict = False, | |
inner_conditioning_scale = inner_conditioning_scale, # Inner conditioning scale | |
conditioning_scale = cond_scale, # Outer conditioning scale | |
guess_mode = guess_mode, | |
) | |
if guess_mode and do_classifier_free_guidance: # Won't consider this one, since we don't use guess mode | |
# Infered ControlNet only for the conditional batch. | |
# To apply the output of ControlNet to both the unconditional and conditional batches, | |
# add 0 to the unconditional batch to keep it unchanged. | |
down_block_res_samples = [torch.cat([torch.zeros_like(d), d]) for d in down_block_res_samples] | |
mid_block_res_sample = torch.cat([torch.zeros_like(mid_block_res_sample), mid_block_res_sample]) | |
# predict the noise residual | |
noise_pred = self.unet( | |
sample = latent_model_input, # [batch, frames, 4*2, height, width] | |
timestep = t, | |
encoder_hidden_states = encoder_hidden_states, | |
added_time_ids = added_time_ids, | |
down_block_additional_residuals = down_block_res_samples, | |
mid_block_additional_residual = mid_block_res_sample, | |
return_dict = False, | |
)[0] # image_embeddings is used for cross attention metioned in the paper | |
# perform guidance | |
if do_classifier_free_guidance: | |
if use_instructpix2pix: | |
noise_pred_1st_frame, noise_pred_cond, noise_pred_uncond = noise_pred.chunk(3) # There are two noises here: one is unconditional and one is conditional | |
noise_pred = noise_pred_uncond + \ | |
self.guidance_scale * (noise_pred_cond - noise_pred_uncond) + \ | |
image_guidance_scale * (noise_pred_cond - noise_pred_1st_frame) # InstructPix2Pix is (noise_pred_1st_frame - noise_pred_cond) | |
else: | |
noise_pred_uncond, noise_pred_cond = noise_pred.chunk(2) # There are two noises here: one is unconditional and one is conditional | |
noise_pred = noise_pred_uncond + self.guidance_scale * (noise_pred_cond - noise_pred_uncond) | |
# compute the previous noisy sample x_t -> x_t-1 | |
latents = self.scheduler.step(noise_pred, t, latents).prev_sample | |
if callback_on_step_end is not None: | |
callback_kwargs = {} | |
for k in callback_on_step_end_tensor_inputs: | |
callback_kwargs[k] = locals()[k] | |
callback_outputs = callback_on_step_end(self, i, t, callback_kwargs) | |
latents = callback_outputs.pop("latents", latents) | |
if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0): | |
progress_bar.update() | |
if not output_type == "latent": | |
# cast back to fp16 if needed | |
if needs_upcasting: | |
self.vae.to(dtype=torch.float16) | |
frames = self.decode_latents(latents, num_frames, decode_chunk_size) | |
frames = tensor2vid(frames, self.image_processor, output_type=output_type) | |
else: | |
frames = latents | |
self.maybe_free_model_hooks() | |
if not return_dict: | |
return frames | |
return StableVideoDiffusionPipelineOutput(frames=frames) | |
# resizing utils | |
# TODO: clean up later (put to shared utils file) | |
def _resize_with_antialiasing(input, size, interpolation="bicubic", align_corners=True): | |
h, w = input.shape[-2:] | |
factors = (h / size[0], w / size[1]) | |
# First, we have to determine sigma | |
# Taken from skimage: https://github.com/scikit-image/scikit-image/blob/v0.19.2/skimage/transform/_warps.py#L171 | |
sigmas = ( | |
max((factors[0] - 1.0) / 2.0, 0.001), | |
max((factors[1] - 1.0) / 2.0, 0.001), | |
) | |
# Now kernel size. Good results are for 3 sigma, but that is kind of slow. Pillow uses 1 sigma | |
# https://github.com/python-pillow/Pillow/blob/master/src/libImaging/Resample.c#L206 | |
# But they do it in the 2 passes, which gives better results. Let's try 2 sigmas for now | |
ks = int(max(2.0 * 2 * sigmas[0], 3)), int(max(2.0 * 2 * sigmas[1], 3)) | |
# Make sure it is odd | |
if (ks[0] % 2) == 0: | |
ks = ks[0] + 1, ks[1] | |
if (ks[1] % 2) == 0: | |
ks = ks[0], ks[1] + 1 | |
input = _gaussian_blur2d(input, ks, sigmas) | |
output = torch.nn.functional.interpolate(input, size=size, mode=interpolation, align_corners=align_corners) | |
return output | |
def _compute_padding(kernel_size): | |
"""Compute padding tuple.""" | |
# 4 or 6 ints: (padding_left, padding_right,padding_top,padding_bottom) | |
# https://pytorch.org/docs/stable/nn.html#torch.nn.functional.pad | |
if len(kernel_size) < 2: | |
raise AssertionError(kernel_size) | |
computed = [k - 1 for k in kernel_size] | |
# for even kernels we need to do asymmetric padding :( | |
out_padding = 2 * len(kernel_size) * [0] | |
for i in range(len(kernel_size)): | |
computed_tmp = computed[-(i + 1)] | |
pad_front = computed_tmp // 2 | |
pad_rear = computed_tmp - pad_front | |
out_padding[2 * i + 0] = pad_front | |
out_padding[2 * i + 1] = pad_rear | |
return out_padding | |
def _filter2d(input, kernel): | |
# prepare kernel | |
b, c, h, w = input.shape | |
tmp_kernel = kernel[:, None, ...].to(device=input.device, dtype=input.dtype) | |
tmp_kernel = tmp_kernel.expand(-1, c, -1, -1) | |
height, width = tmp_kernel.shape[-2:] | |
padding_shape: list[int] = _compute_padding([height, width]) | |
input = torch.nn.functional.pad(input, padding_shape, mode="reflect") | |
# kernel and input tensor reshape to align element-wise or batch-wise params | |
tmp_kernel = tmp_kernel.reshape(-1, 1, height, width) | |
input = input.view(-1, tmp_kernel.size(0), input.size(-2), input.size(-1)) | |
# convolve the tensor with the kernel. | |
output = torch.nn.functional.conv2d(input, tmp_kernel, groups=tmp_kernel.size(0), padding=0, stride=1) | |
out = output.view(b, c, h, w) | |
return out | |
def _gaussian(window_size: int, sigma): | |
if isinstance(sigma, float): | |
sigma = torch.tensor([[sigma]]) | |
batch_size = sigma.shape[0] | |
x = (torch.arange(window_size, device=sigma.device, dtype=sigma.dtype) - window_size // 2).expand(batch_size, -1) | |
if window_size % 2 == 0: | |
x = x + 0.5 | |
gauss = torch.exp(-x.pow(2.0) / (2 * sigma.pow(2.0))) | |
return gauss / gauss.sum(-1, keepdim=True) | |
def _gaussian_blur2d(input, kernel_size, sigma): | |
if isinstance(sigma, tuple): | |
sigma = torch.tensor([sigma], dtype=input.dtype) | |
else: | |
sigma = sigma.to(dtype=input.dtype) | |
ky, kx = int(kernel_size[0]), int(kernel_size[1]) | |
bs = sigma.shape[0] | |
kernel_x = _gaussian(kx, sigma[:, 1].view(bs, 1)) | |
kernel_y = _gaussian(ky, sigma[:, 0].view(bs, 1)) | |
out_x = _filter2d(input, kernel_x[..., None, :]) | |
out = _filter2d(out_x, kernel_y[..., None]) | |
return out | |
def is_compiled_module(module) -> bool: | |
"""Check whether the module was compiled with torch.compile()""" | |
if is_torch_version("<", "2.0.0") or not hasattr(torch, "_dynamo"): | |
return False | |
return isinstance(module, torch._dynamo.eval_frame.OptimizedModule) |