File size: 11,813 Bytes
be9fa39 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 |
# Copyright 2025 Lightricks and 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.
from typing import List, Optional, Union
import torch
from ...image_processor import PipelineImageInput
from ...models import AutoencoderKLLTXVideo
from ...utils import get_logger
from ...utils.torch_utils import randn_tensor
from ...video_processor import VideoProcessor
from ..pipeline_utils import DiffusionPipeline
from .modeling_latent_upsampler import LTXLatentUpsamplerModel
from .pipeline_output import LTXPipelineOutput
logger = get_logger(__name__) # pylint: disable=invalid-name
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.retrieve_latents
def retrieve_latents(
encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = "sample"
):
if hasattr(encoder_output, "latent_dist") and sample_mode == "sample":
return encoder_output.latent_dist.sample(generator)
elif hasattr(encoder_output, "latent_dist") and sample_mode == "argmax":
return encoder_output.latent_dist.mode()
elif hasattr(encoder_output, "latents"):
return encoder_output.latents
else:
raise AttributeError("Could not access latents of provided encoder_output")
class LTXLatentUpsamplePipeline(DiffusionPipeline):
model_cpu_offload_seq = ""
def __init__(
self,
vae: AutoencoderKLLTXVideo,
latent_upsampler: LTXLatentUpsamplerModel,
) -> None:
super().__init__()
self.register_modules(vae=vae, latent_upsampler=latent_upsampler)
self.vae_spatial_compression_ratio = (
self.vae.spatial_compression_ratio if getattr(self, "vae", None) is not None else 32
)
self.vae_temporal_compression_ratio = (
self.vae.temporal_compression_ratio if getattr(self, "vae", None) is not None else 8
)
self.video_processor = VideoProcessor(vae_scale_factor=self.vae_spatial_compression_ratio)
def prepare_latents(
self,
video: Optional[torch.Tensor] = None,
batch_size: int = 1,
dtype: Optional[torch.dtype] = None,
device: Optional[torch.device] = None,
generator: Optional[torch.Generator] = None,
latents: Optional[torch.Tensor] = None,
) -> torch.Tensor:
if latents is not None:
return latents.to(device=device, dtype=dtype)
video = video.to(device=device, dtype=self.vae.dtype)
if isinstance(generator, list):
if 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."
)
init_latents = [
retrieve_latents(self.vae.encode(video[i].unsqueeze(0)), generator[i]) for i in range(batch_size)
]
else:
init_latents = [retrieve_latents(self.vae.encode(vid.unsqueeze(0)), generator) for vid in video]
init_latents = torch.cat(init_latents, dim=0).to(dtype)
init_latents = self._normalize_latents(init_latents, self.vae.latents_mean, self.vae.latents_std)
return init_latents
def adain_filter_latent(self, latents: torch.Tensor, reference_latents: torch.Tensor, factor: float = 1.0):
"""
Applies Adaptive Instance Normalization (AdaIN) to a latent tensor based on statistics from a reference latent
tensor.
Args:
latent (`torch.Tensor`):
Input latents to normalize
reference_latents (`torch.Tensor`):
The reference latents providing style statistics.
factor (`float`):
Blending factor between original and transformed latent. Range: -10.0 to 10.0, Default: 1.0
Returns:
torch.Tensor: The transformed latent tensor
"""
result = latents.clone()
for i in range(latents.size(0)):
for c in range(latents.size(1)):
r_sd, r_mean = torch.std_mean(reference_latents[i, c], dim=None) # index by original dim order
i_sd, i_mean = torch.std_mean(result[i, c], dim=None)
result[i, c] = ((result[i, c] - i_mean) / i_sd) * r_sd + r_mean
result = torch.lerp(latents, result, factor)
return result
@staticmethod
# Copied from diffusers.pipelines.ltx.pipeline_ltx.LTXPipeline._normalize_latents
def _normalize_latents(
latents: torch.Tensor, latents_mean: torch.Tensor, latents_std: torch.Tensor, scaling_factor: float = 1.0
) -> torch.Tensor:
# Normalize latents across the channel dimension [B, C, F, H, W]
latents_mean = latents_mean.view(1, -1, 1, 1, 1).to(latents.device, latents.dtype)
latents_std = latents_std.view(1, -1, 1, 1, 1).to(latents.device, latents.dtype)
latents = (latents - latents_mean) * scaling_factor / latents_std
return latents
@staticmethod
# Copied from diffusers.pipelines.ltx.pipeline_ltx.LTXPipeline._denormalize_latents
def _denormalize_latents(
latents: torch.Tensor, latents_mean: torch.Tensor, latents_std: torch.Tensor, scaling_factor: float = 1.0
) -> torch.Tensor:
# Denormalize latents across the channel dimension [B, C, F, H, W]
latents_mean = latents_mean.view(1, -1, 1, 1, 1).to(latents.device, latents.dtype)
latents_std = latents_std.view(1, -1, 1, 1, 1).to(latents.device, latents.dtype)
latents = latents * latents_std / scaling_factor + latents_mean
return latents
def enable_vae_slicing(self):
r"""
Enable sliced VAE decoding. When this option is enabled, the VAE will split the input tensor in slices to
compute decoding in several steps. This is useful to save some memory and allow larger batch sizes.
"""
self.vae.enable_slicing()
def disable_vae_slicing(self):
r"""
Disable sliced VAE decoding. If `enable_vae_slicing` was previously enabled, this method will go back to
computing decoding in one step.
"""
self.vae.disable_slicing()
def enable_vae_tiling(self):
r"""
Enable tiled VAE decoding. When this option is enabled, the VAE will split the input tensor into tiles to
compute decoding and encoding in several steps. This is useful for saving a large amount of memory and to allow
processing larger images.
"""
self.vae.enable_tiling()
def disable_vae_tiling(self):
r"""
Disable tiled VAE decoding. If `enable_vae_tiling` was previously enabled, this method will go back to
computing decoding in one step.
"""
self.vae.disable_tiling()
def check_inputs(self, video, height, width, latents):
if height % self.vae_spatial_compression_ratio != 0 or width % self.vae_spatial_compression_ratio != 0:
raise ValueError(f"`height` and `width` have to be divisible by 32 but are {height} and {width}.")
if video is not None and latents is not None:
raise ValueError("Only one of `video` or `latents` can be provided.")
if video is None and latents is None:
raise ValueError("One of `video` or `latents` has to be provided.")
@torch.no_grad()
def __call__(
self,
video: Optional[List[PipelineImageInput]] = None,
height: int = 512,
width: int = 704,
latents: Optional[torch.Tensor] = None,
decode_timestep: Union[float, List[float]] = 0.0,
decode_noise_scale: Optional[Union[float, List[float]]] = None,
adain_factor: float = 0.0,
generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
output_type: Optional[str] = "pil",
return_dict: bool = True,
):
self.check_inputs(
video=video,
height=height,
width=width,
latents=latents,
)
if video is not None:
# Batched video input is not yet tested/supported. TODO: take a look later
batch_size = 1
else:
batch_size = latents.shape[0]
device = self._execution_device
if video is not None:
num_frames = len(video)
if num_frames % self.vae_temporal_compression_ratio != 1:
num_frames = (
num_frames // self.vae_temporal_compression_ratio * self.vae_temporal_compression_ratio + 1
)
video = video[:num_frames]
logger.warning(
f"Video length expected to be of the form `k * {self.vae_temporal_compression_ratio} + 1` but is {len(video)}. Truncating to {num_frames} frames."
)
video = self.video_processor.preprocess_video(video, height=height, width=width)
video = video.to(device=device, dtype=torch.float32)
latents = self.prepare_latents(
video=video,
batch_size=batch_size,
dtype=torch.float32,
device=device,
generator=generator,
latents=latents,
)
latents = self._denormalize_latents(
latents, self.vae.latents_mean, self.vae.latents_std, self.vae.config.scaling_factor
)
latents = latents.to(self.latent_upsampler.dtype)
latents_upsampled = self.latent_upsampler(latents)
if adain_factor > 0.0:
latents = self.adain_filter_latent(latents_upsampled, latents, adain_factor)
else:
latents = latents_upsampled
if output_type == "latent":
latents = self._normalize_latents(
latents, self.vae.latents_mean, self.vae.latents_std, self.vae.config.scaling_factor
)
video = latents
else:
if not self.vae.config.timestep_conditioning:
timestep = None
else:
noise = randn_tensor(latents.shape, generator=generator, device=device, dtype=latents.dtype)
if not isinstance(decode_timestep, list):
decode_timestep = [decode_timestep] * batch_size
if decode_noise_scale is None:
decode_noise_scale = decode_timestep
elif not isinstance(decode_noise_scale, list):
decode_noise_scale = [decode_noise_scale] * batch_size
timestep = torch.tensor(decode_timestep, device=device, dtype=latents.dtype)
decode_noise_scale = torch.tensor(decode_noise_scale, device=device, dtype=latents.dtype)[
:, None, None, None, None
]
latents = (1 - decode_noise_scale) * latents + decode_noise_scale * noise
video = self.vae.decode(latents, timestep, return_dict=False)[0]
video = self.video_processor.postprocess_video(video, output_type=output_type)
# Offload all models
self.maybe_free_model_hooks()
if not return_dict:
return (video,)
return LTXPipelineOutput(frames=video)
|