File size: 14,617 Bytes
c98db2a | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 | from typing import Any, Callable, Dict, List, Optional, Union
import numpy as np
import torch
try:
from diffusers import QwenImageEditPlusPipeline
from diffusers.pipelines.qwenimage.pipeline_qwenimage_edit_plus import (
CONDITION_IMAGE_SIZE,
VAE_IMAGE_SIZE,
XLA_AVAILABLE,
logger,
calculate_dimensions,
calculate_shift,
retrieve_timesteps,
)
except ImportError:
raise ImportError(
"Diffusers is out of date. Update diffusers to the latest version by doing 'pip uninstall diffusers' and then 'pip install -r requirements.txt'"
)
from diffusers.image_processor import PipelineImageInput
from diffusers.pipelines.qwenimage.pipeline_output import QwenImagePipelineOutput
class QwenImageEditPlusCustomPipeline(QwenImageEditPlusPipeline):
@torch.no_grad()
def __call__(
self,
image: Optional[PipelineImageInput] = None,
prompt: Union[str, List[str]] = None,
negative_prompt: Union[str, List[str]] = None,
true_cfg_scale: float = 4.0,
height: Optional[int] = None,
width: Optional[int] = None,
num_inference_steps: int = 50,
sigmas: Optional[List[float]] = None,
guidance_scale: Optional[float] = None,
num_images_per_prompt: int = 1,
generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
latents: Optional[torch.Tensor] = None,
prompt_embeds: Optional[torch.Tensor] = None,
prompt_embeds_mask: Optional[torch.Tensor] = None,
negative_prompt_embeds: Optional[torch.Tensor] = None,
negative_prompt_embeds_mask: Optional[torch.Tensor] = None,
output_type: Optional[str] = "pil",
return_dict: bool = True,
attention_kwargs: Optional[Dict[str, Any]] = None,
callback_on_step_end: Optional[Callable[[int, int, Dict], None]] = None,
callback_on_step_end_tensor_inputs: List[str] = ["latents"],
max_sequence_length: int = 512,
do_cfg_norm: bool = False,
):
image_size = image[-1].size if isinstance(image, list) else image.size
calculated_width, calculated_height = calculate_dimensions(
1024 * 1024, image_size[0] / image_size[1]
)
height = height or calculated_height
width = width or calculated_width
multiple_of = self.vae_scale_factor * 2
width = width // multiple_of * multiple_of
height = height // multiple_of * multiple_of
# 1. Check inputs. Raise error if not correct
self.check_inputs(
prompt,
height,
width,
negative_prompt=negative_prompt,
prompt_embeds=prompt_embeds,
negative_prompt_embeds=negative_prompt_embeds,
prompt_embeds_mask=prompt_embeds_mask,
negative_prompt_embeds_mask=negative_prompt_embeds_mask,
callback_on_step_end_tensor_inputs=callback_on_step_end_tensor_inputs,
max_sequence_length=max_sequence_length,
)
self._guidance_scale = guidance_scale
self._attention_kwargs = attention_kwargs
self._current_timestep = None
self._interrupt = False
# 2. Define call parameters
if prompt is not None and isinstance(prompt, str):
batch_size = 1
elif prompt is not None and isinstance(prompt, list):
batch_size = len(prompt)
else:
batch_size = prompt_embeds.shape[0]
device = self._execution_device
# 3. Preprocess image
if image is not None and not (
isinstance(image, torch.Tensor) and image.size(1) == self.latent_channels
):
if not isinstance(image, list):
image = [image]
condition_image_sizes = []
condition_images = []
vae_image_sizes = []
vae_images = []
for img in image:
image_width, image_height = img.size
condition_width, condition_height = calculate_dimensions(
CONDITION_IMAGE_SIZE, image_width / image_height
)
vae_width, vae_height = calculate_dimensions(
VAE_IMAGE_SIZE, image_width / image_height
)
condition_image_sizes.append((condition_width, condition_height))
vae_image_sizes.append((vae_width, vae_height))
condition_images.append(
self.image_processor.resize(img, condition_height, condition_width)
)
vae_images.append(
self.image_processor.preprocess(
img, vae_height, vae_width
).unsqueeze(2)
)
has_neg_prompt = negative_prompt is not None or (
negative_prompt_embeds is not None
and negative_prompt_embeds_mask is not None
)
if true_cfg_scale > 1 and not has_neg_prompt:
logger.warning(
f"true_cfg_scale is passed as {true_cfg_scale}, but classifier-free guidance is not enabled since no negative_prompt is provided."
)
elif true_cfg_scale <= 1 and has_neg_prompt:
logger.warning(
" negative_prompt is passed but classifier-free guidance is not enabled since true_cfg_scale <= 1"
)
do_true_cfg = true_cfg_scale > 1 and has_neg_prompt
prompt_embeds, prompt_embeds_mask = self.encode_prompt(
image=condition_images,
prompt=prompt,
prompt_embeds=prompt_embeds,
prompt_embeds_mask=prompt_embeds_mask,
device=device,
num_images_per_prompt=num_images_per_prompt,
max_sequence_length=max_sequence_length,
)
if do_true_cfg:
negative_prompt_embeds, negative_prompt_embeds_mask = self.encode_prompt(
image=condition_images,
prompt=negative_prompt,
prompt_embeds=negative_prompt_embeds,
prompt_embeds_mask=negative_prompt_embeds_mask,
device=device,
num_images_per_prompt=num_images_per_prompt,
max_sequence_length=max_sequence_length,
)
# 4. Prepare latent variables
num_channels_latents = self.transformer.config.in_channels // 4
latents, image_latents = self.prepare_latents(
vae_images,
batch_size * num_images_per_prompt,
num_channels_latents,
height,
width,
prompt_embeds.dtype,
device,
generator,
latents,
)
img_shapes = [
[
(
1,
height // self.vae_scale_factor // 2,
width // self.vae_scale_factor // 2,
),
*[
(
1,
vae_height // self.vae_scale_factor // 2,
vae_width // self.vae_scale_factor // 2,
)
for vae_width, vae_height in vae_image_sizes
],
]
] * batch_size
# 5. Prepare timesteps
sigmas = (
np.linspace(1.0, 1 / num_inference_steps, num_inference_steps)
if sigmas is None
else sigmas
)
image_seq_len = latents.shape[1]
mu = calculate_shift(
image_seq_len,
self.scheduler.config.get("base_image_seq_len", 256),
self.scheduler.config.get("max_image_seq_len", 4096),
self.scheduler.config.get("base_shift", 0.5),
self.scheduler.config.get("max_shift", 1.15),
)
timesteps, num_inference_steps = retrieve_timesteps(
self.scheduler,
num_inference_steps,
device,
sigmas=sigmas,
mu=mu,
)
num_warmup_steps = max(
len(timesteps) - num_inference_steps * self.scheduler.order, 0
)
self._num_timesteps = len(timesteps)
# handle guidance
if self.transformer.config.guidance_embeds and guidance_scale is None:
raise ValueError("guidance_scale is required for guidance-distilled model.")
elif self.transformer.config.guidance_embeds:
guidance = torch.full(
[1], guidance_scale, device=device, dtype=torch.float32
)
guidance = guidance.expand(latents.shape[0])
elif not self.transformer.config.guidance_embeds and guidance_scale is not None:
logger.warning(
f"guidance_scale is passed as {guidance_scale}, but ignored since the model is not guidance-distilled."
)
guidance = None
elif not self.transformer.config.guidance_embeds and guidance_scale is None:
guidance = None
if self.attention_kwargs is None:
self._attention_kwargs = {}
txt_seq_lens = (
prompt_embeds_mask.sum(dim=1).tolist()
if prompt_embeds_mask is not None
else None
)
negative_txt_seq_lens = (
negative_prompt_embeds_mask.sum(dim=1).tolist()
if negative_prompt_embeds_mask is not None
else None
)
# 6. Denoising loop
self.scheduler.set_begin_index(0)
with self.progress_bar(total=num_inference_steps) as progress_bar:
for i, t in enumerate(timesteps):
if self.interrupt:
continue
self._current_timestep = t
latent_model_input = latents
if image_latents is not None:
latent_model_input = torch.cat([latents, image_latents], dim=1)
# broadcast to batch dimension in a way that's compatible with ONNX/Core ML
timestep = t.expand(latents.shape[0]).to(latents.dtype)
with self.transformer.cache_context("cond"):
noise_pred = self.transformer(
hidden_states=latent_model_input,
timestep=timestep / 1000,
guidance=guidance,
encoder_hidden_states_mask=prompt_embeds_mask,
encoder_hidden_states=prompt_embeds,
img_shapes=img_shapes,
txt_seq_lens=txt_seq_lens,
attention_kwargs=self.attention_kwargs,
return_dict=False,
)[0]
noise_pred = noise_pred[:, : latents.size(1)]
if do_true_cfg:
with self.transformer.cache_context("uncond"):
neg_noise_pred = self.transformer(
hidden_states=latent_model_input,
timestep=timestep / 1000,
guidance=guidance,
encoder_hidden_states_mask=negative_prompt_embeds_mask,
encoder_hidden_states=negative_prompt_embeds,
img_shapes=img_shapes,
txt_seq_lens=negative_txt_seq_lens,
attention_kwargs=self.attention_kwargs,
return_dict=False,
)[0]
neg_noise_pred = neg_noise_pred[:, : latents.size(1)]
comb_pred = neg_noise_pred + true_cfg_scale * (
noise_pred - neg_noise_pred
)
if do_cfg_norm:
# the official code does this, but I find it hurts more often than it helps, leaving it optional but off by default
cond_norm = torch.norm(noise_pred, dim=-1, keepdim=True)
noise_norm = torch.norm(comb_pred, dim=-1, keepdim=True)
noise_pred = comb_pred * (cond_norm / noise_norm)
else:
noise_pred = comb_pred
# compute the previous noisy sample x_t -> x_t-1
latents_dtype = latents.dtype
latents = self.scheduler.step(
noise_pred, t, latents, return_dict=False
)[0]
if latents.dtype != latents_dtype:
if torch.backends.mps.is_available():
# some platforms (eg. apple mps) misbehave due to a pytorch bug: https://github.com/pytorch/pytorch/pull/99272
latents = latents.to(latents_dtype)
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)
prompt_embeds = callback_outputs.pop("prompt_embeds", prompt_embeds)
# call the callback, if provided
if i == len(timesteps) - 1 or (
(i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0
):
progress_bar.update()
if XLA_AVAILABLE:
xm.mark_step()
self._current_timestep = None
if output_type == "latent":
image = latents
else:
latents = self._unpack_latents(
latents, height, width, self.vae_scale_factor
)
latents = latents.to(self.vae.dtype)
latents_mean = (
torch.tensor(self.vae.config.latents_mean)
.view(1, self.vae.config.z_dim, 1, 1, 1)
.to(latents.device, latents.dtype)
)
latents_std = 1.0 / torch.tensor(self.vae.config.latents_std).view(
1, self.vae.config.z_dim, 1, 1, 1
).to(latents.device, latents.dtype)
latents = latents / latents_std + latents_mean
image = self.vae.decode(latents, return_dict=False)[0][:, :, 0]
image = self.image_processor.postprocess(image, output_type=output_type)
# Offload all models
self.maybe_free_model_hooks()
if not return_dict:
return (image,)
return QwenImagePipelineOutput(images=image)
|