Spaces:
Paused
Paused
File size: 12,356 Bytes
cd458ae | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 | # Copyright 2026 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 ...utils import logging
from ..modular_pipeline import AutoPipelineBlocks, SequentialPipelineBlocks
from ..modular_pipeline_utils import InsertableDict, OutputParam
from .before_denoise import (
Flux2PrepareGuidanceStep,
Flux2PrepareImageLatentsStep,
Flux2PrepareLatentsStep,
Flux2RoPEInputsStep,
Flux2SetTimestepsStep,
)
from .decoders import Flux2DecodeStep, Flux2UnpackLatentsStep
from .denoise import Flux2DenoiseStep
from .encoders import (
Flux2TextEncoderStep,
Flux2VaeEncoderStep,
)
from .inputs import (
Flux2ProcessImagesInputStep,
Flux2TextInputStep,
)
logger = logging.get_logger(__name__) # pylint: disable=invalid-name
# auto_docstring
class Flux2VaeEncoderSequentialStep(SequentialPipelineBlocks):
"""
VAE encoder step that preprocesses, encodes, and prepares image latents for Flux2 conditioning.
Components:
image_processor (`Flux2ImageProcessor`) vae (`AutoencoderKLFlux2`)
Inputs:
image (`None`, *optional*):
TODO: Add description.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
Outputs:
condition_images (`list`):
TODO: Add description.
image_latents (`list`):
List of latent representations for each reference image
"""
model_name = "flux2"
block_classes = [Flux2ProcessImagesInputStep(), Flux2VaeEncoderStep()]
block_names = ["preprocess", "encode"]
@property
def description(self) -> str:
return "VAE encoder step that preprocesses, encodes, and prepares image latents for Flux2 conditioning."
# auto_docstring
class Flux2AutoVaeEncoderStep(AutoPipelineBlocks):
"""
VAE encoder step that encodes the image inputs into their latent representations.
This is an auto pipeline block that works for image conditioning tasks.
- `Flux2VaeEncoderSequentialStep` is used when `image` is provided.
- If `image` is not provided, step will be skipped.
Components:
image_processor (`Flux2ImageProcessor`) vae (`AutoencoderKLFlux2`)
Inputs:
image (`None`, *optional*):
TODO: Add description.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
Outputs:
condition_images (`list`):
TODO: Add description.
image_latents (`list`):
List of latent representations for each reference image
"""
block_classes = [Flux2VaeEncoderSequentialStep]
block_names = ["img_conditioning"]
block_trigger_inputs = ["image"]
@property
def description(self):
return (
"VAE encoder step that encodes the image inputs into their latent representations.\n"
"This is an auto pipeline block that works for image conditioning tasks.\n"
" - `Flux2VaeEncoderSequentialStep` is used when `image` is provided.\n"
" - If `image` is not provided, step will be skipped."
)
Flux2CoreDenoiseBlocks = InsertableDict(
[
("input", Flux2TextInputStep()),
("prepare_latents", Flux2PrepareLatentsStep()),
("set_timesteps", Flux2SetTimestepsStep()),
("prepare_guidance", Flux2PrepareGuidanceStep()),
("prepare_rope_inputs", Flux2RoPEInputsStep()),
("denoise", Flux2DenoiseStep()),
("after_denoise", Flux2UnpackLatentsStep()),
]
)
# auto_docstring
class Flux2CoreDenoiseStep(SequentialPipelineBlocks):
"""
Core denoise step that performs the denoising process for Flux2-dev.
Components:
scheduler (`FlowMatchEulerDiscreteScheduler`) transformer (`Flux2Transformer2DModel`)
Inputs:
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
prompt_embeds (`Tensor`):
Pre-generated text embeddings. Can be generated from text_encoder step.
height (`int`, *optional*):
TODO: Add description.
width (`int`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_inference_steps (`None`, *optional*, defaults to 50):
TODO: Add description.
timesteps (`None`, *optional*):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
guidance_scale (`None`, *optional*, defaults to 4.0):
TODO: Add description.
joint_attention_kwargs (`None`, *optional*):
TODO: Add description.
image_latents (`Tensor`, *optional*):
Packed image latents for conditioning. Shape: (B, img_seq_len, C)
image_latent_ids (`Tensor`, *optional*):
Position IDs for image latents. Shape: (B, img_seq_len, 4)
Outputs:
latents (`Tensor`):
Denoised latents.
"""
model_name = "flux2"
block_classes = Flux2CoreDenoiseBlocks.values()
block_names = Flux2CoreDenoiseBlocks.keys()
@property
def description(self):
return "Core denoise step that performs the denoising process for Flux2-dev."
@property
def outputs(self):
return [
OutputParam.template("latents"),
]
Flux2ImageConditionedCoreDenoiseBlocks = InsertableDict(
[
("input", Flux2TextInputStep()),
("prepare_image_latents", Flux2PrepareImageLatentsStep()),
("prepare_latents", Flux2PrepareLatentsStep()),
("set_timesteps", Flux2SetTimestepsStep()),
("prepare_guidance", Flux2PrepareGuidanceStep()),
("prepare_rope_inputs", Flux2RoPEInputsStep()),
("denoise", Flux2DenoiseStep()),
("after_denoise", Flux2UnpackLatentsStep()),
]
)
# auto_docstring
class Flux2ImageConditionedCoreDenoiseStep(SequentialPipelineBlocks):
"""
Core denoise step that performs the denoising process for Flux2-dev with image conditioning.
Components:
scheduler (`FlowMatchEulerDiscreteScheduler`) transformer (`Flux2Transformer2DModel`)
Inputs:
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
prompt_embeds (`Tensor`):
Pre-generated text embeddings. Can be generated from text_encoder step.
image_latents (`list`, *optional*):
TODO: Add description.
height (`int`, *optional*):
TODO: Add description.
width (`int`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_inference_steps (`None`, *optional*, defaults to 50):
TODO: Add description.
timesteps (`None`, *optional*):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
guidance_scale (`None`, *optional*, defaults to 4.0):
TODO: Add description.
joint_attention_kwargs (`None`, *optional*):
TODO: Add description.
Outputs:
latents (`Tensor`):
Denoised latents.
"""
model_name = "flux2"
block_classes = Flux2ImageConditionedCoreDenoiseBlocks.values()
block_names = Flux2ImageConditionedCoreDenoiseBlocks.keys()
@property
def description(self):
return "Core denoise step that performs the denoising process for Flux2-dev with image conditioning."
@property
def outputs(self):
return [
OutputParam.template("latents"),
]
class Flux2AutoCoreDenoiseStep(AutoPipelineBlocks):
model_name = "flux2"
block_classes = [Flux2ImageConditionedCoreDenoiseStep, Flux2CoreDenoiseStep]
block_names = ["image_conditioned", "text2image"]
block_trigger_inputs = ["image_latents", None]
@property
def description(self):
return (
"Auto core denoise step that performs the denoising process for Flux2-dev."
"This is an auto pipeline block that works for text-to-image and image-conditioned generation."
" - `Flux2CoreDenoiseStep` is used for text-to-image generation.\n"
" - `Flux2ImageConditionedCoreDenoiseStep` is used for image-conditioned generation.\n"
)
AUTO_BLOCKS = InsertableDict(
[
("text_encoder", Flux2TextEncoderStep()),
("vae_encoder", Flux2AutoVaeEncoderStep()),
("denoise", Flux2AutoCoreDenoiseStep()),
("decode", Flux2DecodeStep()),
]
)
# auto_docstring
class Flux2AutoBlocks(SequentialPipelineBlocks):
"""
Auto Modular pipeline for text-to-image and image-conditioned generation using Flux2.
Supported workflows:
- `text2image`: requires `prompt`
- `image_conditioned`: requires `image`, `prompt`
Components:
text_encoder (`Mistral3ForConditionalGeneration`) tokenizer (`AutoProcessor`) image_processor
(`Flux2ImageProcessor`) vae (`AutoencoderKLFlux2`) scheduler (`FlowMatchEulerDiscreteScheduler`) transformer
(`Flux2Transformer2DModel`)
Inputs:
prompt (`None`, *optional*):
TODO: Add description.
max_sequence_length (`int`, *optional*, defaults to 512):
TODO: Add description.
text_encoder_out_layers (`tuple`, *optional*, defaults to (10, 20, 30)):
TODO: Add description.
image (`None`, *optional*):
TODO: Add description.
height (`None`, *optional*):
TODO: Add description.
width (`None`, *optional*):
TODO: Add description.
generator (`None`, *optional*):
TODO: Add description.
num_images_per_prompt (`None`, *optional*, defaults to 1):
TODO: Add description.
image_latents (`list`, *optional*):
TODO: Add description.
latents (`Tensor | NoneType`):
TODO: Add description.
num_inference_steps (`None`):
TODO: Add description.
timesteps (`None`):
TODO: Add description.
sigmas (`None`, *optional*):
TODO: Add description.
guidance_scale (`None`, *optional*, defaults to 4.0):
TODO: Add description.
joint_attention_kwargs (`None`, *optional*):
TODO: Add description.
image_latent_ids (`Tensor`, *optional*):
Position IDs for image latents. Shape: (B, img_seq_len, 4)
output_type (`None`, *optional*, defaults to pil):
TODO: Add description.
Outputs:
images (`list`):
Generated images.
"""
model_name = "flux2"
block_classes = AUTO_BLOCKS.values()
block_names = AUTO_BLOCKS.keys()
_workflow_map = {
"text2image": {"prompt": True},
"image_conditioned": {"image": True, "prompt": True},
}
@property
def description(self):
return "Auto Modular pipeline for text-to-image and image-conditioned generation using Flux2."
@property
def outputs(self):
return [
OutputParam.template("images"),
]
|