diffusers 0.30.3__py3-none-any.whl → 0.32.0__py3-none-any.whl

diffusers/pipelines/mochi/pipeline_output.py

@@ -0,0 +1,20 @@
+from dataclasses import dataclass
+
+import torch
+
+from diffusers.utils import BaseOutput
+
+
+@dataclass
+class MochiPipelineOutput(BaseOutput):
+    r"""
+    Output class for Mochi pipelines.
+
+    Args:
+        frames (`torch.Tensor`, `np.ndarray`, or List[List[PIL.Image.Image]]):
+            List of video outputs - It can be a nested list of length `batch_size,` with each sub-list containing
+            denoised PIL image sequences of length `num_frames.` It can also be a NumPy array or Torch tensor of shape
+            `(batch_size, num_frames, channels, height, width)`.
+    """
+
+    frames: torch.Tensor

diffusers/pipelines/pag/__init__.py

@@ -23,13 +23,20 @@ except OptionalDependencyNotAvailable:
     _dummy_objects.update(get_objects_from_module(dummy_torch_and_transformers_objects))
 else:
     _import_structure["pipeline_pag_controlnet_sd"] = ["StableDiffusionControlNetPAGPipeline"]
+    _import_structure["pipeline_pag_controlnet_sd_inpaint"] = ["StableDiffusionControlNetPAGInpaintPipeline"]
     _import_structure["pipeline_pag_controlnet_sd_xl"] = ["StableDiffusionXLControlNetPAGPipeline"]
+    _import_structure["pipeline_pag_controlnet_sd_xl_img2img"] = ["StableDiffusionXLControlNetPAGImg2ImgPipeline"]
     _import_structure["pipeline_pag_hunyuandit"] = ["HunyuanDiTPAGPipeline"]
     _import_structure["pipeline_pag_kolors"] = ["KolorsPAGPipeline"]
     _import_structure["pipeline_pag_pixart_sigma"] = ["PixArtSigmaPAGPipeline"]
+    _import_structure["pipeline_pag_sana"] = ["SanaPAGPipeline"]
     _import_structure["pipeline_pag_sd"] = ["StableDiffusionPAGPipeline"]
     _import_structure["pipeline_pag_sd_3"] = ["StableDiffusion3PAGPipeline"]
+    _import_structure["pipeline_pag_sd_3_img2img"] = ["StableDiffusion3PAGImg2ImgPipeline"]
     _import_structure["pipeline_pag_sd_animatediff"] = ["AnimateDiffPAGPipeline"]
+    _import_structure["pipeline_pag_sd_img2img"] = ["StableDiffusionPAGImg2ImgPipeline"]
+    _import_structure["pipeline_pag_sd_inpaint"] = ["StableDiffusionPAGInpaintPipeline"]
+
     _import_structure["pipeline_pag_sd_xl"] = ["StableDiffusionXLPAGPipeline"]
     _import_structure["pipeline_pag_sd_xl_img2img"] = ["StableDiffusionXLPAGImg2ImgPipeline"]
     _import_structure["pipeline_pag_sd_xl_inpaint"] = ["StableDiffusionXLPAGInpaintPipeline"]
@@ -43,13 +50,19 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
         from ...utils.dummy_torch_and_transformers_objects import *
     else:
         from .pipeline_pag_controlnet_sd import StableDiffusionControlNetPAGPipeline
+        from .pipeline_pag_controlnet_sd_inpaint import StableDiffusionControlNetPAGInpaintPipeline
         from .pipeline_pag_controlnet_sd_xl import StableDiffusionXLControlNetPAGPipeline
+        from .pipeline_pag_controlnet_sd_xl_img2img import StableDiffusionXLControlNetPAGImg2ImgPipeline
         from .pipeline_pag_hunyuandit import HunyuanDiTPAGPipeline
         from .pipeline_pag_kolors import KolorsPAGPipeline
         from .pipeline_pag_pixart_sigma import PixArtSigmaPAGPipeline
+        from .pipeline_pag_sana import SanaPAGPipeline
         from .pipeline_pag_sd import StableDiffusionPAGPipeline
         from .pipeline_pag_sd_3 import StableDiffusion3PAGPipeline
+        from .pipeline_pag_sd_3_img2img import StableDiffusion3PAGImg2ImgPipeline
         from .pipeline_pag_sd_animatediff import AnimateDiffPAGPipeline
+        from .pipeline_pag_sd_img2img import StableDiffusionPAGImg2ImgPipeline
+        from .pipeline_pag_sd_inpaint import StableDiffusionPAGInpaintPipeline
         from .pipeline_pag_sd_xl import StableDiffusionXLPAGPipeline
         from .pipeline_pag_sd_xl_img2img import StableDiffusionXLPAGImg2ImgPipeline
         from .pipeline_pag_sd_xl_inpaint import StableDiffusionXLPAGInpaintPipeline

diffusers/pipelines/pag/pag_utils.py

@@ -98,7 +98,9 @@ class PAGMixin:
         else:
             return self.pag_scale
 
-    def _apply_perturbed_attention_guidance(self, noise_pred, do_classifier_free_guidance, guidance_scale, t):
+    def _apply_perturbed_attention_guidance(
+        self, noise_pred, do_classifier_free_guidance, guidance_scale, t, return_pred_text=False
+    ):
         r"""
         Apply perturbed attention guidance to the noise prediction.
 
@@ -107,9 +109,11 @@ class PAGMixin:
             do_classifier_free_guidance (bool): Whether to apply classifier-free guidance.
             guidance_scale (float): The scale factor for the guidance term.
             t (int): The current time step.
+            return_pred_text (bool): Whether to return the text noise prediction.
 
         Returns:
-            torch.Tensor: The updated noise prediction tensor after applying perturbed attention guidance.
+            Union[torch.Tensor, Tuple[torch.Tensor, torch.Tensor]]: The updated noise prediction tensor after applying
+            perturbed attention guidance and the text noise prediction.
         """
         pag_scale = self._get_pag_scale(t)
         if do_classifier_free_guidance:
@@ -122,6 +126,8 @@ class PAGMixin:
         else:
             noise_pred_text, noise_pred_perturb = noise_pred.chunk(2)
             noise_pred = noise_pred_text + pag_scale * (noise_pred_text - noise_pred_perturb)
+        if return_pred_text:
+            return noise_pred, noise_pred_text
         return noise_pred
 
     def _prepare_perturbed_attention_guidance(self, cond, uncond, do_classifier_free_guidance):

diffusers/pipelines/pag/pipeline_pag_controlnet_sd.py

@@ -25,7 +25,7 @@ from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer, CLIPV
 from ...callbacks import MultiPipelineCallbacks, PipelineCallback
 from ...image_processor import PipelineImageInput, VaeImageProcessor
 from ...loaders import FromSingleFileMixin, IPAdapterMixin, StableDiffusionLoraLoaderMixin, TextualInversionLoaderMixin
-from ...models import AutoencoderKL, ControlNetModel, ImageProjection, UNet2DConditionModel
+from ...models import AutoencoderKL, ControlNetModel, ImageProjection, MultiControlNetModel, UNet2DConditionModel
 from ...models.lora import adjust_lora_scale_text_encoder
 from ...schedulers import KarrasDiffusionSchedulers
 from ...utils import (
@@ -36,7 +36,6 @@ from ...utils import (
     unscale_lora_layers,
 )
 from ...utils.torch_utils import is_compiled_module, is_torch_version, randn_tensor
-from ..controlnet.multicontrolnet import MultiControlNetModel
 from ..pipeline_utils import DiffusionPipeline, StableDiffusionMixin
 from ..stable_diffusion.pipeline_output import StableDiffusionPipelineOutput
 from ..stable_diffusion.safety_checker import StableDiffusionSafetyChecker
@@ -104,7 +103,7 @@ def retrieve_timesteps(
     sigmas: Optional[List[float]] = None,
     **kwargs,
 ):
-    """
+    r"""
     Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles
     custom timesteps. Any kwargs will be supplied to `scheduler.set_timesteps`.