diffusers 0.23.1__py3-none-any.whl → 0.25.0__py3-none-any.whl

diffusers/pipelines/{alt_diffusion → deprecated/alt_diffusion}/pipeline_alt_diffusion_img2img.py

@@ -19,15 +19,16 @@ import numpy as np
 import PIL.Image
 import torch
 from packaging import version
-from transformers import CLIPImageProcessor, XLMRobertaTokenizer
-
-from ...configuration_utils import FrozenDict
-from ...image_processor import PipelineImageInput, VaeImageProcessor
-from ...loaders import FromSingleFileMixin, LoraLoaderMixin, TextualInversionLoaderMixin
-from ...models import AutoencoderKL, UNet2DConditionModel
-from ...models.lora import adjust_lora_scale_text_encoder
-from ...schedulers import KarrasDiffusionSchedulers
-from ...utils import (
+from transformers import CLIPImageProcessor, CLIPVisionModelWithProjection, XLMRobertaTokenizer
+
+from ....configuration_utils import FrozenDict
+from ....image_processor import PipelineImageInput, VaeImageProcessor
+from ....loaders import FromSingleFileMixin, IPAdapterMixin, LoraLoaderMixin, TextualInversionLoaderMixin
+from ....models import AutoencoderKL, ImageProjection, UNet2DConditionModel
+from ....models.attention_processor import FusedAttnProcessor2_0
+from ....models.lora import adjust_lora_scale_text_encoder
+from ....schedulers import KarrasDiffusionSchedulers
+from ....utils import (
     PIL_INTERPOLATION,
     USE_PEFT_BACKEND,
     deprecate,
@@ -36,9 +37,9 @@ from ...utils import (
     scale_lora_layers,
     unscale_lora_layers,
 )
-from ...utils.torch_utils import randn_tensor
-from ..pipeline_utils import DiffusionPipeline
-from ..stable_diffusion.safety_checker import StableDiffusionSafetyChecker
+from ....utils.torch_utils import randn_tensor
+from ...pipeline_utils import DiffusionPipeline
+from ...stable_diffusion.safety_checker import StableDiffusionSafetyChecker
 from .modeling_roberta_series import RobertaSeriesModelWithTransformation
 from .pipeline_output import AltDiffusionPipelineOutput
 
@@ -76,9 +77,13 @@ EXAMPLE_DOC_STRING = """
 
 
 # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.retrieve_latents
-def retrieve_latents(encoder_output, generator):
-    if hasattr(encoder_output, "latent_dist"):
+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:
@@ -109,9 +114,53 @@ def preprocess(image):
     return image
 
 
-# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.StableDiffusionImg2ImgPipeline with Stable->Alt, CLIPTextModel->RobertaSeriesModelWithTransformation, CLIPTokenizer->XLMRobertaTokenizer, AltDiffusionSafetyChecker->StableDiffusionSafetyChecker
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.retrieve_timesteps
+def retrieve_timesteps(
+    scheduler,
+    num_inference_steps: Optional[int] = None,
+    device: Optional[Union[str, torch.device]] = None,
+    timesteps: Optional[List[int]] = None,
+    **kwargs,
+):
+    """
+    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`.
+
+    Args:
+        scheduler (`SchedulerMixin`):
+            The scheduler to get timesteps from.
+        num_inference_steps (`int`):
+            The number of diffusion steps used when generating samples with a pre-trained model. If used,
+            `timesteps` must be `None`.
+        device (`str` or `torch.device`, *optional*):
+            The device to which the timesteps should be moved to. If `None`, the timesteps are not moved.
+        timesteps (`List[int]`, *optional*):
+                Custom timesteps used to support arbitrary spacing between timesteps. If `None`, then the default
+                timestep spacing strategy of the scheduler is used. If `timesteps` is passed, `num_inference_steps`
+                must be `None`.
+
+    Returns:
+        `Tuple[torch.Tensor, int]`: A tuple where the first element is the timestep schedule from the scheduler and the
+        second element is the number of inference steps.
+    """
+    if timesteps is not None:
+        accepts_timesteps = "timesteps" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accepts_timesteps:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" timestep schedules. Please check whether you are using the correct scheduler."
+            )
+        scheduler.set_timesteps(timesteps=timesteps, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+        num_inference_steps = len(timesteps)
+    else:
+        scheduler.set_timesteps(num_inference_steps, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+    return timesteps, num_inference_steps
+
+
 class AltDiffusionImg2ImgPipeline(
-    DiffusionPipeline, TextualInversionLoaderMixin, LoraLoaderMixin, FromSingleFileMixin
+    DiffusionPipeline, TextualInversionLoaderMixin, IPAdapterMixin, LoraLoaderMixin, FromSingleFileMixin
 ):
     r"""
     Pipeline for text-guided image-to-image generation using Alt Diffusion.
@@ -124,6 +173,7 @@ class AltDiffusionImg2ImgPipeline(
         - [`~loaders.LoraLoaderMixin.load_lora_weights`] for loading LoRA weights
         - [`~loaders.LoraLoaderMixin.save_lora_weights`] for saving LoRA weights
         - [`~loaders.FromSingleFileMixin.from_single_file`] for loading `.ckpt` files
+        - [`~loaders.IPAdapterMixin.load_ip_adapter`] for loading IP Adapters
 
     Args:
         vae ([`AutoencoderKL`]):
@@ -145,8 +195,8 @@ class AltDiffusionImg2ImgPipeline(
             A `CLIPImageProcessor` to extract features from generated images; used as inputs to the `safety_checker`.
     """
 
-    model_cpu_offload_seq = "text_encoder->unet->vae"
-    _optional_components = ["safety_checker", "feature_extractor"]
+    model_cpu_offload_seq = "text_encoder->image_encoder->unet->vae"
+    _optional_components = ["safety_checker", "feature_extractor", "image_encoder"]
     _exclude_from_cpu_offload = ["safety_checker"]
     _callback_tensor_inputs = ["latents", "prompt_embeds", "negative_prompt_embeds"]
 
@@ -159,6 +209,7 @@ class AltDiffusionImg2ImgPipeline(
         scheduler: KarrasDiffusionSchedulers,
         safety_checker: StableDiffusionSafetyChecker,
         feature_extractor: CLIPImageProcessor,
+        image_encoder: CLIPVisionModelWithProjection = None,
         requires_safety_checker: bool = True,
     ):
         super().__init__()
@@ -235,6 +286,7 @@ class AltDiffusionImg2ImgPipeline(
             scheduler=scheduler,
             safety_checker=safety_checker,
             feature_extractor=feature_extractor,
+            image_encoder=image_encoder,
         )
         self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
@@ -252,10 +304,7 @@ class AltDiffusionImg2ImgPipeline(
         lora_scale: Optional[float] = None,
         **kwargs,
     ):
-        deprecation_message = (
-            "`_encode_prompt()` is deprecated and it will be removed in a future version. Use `encode_prompt()`"
-            " instead. Also, be aware that the output format changed from a concatenated tensor to a tuple."
-        )
+        deprecation_message = "`_encode_prompt()` is deprecated and it will be removed in a future version. Use `encode_prompt()` instead. Also, be aware that the output format changed from a concatenated tensor to a tuple."
         deprecate("_encode_prompt()", "1.0.0", deprecation_message, standard_warn=False)
 
         prompt_embeds_tuple = self.encode_prompt(
@@ -456,6 +505,30 @@ class AltDiffusionImg2ImgPipeline(
 
         return prompt_embeds, negative_prompt_embeds
 
+    def encode_image(self, image, device, num_images_per_prompt, output_hidden_states=None):
+        dtype = next(self.image_encoder.parameters()).dtype
+
+        if not isinstance(image, torch.Tensor):
+            image = self.feature_extractor(image, return_tensors="pt").pixel_values
+
+        image = image.to(device=device, dtype=dtype)
+        if output_hidden_states:
+            image_enc_hidden_states = self.image_encoder(image, output_hidden_states=True).hidden_states[-2]
+            image_enc_hidden_states = image_enc_hidden_states.repeat_interleave(num_images_per_prompt, dim=0)
+            uncond_image_enc_hidden_states = self.image_encoder(
+                torch.zeros_like(image), output_hidden_states=True
+            ).hidden_states[-2]
+            uncond_image_enc_hidden_states = uncond_image_enc_hidden_states.repeat_interleave(
+                num_images_per_prompt, dim=0
+            )
+            return image_enc_hidden_states, uncond_image_enc_hidden_states
+        else:
+            image_embeds = self.image_encoder(image).image_embeds
+            image_embeds = image_embeds.repeat_interleave(num_images_per_prompt, dim=0)
+            uncond_image_embeds = torch.zeros_like(image_embeds)
+
+            return image_embeds, uncond_image_embeds
+
     def run_safety_checker(self, image, device, dtype):
         if self.safety_checker is None:
             has_nsfw_concept = None
@@ -471,10 +544,7 @@ class AltDiffusionImg2ImgPipeline(
         return image, has_nsfw_concept
 
     def decode_latents(self, latents):
-        deprecation_message = (
-            "The decode_latents method is deprecated and will be removed in 1.0.0. Please use"
-            " VaeImageProcessor.postprocess(...) instead"
-        )
+        deprecation_message = "The decode_latents method is deprecated and will be removed in 1.0.0. Please use VaeImageProcessor.postprocess(...) instead"
         deprecate("decode_latents", "1.0.0", deprecation_message, standard_warn=False)
 
         latents = 1 / self.vae.config.scaling_factor * latents
@@ -524,8 +594,7 @@ class AltDiffusionImg2ImgPipeline(
             k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs
         ):
             raise ValueError(
-                f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found"
-                f" {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}"
+                f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}"
             )
         if prompt is not None and prompt_embeds is not None:
             raise ValueError(
@@ -578,8 +647,8 @@ class AltDiffusionImg2ImgPipeline(
         else:
             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"
-                    f" batch size of {batch_size}. Make sure the batch size matches the length of the generators."
+                    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."
                 )
 
             elif isinstance(generator, list):
@@ -646,6 +715,65 @@ class AltDiffusionImg2ImgPipeline(
         """Disables the FreeU mechanism if enabled."""
         self.unet.disable_freeu()
 
+    def fuse_qkv_projections(self, unet: bool = True, vae: bool = True):
+        """
+        Enables fused QKV projections. For self-attention modules, all projection matrices (i.e., query,
+        key, value) are fused. For cross-attention modules, key and value projection matrices are fused.
+
+        <Tip warning={true}>
+
+        This API is 🧪 experimental.
+
+        </Tip>
+
+        Args:
+            unet (`bool`, defaults to `True`): To apply fusion on the UNet.
+            vae (`bool`, defaults to `True`): To apply fusion on the VAE.
+        """
+        self.fusing_unet = False
+        self.fusing_vae = False
+
+        if unet:
+            self.fusing_unet = True
+            self.unet.fuse_qkv_projections()
+            self.unet.set_attn_processor(FusedAttnProcessor2_0())
+
+        if vae:
+            if not isinstance(self.vae, AutoencoderKL):
+                raise ValueError("`fuse_qkv_projections()` is only supported for the VAE of type `AutoencoderKL`.")
+
+            self.fusing_vae = True
+            self.vae.fuse_qkv_projections()
+            self.vae.set_attn_processor(FusedAttnProcessor2_0())
+
+    def unfuse_qkv_projections(self, unet: bool = True, vae: bool = True):
+        """Disable QKV projection fusion if enabled.
+
+        <Tip warning={true}>
+
+        This API is 🧪 experimental.
+
+        </Tip>
+
+        Args:
+            unet (`bool`, defaults to `True`): To apply fusion on the UNet.
+            vae (`bool`, defaults to `True`): To apply fusion on the VAE.
+
+        """
+        if unet:
+            if not self.fusing_unet:
+                logger.warning("The UNet was not initially fused for QKV projections. Doing nothing.")
+            else:
+                self.unet.unfuse_qkv_projections()
+                self.fusing_unet = False
+
+        if vae:
+            if not self.fusing_vae:
+                logger.warning("The VAE was not initially fused for QKV projections. Doing nothing.")
+            else:
+                self.vae.unfuse_qkv_projections()
+                self.fusing_vae = False
+
     def get_guidance_scale_embedding(self, w, embedding_dim=512, dtype=torch.float32):
         """
         See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
@@ -705,6 +833,7 @@ class AltDiffusionImg2ImgPipeline(
         image: PipelineImageInput = None,
         strength: float = 0.8,
         num_inference_steps: Optional[int] = 50,
+        timesteps: List[int] = None,
         guidance_scale: Optional[float] = 7.5,
         negative_prompt: Optional[Union[str, List[str]]] = None,
         num_images_per_prompt: Optional[int] = 1,
@@ -712,6 +841,7 @@ class AltDiffusionImg2ImgPipeline(
         generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
         prompt_embeds: Optional[torch.FloatTensor] = None,
         negative_prompt_embeds: Optional[torch.FloatTensor] = None,
+        ip_adapter_image: Optional[PipelineImageInput] = None,
         output_type: Optional[str] = "pil",
         return_dict: bool = True,
         cross_attention_kwargs: Optional[Dict[str, Any]] = None,
@@ -741,6 +871,10 @@ class AltDiffusionImg2ImgPipeline(
             num_inference_steps (`int`, *optional*, defaults to 50):
                 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`.
+            timesteps (`List[int]`, *optional*):
+                Custom timesteps to use for the denoising process with schedulers which support a `timesteps` argument
+                in their `set_timesteps` method. If not defined, the default behavior when `num_inference_steps` is
+                passed will be used. Must be in descending order.
             guidance_scale (`float`, *optional*, defaults to 7.5):
                 A higher guidance scale value encourages the model to generate images closely linked to the text
                 `prompt` at the expense of lower image quality. Guidance scale is enabled when `guidance_scale > 1`.
@@ -761,6 +895,7 @@ class AltDiffusionImg2ImgPipeline(
             negative_prompt_embeds (`torch.FloatTensor`, *optional*):
                 Pre-generated negative text embeddings. Can be used to easily tweak text inputs (prompt weighting). If
                 not provided, `negative_prompt_embeds` are generated from the `negative_prompt` input argument.
+            ip_adapter_image: (`PipelineImageInput`, *optional*): Optional image input to work with IP Adapters.
             output_type (`str`, *optional*, defaults to `"pil"`):
                 The output format of the generated image. Choose between `PIL.Image` or `np.array`.
             return_dict (`bool`, *optional*, defaults to `True`):
@@ -780,7 +915,7 @@ class AltDiffusionImg2ImgPipeline(
             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 pipeine class.
+                `._callback_tensor_inputs` attribute of your pipeline class.
         Examples:
 
         Returns:
@@ -798,15 +933,13 @@ class AltDiffusionImg2ImgPipeline(
             deprecate(
                 "callback",
                 "1.0.0",
-                "Passing `callback` as an input argument to `__call__` is deprecated, consider use"
-                " `callback_on_step_end`",
+                "Passing `callback` as an input argument to `__call__` is deprecated, consider use `callback_on_step_end`",
             )
         if callback_steps is not None:
             deprecate(
                 "callback_steps",
                 "1.0.0",
-                "Passing `callback_steps` as an input argument to `__call__` is deprecated, consider use"
-                " `callback_on_step_end`",
+                "Passing `callback_steps` as an input argument to `__call__` is deprecated, consider use `callback_on_step_end`",
             )
 
         # 1. Check inputs. Raise error if not correct
@@ -855,11 +988,19 @@ class AltDiffusionImg2ImgPipeline(
         if self.do_classifier_free_guidance:
             prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])
 
+        if ip_adapter_image is not None:
+            output_hidden_state = False if isinstance(self.unet.encoder_hid_proj, ImageProjection) else True
+            image_embeds, negative_image_embeds = self.encode_image(
+                ip_adapter_image, device, num_images_per_prompt, output_hidden_state
+            )
+            if self.do_classifier_free_guidance:
+                image_embeds = torch.cat([negative_image_embeds, image_embeds])
+
         # 4. Preprocess image
         image = self.image_processor.preprocess(image)
 
         # 5. set timesteps
-        self.scheduler.set_timesteps(num_inference_steps, device=device)
+        timesteps, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, timesteps)
         timesteps, num_inference_steps = self.get_timesteps(num_inference_steps, strength, device)
         latent_timestep = timesteps[:1].repeat(batch_size * num_images_per_prompt)
 
@@ -877,7 +1018,10 @@ class AltDiffusionImg2ImgPipeline(
         # 7. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
         extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta)
 
-        # 7.5 Optionally get Guidance Scale Embedding
+        # 7.1 Add image embeds for IP-Adapter
+        added_cond_kwargs = {"image_embeds": image_embeds} if ip_adapter_image is not None else None
+
+        # 7.2 Optionally get Guidance Scale Embedding
         timestep_cond = None
         if self.unet.config.time_cond_proj_dim is not None:
             guidance_scale_tensor = torch.tensor(self.guidance_scale - 1).repeat(batch_size * num_images_per_prompt)
@@ -901,6 +1045,7 @@ class AltDiffusionImg2ImgPipeline(
                     encoder_hidden_states=prompt_embeds,
                     timestep_cond=timestep_cond,
                     cross_attention_kwargs=self.cross_attention_kwargs,
+                    added_cond_kwargs=added_cond_kwargs,
                     return_dict=False,
                 )[0]
 

diffusers/pipelines/{alt_diffusion → deprecated/alt_diffusion}/pipeline_output.py

@@ -4,7 +4,7 @@ from typing import List, Optional, Union
 import numpy as np
 import PIL.Image
 
-from ...utils import (
+from ....utils import (
     BaseOutput,
 )
 

diffusers/pipelines/{audio_diffusion → deprecated/audio_diffusion}/__init__.py

@@ -1,6 +1,6 @@
 from typing import TYPE_CHECKING
 
-from ...utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
+from ....utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
 
 
 _import_structure = {

diffusers/pipelines/{audio_diffusion → deprecated/audio_diffusion}/mel.py

@@ -15,8 +15,8 @@
 
 import numpy as np  # noqa: E402
 
-from ...configuration_utils import ConfigMixin, register_to_config
-from ...schedulers.scheduling_utils import SchedulerMixin
+from ....configuration_utils import ConfigMixin, register_to_config
+from ....schedulers.scheduling_utils import SchedulerMixin
 
 
 try:

diffusers/pipelines/{audio_diffusion → deprecated/audio_diffusion}/pipeline_audio_diffusion.py

@@ -20,10 +20,10 @@ import numpy as np
 import torch
 from PIL import Image
 
-from ...models import AutoencoderKL, UNet2DConditionModel
-from ...schedulers import DDIMScheduler, DDPMScheduler
-from ...utils.torch_utils import randn_tensor
-from ..pipeline_utils import AudioPipelineOutput, BaseOutput, DiffusionPipeline, ImagePipelineOutput
+from ....models import AutoencoderKL, UNet2DConditionModel
+from ....schedulers import DDIMScheduler, DDPMScheduler
+from ....utils.torch_utils import randn_tensor
+from ...pipeline_utils import AudioPipelineOutput, BaseOutput, DiffusionPipeline, ImagePipelineOutput
 from .mel import Mel
 
 

diffusers/pipelines/{latent_diffusion_uncond → deprecated/latent_diffusion_uncond}/__init__.py

@@ -1,6 +1,6 @@
 from typing import TYPE_CHECKING
 
-from ...utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
+from ....utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
 
 
 _import_structure = {"pipeline_latent_diffusion_uncond": ["LDMPipeline"]}

diffusers/pipelines/{latent_diffusion_uncond → deprecated/latent_diffusion_uncond}/pipeline_latent_diffusion_uncond.py

@@ -17,10 +17,10 @@ from typing import List, Optional, Tuple, Union
 
 import torch
 
-from ...models import UNet2DModel, VQModel
-from ...schedulers import DDIMScheduler
-from ...utils.torch_utils import randn_tensor
-from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
+from ....models import UNet2DModel, VQModel
+from ....schedulers import DDIMScheduler
+from ....utils.torch_utils import randn_tensor
+from ...pipeline_utils import DiffusionPipeline, ImagePipelineOutput
 
 
 class LDMPipeline(DiffusionPipeline):

diffusers/pipelines/{pndm → deprecated/pndm}/__init__.py

@@ -1,6 +1,6 @@
 from typing import TYPE_CHECKING
 
-from ...utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
+from ....utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
 
 
 _import_structure = {"pipeline_pndm": ["PNDMPipeline"]}

diffusers/pipelines/{pndm → deprecated/pndm}/pipeline_pndm.py

@@ -17,10 +17,10 @@ from typing import List, Optional, Tuple, Union
 
 import torch
 
-from ...models import UNet2DModel
-from ...schedulers import PNDMScheduler
-from ...utils.torch_utils import randn_tensor
-from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
+from ....models import UNet2DModel
+from ....schedulers import PNDMScheduler
+from ....utils.torch_utils import randn_tensor
+from ...pipeline_utils import DiffusionPipeline, ImagePipelineOutput
 
 
 class PNDMPipeline(DiffusionPipeline):

diffusers/pipelines/{repaint → deprecated/repaint}/__init__.py

@@ -1,6 +1,6 @@
 from typing import TYPE_CHECKING
 
-from ...utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
+from ....utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
 
 
 _import_structure = {"pipeline_repaint": ["RePaintPipeline"]}

diffusers/pipelines/{repaint → deprecated/repaint}/pipeline_repaint.py

@@ -19,11 +19,11 @@ import numpy as np
 import PIL.Image
 import torch
 
-from ...models import UNet2DModel
-from ...schedulers import RePaintScheduler
-from ...utils import PIL_INTERPOLATION, deprecate, logging
-from ...utils.torch_utils import randn_tensor
-from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
+from ....models import UNet2DModel
+from ....schedulers import RePaintScheduler
+from ....utils import PIL_INTERPOLATION, deprecate, logging
+from ....utils.torch_utils import randn_tensor
+from ...pipeline_utils import DiffusionPipeline, ImagePipelineOutput
 
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name

diffusers/pipelines/{score_sde_ve → deprecated/score_sde_ve}/__init__.py

@@ -1,6 +1,6 @@
 from typing import TYPE_CHECKING
 
-from ...utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
+from ....utils import DIFFUSERS_SLOW_IMPORT, _LazyModule
 
 
 _import_structure = {"pipeline_score_sde_ve": ["ScoreSdeVePipeline"]}

diffusers/pipelines/{score_sde_ve → deprecated/score_sde_ve}/pipeline_score_sde_ve.py

@@ -16,10 +16,10 @@ from typing import List, Optional, Tuple, Union
 
 import torch
 
-from ...models import UNet2DModel
-from ...schedulers import ScoreSdeVeScheduler
-from ...utils.torch_utils import randn_tensor
-from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
+from ....models import UNet2DModel
+from ....schedulers import ScoreSdeVeScheduler
+from ....utils.torch_utils import randn_tensor
+from ...pipeline_utils import DiffusionPipeline, ImagePipelineOutput
 
 
 class ScoreSdeVePipeline(DiffusionPipeline):
@@ -35,6 +35,7 @@ class ScoreSdeVePipeline(DiffusionPipeline):
         scheduler ([`ScoreSdeVeScheduler`]):
             A `ScoreSdeVeScheduler` to be used in combination with `unet` to denoise the encoded image.
     """
+
     unet: UNet2DModel
     scheduler: ScoreSdeVeScheduler
 

diffusers/pipelines/{spectrogram_diffusion → deprecated/spectrogram_diffusion}/__init__.py

@@ -1,7 +1,7 @@
 # flake8: noqa
 from typing import TYPE_CHECKING
-from ...utils import DIFFUSERS_SLOW_IMPORT
-from ...utils import (
+from ....utils import (
+    DIFFUSERS_SLOW_IMPORT,
     _LazyModule,
     is_note_seq_available,
     OptionalDependencyNotAvailable,
@@ -17,7 +17,7 @@ try:
     if not (is_transformers_available() and is_torch_available()):
         raise OptionalDependencyNotAvailable()
 except OptionalDependencyNotAvailable:
-    from ...utils import dummy_torch_and_transformers_objects  # noqa F403
+    from ....utils import dummy_torch_and_transformers_objects  # noqa F403
 
     _dummy_objects.update(get_objects_from_module(dummy_torch_and_transformers_objects))
 else:
@@ -32,7 +32,7 @@ try:
     if not (is_transformers_available() and is_torch_available() and is_note_seq_available()):
         raise OptionalDependencyNotAvailable()
 except OptionalDependencyNotAvailable:
-    from ...utils import dummy_transformers_and_torch_and_note_seq_objects
+    from ....utils import dummy_transformers_and_torch_and_note_seq_objects
 
     _dummy_objects.update(get_objects_from_module(dummy_transformers_and_torch_and_note_seq_objects))
 else:
@@ -45,7 +45,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
             raise OptionalDependencyNotAvailable()
 
     except OptionalDependencyNotAvailable:
-        from ...utils.dummy_torch_and_transformers_objects import *
+        from ....utils.dummy_torch_and_transformers_objects import *
     else:
         from .pipeline_spectrogram_diffusion import SpectrogramDiffusionPipeline
         from .pipeline_spectrogram_diffusion import SpectrogramContEncoder
@@ -56,7 +56,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
         if not (is_transformers_available() and is_torch_available() and is_note_seq_available()):
             raise OptionalDependencyNotAvailable()
     except OptionalDependencyNotAvailable:
-        from ...utils.dummy_transformers_and_torch_and_note_seq_objects import *
+        from ....utils.dummy_transformers_and_torch_and_note_seq_objects import *
 
     else:
         from .midi_utils import MidiProcessor

diffusers/pipelines/{spectrogram_diffusion/continous_encoder.py → deprecated/spectrogram_diffusion/continuous_encoder.py}

@@ -22,8 +22,8 @@ from transformers.models.t5.modeling_t5 import (
     T5LayerNorm,
 )
 
-from ...configuration_utils import ConfigMixin, register_to_config
-from ...models import ModelMixin
+from ....configuration_utils import ConfigMixin, register_to_config
+from ....models import ModelMixin
 
 
 class SpectrogramContEncoder(ModelMixin, ConfigMixin, ModuleUtilsMixin):

diffusers/pipelines/{spectrogram_diffusion → deprecated/spectrogram_diffusion}/midi_utils.py

@@ -22,7 +22,7 @@ import numpy as np
 import torch
 import torch.nn.functional as F
 
-from ...utils import is_note_seq_available
+from ....utils import is_note_seq_available
 from .pipeline_spectrogram_diffusion import TARGET_FEATURE_LENGTH
 
 

diffusers/pipelines/{spectrogram_diffusion → deprecated/spectrogram_diffusion}/notes_encoder.py

@@ -18,8 +18,8 @@ import torch.nn as nn
 from transformers.modeling_utils import ModuleUtilsMixin
 from transformers.models.t5.modeling_t5 import T5Block, T5Config, T5LayerNorm
 
-from ...configuration_utils import ConfigMixin, register_to_config
-from ...models import ModelMixin
+from ....configuration_utils import ConfigMixin, register_to_config
+from ....models import ModelMixin
 
 
 class SpectrogramNotesEncoder(ModelMixin, ConfigMixin, ModuleUtilsMixin):

diffusers/pipelines/{spectrogram_diffusion → deprecated/spectrogram_diffusion}/pipeline_spectrogram_diffusion.py

@@ -19,17 +19,17 @@ from typing import Any, Callable, List, Optional, Tuple, Union
 import numpy as np
 import torch
 
-from ...models import T5FilmDecoder
-from ...schedulers import DDPMScheduler
-from ...utils import is_onnx_available, logging
-from ...utils.torch_utils import randn_tensor
+from ....models import T5FilmDecoder
+from ....schedulers import DDPMScheduler
+from ....utils import is_onnx_available, logging
+from ....utils.torch_utils import randn_tensor
 
 
 if is_onnx_available():
-    from ..onnx_utils import OnnxRuntimeModel
+    from ...onnx_utils import OnnxRuntimeModel
 
-from ..pipeline_utils import AudioPipelineOutput, DiffusionPipeline
-from .continous_encoder import SpectrogramContEncoder
+from ...pipeline_utils import AudioPipelineOutput, DiffusionPipeline
+from .continuous_encoder import SpectrogramContEncoder
 from .notes_encoder import SpectrogramNotesEncoder
 
 
@@ -54,6 +54,7 @@ class SpectrogramDiffusionPipeline(DiffusionPipeline):
             A scheduler to be used in combination with `decoder` to denoise the encoded audio latents.
         melgan ([`OnnxRuntimeModel`]):
     """
+
     _optional_components = ["melgan"]
 
     def __init__(