Merge branch 'main' into guidance-scale-docs

Author: sayakpaul

src/diffusers/configuration_utils.py

@@ -763,4 +763,7 @@ def from_config(cls, config: Union[FrozenDict, Dict[str, Any]] = None, return_un
         # resolve remapping
         remapped_class = _fetch_remapped_cls_from_config(config, cls)
 
-        return remapped_class.from_config(config, return_unused_kwargs, **kwargs)
+        if remapped_class is cls:
+            return super(LegacyConfigMixin, remapped_class).from_config(config, return_unused_kwargs, **kwargs)
+        else:
+            return remapped_class.from_config(config, return_unused_kwargs, **kwargs)

src/diffusers/models/modeling_utils.py

@@ -1877,4 +1877,9 @@ def from_pretrained(cls, pretrained_model_name_or_path: Optional[Union[str, os.P
         # resolve remapping
         remapped_class = _fetch_remapped_cls_from_config(config, cls)
 
-        return remapped_class.from_pretrained(pretrained_model_name_or_path, **kwargs_copy)
+        if remapped_class is cls:
+            return super(LegacyModelMixin, remapped_class).from_pretrained(
+                pretrained_model_name_or_path, **kwargs_copy
+            )
+        else:
+            return remapped_class.from_pretrained(pretrained_model_name_or_path, **kwargs_copy)

src/diffusers/pipelines/controlnet/pipeline_controlnet_union_sd_xl_img2img.py

@@ -844,6 +844,8 @@ def __call__(
 
         # 5. Prepare timesteps
         sigmas = np.linspace(1.0, 1 / num_inference_steps, num_inference_steps) if sigmas is None else sigmas
+        if hasattr(self.scheduler.config, "use_flow_sigmas") and self.scheduler.config.use_flow_sigmas:
+            sigmas = None
         image_seq_len = latents.shape[1]
         mu = calculate_shift(
             image_seq_len,

src/diffusers/pipelines/flux/pipeline_flux.py

@@ -383,7 +383,8 @@ def __call__(
         # set timesteps
         self.scheduler.set_timesteps(num_inference_steps)
 
-        latents = latents * np.float64(self.scheduler.init_noise_sigma)
+        # scale the initial noise by the standard deviation required by the scheduler
+        latents = latents * self.scheduler.init_noise_sigma
 
         # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature
         # eta (η) is only used with the DDIMScheduler, it will be ignored for other schedulers.

src/diffusers/pipelines/stable_diffusion/pipeline_onnx_stable_diffusion.py

@@ -483,7 +483,7 @@ def __call__(
         self.scheduler.set_timesteps(num_inference_steps)
 
         # scale the initial noise by the standard deviation required by the scheduler
-        latents = latents * np.float64(self.scheduler.init_noise_sigma)
+        latents = latents * self.scheduler.init_noise_sigma
 
         # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature
         # eta (η) is only used with the DDIMScheduler, it will be ignored for other schedulers.

src/diffusers/pipelines/stable_diffusion/pipeline_onnx_stable_diffusion_inpaint.py

@@ -481,7 +481,7 @@ def __call__(
         timesteps = self.scheduler.timesteps
 
         # Scale the initial noise by the standard deviation required by the scheduler
-        latents = latents * np.float64(self.scheduler.init_noise_sigma)
+        latents = latents * self.scheduler.init_noise_sigma
 
         # 5. Add noise to image
         noise_level = np.array([noise_level]).astype(np.int64)

src/diffusers/pipelines/stable_diffusion/pipeline_onnx_stable_diffusion_upscale.py

@@ -153,6 +153,8 @@ def __init__(
         flow_shift: Optional[float] = 1.0,
         timestep_spacing: str = "linspace",
         steps_offset: int = 0,
+        use_dynamic_shifting: bool = False,
+        time_shift_type: str = "exponential",
     ):
         if self.config.use_beta_sigmas and not is_scipy_available():
             raise ImportError("Make sure to install scipy if you want to use beta sigmas.")
@@ -232,7 +234,9 @@ def set_begin_index(self, begin_index: int = 0):
         """
         self._begin_index = begin_index
 
-    def set_timesteps(self, num_inference_steps: int, device: Union[str, torch.device] = None):
+    def set_timesteps(
+        self, num_inference_steps: int, device: Union[str, torch.device] = None, mu: Optional[float] = None
+    ):
         """
         Sets the discrete timesteps used for the diffusion chain (to be run before inference).
 
@@ -242,6 +246,9 @@ def set_timesteps(self, num_inference_steps: int, device: Union[str, torch.devic
             device (`str` or `torch.device`, *optional*):
                 The device to which the timesteps should be moved to. If `None`, the timesteps are not moved.
         """
+        if mu is not None:
+            assert self.config.use_dynamic_shifting and self.config.time_shift_type == "exponential"
+            self.config.flow_shift = np.exp(mu)
         # "linspace", "leading", "trailing" corresponds to annotation of Table 2. of https://huggingface.co/papers/2305.08891
         if self.config.timestep_spacing == "linspace":
             timesteps = (

src/diffusers/schedulers/scheduling_deis_multistep.py

@@ -230,6 +230,8 @@ def __init__(
         timestep_spacing: str = "linspace",
         steps_offset: int = 0,
         rescale_betas_zero_snr: bool = False,
+        use_dynamic_shifting: bool = False,
+        time_shift_type: str = "exponential",
     ):
         if self.config.use_beta_sigmas and not is_scipy_available():
             raise ImportError("Make sure to install scipy if you want to use beta sigmas.")
@@ -330,6 +332,7 @@ def set_timesteps(
         self,
         num_inference_steps: int = None,
         device: Union[str, torch.device] = None,
+        mu: Optional[float] = None,
         timesteps: Optional[List[int]] = None,
     ):
         """
@@ -345,6 +348,9 @@ def set_timesteps(
                 based on the `timestep_spacing` attribute. If `timesteps` is passed, `num_inference_steps` and `sigmas`
                 must be `None`, and `timestep_spacing` attribute will be ignored.
         """
+        if mu is not None:
+            assert self.config.use_dynamic_shifting and self.config.time_shift_type == "exponential"
+            self.config.flow_shift = np.exp(mu)
         if num_inference_steps is None and timesteps is None:
             raise ValueError("Must pass exactly one of `num_inference_steps` or `timesteps`.")
         if num_inference_steps is not None and timesteps is not None:

src/diffusers/schedulers/scheduling_dpmsolver_multistep.py

@@ -169,6 +169,8 @@ def __init__(
         final_sigmas_type: Optional[str] = "zero",  # "zero", "sigma_min"
         lambda_min_clipped: float = -float("inf"),
         variance_type: Optional[str] = None,
+        use_dynamic_shifting: bool = False,
+        time_shift_type: str = "exponential",
     ):
         if self.config.use_beta_sigmas and not is_scipy_available():
             raise ImportError("Make sure to install scipy if you want to use beta sigmas.")
@@ -301,6 +303,7 @@ def set_timesteps(
         self,
         num_inference_steps: int = None,
         device: Union[str, torch.device] = None,
+        mu: Optional[float] = None,
         timesteps: Optional[List[int]] = None,
     ):
         """
@@ -316,6 +319,9 @@ def set_timesteps(
                 timestep spacing strategy of equal spacing between timesteps schedule is used. If `timesteps` is
                 passed, `num_inference_steps` must be `None`.
         """
+        if mu is not None:
+            assert self.config.use_dynamic_shifting and self.config.time_shift_type == "exponential"
+            self.config.flow_shift = np.exp(mu)
         if num_inference_steps is None and timesteps is None:
             raise ValueError("Must pass exactly one of  `num_inference_steps` or `timesteps`.")
         if num_inference_steps is not None and timesteps is not None: