k-diffusion-euler (open-mmlab#1019)

* k-diffusion-euler

* make style make quality

* make fix-copies

* fix tests for euler a

* Update src/diffusers/schedulers/scheduling_euler_ancestral_discrete.py

Co-authored-by: Anton Lozhkov <[email protected]>

* Update src/diffusers/schedulers/scheduling_euler_ancestral_discrete.py

Co-authored-by: Anton Lozhkov <[email protected]>

* Update src/diffusers/schedulers/scheduling_euler_discrete.py

Co-authored-by: Anton Lozhkov <[email protected]>

* Update src/diffusers/schedulers/scheduling_euler_discrete.py

Co-authored-by: Anton Lozhkov <[email protected]>

* remove unused arg and method

* update doc

* quality

* make flake happy

* use logger instead of warn

* raise error instead of deprication

* don't require scipy

* pass generator in step

* fix tests

* Apply suggestions from code review

Co-authored-by: Pedro Cuenca <[email protected]>

* Update tests/test_scheduler.py

Co-authored-by: Patrick von Platen <[email protected]>

* remove unused generator

* pass generator as extra_step_kwargs

* update tests

* pass generator as kwarg

* pass generator as kwarg

* quality

* fix test for lms

* fix tests

Co-authored-by: patil-suraj <[email protected]>
Co-authored-by: Anton Lozhkov <[email protected]>
Co-authored-by: Pedro Cuenca <[email protected]>
Co-authored-by: Patrick von Platen <[email protected]>

src/diffusers/__init__.py

@@ -41,6 +41,8 @@
     from .schedulers import (
         DDIMScheduler,
         DDPMScheduler,
+        EulerAncestralDiscreteScheduler,
+        EulerDiscreteScheduler,
         IPNDMScheduler,
         KarrasVeScheduler,
         PNDMScheduler,

src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py

@@ -9,7 +9,13 @@
 from ...configuration_utils import FrozenDict
 from ...models import AutoencoderKL, UNet2DConditionModel
 from ...pipeline_utils import DiffusionPipeline
-from ...schedulers import DDIMScheduler, LMSDiscreteScheduler, PNDMScheduler
+from ...schedulers import (
+    DDIMScheduler,
+    EulerAncestralDiscreteScheduler,
+    EulerDiscreteScheduler,
+    LMSDiscreteScheduler,
+    PNDMScheduler,
+)
 from ...utils import deprecate, logging
 from . import StableDiffusionPipelineOutput
 from .safety_checker import StableDiffusionSafetyChecker
@@ -52,7 +58,9 @@ def __init__(
         text_encoder: CLIPTextModel,
         tokenizer: CLIPTokenizer,
         unet: UNet2DConditionModel,
-        scheduler: Union[DDIMScheduler, PNDMScheduler, LMSDiscreteScheduler],
+        scheduler: Union[
+            DDIMScheduler, PNDMScheduler, LMSDiscreteScheduler, EulerDiscreteScheduler, EulerAncestralDiscreteScheduler
+        ],
         safety_checker: StableDiffusionSafetyChecker,
         feature_extractor: CLIPFeatureExtractor,
     ):
@@ -334,6 +342,11 @@ def __call__(
         if accepts_eta:
             extra_step_kwargs["eta"] = eta
 
+        # check if the scheduler accepts generator
+        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        if accepts_generator:
+            extra_step_kwargs["generator"] = generator
+
         for i, t in enumerate(self.progress_bar(timesteps_tensor)):
             # expand the latents if we are doing classifier free guidance
             latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents

src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_img2img.py

@@ -10,7 +10,13 @@
 from ...configuration_utils import FrozenDict
 from ...models import AutoencoderKL, UNet2DConditionModel
 from ...pipeline_utils import DiffusionPipeline
-from ...schedulers import DDIMScheduler, LMSDiscreteScheduler, PNDMScheduler
+from ...schedulers import (
+    DDIMScheduler,
+    EulerAncestralDiscreteScheduler,
+    EulerDiscreteScheduler,
+    LMSDiscreteScheduler,
+    PNDMScheduler,
+)
 from ...utils import deprecate, logging
 from . import StableDiffusionPipelineOutput
 from .safety_checker import StableDiffusionSafetyChecker
@@ -63,7 +69,9 @@ def __init__(
         text_encoder: CLIPTextModel,
         tokenizer: CLIPTokenizer,
         unet: UNet2DConditionModel,
-        scheduler: Union[DDIMScheduler, PNDMScheduler, LMSDiscreteScheduler],
+        scheduler: Union[
+            DDIMScheduler, PNDMScheduler, LMSDiscreteScheduler, EulerDiscreteScheduler, EulerAncestralDiscreteScheduler
+        ],
         safety_checker: StableDiffusionSafetyChecker,
         feature_extractor: CLIPFeatureExtractor,
     ):
@@ -335,6 +343,11 @@ def __call__(
         if accepts_eta:
             extra_step_kwargs["eta"] = eta
 
+        # check if the scheduler accepts generator
+        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        if accepts_generator:
+            extra_step_kwargs["generator"] = generator
+
         latents = init_latents
 
         t_start = max(num_inference_steps - init_timestep + offset, 0)

src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_inpaint.py

@@ -379,6 +379,11 @@ def __call__(
         if accepts_eta:
             extra_step_kwargs["eta"] = eta
 
+        # check if the scheduler accepts generator
+        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        if accepts_generator:
+            extra_step_kwargs["generator"] = generator
+
         for i, t in enumerate(self.progress_bar(timesteps_tensor)):
             # expand the latents if we are doing classifier free guidance
             latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents

src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion_inpaint_legacy.py

@@ -352,6 +352,11 @@ def __call__(
         if accepts_eta:
             extra_step_kwargs["eta"] = eta
 
+        # check if the scheduler accepts generator
+        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        if accepts_generator:
+            extra_step_kwargs["generator"] = generator
+
         latents = init_latents
 
         t_start = max(num_inference_steps - init_timestep + offset, 0)

src/diffusers/schedulers/__init__.py

@@ -19,6 +19,8 @@
 if is_torch_available():
     from .scheduling_ddim import DDIMScheduler
     from .scheduling_ddpm import DDPMScheduler
+    from .scheduling_euler_ancestral_discrete import EulerAncestralDiscreteScheduler
+    from .scheduling_euler_discrete import EulerDiscreteScheduler
     from .scheduling_ipndm import IPNDMScheduler
     from .scheduling_karras_ve import KarrasVeScheduler
     from .scheduling_pndm import PNDMScheduler

src/diffusers/schedulers/scheduling_euler_ancestral_discrete.py

@@ -0,0 +1,261 @@
+# Copyright 2022 Katherine Crowson and 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 dataclasses import dataclass
+from typing import Optional, Tuple, Union
+
+import numpy as np
+import torch
+
+from ..configuration_utils import ConfigMixin, register_to_config
+from ..utils import BaseOutput, deprecate, logging
+from .scheduling_utils import SchedulerMixin
+
+
+logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
+
+
+@dataclass
+# Copied from diffusers.schedulers.scheduling_ddpm.DDPMSchedulerOutput with DDPM->EulerAncestralDiscrete
+class EulerAncestralDiscreteSchedulerOutput(BaseOutput):
+    """
+    Output class for the scheduler's step function output.
+
+    Args:
+        prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):
+            Computed sample (x_{t-1}) of previous timestep. `prev_sample` should be used as next model input in the
+            denoising loop.
+        pred_original_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images):
+            The predicted denoised sample (x_{0}) based on the model output from the current timestep.
+            `pred_original_sample` can be used to preview progress or for guidance.
+    """
+
+    prev_sample: torch.FloatTensor
+    pred_original_sample: Optional[torch.FloatTensor] = None
+
+
+class EulerAncestralDiscreteScheduler(SchedulerMixin, ConfigMixin):
+    """
+    Ancestral sampling with Euler method steps. Based on the original k-diffusion implementation by Katherine Crowson:
+    https://github.com/crowsonkb/k-diffusion/blob/481677d114f6ea445aa009cf5bd7a9cdee909e47/k_diffusion/sampling.py#L72
+
+    [`~ConfigMixin`] takes care of storing all config attributes that are passed in the scheduler's `__init__`
+    function, such as `num_train_timesteps`. They can be accessed via `scheduler.config.num_train_timesteps`.
+    [`~ConfigMixin`] also provides general loading and saving functionality via the [`~ConfigMixin.save_config`] and
+    [`~ConfigMixin.from_config`] functions.
+
+    Args:
+        num_train_timesteps (`int`): number of diffusion steps used to train the model.
+        beta_start (`float`): the starting `beta` value of inference.
+        beta_end (`float`): the final `beta` value.
+        beta_schedule (`str`):
+            the beta schedule, a mapping from a beta range to a sequence of betas for stepping the model. Choose from
+            `linear` or `scaled_linear`.
+        trained_betas (`np.ndarray`, optional):
+            option to pass an array of betas directly to the constructor to bypass `beta_start`, `beta_end` etc.
+
+    """
+
+    @register_to_config
+    def __init__(
+        self,
+        num_train_timesteps: int = 1000,
+        beta_start: float = 0.0001,
+        beta_end: float = 0.02,
+        beta_schedule: str = "linear",
+        trained_betas: Optional[np.ndarray] = None,
+    ):
+        if trained_betas is not None:
+            self.betas = torch.from_numpy(trained_betas)
+        elif beta_schedule == "linear":
+            self.betas = torch.linspace(beta_start, beta_end, num_train_timesteps, dtype=torch.float32)
+        elif beta_schedule == "scaled_linear":
+            # this schedule is very specific to the latent diffusion model.
+            self.betas = (
+                torch.linspace(beta_start**0.5, beta_end**0.5, num_train_timesteps, dtype=torch.float32) ** 2
+            )
+        else:
+            raise NotImplementedError(f"{beta_schedule} does is not implemented for {self.__class__}")
+
+        self.alphas = 1.0 - self.betas
+        self.alphas_cumprod = torch.cumprod(self.alphas, dim=0)
+
+        sigmas = np.array(((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5)
+        sigmas = np.concatenate([sigmas[::-1], [0.0]]).astype(np.float32)
+        self.sigmas = torch.from_numpy(sigmas)
+
+        # standard deviation of the initial noise distribution
+        self.init_noise_sigma = self.sigmas.max()
+
+        # setable values
+        self.num_inference_steps = None
+        timesteps = np.linspace(0, num_train_timesteps - 1, num_train_timesteps, dtype=float)[::-1].copy()
+        self.timesteps = torch.from_numpy(timesteps)
+        self.is_scale_input_called = False
+
+    def scale_model_input(
+        self, sample: torch.FloatTensor, timestep: Union[float, torch.FloatTensor]
+    ) -> torch.FloatTensor:
+        """
+        Scales the denoising model input by `(sigma**2 + 1) ** 0.5` to match the Euler algorithm.
+
+        Args:
+            sample (`torch.FloatTensor`): input sample
+            timestep (`float` or `torch.FloatTensor`): the current timestep in the diffusion chain
+
+        Returns:
+            `torch.FloatTensor`: scaled input sample
+        """
+        if isinstance(timestep, torch.Tensor):
+            timestep = timestep.to(self.timesteps.device)
+        step_index = (self.timesteps == timestep).nonzero().item()
+        sigma = self.sigmas[step_index]
+        sample = sample / ((sigma**2 + 1) ** 0.5)
+        self.is_scale_input_called = True
+        return sample
+
+    def set_timesteps(self, num_inference_steps: int, device: Union[str, torch.device] = None):
+        """
+        Sets the timesteps used for the diffusion chain. Supporting function to be run before inference.
+
+        Args:
+            num_inference_steps (`int`):
+                the number of diffusion steps used when generating samples with a pre-trained model.
+            device (`str` or `torch.device`, optional):
+                the device to which the timesteps should be moved to. If `None`, the timesteps are not moved.
+        """
+        self.num_inference_steps = num_inference_steps
+
+        timesteps = np.linspace(0, self.config.num_train_timesteps - 1, num_inference_steps, dtype=float)[::-1].copy()
+        sigmas = np.array(((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5)
+        sigmas = np.interp(timesteps, np.arange(0, len(sigmas)), sigmas)
+        sigmas = np.concatenate([sigmas, [0.0]]).astype(np.float32)
+        self.sigmas = torch.from_numpy(sigmas).to(device=device)
+        self.timesteps = torch.from_numpy(timesteps).to(device=device)
+
+    def step(
+        self,
+        model_output: torch.FloatTensor,
+        timestep: Union[float, torch.FloatTensor],
+        sample: torch.FloatTensor,
+        generator: Optional[torch.Generator] = None,
+        return_dict: bool = True,
+    ) -> Union[EulerAncestralDiscreteSchedulerOutput, Tuple]:
+        """
+        Predict the sample at the previous timestep by reversing the SDE. Core function to propagate the diffusion
+        process from the learned model outputs (most often the predicted noise).
+
+        Args:
+            model_output (`torch.FloatTensor`): direct output from learned diffusion model.
+            timestep (`float`): current timestep in the diffusion chain.
+            sample (`torch.FloatTensor`):
+                current instance of sample being created by diffusion process.
+            generator (`torch.Generator`, optional): Random number generator.
+            return_dict (`bool`): option for returning tuple rather than EulerAncestralDiscreteSchedulerOutput class
+
+        Returns:
+            [`~schedulers.scheduling_utils.EulerAncestralDiscreteSchedulerOutput`] or `tuple`:
+            [`~schedulers.scheduling_utils.EulerAncestralDiscreteSchedulerOutput`] if `return_dict` is True, otherwise
+            a `tuple`. When returning a tuple, the first element is the sample tensor.
+
+        """
+
+        if (
+            isinstance(timestep, int)
+            or isinstance(timestep, torch.IntTensor)
+            or isinstance(timestep, torch.LongTensor)
+        ):
+            raise ValueError(
+                "Passing integer indices (e.g. from `enumerate(timesteps)`) as timesteps to"
+                " `EulerDiscreteScheduler.step()` is not supported. Make sure to pass"
+                " one of the `scheduler.timesteps` as a timestep.",
+            )
+
+        if not self.is_scale_input_called:
+            logger.warn(
+                "The `scale_model_input` function should be called before `step` to ensure correct denoising. "
+                "See `StableDiffusionPipeline` for a usage example."
+            )
+
+        if isinstance(timestep, torch.Tensor):
+            timestep = timestep.to(self.timesteps.device)
+
+        step_index = (self.timesteps == timestep).nonzero().item()
+        sigma = self.sigmas[step_index]
+
+        # 1. compute predicted original sample (x_0) from sigma-scaled predicted noise
+        pred_original_sample = sample - sigma * model_output
+        sigma_from = self.sigmas[step_index]
+        sigma_to = self.sigmas[step_index + 1]
+        sigma_up = (sigma_to**2 * (sigma_from**2 - sigma_to**2) / sigma_from**2) ** 0.5
+        sigma_down = (sigma_to**2 - sigma_up**2) ** 0.5
+
+        # 2. Convert to an ODE derivative
+        derivative = (sample - pred_original_sample) / sigma
+
+        dt = sigma_down - sigma
+
+        prev_sample = sample + derivative * dt
+
+        device = model_output.device if torch.is_tensor(model_output) else "cpu"
+        noise = torch.randn(model_output.shape, dtype=model_output.dtype, generator=generator).to(device)
+        prev_sample = prev_sample + noise * sigma_up
+
+        if not return_dict:
+            return (prev_sample,)
+
+        return EulerAncestralDiscreteSchedulerOutput(
+            prev_sample=prev_sample, pred_original_sample=pred_original_sample
+        )
+
+    def add_noise(
+        self,
+        original_samples: torch.FloatTensor,
+        noise: torch.FloatTensor,
+        timesteps: torch.FloatTensor,
+    ) -> torch.FloatTensor:
+        # Make sure sigmas and timesteps have the same device and dtype as original_samples
+        self.sigmas = self.sigmas.to(device=original_samples.device, dtype=original_samples.dtype)
+        if original_samples.device.type == "mps" and torch.is_floating_point(timesteps):
+            # mps does not support float64
+            self.timesteps = self.timesteps.to(original_samples.device, dtype=torch.float32)
+            timesteps = timesteps.to(original_samples.device, dtype=torch.float32)
+        else:
+            self.timesteps = self.timesteps.to(original_samples.device)
+            timesteps = timesteps.to(original_samples.device)
+
+        schedule_timesteps = self.timesteps
+
+        if isinstance(timesteps, torch.IntTensor) or isinstance(timesteps, torch.LongTensor):
+            deprecate(
+                "timesteps as indices",
+                "0.8.0",
+                "Passing integer indices  (e.g. from `enumerate(timesteps)`) as timesteps to"
+                " `EulerAncestralDiscreteScheduler.add_noise()` will not be supported in future versions. Make sure to"
+                " pass values from `scheduler.timesteps` as timesteps.",
+                standard_warn=False,
+            )
+            step_indices = timesteps
+        else:
+            step_indices = [(schedule_timesteps == t).nonzero().item() for t in timesteps]
+
+        sigma = self.sigmas[step_indices].flatten()
+        while len(sigma.shape) < len(original_samples.shape):
+            sigma = sigma.unsqueeze(-1)
+
+        noisy_samples = original_samples + noise * sigma
+        return noisy_samples
+
+    def __len__(self):
+        return self.config.num_train_timesteps