Diffusers documentation

UniPC

You are viewing v0.16.0 version. A newer version v0.31.0 is available.
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

UniPC

Overview

UniPC is a training-free framework designed for the fast sampling of diffusion models, which consists of a corrector (UniC) and a predictor (UniP) that share a unified analytical form and support arbitrary orders.

For more details about the method, please refer to the paper and the code.

Fast Sampling of Diffusion Models with Exponential Integrator.

UniPCMultistepScheduler

class diffusers.UniPCMultistepScheduler

< >

( num_train_timesteps: int = 1000 beta_start: float = 0.0001 beta_end: float = 0.02 beta_schedule: str = 'linear' trained_betas: typing.Union[numpy.ndarray, typing.List[float], NoneType] = None solver_order: int = 2 prediction_type: str = 'epsilon' thresholding: bool = False dynamic_thresholding_ratio: float = 0.995 sample_max_value: float = 1.0 predict_x0: bool = True solver_type: str = 'bh2' lower_order_final: bool = True disable_corrector: typing.List[int] = [] solver_p: SchedulerMixin = None )

Parameters

  • 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, scaled_linear, or squaredcos_cap_v2.
  • trained_betas (np.ndarray, optional) — option to pass an array of betas directly to the constructor to bypass beta_start, beta_end etc.
  • solver_order (int, default 2) — the order of UniPC, also the p in UniPC-p; can be any positive integer. Note that the effective order of accuracy is solver_order + 1 due to the UniC. We recommend to use solver_order=2 for guided sampling, and solver_order=3 for unconditional sampling.
  • prediction_type (str, default epsilon, optional) — prediction type of the scheduler function, one of epsilon (predicting the noise of the diffusion process), sample (directly predicting the noisy sample) or v_prediction` (see section 2.4 https://imagen.research.google/video/paper.pdf)
  • thresholding (bool, default False) — whether to use the “dynamic thresholding” method (introduced by Imagen, https://arxiv.org/abs/2205.11487). For pixel-space diffusion models, you can set both predict_x0=True and thresholding=True to use the dynamic thresholding. Note that the thresholding method is unsuitable for latent-space diffusion models (such as stable-diffusion).
  • dynamic_thresholding_ratio (float, default 0.995) — the ratio for the dynamic thresholding method. Default is 0.995, the same as Imagen (https://arxiv.org/abs/2205.11487).
  • sample_max_value (float, default 1.0) — the threshold value for dynamic thresholding. Valid only when thresholding=True and predict_x0=True.
  • predict_x0 (bool, default True) — whether to use the updating algrithm on the predicted x0. See https://arxiv.org/abs/2211.01095 for details
  • solver_type (str, default bh2) — the solver type of UniPC. We recommend use bh1 for unconditional sampling when steps < 10, and use bh2 otherwise.
  • lower_order_final (bool, default True) — whether to use lower-order solvers in the final steps. Only valid for < 15 inference steps. We empirically find this trick can stabilize the sampling of DPM-Solver for steps < 15, especially for steps <= 10.
  • disable_corrector (list, default []) — decide which step to disable the corrector. For large guidance scale, the misalignment between the epsilon_theta(x_t, c)and epsilon_theta(x_t^c, c) might influence the convergence. This can be mitigated by disable the corrector at the first few steps (e.g., disable_corrector=[0])
  • solver_p (SchedulerMixin, default None) — can be any other scheduler. If specified, the algorithm will become solver_p + UniC.

UniPC is a training-free framework designed for the fast sampling of diffusion models, which consists of a corrector (UniC) and a predictor (UniP) that share a unified analytical form and support arbitrary orders. UniPC is by desinged model-agnostic, supporting pixel-space/latent-space DPMs on unconditional/conditional sampling. It can also be applied to both noise prediction model and data prediction model. The corrector UniC can be also applied after any off-the-shelf solvers to increase the order of accuracy.

For more details, see the original paper: https://arxiv.org/abs/2302.04867

Currently, we support the multistep UniPC for both noise prediction models and data prediction models. We recommend to use solver_order=2 for guided sampling, and solver_order=3 for unconditional sampling.

We also support the “dynamic thresholding” method in Imagen (https://arxiv.org/abs/2205.11487). For pixel-space diffusion models, you can set both predict_x0=True and thresholding=True to use the dynamic thresholding. Note that the thresholding method is unsuitable for latent-space diffusion models (such as stable-diffusion).

~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. SchedulerMixin provides general loading and saving functionality via the SchedulerMixin.save_pretrained() and from_pretrained() functions.

convert_model_output

< >

( model_output: FloatTensor timestep: int sample: FloatTensor ) torch.FloatTensor

Parameters

  • model_output (torch.FloatTensor) — direct output from learned diffusion model.
  • timestep (int) — current discrete timestep in the diffusion chain.
  • sample (torch.FloatTensor) — current instance of sample being created by diffusion process.

Returns

torch.FloatTensor

the converted model output.

Convert the model output to the corresponding type that the algorithm PC needs.

multistep_uni_c_bh_update

< >

( this_model_output: FloatTensor this_timestep: int last_sample: FloatTensor this_sample: FloatTensor order: int ) torch.FloatTensor

Parameters

  • this_model_output (torch.FloatTensor) — the model outputs at x_t
  • this_timestep (int) — the current timestep t
  • last_sample (torch.FloatTensor) — the generated sample before the last predictor: x_{t-1}
  • this_sample (torch.FloatTensor) — the generated sample after the last predictor: x_{t}
  • order (int) — the p of UniC-p at this step. Note that the effective order of accuracy should be order + 1

Returns

torch.FloatTensor

the corrected sample tensor at the current timestep.

One step for the UniC (B(h) version).

multistep_uni_p_bh_update

< >

( model_output: FloatTensor prev_timestep: int sample: FloatTensor order: int ) torch.FloatTensor

Parameters

  • model_output (torch.FloatTensor) — direct outputs from learned diffusion model at the current timestep.
  • prev_timestep (int) — previous discrete timestep in the diffusion chain.
  • sample (torch.FloatTensor) — current instance of sample being created by diffusion process.
  • order (int) — the order of UniP at this step, also the p in UniPC-p.

Returns

torch.FloatTensor

the sample tensor at the previous timestep.

One step for the UniP (B(h) version). Alternatively, self.solver_p is used if is specified.

scale_model_input

< >

( sample: FloatTensor *args **kwargs ) torch.FloatTensor

Parameters

  • sample (torch.FloatTensor) — input sample

Returns

torch.FloatTensor

scaled input sample

Ensures interchangeability with schedulers that need to scale the denoising model input depending on the current timestep.

set_timesteps

< >

( num_inference_steps: int device: typing.Union[str, torch.device] = None )

Parameters

  • 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.

Sets the timesteps used for the diffusion chain. Supporting function to be run before inference.

step

< >

( model_output: FloatTensor timestep: int sample: FloatTensor return_dict: bool = True ) ~scheduling_utils.SchedulerOutput or tuple

Parameters

  • model_output (torch.FloatTensor) — direct output from learned diffusion model.
  • timestep (int) — current discrete timestep in the diffusion chain.
  • sample (torch.FloatTensor) — current instance of sample being created by diffusion process.
  • return_dict (bool) — option for returning tuple rather than SchedulerOutput class

Returns

~scheduling_utils.SchedulerOutput or tuple

~scheduling_utils.SchedulerOutput if return_dict is True, otherwise a tuple. When returning a tuple, the first element is the sample tensor.

Step function propagating the sample with the multistep UniPC.