Diffusers contains multiple pre-built schedule functions for the diffusion process.
The schedule functions, denoted Schedulers in the library take in the output of a trained model, a sample which the diffusion process is iterating on, and a timestep to return a denoised sample.
All schedulers take in a timestep to predict the updated version of the sample being diffused.
The timesteps dictate where in the diffusion process the step is, where data is generated by iterating forward in time and inference is executed by propagating backwards through timesteps.
Different algorithms use timesteps that both discrete (accepting int inputs), such as the DDPMScheduler or PNDMScheduler, and continuous (accepting float inputs), such as the score-based schedulers ScoreSdeVeScheduler or ScoreSdeVpScheduler.
The core design principle between the schedule functions is to be model, system, and framework independent. This allows for rapid experimentation and cleaner abstractions in the code, where the model prediction is separated from the sample update. To this end, the design of schedulers is such that:
The core API for any new scheduler must follow a limited structure.
def step(...) functions that should be called to update the generated sample iteratively.set_timesteps(...) method that configures the parameters of a schedule function for a specific inference task.set_format(...) method.The base class SchedulerMixin implements low level utilities used by multiple schedulers.
Mixin containing common functions for the schedulers.
( values: typing.Union[numpy.ndarray, torch.Tensor] broadcast_array: typing.Union[numpy.ndarray, torch.Tensor] )
Turns a 1-D array into an array or tensor with len(broadcast_array.shape) dims.
( prev_sample: FloatTensor )
Base class for the scheduler’s step function output.
Original paper can be found here.
( num_train_timesteps: int = 1000 beta_start: float = 0.0001 beta_end: float = 0.02 beta_schedule: str = 'linear' trained_betas: typing.Optional[numpy.ndarray] = None timestep_values: typing.Optional[numpy.ndarray] = None clip_sample: bool = True set_alpha_to_one: bool = True tensor_format: str = 'pt' )
Parameters
int) — number of diffusion steps used to train the model.
float) — the starting beta value of inference.
float) — the final beta value.
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.
np.ndarray, optional) — TODO
np.ndarray, optional) — TODO
bool, default True) —
option to clip predicted sample between -1 and 1 for numerical stability.
bool, default True) —
if alpha for final step is 1 or the final alpha of the “non-previous” one.
str) — whether the scheduler expects pytorch or numpy arrays.
Denoising diffusion implicit models is a scheduler that extends the denoising procedure introduced in denoising diffusion probabilistic models (DDPMs) with non-Markovian guidance.
~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 save_config() and
from_config() functios.
For more details, see the original paper: https://arxiv.org/abs/2010.02502
( num_inference_steps: int offset: int = 0 )
Sets the discrete timesteps used for the diffusion chain. Supporting function to be run before inference.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
eta: float = 0.0
use_clipped_model_output: bool = False
generator = None
return_dict: bool = True
)
→
SchedulerOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
float) — weight of noise for added noise in diffusion step.
bool) — TODO
generator — random number generator.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SchedulerOutput or tuple
SchedulerOutput if return_dict is True, otherwise a tuple. When
returning a tuple, the first element is the sample tensor.
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).
Original paper can be found here.
( num_train_timesteps: int = 1000 beta_start: float = 0.0001 beta_end: float = 0.02 beta_schedule: str = 'linear' trained_betas: typing.Optional[numpy.ndarray] = None variance_type: str = 'fixed_small' clip_sample: bool = True tensor_format: str = 'pt' )
Parameters
int) — number of diffusion steps used to train the model.
float) — the starting beta value of inference.
float) — the final beta value.
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.
np.ndarray, optional) — TODO
str) —
options to clip the variance used when adding noise to the denoised sample. Choose from fixed_small,
fixed_small_log, fixed_large, fixed_large_log, learned or learned_range.
bool, default True) —
option to clip predicted sample between -1 and 1 for numerical stability.
str) — whether the scheduler expects pytorch or numpy arrays.
Denoising diffusion probabilistic models (DDPMs) explores the connections between denoising score matching and Langevin dynamics sampling.
~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 save_config() and
from_config() functios.
For more details, see the original paper: https://arxiv.org/abs/2006.11239
( num_inference_steps: int )
Sets the discrete timesteps used for the diffusion chain. Supporting function to be run before inference.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
predict_epsilon = True
generator = None
return_dict: bool = True
)
→
SchedulerOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
float) — weight of noise for added noise in diffusion step.
bool) —
optional flag to use when model predicts the samples directly instead of the noise, epsilon.
generator — random number generator.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SchedulerOutput or tuple
SchedulerOutput if return_dict is True, otherwise a tuple. When
returning a tuple, the first element is the sample tensor.
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).
Original paper can be found here.
( sigma_min: float = 0.02 sigma_max: float = 100 s_noise: float = 1.007 s_churn: float = 80 s_min: float = 0.05 s_max: float = 50 tensor_format: str = 'pt' )
Parameters
float) — minimum noise magnitude
float) — maximum noise magnitude
float) — the amount of additional noise to counteract loss of detail during sampling.
A reasonable range is [1.000, 1.011].
float) — the parameter controlling the overall amount of stochasticity.
A reasonable range is [0, 100].
float) — the start value of the sigma range where we add noise (enable stochasticity).
A reasonable range is [0, 10].
float) — the end value of the sigma range where we add noise.
A reasonable range is [0.2, 80].
str) — whether the scheduler expects pytorch or numpy arrays.
Stochastic sampling from Karras et al. [1] tailored to the Variance-Expanding (VE) models [2]. Use Algorithm 2 and the VE column of Table 1 from [1] for reference.
[1] Karras, Tero, et al. “Elucidating the Design Space of Diffusion-Based Generative Models.” https://arxiv.org/abs/2206.00364 [2] Song, Yang, et al. “Score-based generative modeling through stochastic differential equations.” https://arxiv.org/abs/2011.13456
~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 save_config() and
from_config() functios.
For more details on the parameters, see the original paper’s Appendix E.: “Elucidating the Design Space of Diffusion-Based Generative Models.” https://arxiv.org/abs/2206.00364. The grid search values used to find the optimal {s_noise, s_churn, s_min, s_max} for a specific model are described in Table 5 of the paper.
( sample: typing.Union[torch.FloatTensor, numpy.ndarray] sigma: float generator: typing.Optional[torch._C.Generator] = None )
Explicit Langevin-like “churn” step of adding noise to the sample according to a factor gamma_i ≥ 0 to reach a higher noise level sigma_hat = sigma_i + gamma_i*sigma_i.
TODO Args:
( num_inference_steps: int )
Sets the continuous timesteps used for the diffusion chain. Supporting function to be run before inference.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
sigma_hat: float
sigma_prev: float
sample_hat: typing.Union[torch.FloatTensor, numpy.ndarray]
return_dict: bool = True
)
→
KarrasVeOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
float) — TODO
float) — TODO
torch.FloatTensor or np.ndarray) — TODO
bool) — option for returning tuple rather than SchedulerOutput class
KarrasVeOutput — updated sample in the diffusion chain and derivative (TODO double check).
Returns
KarrasVeOutput or tuple
KarrasVeOutput if return_dict is True, otherwise a tuple. When
returning a tuple, the first element is the sample tensor.
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).
( model_output: typing.Union[torch.FloatTensor, numpy.ndarray] sigma_hat: float sigma_prev: float sample_hat: typing.Union[torch.FloatTensor, numpy.ndarray] sample_prev: typing.Union[torch.FloatTensor, numpy.ndarray] derivative: typing.Union[torch.FloatTensor, numpy.ndarray] return_dict: bool = True ) → prev_sample (TODO)
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
float) — TODO
float) — TODO
torch.FloatTensor or np.ndarray) — TODO
torch.FloatTensor or np.ndarray) — TODO
torch.FloatTensor or np.ndarray) — TODO
bool) — option for returning tuple rather than SchedulerOutput class
Returns
prev_sample (TODO)
updated sample in the diffusion chain. derivative (TODO): TODO
Correct the predicted sample based on the output model_output of the network. TODO complete description
Original implementation can be found here.
( num_train_timesteps: int = 1000 beta_start: float = 0.0001 beta_end: float = 0.02 beta_schedule: str = 'linear' trained_betas: typing.Optional[numpy.ndarray] = None timestep_values: typing.Optional[numpy.ndarray] = None tensor_format: str = 'pt' )
Parameters
int) — number of diffusion steps used to train the model.
float) — the starting beta value of inference.
float) — the final beta value.
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.
np.ndarray, optional) — TODO
options to clip the variance used when adding noise to the denoised sample. Choose from fixed_small,
fixed_small_log, fixed_large, fixed_large_log, learned or learned_range.
np.ndarry, optional) — TODO
str) — whether the scheduler expects pytorch or numpy arrays.
Linear Multistep Scheduler for discrete beta schedules. Based on the original k-diffusion implementation by Katherine Crowson: https://github.com/crowsonkb/k-diffusion/blob/481677d114f6ea445aa009cf5bd7a9cdee909e47/k_diffusion/sampling.py#L181
~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 save_config() and
from_config() functios.
( order t current_order )
Compute a linear multistep coefficient.
( num_inference_steps: int )
Sets the timesteps used for the diffusion chain. Supporting function to be run before inference.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
order: int = 4
return_dict: bool = True
)
→
SchedulerOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
order — coefficient for multi-step inference.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SchedulerOutput or tuple
SchedulerOutput if return_dict is True, otherwise a tuple. When
returning a tuple, the first element is the sample tensor.
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).
Original implementation can be found here.
( num_train_timesteps: int = 1000 beta_start: float = 0.0001 beta_end: float = 0.02 beta_schedule: str = 'linear' trained_betas: typing.Optional[numpy.ndarray] = None tensor_format: str = 'pt' skip_prk_steps: bool = False )
Parameters
int) — number of diffusion steps used to train the model.
float) — the starting beta value of inference.
float) — the final beta value.
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.
np.ndarray, optional) — TODO
str) — whether the scheduler expects pytorch or numpy arrays
bool) —
allows the scheduler to skip the Runge-Kutta steps that are defined in the original paper as being required
before plms steps; defaults to False.
Pseudo numerical methods for diffusion models (PNDM) proposes using more advanced ODE integration techniques, namely Runge-Kutta method and a linear multi-step method.
~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 save_config() and
from_config() functios.
For more details, see the original paper: https://arxiv.org/abs/2202.09778
( num_inference_steps: int offset: int = 0 )
Sets the discrete timesteps used for the diffusion chain. Supporting function to be run before inference.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
return_dict: bool = True
)
→
SchedulerOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SchedulerOutput or tuple
SchedulerOutput if return_dict is True, otherwise a tuple. When
returning a tuple, the first element is the sample tensor.
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).
This function calls step_prk() or step_plms() depending on the internal variable counter.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
return_dict: bool = True
)
→
SchedulerOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SchedulerOutput or tuple
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 linear multi-step method. This has one forward pass with multiple times to approximate the solution.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
return_dict: bool = True
)
→
SchedulerOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SchedulerOutput or tuple
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 Runge-Kutta method. RK takes 4 forward passes to approximate the solution to the differential equation.
Original paper can be found here.
( num_train_timesteps: int = 2000 snr: float = 0.15 sigma_min: float = 0.01 sigma_max: float = 1348.0 sampling_eps: float = 1e-05 correct_steps: int = 1 tensor_format: str = 'pt' )
Parameters
float) —
coefficient weighting the step from the model_output sample (from the network) to the random noise.
float) —
initial noise scale for sigma sequence in sampling procedure. The minimum sigma should mirror the
distribution of the data.
float) — maximum value used for the range of continuous timesteps passed into the model.
float) — the end value of sampling, where timesteps decrease progessively from 1 to
epsilon. —
int) — number of correction steps performed on a produced sample.
str) — “np” or “pt” for the expected format of samples passed to the Scheduler.
The variance exploding stochastic differential equation (SDE) scheduler.
For more information, see the original paper: https://arxiv.org/abs/2011.13456
~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 save_config() and
from_config() functios.
( num_inference_steps: int sigma_min: float = None sigma_max: float = None sampling_eps: float = None )
Parameters
int) —
the number of diffusion steps used when generating samples with a pre-trained model.
float, optional) —
initial noise scale value (overrides value given at Scheduler instantiation).
float, optional) — final noise scale value (overrides value given at Scheduler instantiation).
float, optional) — final timestep value (overrides value given at Scheduler instantiation).
Sets the noise scales used for the diffusion chain. Supporting function to be run before inference.
The sigmas control the weight of the drift and diffusion components of sample update.
( num_inference_steps: int sampling_eps: float = None )
Sets the continuous timesteps used for the diffusion chain. Supporting function to be run before inference.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
generator: typing.Optional[torch._C.Generator] = None
return_dict: bool = True
**kwargs
)
→
SdeVeOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
generator — random number generator.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SdeVeOutput or tuple
SdeVeOutput if
return_dict is True, otherwise a tuple. When returning a tuple, the first element is the sample tensor.
Correct the predicted sample based on the output model_output of the network. This is often run repeatedly after making the prediction for the previous timestep.
(
model_output: typing.Union[torch.FloatTensor, numpy.ndarray]
timestep: int
sample: typing.Union[torch.FloatTensor, numpy.ndarray]
generator: typing.Optional[torch._C.Generator] = None
return_dict: bool = True
**kwargs
)
→
SdeVeOutput or tuple
Parameters
torch.FloatTensor or np.ndarray) — direct output from learned diffusion model.
int) — current discrete timestep in the diffusion chain.
torch.FloatTensor or np.ndarray) —
current instance of sample being created by diffusion process.
generator — random number generator.
bool) — option for returning tuple rather than SchedulerOutput class
Returns
SdeVeOutput or tuple
SdeVeOutput if
return_dict is True, otherwise a tuple. When returning a tuple, the first element is the sample tensor.
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).
Original paper can be found here.
Score SDE-VP is under construction.
( num_train_timesteps = 2000 beta_min = 0.1 beta_max = 20 sampling_eps = 0.001 tensor_format = 'np' )
The variance preserving stochastic differential equation (SDE) scheduler.
~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 save_config() and
from_config() functios.
For more information, see the original paper: https://arxiv.org/abs/2011.13456
UNDER CONSTRUCTION