+h% S SK Jr S SKJrJrJr S SKrS SKrSSK J
r
J r SSK J
r
SSKJr SS KJr \ " S
S \
5 5 r " S S
\\
5 rg) ) dataclass)OptionalTupleUnionN ) ConfigMixinregister_to_config)
BaseOutput) randn_tensor )SchedulerMixinc ~ \ rS rSr% Sr\R \S' \R \S' Sr\ \R \S' Sr
g)KarrasVeOutput a
Output class for the scheduler's step function output.
Args:
prev_sample (`torch.Tensor` 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.
derivative (`torch.Tensor` of shape `(batch_size, num_channels, height, width)` for images):
Derivative of predicted original image sample (x_0).
pred_original_sample (`torch.Tensor` 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
derivativeNpred_original_sample ) __name__
__module__ __qualname____firstlineno____doc__torchTensor__annotations__r r __static_attributes__r n/home/james-whalen/.local/lib/python3.13/site-packages/diffusers/schedulers/deprecated/scheduling_karras_ve.pyr r s1 37(5<<07r r c * \ rS rSrSrSr\ S!S\S\S\S\S\S \4 S
jj5 rS"S \ R S
\ \ S\ R 4S jjr
S"S\ S\\\ R 4 4S jjr S"S \ R S\S\ \ R$ S\\ R \4 4S jjr S#S\ R S\S\S\ R S\S\\\4 4 S jjr S#S\ R S\S\S\ R S \ R S \ R S\S\\\4 4S jjrS rS rg )$KarrasVeScheduler1 a
A stochastic scheduler tailored to variance-expanding models.
This model inherits from [`SchedulerMixin`] and [`ConfigMixin`]. Check the superclass documentation for the generic
methods the library implements for all schedulers such as loading and saving.
For more details on the parameters, see [Appendix E](https://huggingface.co/papers/2206.00364). The grid search
values used to find the optimal `{s_noise, s_churn, s_min, s_max}` for a specific model are described in Table 5 of
the paper.
Args:
sigma_min (`float`, defaults to 0.02):
The minimum noise magnitude.
sigma_max (`float`, defaults to 100):
The maximum noise magnitude.
s_noise (`float`, defaults to 1.007):
The amount of additional noise to counteract loss of detail during sampling. A reasonable range is [1.000,
1.011].
s_churn (`float`, defaults to 80):
The parameter controlling the overall amount of stochasticity. A reasonable range is [0, 100].
s_min (`float`, defaults to 0.05):
The start value of the sigma range to add noise (enable stochasticity). A reasonable range is [0, 10].
s_max (`float`, defaults to 50):
The end value of the sigma range to add noise. A reasonable range is [0.2, 80].
r sigma_min sigma_maxs_noises_churns_mins_maxc : X l S U l S U l S U l g N)init_noise_sigmanum_inference_steps timestepsschedule)selfr# r$ r% r& r' r( s r __init__KarrasVeScheduler.__init__R s" !* )- '+&*
r Nsampletimestepreturnc U$ )aN
Ensures interchangeability with schedulers that need to scale the denoising model input depending on the
current timestep.
Args:
sample (`torch.Tensor`):
The input sample.
timestep (`int`, *optional*):
The current timestep in the diffusion chain.
Returns:
`torch.Tensor`:
A scaled input sample.
r )r/ r2 r3 s r scale_model_input#KarrasVeScheduler.scale_model_inputd s
r r, devicec Xl [ R " SU R 5 SSS2 R 5 n[ R
" U5 R
U5 U l U R Vs/ s HX nU R R S- U R R S- U R R S- - XAS-
- - - PMZ nn[ R " U[ R US9U l
gs snf )au
Sets the discrete timesteps used for the diffusion chain (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.
r Nr )dtyper8 )r, nparangecopyr
from_numpytor- configr$ r# tensorfloat32r. )r/ r, r8 r- ir. s r
set_timesteps KarrasVeScheduler.set_timestepsu s $7 IIa!9!9:4R4@EEG )))477? ^^
$ %%q(;;((!+dkk.C.CQ.FFAghQhLijk $
XU]]6R
s 4A C?sigma generatorc U R R Us=:: a U R R :: a1 O O.[ U R R U R
- S5 nOSnU R R [ UR US9R UR 5 - nX$U- - nXS- US- -
S- U- - nXv4$ )uN
Explicit Langevin-like "churn" step of adding noise to the sample according to a `gamma_i ≥ 0` to reach a
higher noise level `sigma_hat = sigma_i + gamma_i*sigma_i`.
Args:
sample (`torch.Tensor`):
The input sample.
sigma (`float`):
generator (`torch.Generator`, *optional*):
A random number generator.
g4y?r )rI r ?) rB r' r( minr& r, r% r shaperA r8 )r/ r2 rH rI gammaeps sigma_hat
sample_hats r add_noise_to_input$KarrasVeScheduler.add_noise_to_input s ;; :):): : ++d.F.FF
SEE kk!!L $S$V$VW]WdWd$eeEM) 1 uax 7C ?# EF
$$r model_outputrP
sigma_prevrQ return_dictc Z XBU- - nXF-
U- nXCU-
U- - nU( d X4$ [ XUS9$ )a)
Predict the sample from the previous timestep by reversing the SDE. This function propagates the diffusion
process from the learned model outputs (most often the predicted noise).
Args:
model_output (`torch.Tensor`):
The direct output from learned diffusion model.
sigma_hat (`float`):
sigma_prev (`float`):
sample_hat (`torch.Tensor`):
return_dict (`bool`, *optional*, defaults to `True`):
Whether or not to return a [`~schedulers.scheduling_karras_ve.KarrasVESchedulerOutput`] or `tuple`.
Returns:
[`~schedulers.scheduling_karras_ve.KarrasVESchedulerOutput`] or `tuple`:
If return_dict is `True`, [`~schedulers.scheduling_karras_ve.KarrasVESchedulerOutput`] is returned,
otherwise a tuple is returned where the first element is the sample tensor.
r r r r ) r/ rT rP rU rQ rV r r sample_prevs r stepKarrasVeScheduler.step sP 8 * ,D D 79D
$:j#HH , , #Qe
r rZ r c l XSU- - nXX-
U- n XCU-
SU- SU - - - - nU( d XV4$ [ XVUS9$ )a\
Corrects the predicted sample based on the `model_output` of the network.
Args:
model_output (`torch.Tensor`):
The direct output from learned diffusion model.
sigma_hat (`float`): TODO
sigma_prev (`float`): TODO
sample_hat (`torch.Tensor`): TODO
sample_prev (`torch.Tensor`): TODO
derivative (`torch.Tensor`): TODO
return_dict (`bool`, *optional*, defaults to `True`):
Whether or not to return a [`~schedulers.scheduling_ddpm.DDPMSchedulerOutput`] or `tuple`.
Returns:
prev_sample (TODO): updated sample in the diffusion chain. derivative (TODO): TODO
rK rX rY )
r/ rT rP rU rQ rZ r rV r derivative_corrs
r step_correct KarrasVeScheduler.step_correct sa 8 +,-F F &=K $:sZ?ORUXgRg?g#hh , , #Qe
r c [ 5 er* )NotImplementedError)r/ original_samplesnoiser- s r add_noiseKarrasVeScheduler.add_noise s !##r )r+ r, r. r- )g{Gz?d g&1 ?P g?2 r* )T)r r r r r orderr floatr0 r r r intr6 r strr8 rF Generatorr rR boolr r[ r_ re r r r r r! r! 1 s <
E ++ + +
+
+ + +"
Y^YeYe "S SeC DU>V S. Z^%ll%+0%=Eeoo=V% u||U" #%@ !
%
ll%
%
%
LL %
%
~u$ %%
^ !%
ll%
%
%
LL %
\\
%
LL%
%
~u$ %%
N$r r! ) dataclassesr typingr r r numpyr= r configuration_utilsr r utilsr
utils.torch_utilsr scheduling_utilsr
r r! r r r rw sM " ) ) B - - 8Z 8 8(C$ C$r