+h"SSKrSSKJr SSKJrJrJrJr SSKrSSK J r J r SSK J r SSKJr SS KJr \"S S \ 55rSS jr"S S\\ 5rg)N) dataclass)ListOptionalTupleUnion) ConfigMixinregister_to_config) BaseOutput) randn_tensor)SchedulerMixinc8\rSrSr%Sr\R \S'Srg)DDPMWuerstchenSchedulerOutputa) 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. prev_sampleN) __name__ __module__ __qualname____firstlineno____doc__torchTensor__annotations____static_attributes__ri/home/james-whalen/.local/lib/python3.13/site-packages/diffusers/schedulers/scheduling_ddpm_wuerstchen.pyrrsrrc &US:XaSnOUS:XaSnO[SU35e/n[U5H<nXP- nUS-U- nUR[SU"U5U"U5- - U55 M> [R "U[R S9$)a Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t = [0,1]. Contains a function alpha_bar that takes an argument t and transforms it to the cumulative product of (1-beta) up to that part of the diffusion process. Args: num_diffusion_timesteps (`int`): the number of betas to produce. max_beta (`float`): the maximum beta to use; use values lower than 1 to prevent singularities. alpha_transform_type (`str`, *optional*, default to `cosine`): the type of noise schedule for alpha_bar. Choose from `cosine` or `exp` Returns: betas (`np.ndarray`): the betas used by the scheduler to step the model outputs cosinech[R"US-S- [R-S- 5S-$)NMb?gT㥛 ?r)mathcospits r alpha_bar_fn)betas_for_alpha_bar..alpha_bar_fnEs-88QY%/$''9A=>!C Crexpc4[R"US-5$)Ng()r#r*r&s rr(r)Js88AI& &rz"Unsupported alpha_transform_type: r dtype) ValueErrorrangeappendminrtensorfloat32)num_diffusion_timestepsmax_betaalpha_transform_typer(betasit1t2s rbetas_for_alpha_barr;,s.x' D  & '=>R=STUU E * +  (!e. . S\"- R0@@@(KL, <<U]] 33rc \rSrSrSr\SS\S\4Sjj5rSrSS\ RS \ \ S \ R4S jjr SS \ S \ \\ S\\\ R"44SjjrSS\ RS \ S\ RS\S \\\44 SjjrS\ RS\ RS \ RS \ R4SjrSrSrSrg)DDPMWuerstchenSchedulerXa 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`. [`SchedulerMixin`] provides general loading and saving functionality via the [`SchedulerMixin.save_pretrained`] and [`~SchedulerMixin.from_pretrained`] functions. For more details, see the original paper: https://huggingface.co/papers/2006.11239 Args: scaler (`float`): .... s (`float`): .... scalerscXl[R"U/5Ul[R"URSUR-- [R -S-5S-UlSUlg)Nr ?r?)r?rr2r@r$r%_init_alpha_cumprodinit_noise_sigma)selfr?r@s r__init__ DDPMWuerstchenScheduler.__init__is\  qc"#(99TVVq466z-BUXX-MPS-S#TXY#Y !$rcURS:aSSU- UR-- nOURS:aXR-n[R"XRR U5-SURR U5-- [R -S-5S-UR R U5- nURSS5$)Nr rBrg-C6?gH.?)r?rr$r@tor%rDclamp)rFr'device alpha_cumprods r_alpha_cumprod&DDPMWuerstchenScheduler._alpha_cumprodvs ;;?QUt{{**A [[1_;;A 6" "q46699V+<'< = H3 N )),,V45 ""6622rNsampletimestepreturncU$)a Ensures interchangeability with schedulers that need to scale the denoising model input depending on the current timestep. Args: sample (`torch.Tensor`): input sample timestep (`int`, optional): current timestep Returns: `torch.Tensor`: scaled input sample r)rFrPrQs rscale_model_input)DDPMWuerstchenScheduler.scale_model_inputs  rnum_inference_steps timestepsrLcUc[R"SSUS-US9n[U[R5(d%[R"U5R U5nX lg)a Sets the discrete timesteps used for the diffusion chain. Supporting function to be run before inference. Args: num_inference_steps (`Dict[float, int]`): the number of diffusion steps used when generating samples with a pre-trained model. If passed, then `timesteps` must be `None`. device (`str` or `torch.device`, optional): the device to which the timesteps are moved to. {2 / 3: 20, 0.0: 10} NrCgr rL)rlinspace isinstancerrJrW)rFrVrWrLs r set_timesteps%DDPMWuerstchenScheduler.set_timestepssT  sC1Dq1HQWXI)U\\22 Y/226:I"r model_output return_dictc URnURnUnURU5n URX5R"UR S5/UR SSV s/sHn SPM sn Q76n URX5R"U R S5/UR SSV s/sHn SPM sn Q76n X- n SU - R5USU - U-SU - R5- - -n[UR XARURS9nSU - SU - -SU - - R5U-nUUU S:gR5R"U R S5/UR SSV s/sHn SPM sn Q76--nU(dURU54$[URU5S9$s sn fs sn fs sn f)a 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.Tensor`): direct output from learned diffusion model. timestep (`int`): current discrete timestep in the diffusion chain. sample (`torch.Tensor`): current instance of sample being created by diffusion process. generator: random number generator. return_dict (`bool`): option for returning tuple rather than DDPMWuerstchenSchedulerOutput class Returns: [`DDPMWuerstchenSchedulerOutput`] or `tuple`: [`DDPMWuerstchenSchedulerOutput`] if `return_dict` is True, otherwise a `tuple`. When returning a tuple, the first element is the sample tensor. rr NrC) generatorrLr-)r) r-rLprevious_timesteprNviewsizeshapesqrtr floatrJr)rFr^rQrPrar_r-rLr'prev_t_rMalpha_cumprod_prevalphamu std_noisestdpreds rstepDDPMWuerstchenScheduler.steps2""$$ ''*++A6;;AFF1IfTZT`T`abacTdHeTdqTdHef !00@EEfkkRSnucicocopqprcsWtcs^_XYcsWtu2Ek   !Vq5yL.HAP]L]KcKcKe.e%e f YGZGZbnbtbtu E c$6673;NOUUWZccC6Q;--/44V[[^dRXR^R^_`_aRbFcRbQqRbFcdddGGEN$ $,HHIfWtGds) G 8 G Goriginal_samplesnoisecHURnURnURX4S9R"UR S5/UR SSVs/sHnSPM snQ76nUR 5U-SU- R 5U--nURUS9$s snf)NrYrr r,)rLr-rNrcrdrerfrJ) rFrrrsrWrLr-rirM noisy_sampless r add_noise!DDPMWuerstchenScheduler.add_noises "(( &&++I+EJJ NN1  ,<,B,B12,F G,Fq,F G &**,/??1}CTBZBZB\_dBdd e,,!Hs B c.URR$N)confignum_train_timesteps)rFs r__len__DDPMWuerstchenScheduler.__len__s{{...rcURUS- R5R5R5nURUS-SR UR S5nU$)Nrr )rWabsargminitemexpandre)rFrQindexrhs rrb)DDPMWuerstchenScheduler.previous_timesteps_(1+-224;;=BBD *4077q8IJ r)rDrEr@r?rW)rCr"ry)NNN)NT)rrrrrr rgrGrNrrrintrTrrstrrLr\boolrrrprvr|rbrrrrr=r=XsB  $ $  $ $3   Y^YeYe  $()-+/ # #DI&#c5<<'( #6 ,Ill,I,I ,I  ,I ,e3 4,I\ -,, -|| -<< -  -/rr=)g+?r )r# dataclassesrtypingrrrrrconfiguration_utilsr r utilsr utils.torch_utilsr scheduling_utilsrrr;r=rrrrs[$ !// A,,  J   !)4XNnkNr