3j3SSKJr SSKrSSKrSSKJrJr SSKJ r J r SSK J r SSK Jr \ R"\5r\"S S \ 55r"S S \\5rg) ) dataclassN) ConfigMixinregister_to_config) BaseOutputlogging) randn_tensor)SchedulerMixinc8\rSrSr%Sr\R \S'Srg)$FlowMatchHeunDiscreteSchedulerOutputa2 Output class for the scheduler's `step` function output. Args: prev_sample (`torch.FloatTensor` of shape `(batch_size, num_channels, height, width)` for images): Computed sample `(x_{t-1})` of previous timestep. `prev_sample` should be used as next model input in the denoising loop. prev_sampleN) __name__ __module__ __qualname____firstlineno____doc__torch FloatTensor__annotations____static_attributes__rr/home/wildlama/miniconda3/lib/python3.13/site-packages/diffusers/schedulers/scheduling_flow_match_heun_discrete.pyr r s"""rr c\rSrSrSr/rSr\S(S\S\ 4Sjj5r \ S5r \ S 5r S)S \4S jjrS \R S \ \R -S\R S\R 4SjrS\ S\ 4SjrS*S\S\\R(-SS4SjjrS*S \ \R -S\R S-S\4SjjrS \ \R -SS4Sjr\ S\4Sj5rSS\ "S5SSS4S\R S \ \R -S \R S\ S \ S!\ S"\ S#\R4S-S$\S\\-4S%jjrS\4S&jrS'rg)+FlowMatchHeunDiscreteScheduler+a Heun scheduler. 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. Args: num_train_timesteps (`int`, defaults to 1000): The number of diffusion steps to train the model. shift (`float`, defaults to 1.0): The shift value for the timestep schedule. r?num_train_timestepsshiftc[R"SX[RS9SSS2R5n[R "U5R [RS9nX1- nX$-SUS- U--- nXA-UlSUlSUl UR S5Ul URSR5Ul URSR5Ul g)Nr )dtypecpur)nplinspacefloat32copyr from_numpyto timesteps _step_index _begin_indexsigmasitem sigma_min sigma_max)selfr r!r,r/s r__init__'FlowMatchHeunDiscreteScheduler.__init__<s KK#6SUS]S]^_cac_cdiik $$Y/222G 01 V';#;<5 ii& R--/Q,,.rcUR$)zW The index counter for current timestep. It will increase 1 after each scheduler step. )r-r3s r step_index)FlowMatchHeunDiscreteScheduler.step_indexQs rcUR$)za The index for the first timestep. It should be set from pipeline with `set_begin_index` method. r.r7s r begin_index*FlowMatchHeunDiscreteScheduler.begin_indexXs    rr<cXlg)z Sets the begin index for the scheduler. This function should be run from pipeline before the inference. Args: begin_index (`int`, defaults to `0`): The begin index for the scheduler. Nr;)r3r<s rset_begin_index.FlowMatchHeunDiscreteScheduler.set_begin_index`s (rsampletimestepnoisereturncURcURU5 URURnXC-SU- U--nU$)aF Forward process in flow-matching Args: sample (`torch.FloatTensor`): The input sample. timestep (`float` or `torch.FloatTensor`): The current timestep in the diffusion chain. noise (`torch.FloatTensor`): The noise tensor. Returns: `torch.FloatTensor`: A scaled input sample. r)r8_init_step_indexr/)r3rArBrCsigmas r scale_noise*FlowMatchHeunDiscreteScheduler.scale_noisejsH* ?? "  ! !( + DOO,#+!77 rrGc2XRR-$Nconfigr )r3rGs r _sigma_to_t*FlowMatchHeunDiscreteScheduler._sigma_to_ts{{6666rNnum_inference_stepsdevicecjXl[R"URUR5URUR 5U5nX0R R- nUR RU-SUR RS- U--- n[R"U5R[RUS9nX@R R-n[R"USSUSSRS5/5nURUS9Ul[R"U[R "SUR"S9/5n[R"USSUSSRS5USS/5UlSUlSUlSUlSUlg)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 )r#rQNr)rQr$)rPr&r'rNr2r1rMr r!rr*r+r(catrepeat_interleaver,zerosrQr/prev_derivativedtr-r.)r3rPrQr,r/s r set_timesteps,FlowMatchHeunDiscreteScheduler.set_timestepssr$7 KK   T^^ ,   T^^ ,  [[<<<""V+qDKK4E4E4IV3S/ST!!&),,5==,P[[<<< IIy!}im.M.Ma.PQR "V4FEKK&--$HIJiiVAb\-K-KA-NPVWYWZP[ \]  $ rschedule_timestepscUc URnX!:HR5n[U5S:aSOSnX4R5$)ax Find the index of a given timestep in the timestep schedule. Args: timestep (`float` or `torch.FloatTensor`): The timestep value to find in the schedule. schedule_timesteps (`torch.FloatTensor`, *optional*): The timestep schedule to search in. If `None`, uses `self.timesteps`. Returns: `int`: The index of the timestep in the schedule. r r)r,nonzerolenr0)r3rBrZindicesposs rindex_for_timestep1FlowMatchHeunDiscreteScheduler.index_for_timestepsJ$  %!% %1::< w[U[R5(d[U[R5(a [ S5eUR cURU5 UR[R5nUR(a6URUR n URUR S-n O5URUR S- n URUR n XZs=::aU::a(O O%[U[UR5S- - S5OSn XS--n U S:aA[URUR UR"US9nX-nX?U S-U S-- S ---nUR(a(X1U -- nUU- U - nX- nUUlUUlX0lONX1U -- nUU- U - nS UR$U--nUR&nUR(nSUlSUlSUlUUU--nURUR 5nU=R*S- slU (dU4$[-US 9$) ac 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.FloatTensor`): The direct output from learned diffusion model. timestep (`float` or `torch.FloatTensor`): The current discrete timestep in the diffusion chain. sample (`torch.FloatTensor`): A current instance of a sample created by the diffusion process. s_churn (`float`): Stochasticity parameter that controls the amount of noise added during sampling. Higher values increase randomness. s_tmin (`float`): Minimum timestep threshold for applying stochasticity. Only timesteps above this value will have noise added. s_tmax (`float`): Maximum timestep threshold for applying stochasticity. Only timesteps below this value will have noise added. s_noise (`float`, defaults to 1.0): Scaling factor for noise added to the sample. generator (`torch.Generator`, *optional*): A random number generator. return_dict (`bool`): Whether or not to return a [`~schedulers.scheduling_flow_match_heun_discrete.FlowMatchHeunDiscreteSchedulerOutput`] tuple. Returns: [`~schedulers.scheduling_flow_match_heun_discrete.FlowMatchHeunDiscreteSchedulerOutput`] or `tuple`: If return_dict is `True`, [`~schedulers.scheduling_flow_match_heun_discrete.FlowMatchHeunDiscreteSchedulerOutput`] is returned, otherwise a tuple is returned where the first element is the sample tensor. zPassing integer indices (e.g. from `enumerate(timesteps)`) as timesteps to `FlowMatchHeunDiscreteScheduler.step()` is not supported. Make sure to pass one of the `scheduler.timesteps` as a timestep.Nr g4y?rir)r#rQrprg?)r)rcintr IntTensor LongTensor ValueErrorr8rFr+r(rgr/minr]r shaper#rQrVrWrAr-r )r3rkrBrArlrmrnrorprqrG sigma_nextgamma sigma_hatrCepsdenoised derivativerWrs rstep#FlowMatchHeunDiscreteScheduler.stepsQ` x % %(EOO44(E$4$455G  ?? "  ! !( +5==)  $ $KK0ET__q%89JKK! 34ET__5JEKE^X^E^Gs4;;/!34jAdgQY' 19 """((#**# E /CY\E1H%<$DDDF  $ $u 44H 8+y8J'B$.D DG Kz 99H 8+z9J 4 4z ABJB[[F$(D DGDKzB. !nn\%7%78  A> !3 LLrc.URR$rKrLr7s r__len__&FlowMatchHeunDiscreteScheduler.__len__bs{{...r) r.r-rWrPrVrAr2r1r/r,)ir)rrK) rrrrr _compatiblesorderrrsfloatr4propertyr8r<r?rrrHrNstrrQrXr`rFboolrg Generatorr tuplerrrrrrrr+s: L E$(/ ///(  !!(3(!!%+++      <7757 &*&! &!ell"&!  &!V8<#%+++#"--4# #>11B1B)B1t1de ,0 M''M%+++M!! M  M  MMM??T)MM . 5MB//rr) dataclassesrnumpyr&rconfiguration_utilsrrutilsrrutils.torch_utilsr scheduling_utilsr get_loggerrloggerr rrrrrsY" A',,   H %  #: #  #x/^[x/r