3j:u SSKrSSKJr SSKJr SSKrSSKrSSKJ r J r SSK J r J r SSKJrJr \ "5(aSSKr\"S S \ 55rSS \S \S \SS\R,4Sjjr"SS\\ 5rg)N) dataclass)Literal) ConfigMixinregister_to_config) BaseOutputis_scipy_available)KarrasDiffusionSchedulersSchedulerMixinc`\rSrSr%Sr\R \S'Sr\R S-\S'Sr g)KDPM2DiscreteSchedulerOutputaM 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. 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_sampleNpred_original_sample) __name__ __module__ __qualname____firstlineno____doc__torchTensor__annotations__r__static_attributes__rj/home/wildlama/miniconda3/lib/python3.13/site-packages/diffusers/schedulers/scheduling_k_dpm_2_discrete.pyrrs' 04%,,-4rrnum_diffusion_timestepsmax_betaalpha_transform_type)cosineexplaplacereturnc :US:XaSnO"US:XaSnOUS:XaSnO[SU35e/n[U5H<nXP- nUS-U- nUR[SU"U5U"U5- - U55 M> [R "U[R S 9$) a Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t = [0,1]. Contains a function alpha_bar that takes an argument t and transforms it to the cumulative product of (1-beta) up to that part of the diffusion process. Args: num_diffusion_timesteps (`int`): The number of betas to produce. max_beta (`float`, defaults to `0.999`): The maximum beta to use; use values lower than 1 to avoid numerical instability. alpha_transform_type (`str`, defaults to `"cosine"`): The type of noise schedule for `alpha_bar`. Choose from `cosine`, `exp`, or `laplace`. Returns: `torch.Tensor`: The betas used by the scheduler to step the model outputs. r!ch[R"US-S- [R-S- 5S-$)NgMb?gT㥛 ?r)mathcospits r alpha_bar_fn)betas_for_alpha_bar..alpha_bar_fnMs-88QY%/$''9A=>!C Crr#c S[R"SSU- 5-[R"SS[R"SU- 5-- S-5-n[R"U5n[R "USU-- 5$)Ngr ?rgư>)r'copysignlogfabsr"sqrt)r+lmbsnrs rr,r-Rsmq#'22TXXa!diiPSVWPWFXBX>X[_>_5``C((3-C99SAG_- -rr"c4[R"US-5$)Ng()r'r"r*s rr,r-Ys88AI& &rz"Unsupported alpha_transform_type: r dtype) ValueErrorrangeappendminrtensorfloat32)rrr r,betasit1t2s rbetas_for_alpha_barrC3s0x' D  * .  & '=>R=STUU E * +  (!e. . S\"- R0@@@(KL, <<U]] 33rc\rSrSrSr\VVs/sHoR PM snnrSr\ S1S\ S\ S\ S\ S \ R\\ -S-S \S \S \S \ S\ S\ 4Sjj5r\S5r\S5r\S5rS2S\ 4SjjrS\R0S\ \R0-S\R04SjrS3S\ S\ \R4-S\ 4Sjjr\S5rS4S\ \R0-S\R0S-S\ 4SjjrS\ \R0-SS4S jrS!rS"\R0S\R04S#jr S"\R0S\ S\R04S$jr!S5S"\R0S\ S%\ S&\ S\R04 S'jjr"S6S(\R0\ R-S\ \R0-S\R0\ R-S)\S\#\$-4 S*jjr%S+\R0S,\R0S-\R0S\R04S.jr&S/r'S0r(gs snnf)7KDPM2DiscreteSchedulergu KDPM2DiscreteScheduler is inspired by the DPMSolver2 and Algorithm 2 from the [Elucidating the Design Space of Diffusion-Based Generative Models](https://huggingface.co/papers/2206.00364) paper. 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. beta_start (`float`, defaults to 0.00085): The starting `beta` value of inference. beta_end (`float`, defaults to 0.012): The final `beta` value. beta_schedule (`str`, defaults to `"linear"`): The beta schedule, a mapping from a beta range to a sequence of betas for stepping the model. Choose from `linear` or `scaled_linear`. trained_betas (`np.ndarray`, *optional*): Pass an array of betas directly to the constructor to bypass `beta_start` and `beta_end`. use_karras_sigmas (`bool`, *optional*, defaults to `False`): Whether to use Karras sigmas for step sizes in the noise schedule during the sampling process. If `True`, the sigmas are determined according to a sequence of noise levels {σi}. use_exponential_sigmas (`bool`, *optional*, defaults to `False`): Whether to use exponential sigmas for step sizes in the noise schedule during the sampling process. use_beta_sigmas (`bool`, *optional*, defaults to `False`): Whether to use beta sigmas for step sizes in the noise schedule during the sampling process. Refer to [Beta Sampling is All You Need](https://huggingface.co/papers/2407.12173) for more information. prediction_type (`str`, defaults to `epsilon`, *optional*): Prediction type of the scheduler function; can be `epsilon` (predicts the noise of the diffusion process), `sample` (directly predicts the noisy sample`) or `v_prediction` (see section 2.4 of [Imagen Video](https://huggingface.co/papers/2210.02303) paper). timestep_spacing (`str`, defaults to `"linspace"`): The way the timesteps should be scaled. Refer to Table 2 of the [Common Diffusion Noise Schedules and Sample Steps are Flawed](https://huggingface.co/papers/2305.08891) for more information. steps_offset (`int`, defaults to 0): An offset added to the inference steps, as required by some model families. rNnum_train_timesteps beta_startbeta_end beta_schedule trained_betasuse_karras_sigmasuse_exponential_sigmasuse_beta_sigmasprediction_typetimestep_spacing steps_offsetc URR(a[5(d [S5e[ URRURR URR /5S:a [S5eUb)[R"U[RS9Ul OUS:Xa*[R"X#U[RS9Ul OkUS:Xa4[R"US-US-U[RS9S-Ul O1US :Xa[U5Ul O[US UR35eS UR- Ul[R""UR S S 9UlUR'USU5 SUlSUlUR,R/S5Ulg)Nz:Make sure to install scipy if you want to use beta sigmas.r znOnly one of `config.use_beta_sigmas`, `config.use_exponential_sigmas`, `config.use_karras_sigmas` can be used.r7linear scaled_linearr/rsquaredcos_cap_v2z is not implemented for g?rdimcpu)configrNr ImportErrorsumrMrLr9rr=r>r?linspacerCNotImplementedError __class__alphascumprodalphas_cumprod set_timesteps _step_index _begin_indexsigmasto) selfrGrHrIrJrKrLrMrNrOrPrQs r__init__KDPM2DiscreteScheduler.__init__s{ ;; & &/A/C/CZ[ [  ++T[[-O-OQUQ\Q\QnQno pst tA   $m5==IDJ h & >QY^YfYfgDJ o - C3H[chcpcpquvvDJ 1 1,-@ADJ%7OPTP^P^O_&`a aDJJ& #mmDKKQ? .6IJ kknnU+ rcURRS;aURR5$URR5S-S-S-$)N)r\trailingrr r/)rYrPremaxrgs rinit_noise_sigma'KDPM2DiscreteScheduler.init_noise_sigmasH ;; ' '+C C;;??$ $ !Q&*s22rcUR$)zW The index counter for current timestep. It will increase 1 after each scheduler step. )rcrms r step_index!KDPM2DiscreteScheduler.step_indexs rcUR$)za The index for the first timestep. It should be set from pipeline with `set_begin_index` method. rdrms r begin_index"KDPM2DiscreteScheduler.begin_indexs    rrucXlg)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. Nrt)rgrus rset_begin_index&KDPM2DiscreteScheduler.set_begin_indexs (rsampletimestepr$cURcURU5 UR(aURURnOURURnXS-S-S-- nU$)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. rr r/)rq_init_step_indexstate_in_first_orderresigmas_interpol)rgrzr{sigmas rscale_model_input(KDPM2DiscreteScheduler.scale_model_inputse& ?? "  ! !( +  $ $KK0E((9E1HqLS01 rnum_inference_stepsdevicec  XlU=(d URRnURRS:Xa>[R "SUS- U[R S9SSS2R5nGOcopyarangeroundastyperQr9arrayrar1interplenrL_convert_to_karras _sigma_to_trM_convert_to_exponentialrN_convert_to_betar from_numpyrf log_sigmas concatenatelerprollr"catrepeat_interleavererrXr8stackflatten timestepsrzrcrd) rgrrrGr step_ratiorerrrsigma_interpoltimesteps_interpolinterleaved_timestepss rrb$KDPM2DiscreteScheduler.set_timestepss$7 1TT[[5T5T ;; ' ': 5 A':Q'>@S[][e[efgkikgklqqsI [[ ) )Y 6,0H0HHJ1&9:ZGNNPQUSUQUV[[]ddegeoeopI 11 1I [[ ) )Z 7,/G/GGJ#6J;GNNPUUW^^_a_i_ijI NI;;//01JK A 3 33t7J7JJsRSVVF^ 9bii3v;&?H ;; ( (,,v,gFSY!ZSY%"2"25"ESY!Z[aacI [[ / /11F1lFSY!ZSY%"2"25"ESY!Z[I [[ ( (**V*eFSY!ZSY%"2"25"ESY!Z[I**:6999H#077 C!!&),,F,;!**,++FKKN,>,>,@#FJJLiiVABZ-I-I!-LfUWUXk Z[ $yy Ra /!""5"G"G"JO\^\_L` a $$Y/226: *--/__((* XXP_ `P_n  n 9P_ ` #--.@ADDVSbSbDc % -?"d -KYWXWY[_W_M`,agi j r r tIbqM3H#IJ  kknnU+ K"["["[( as7W%WW"W'cURSL$N)rzrms rr~+KDPM2DiscreteScheduler.state_in_first_orderPs{{d""rschedule_timestepscUc URnX!:HR5n[U5S:aSOSnX4R5$)a Find the index of a given timestep in the timestep schedule. Args: timestep (`float` or `torch.Tensor`): The timestep value to find in the schedule. schedule_timesteps (`torch.Tensor`, *optional*): The timestep schedule to search in. If `None`, uses `self.timesteps`. Returns: `int`: The index of the timestep in the schedule. For the very first step, returns the second index if multiple matches exist to avoid skipping a sigma when starting mid-schedule (e.g., for image-to-image). r r)rnonzeroritem)rgr{rindicesposs rindex_for_timestep)KDPM2DiscreteScheduler.index_for_timestepUsJ"  %!% %1::< wTUCUZ]B]3]#^q.1,-$ [[ ( (H 4%&ST T+DKK,G,G+HHtu   $ $ 79DJ+B!K!7>IJ'B[[FDK AB. $  , oororiginal_samplesnoisercURRURURS9nURRS:Xav[ R "U5(a[URRUR[ RS9nURUR[ RS9nO@URRUR5nURUR5nURc!UVs/sHo`RXe5PM nnOHURbUR/URS-nOUR/URS-nXGR5n[UR5[UR5:a?URS5n[UR5[UR5:aM?XU--n U $s snf)a  Add noise to the original samples according to the noise schedule at the specified timesteps. Args: original_samples (`torch.Tensor`): The original samples to which noise will be added. noise (`torch.Tensor`): The noise tensor to add to the original samples. timesteps (`torch.Tensor`): The timesteps at which to add noise, determining the noise level from the schedule. Returns: `torch.Tensor`: The noisy samples with added noise scaled according to the timestep schedule. )rr8mpsr7rr)rerfrr8typeris_floating_pointrr>rurrqrrr unsqueeze) rgrrrrerr+ step_indicesr noisy_sampless r add_noise KDPM2DiscreteScheduler.add_noises,'7'>'>FVF\F\]  " " ' '5 0U5L5LY5W5W!%!2!23C3J3JRWR_R_!2!` ! %5%<%>$=qFF$=>L E$(#%9="'', %( *), ),), ),  ), zzDK/$6 ), ),!%),),),),),),V33  !!(3( %,,&  B&*#' R, R,ell"R,! R,h## Y]# ,#BG,,QUBU# #> 1)= 1$ 1 "J$ELL$RWR^R^$NTW\a\h\hFdg..rs ! A2G 5:5 5(@H14 1414""<=14 \\ 14hK /^[K /r