3j) SSKrSSKJr SSKJr SSKrSSKJrJr SSK J r SSK J r SS K Jr SS \S \S \S S\R"4Sjjr\"SS\ 55r"SS\\5rg)N) dataclass)Literal) ConfigMixinregister_to_config) BaseOutput) randn_tensor)SchedulerMixinnum_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$) a Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t = [0,1]. Contains a function alpha_bar that takes an argument t and transforms it to the cumulative product of (1-beta) up to that part of the diffusion process. Args: num_diffusion_timesteps (`int`): The number of betas to produce. max_beta (`float`, defaults to `0.999`): The maximum beta to use; use values lower than 1 to avoid numerical instability. alpha_transform_type (`str`, defaults to `"cosine"`): The type of noise schedule for `alpha_bar`. Choose from `cosine`, `exp`, or `laplace`. Returns: `torch.Tensor`: The betas used by the scheduler to step the model outputs. rch[R"US-S- [R-S- 5S-$)NgMb?gT㥛 ?r)mathcospits m/home/wildlama/miniconda3/lib/python3.13/site-packages/diffusers/schedulers/scheduling_consistency_decoder.py alpha_bar_fn)betas_for_alpha_bar..alpha_bar_fn(s-88QY%/$''9A=>!C Crc S[R"SSU- 5-[R"SS[R"SU- 5-- S-5-n[R"U5n[R "USU-- 5$)Ngr ?rgư>)rcopysignlogfabsrsqrt)rlmbsnrs rrr-smq#'22TXXa!diiPSVWPWFXBX>X[_>_5``C((3-C99SAG_- -rrc4[R"US-5$)Ng()rrrs rrr4s88AI& &rz"Unsupported alpha_transform_type: r )dtype) ValueErrorrangeappendmintorchtensorfloat32)r r rrbetasit1t2s rbetas_for_alpha_barr3s0x' D  * .  & '=>R=STUU E * +  (!e. . S\"- R0@@@(KL, <<U]] 33rc8\rSrSr%Sr\R \S'Srg)!ConsistencyDecoderSchedulerOutputBa& Output class for the scheduler's `step` function. Args: prev_sample (`torch.Tensor` of shape `(batch_size, num_channels, height, width)` for images): Computed sample `(x_{t-1})` of previous timestep. `prev_sample` should be used as next model input in the denoising loop. prev_sampleN) __name__ __module__ __qualname____firstlineno____doc__r,Tensor__annotations____static_attributes__r8rrr5r5Bsrr5c~\rSrSrSrSr\SS\S\SS4Sjj5r SS \S-S \ \ R-4S jjr \S\ R4S j5rSS \ RS\S-S\ R4SjjrSS\ RS\\ R-S \ RS\ R$S-S\S\\-4 SjjrSrg)ConsistencyDecoderSchedulerPa A scheduler for the consistency decoder used in Stable Diffusion pipelines. This scheduler implements a two-step denoising process using consistency models for decoding latent representations into images. 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`, *optional*, defaults to `1024`): The number of diffusion steps to train the model. sigma_data (`float`, *optional*, defaults to `0.5`): The standard deviation of the data distribution. Used for computing the skip and output scaling factors. r num_train_timesteps sigma_datarNc[U5nSU- n[R"USS9n[R"U5Ul[R"SU- 5Ul[R"SU- S- 5n[R"SU- 5nXrS--US-US--- UlXb-US-US--S-- UlXvS-US--S-- Ulg)Ng?r)dimr rr) r3r,cumprodr#sqrt_alphas_cumprodsqrt_one_minus_alphas_cumprodc_skipc_outc_in)selfrDrEr/alphasalphas_cumprodsigmassqrt_recip_alphas_cumprods r__init__$ConsistencyDecoderScheduler.__init__cs $$78uv15#(::n#= -2ZZn8L-M*C.0145$)JJs^/C$D!/a-?619z[\}C\] (FAI A ,E#+MM -Z]1Js0RR rnum_inference_stepsdevicecUS:wa [S5e[R"SS/[RUS9UlUR R U5UlURR U5UlURR U5UlURR U5Ul URR U5Ul g)Nrz8Currently more than 2 inference steps are not supported.ii)r'rV) r(r,r-long timestepsrItorJrKrLrM)rNrUrVs r set_timesteps)ConsistencyDecoderScheduler.set_timestepsys ! #WX XtSkFS#'#;#;#>#>v#F -1-O-O-R-RSY-Z*kknnV, ZZ]]6* IILL( rc:URURS$)z Return the standard deviation of the initial noise distribution. Returns: `torch.Tensor`: The initial noise sigma value from the precomputed `sqrt_one_minus_alphas_cumprod` at the first timestep. r)rJrY)rNs rinit_noise_sigma,ConsistencyDecoderScheduler.init_noise_sigmas11$..2CDDrsampletimestepc$XRU-$)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. )rM)rNr`ras rscale_model_input-ConsistencyDecoderScheduler.scale_model_inputs (+++r model_output generator return_dictc<URUU-URUU--n[R"URU:H5SnU[ UR5S- :XaUnO[ URXFRURS9n URURUS-RUR5U-URURUS-RUR5U --nU(dU4$[US9$)aQ 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 the learned diffusion model. timestep (`float` or `torch.Tensor`): The current timestep in the diffusion chain. sample (`torch.Tensor`): A current instance of a sample created by the diffusion process. generator (`torch.Generator`, *optional*): A random number generator for reproducibility. return_dict (`bool`, *optional*, defaults to `True`): Whether or not to return a [`~schedulers.scheduling_consistency_decoder.ConsistencyDecoderSchedulerOutput`] or `tuple`. Returns: [`~schedulers.scheduling_consistency_decoder.ConsistencyDecoderSchedulerOutput`] or `tuple`: If `return_dict` is `True`, [`~schedulers.scheduling_consistency_decoder.ConsistencyDecoderSchedulerOutput`] is returned, otherwise a tuple is returned where the first element is the sample tensor. rr )rfr'rV)r7)rLrKr,whererYlenr shaper'rVrIrZrJr5) rNrerar`rfrgx_0 timestep_idxr7noises rstep ConsistencyDecoderScheduler.steps>jj"\1DKK4IF4RR{{4>>X#=>qA 3t~~.2 2K iyyY\YcYcdE(( q8H)IJMMciiX[^^44T^^LSTDT5UVYYZ]ZcZcdgllm  > !0[IIr)rMrLrKrIrJrY)ir)NN)N)NT)r9r:r;r<r=orderrintfloatrSstrr,rVr[propertyr>r^rc Generatorboolr5tupleror@r8rrrBrBPs3 E$(S SS  SS.+/%) ) 4Z )ell" ) E%,, E E, ,d ,V[VbVb,,-1 /Jll/J%,,&/J /J ??T) /J  /J +U 2/J/JrrB)g+?r)r dataclassesrtypingrr,configuration_utilsrrutilsrutils.torch_utilsr scheduling_utilsr rrrsr>r3r5rBr8rrrs ! A,, @H14 1414""<=14 \\ 14h     DJ.+DJr