+h4_SSKrSSKrSSKJr SSKJrJrJrJr SSK r SSK r SSK r SSK Jr SSKJrJr SSKJr SS KJrJr \"S S \55rSS jr"S S\\5rg)N) dataclass)ListOptionalTupleUnion) integrate) ConfigMixinregister_to_config) BaseOutput)KarrasDiffusionSchedulersSchedulerMixinc`\rSrSr%Sr\R \S'Sr\ \R \S'Sr g)LMSDiscreteSchedulerOutputaM 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__rr__static_attributes__rf/home/james-whalen/.local/lib/python3.13/site-packages/diffusers/schedulers/scheduling_lms_discrete.pyrrs' 37(5<<07rrc &US:XaSnOUS:XaSnO[SU35e/n[U5H<nXP- nUS-U- nUR[SU"U5U"U5- - U55 M> [R "U[R S9$)a Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t = [0,1]. Contains a function alpha_bar that takes an argument t and transforms it to the cumulative product of (1-beta) up to that part of the diffusion process. Args: num_diffusion_timesteps (`int`): the number of betas to produce. max_beta (`float`): the maximum beta to use; use values lower than 1 to prevent singularities. alpha_transform_type (`str`, *optional*, default to `cosine`): the type of noise schedule for alpha_bar. Choose from `cosine` or `exp` Returns: betas (`np.ndarray`): the betas used by the scheduler to step the model outputs cosinech[R"US-S- [R-S- 5S-$)NgMb?gT㥛 ?r )mathcospits r alpha_bar_fn)betas_for_alpha_bar..alpha_bar_fnJs-88QY%/$''9A=>!C Crexpc4[R"US-5$)Ng()r$r+r's r r)r*Os88AI& &rz"Unsupported alpha_transform_type: r dtype) ValueErrorrangeappendminrtensorfloat32)num_diffusion_timestepsmax_betaalpha_transform_typer)betasit1t2s r betas_for_alpha_barr<1s.x' D  & '=>R=STUU E * +  (!e. . S\"- R0@@@(KL, <<U]] 33rc\rSrSrSr\VVs/sHoR PM snnrSr\ S1S\ S\ S\ S\ S \ \\R \\ 4S \ \S \ \S \ \S \ S\ S\ 4Sjj5r\S5r\S5r\S5rS2S\ 4SjjrS\R4S\\ \R44S\R44SjrSrS3S\ S\\ \R:44SjjrS3SjrSr S r!S!\R4S\R44S"jr"S!\R4S\ S\R44S#jr#S4S!\R4S\ S$\ S%\ S\R44 S&jjr$S5S'\R4S\\ \R44S\R4S(\ S)\S\\%\&44 S*jjr'S+\R4S,\R4S-\R4S\R44S.jr(S/r)S0r*gs snnf)6LMSDiscreteScheduler]u" A linear multistep scheduler for discrete beta schedules. 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 0.0001): The starting `beta` value of inference. beta_end (`float`, defaults to 0.02): 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://imagen.research.google/video/paper.pdf) 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. r Nnum_train_timesteps beta_startbeta_end beta_schedule trained_betasuse_karras_sigmasuse_exponential_sigmasuse_beta_sigmasprediction_typetimestep_spacing steps_offsetc ~[URRURRURR/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"URS S 9Ul["R$"SUR - UR - S-5n ["R&"U SSS 2S//5R)["R5n [ R*"U 5UlSUlX`lUR1US5 /UlSUlSUlSUlUR,R;S5Ulg)Nr znOnly one of `config.use_beta_sigmas`, `config.use_exponential_sigmas`, `config.use_karras_sigmas` can be used.r-linear scaled_linear?r squaredcos_cap_v2z is not implemented for ?r)dimFcpu)sumconfigrGrFrEr/rr3r4r8linspacer<NotImplementedError __class__alphascumprodalphas_cumprodnparray concatenateastype from_numpysigmasnum_inference_steps set_timesteps derivativesis_scale_input_called _step_index _begin_indexto) selfr@rArBrCrDrErFrGrHrIrJrbs r __init__LMSDiscreteScheduler.__init__s  ++T[[-O-OQUQ\Q\QnQno pst tA   $m5==IDJ h & >QY^YfYfgDJ o - C3H[chcpcpquvvDJ 1 1,-@ADJ%7OPTP^P^O_&`a aDJJ& #mmDKKQ?A 3 33t7J7JJsRS"u 56==bjjI&&v. $( !2 .5%*" kknnU+ rcURRS;aURR5$URR5S-S-S-$)N)rWtrailingr r rN)rVrIrbmaxrjs r init_noise_sigma%LMSDiscreteScheduler.init_noise_sigmasH ;; ' '+C C;;??$ $ !Q&*s22rcUR$)zW The index counter for current timestep. It will increase 1 after each scheduler step. )rgrps r step_indexLMSDiscreteScheduler.step_indexs rcUR$)za The index for the first timestep. It should be set from pipeline with `set_begin_index` method. rhrps r begin_index LMSDiscreteScheduler.begin_indexs    rrxcXlg)z Sets the begin index for the scheduler. This function should be run from pipeline before the inference. Args: begin_index (`int`): The begin index for the scheduler. Nrw)rjrxs r set_begin_index$LMSDiscreteScheduler.set_begin_indexs (rsampletimestepreturncURcURU5 URURnXS-S-S-- nSUlU$)aV 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 (`float` or `torch.Tensor`): The current timestep in the diffusion chain. Returns: `torch.Tensor`: A scaled input sample. r r rNT)rt_init_step_indexrbrf)rjr}r~sigmas r scale_model_input&LMSDiscreteScheduler.scale_model_inputsP ?? "  ! !( + DOO,1HqLS01%)" rc^^^^UUUU4Sjn[R"UTRTTRTS-SS9SnU$)z` Compute the linear multistep coefficient. Args: order (): t (): current_order (): c>Sn[T5HHnTU:XaM XTRTU- - TRTT- TRTU- - - -nMJ U$)NrP)r0rb)tauprodk current_orderorderrjr(s r lms_derivative@LMSDiscreteScheduler.get_lms_coefficient..lms_derivativesmD5\ A%t{{1q511dkk!mBS6TW[WbWbcdghchWi6ijj"Krr -C6?)epsrelr)rquadrb)rjrr(rrintegrated_coeffs```` r get_lms_coefficient(LMSDiscreteScheduler.get_lms_coefficientsI  %>>.$++a.$++VWZ[V[J\eijklmrrcdevicec XlURRS:XaR[R"SURR S- U[R S9SSS2R5nGOzURRS:XaURR UR-n[R"SU5U-R5SSS2R5R[R 5nX0RR- nOURRS:XaURR UR- n[R"URR SU*5R5R5R[R 5nUS-nO"[URRS 35e[R"SUR- UR- S -5n[R"U5n[R "U[R"S[#U55U5nURR$(aEUR'US 9n[R"UVs/sHopR)Xv5PM sn5nOURR*(aEUR-XQS 9n[R"UVs/sHopR)Xv5PM sn5nO_URR.(aDUR1XQS 9n[R"UVs/sHopR)Xv5PM sn5n[R2"US //5R[R 5n[4R6"U5R9US9Ul[4R6"U5R9U[4R S9UlSUlSUl UR:R9S5Ul/Ul!gs snfs snfs 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. rWrr r-NrRleadingrnzY is not supported. Please make sure to choose one of 'linspace', 'leading' or 'trailing'.rN) in_sigmas)rrcrS)rrr.rT)"rcrVrIr]rWr@r4copyarangeroundr`rJr/r^r\loginterplenrE_convert_to_karras _sigma_to_trF_convert_to_exponentialrG_convert_to_betar_rrarirb timestepsrgrhre)rjrcrr step_ratiorb log_sigmasrs r rd"LMSDiscreteScheduler.set_timestepssd$7  ;; ' ': 5 At{{'F'F'JL_gigqgqr"df [[ ) )Y 688DFSY!ZSY%"2"25"ESY!Z[I [[ / /11F1lFSY!ZSY%"2"25"ESY!Z[I [[ ( (**V*eFSY!ZSY%"2"25"ESY!Z[I#077 C&&v.111@ )))477vU]]7[ kknnU+ !"["["[s*Q Q*Q cUc URnX!:HR5n[U5S:aSOSnX4R5$)Nr r)rnonzeroritem)rjr~schedule_timestepsindicesposs r index_for_timestep'LMSDiscreteScheduler.index_for_timestepBsH  %!% %1::< w,LMSDiscreteScheduler.step..s# 8c#45E 8cs)rr)rfwarningswarnrtrrbrVrHr/rer1rpopr2r0rrUzipreversedrgr) rjrr~r}rrrrr curr_order lms_coeffsrs r stepLMSDiscreteScheduler.steps<)) MME  ?? "  ! !( + DOO, ;; & &) 3#)L,@#@ [[ ( (N :#/6UAX\c'>FVF\F\]  " " ' '5 0U5L5LY5W5W!%!2!23C3J3JRWR_R_!2!` ! %5%<%r>]s#J%>>$=qFF$=>L E$("%BF,116*/( */, /,/, /,  /,  bjj$u+&= >? /,$D>/,!)/,"$/,/,/,/,/,b33  !!(3( eU\\FY@Z_d_k_k0 ,88eCDU>V8v #10 ELL U\\ TW\a\h\h.dg)g+?r")r$r dataclassesrtypingrrrrnumpyr] scipy.statsrrrconfiguration_utilsr r utilsr scheduling_utilsrrrr<r>rrr r sd !// AG 88 8(!)4XK/>;K/r