3ji0 SSKJr SSKrSSKrSSKJr SSKJr SSKJ r J r SSK J r SSK JrJrJr \ R""\5r\R(R"S S 55r\"S S \55r"S S\\ 5rg)) dataclassN)random) ConfigMixinregister_to_config)logging)FlaxSchedulerMixinFlaxSchedulerOutputbroadcast_to_shape_from_leftc\rSrSr%Sr\R S-\S'Sr\R S-\S'Sr \R S-\S'\ S5r Sr g)ScoreSdeVeSchedulerState$N timestepsdiscrete_sigmassigmascU"5$N)clss e/home/wildlama/miniconda3/lib/python3.13/site-packages/diffusers/schedulers/scheduling_sde_ve_flax.pycreateScoreSdeVeSchedulerState.create+s u r) __name__ __module__ __qualname____firstlineno__rjnpndarray__annotations__rr classmethodr__static_attributes__rrrrr$sO%)Is{{T!(*.OS[[4'.!%FCKK$ %rrcj\rSrSr%Sr\\S'\R\S'Sr \RS-\S'Sr g)FlaxSdeVeOutput0a( Output class for the ScoreSdeVeScheduler's step function output. Args: state (`ScoreSdeVeSchedulerState`): prev_sample (`jnp.ndarray` 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_sample_mean (`jnp.ndarray` of shape `(batch_size, num_channels, height, width)` for images): Mean averaged `prev_sample`. Same as `prev_sample`, only mean-averaged over previous timesteps. state prev_sampleNprev_sample_meanr) rrrr__doc__rr!rr r)r#rrrr%r%0s-  $#+/ckkD(/rr%c\rSrSrSr\S5r\SS\S\ S\ S\ S\ S \4 S jj5r S r SS\ S\S\ S\ S\ 4 SjjrSS\ S\S\ S\ S\ S\ 4 SjjrSrS S\ S\R$S\S\R$S\R(S\S\\ -4SjjrS S\ S\R$S\R$S\R(S\S\\ -4 SjjrSrS rg )!FlaxScoreSdeVeSchedulerCa The variance exploding stochastic differential equation (SDE) scheduler. For more information, see the original paper: https://huggingface.co/papers/2011.13456 [`~ConfigMixin`] takes care of storing all config attributes that are passed in the scheduler's `__init__` function, such as `num_train_timesteps`. They can be accessed via `scheduler.config.num_train_timesteps`. [`SchedulerMixin`] provides general loading and saving functionality via the [`SchedulerMixin.save_pretrained`] and [`~SchedulerMixin.from_pretrained`] functions. Args: num_train_timesteps (`int`): number of diffusion steps used to train the model. snr (`float`): coefficient weighting the step from the model_output sample (from the network) to the random noise. sigma_min (`float`): initial noise scale for sigma sequence in sampling procedure. The minimum sigma should mirror the distribution of the data. sigma_max (`float`): maximum value used for the range of continuous timesteps passed into the model. sampling_eps (`float`): the end value of sampling, where timesteps decrease progressively from 1 to epsilon. correct_steps (`int`): number of correction steps performed on a produced sample. cg)NTrselfs r has_state!FlaxScoreSdeVeScheduler.has_state[srnum_train_timestepssnr sigma_min sigma_max sampling_eps correct_stepsc.[RS5 g)NzFlax classes are deprecated and will be removed in Diffusers v1.0.0. We recommend migrating to PyTorch classes or pinning your version of Diffusers.)loggerwarning)r0r3r4r5r6r7r8s r__init__ FlaxScoreSdeVeScheduler.__init___s  [ rc[R5nURUURRURR URR URR5$r)rr set_sigmasconfigr3r5r6r7)r0r's r create_state$FlaxScoreSdeVeScheduler.create_statensX(//1  KK + + KK ! ! KK ! ! KK $ $   rrNr'num_inference_stepsshapereturncUbUOURRn[R"SXB5nUR US9$)a Sets the continuous timesteps used for the diffusion chain. Supporting function to be run before inference. Args: state (`ScoreSdeVeSchedulerState`): the `FlaxScoreSdeVeScheduler` state data class instance. num_inference_steps (`int`): the number of diffusion steps used when generating samples with a pre-trained model. sampling_eps (`float`, optional): final timestep value (overrides value given at Scheduler instantiation). r )r)r@r7rlinspacereplace)r0r'rCrDr7rs r set_timesteps%FlaxScoreSdeVeScheduler.set_timestepsxs;$(4'?|T[[E]E] LLLF }}y}11rc"UbUOURRnUbUOURRnUbUOURRnURcUR XU5n[ R"[ R"[ R"U5[ R"U5U55n[ R"URVs/sH osXC- U--PM sn5nURXhS9$s snf)a  Sets the noise scales used for the diffusion chain. Supporting function to be run before inference. The sigmas control the weight of the `drift` and `diffusion` components of sample update. Args: state (`ScoreSdeVeSchedulerState`): the `FlaxScoreSdeVeScheduler` state data class instance. num_inference_steps (`int`): the number of diffusion steps used when generating samples with a pre-trained model. sigma_min (`float`, optional): initial noise scale value (overrides value given at Scheduler instantiation). sigma_max (`float`, optional): final noise scale value (overrides value given at Scheduler instantiation). sampling_eps (`float`, optional): final timestep value (overrides value given at Scheduler instantiation). )rr) r@r5r6r7rrIrexprGlogarrayrH) r0r'rCr5r6r7rtrs rr?"FlaxScoreSdeVeScheduler.set_sigmass0"+!6IDKK1(DDo^_}}_}LL_s#D c[R"US:H[R"U5URUS- 5$)Nrr )rwhere zeros_liker)r0r'rrOs rget_adjacent_sigma*FlaxScoreSdeVeScheduler.get_adjacent_sigmas4yya):E&U):: 59 9;afggrc<URc [S5e[R"USS9n[R"XCR S9n[ RRU5n[ RRU5nURRU-U- S-S-n U [ R"UR S5-n U R5n [XR 5n X9U--n XS-S-U--n U(dX4$[XS9$) a Correct the predicted sample based on the output model_output of the network. This is often run repeatedly after making the prediction for the previous timestep. Args: state (`ScoreSdeVeSchedulerState`): the `FlaxScoreSdeVeScheduler` state data class instance. model_output (`jnp.ndarray`): direct output from learned diffusion model. sample (`jnp.ndarray`): current instance of sample being created by diffusion process. generator: random number generator. return_dict (`bool`): option for returning tuple rather than FlaxSdeVeOutput class Returns: [`FlaxSdeVeOutput`] or `tuple`: [`FlaxSdeVeOutput`] if `return_dict` is True, otherwise a `tuple`. When returning a tuple, the first element is the sample tensor. r\r r^r`rrr])r(r')rrarrfrgrDrlinalgnormr@r4rbrer r%) r0r'rVrXrYrZrl grad_norm noise_norm step_sizer)r(s r step_correct$FlaxScoreSdeVeScheduler.step_corrects2 ?? "l  ll3A& #\\:JJOOL1 ZZ__U+ [[__z1I=!CaG a 99 %%' 0LLI ! $<<&q=S*@E)II ' ';DDrc.URR$r)r@r3r/s r__len__FlaxScoreSdeVeScheduler.__len__%s{{...r)ig333333?g{Gz?g@gh㈵>r )rN)NNN)T)rrrrr*propertyr1rintfloatr<rArtuplerIr?rTrr jaxArrayboolr%rmrurxr#rrrr,r,Cs.$(!"                  " 2'2!2 2  2 " 26 " !M'!M!!M !M  !M  !M "!MFb!:h':hkk:h :h  :h YY :h:h 5 :hD! 2E'2Ekk2E 2E YY 2E  2E 5 2Eh/rr,) dataclassesrflaxr~ jax.numpynumpyrrconfiguration_utilsrrutilsrscheduling_utils_flaxr r r get_loggerrr:structrr%r,rrrrs"" A   H % 0)0 0$c/0+c/r