+h9*SSKJr SSKJrJrJr SSKrSSKJr SSK J r J r SSK JrJrJrJrJr \R&R"SS 55r\"S S \55r"S S \\ 5rg)) dataclass)OptionalTupleUnionN) ConfigMixinregister_to_config)CommonSchedulerStateFlaxKarrasDiffusionSchedulersFlaxSchedulerMixinFlaxSchedulerOutputbroadcast_to_shape_from_leftc \rSrSr%\\S'\R\S'\R\S'\R\S'Sr\ \ \S'\ S\S\RS\RS\R4Sj5r S r g) EulerDiscreteSchedulerStatecommoninit_noise_sigma timestepssigmasNnum_inference_stepscU"XX4S9$)Nrrrr)clsrrrrs m/home/james-whalen/.local/lib/python3.13/site-packages/diffusers/schedulers/scheduling_euler_discrete_flax.pycreate"EulerDiscreteSchedulerState.create)s&yhhr)__name__ __module__ __qualname____firstlineno__r __annotations__jnpndarrayrrint classmethodr__static_attributes__rrrrrs  kk!{{ KK)-#-i)i=@[[iUXU`U`ijmjujuiirrc \rSrSr%\\S'Srg) FlaxEulerDiscreteSchedulerOutput0staterN)r r!r"r#rr$r)rrrr+r+0s &&rr+cf\rSrSr%Sr\VVs/sHoR PM snnr\R\ S'\ S5r \ SSSSS S S \R4S \S \S\S\S\\R&S\S\S\R4Sjj5rS'S\\S\4SjjrS\S\R&S\S\R&4SjrS(S\S\S\S\4SjjrS)S\S\R&S\S\R&S\S\\\44 S jjrS\S!\R&S"\R&S#\R&S\R&4 S$jrS%r S&r!g s snnf)*FlaxEulerDiscreteScheduler5aj Euler scheduler (Algorithm 2) from Karras et al. (2022) https://huggingface.co/papers/2206.00364. . Based on the original k-diffusion implementation by Katherine Crowson: https://github.com/crowsonkb/k-diffusion/blob/481677d114f6ea445aa009cf5bd7a9cdee909e47/k_diffusion/sampling.py#L51 [`~ConfigMixin`] takes care of storing all config attributes that are passed in the scheduler's `__init__` function, such as `num_train_timesteps`. They can be accessed via `scheduler.config.num_train_timesteps`. [`SchedulerMixin`] provides general loading and saving functionality via the [`SchedulerMixin.save_pretrained`] and [`~SchedulerMixin.from_pretrained`] functions. Args: num_train_timesteps (`int`): number of diffusion steps used to train the model. beta_start (`float`): the starting `beta` value of inference. beta_end (`float`): the final `beta` value. beta_schedule (`str`): the beta schedule, a mapping from a beta range to a sequence of betas for stepping the model. Choose from `linear` or `scaled_linear`. trained_betas (`jnp.ndarray`, optional): option to pass an array of betas directly to the constructor to bypass `beta_start`, `beta_end` etc. prediction_type (`str`, default `epsilon`, optional): prediction type of the scheduler function, one of `epsilon` (predicting the noise of the diffusion process), `sample` (directly predicting the noisy sample`) or `v_prediction` (see section 2.4 https://imagen.research.google/video/paper.pdf) dtype (`jnp.dtype`, *optional*, defaults to `jnp.float32`): the `dtype` used for params and computation. dtypecg)NTrselfs r has_state$FlaxEulerDiscreteScheduler.has_stateVsrig-C6?g{Gz?linearNepsilonlinspacenum_train_timesteps beta_startbeta_end beta_schedule trained_betasprediction_typetimestep_spacingc XlgNr1) r4r:r;r<r=r>r?r@r1s r__init__#FlaxEulerDiscreteScheduler.__init__Zs  rrreturnc |Uc[R"U5n[R"SURR 5R 5SSS2nSUR- UR- S-n[R"U[R"S[U55U5n[R"U[R"S/URS9/5nURRS;aUR5nOUR5S-S-S-n[RUUUUS 9$) Nrr ?rCr9trailingrr)r rr%arangeconfigr:roundalphas_cumprodinterplen concatenatearrayr1r@maxr)r4rrrrs r create_state'FlaxEulerDiscreteScheduler.create_statehs >)006FJJq$++"A"ABHHJ4R4P v,,,0E0EE#MIszz!S['A6J&#))SE*L!MN ;; ' '+C C%zz|  &  1A 5#= *11- 2  rr-sampletimestepc[R"URU:HSS9unUSnURUnX%S-S-S-- nU$)a Scales the denoising model input by `(sigma**2 + 1) ** 0.5` to match the Euler algorithm. Args: state (`EulerDiscreteSchedulerState`): the `FlaxEulerDiscreteScheduler` state data class instance. sample (`jnp.ndarray`): current instance of sample being created by diffusion process. timestep (`int`): current discrete timestep in the diffusion chain. Returns: `jnp.ndarray`: scaled input sample r sizerrrI)r%whererr)r4r-rXrY step_indexsigmas rscale_model_input,FlaxEulerDiscreteScheduler.scale_model_input~sP %//X"=AF ]  Z(1HqLS01 rrshapec URRS:Xa8[R"URRS- SX R S9nOURRS:XanURRU-n[R "SU5U-R5SSS2R5R[5nUS- nO"[SURR35eSURR- URR- S -n[R"U[R "S[U55U5n[R "U[R""S /UR S9/5nURRS ;aUR%5nOUR%5S -S-S -nUR'UUUUS 9$)aV Sets the timesteps used for the diffusion chain. Supporting function to be run before inference. Args: state (`EulerDiscreteSchedulerState`): the `FlaxEulerDiscreteScheduler` state data class instance. num_inference_steps (`int`): the number of diffusion steps used when generating samples with a pre-trained model. r9r rrCleadingNrHz=timestep_spacing must be one of ['linspace', 'leading'], got rIrJrKr)rrrr)rNr@r%r9r:r1rMrOcopyastypefloat ValueErrorrrPrQrRrSrTrUreplace)r4r-rrbr step_ratiorrs r set_timesteps(FlaxEulerDiscreteScheduler.set_timestepss ;; ' ': 5 T[[%D%Dq%H!M`hrhrsI [[ ) )Y 688rsv")) A i i i  '':' 'T/!3[T/r