bigSddlZddlmZddlmZmZmZmZddlZ ddl Z ddl m Z m Z ddlmZddlmZmZeGd d eZ dd Zd ZGd dee Zy)N) dataclass)ListOptionalTupleUnion) ConfigMixinregister_to_config) BaseOutput)KarrasDiffusionSchedulersSchedulerMixincXeZdZUdZej ed<dZeej ed<y)DDIMSchedulerOutputaq 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____doc__torchTensor__annotations__rri/home/cdr/jupyterlab/.venv/lib/python3.12/site-packages/diffusers/schedulers/scheduling_ddim_cogvideox.pyrrs' 37(5<<07rrc $|dk(rd}n|dk(rd}ntd|g}t|D]<}||z }|dz|z }|jtd||||z z |>t j |tj S)a Create a beta schedule that discretizes the given alpha_t_bar function, which defines the cumulative product of (1-beta) over time from t = [0,1]. Contains a function alpha_bar that takes an argument t and transforms it to the cumulative product of (1-beta) up to that part of the diffusion process. Args: num_diffusion_timesteps (`int`): the number of betas to produce. max_beta (`float`): the maximum beta to use; use values lower than 1 to prevent singularities. alpha_transform_type (`str`, *optional*, default to `cosine`): the type of noise schedule for alpha_bar. Choose from `cosine` or `exp` Returns: betas (`np.ndarray`): the betas used by the scheduler to step the model outputs cosinecftj|dzdz tjzdz dzS)NgMb?gT㥛 ?r)mathcospits r alpha_bar_fnz)betas_for_alpha_bar..alpha_bar_fnLs-88QY%/$''9A=>!C Crexpc2tj|dzS)Ng()r r&r#s rr%z)betas_for_alpha_bar..alpha_bar_fnQs88AI& &rz"Unsupported alpha_transform_type: r dtype) ValueErrorrangeappendminrtensorfloat32)num_diffusion_timestepsmax_betaalpha_transform_typer%betasit1t2s rbetas_for_alpha_barr73s.x' D  & '=>R=STUU E * +M ( (!e. . S\"- R0@@@(KLM <<U]] 33rc|j}|dj}|dj}||z}||||z z z}|dz}|S)a. Rescales betas to have zero terminal SNR Based on https://huggingface.co/papers/2305.08891 (Algorithm 1) Args: betas (`torch.Tensor`): the betas that the scheduler is being initialized with. Returns: `torch.Tensor`: rescaled betas with zero terminal SNR rr)sqrtclone)alphas_cumprodalphas_bar_sqrtalphas_bar_sqrt_0alphas_bar_sqrt_T alphas_bars rrescale_zero_terminal_snrrA_st%))+O(*002'+113((O(,=@Q,QRRO!!#J rceZdZdZeDcgc]}|j c}}ZdZe d'de de de de de e ejee fd ed ed e d e d e de de dede fdZdZd(dej*de e dej*fdZd(de de e ej.ffdZ d)dej*de dej*de dede ej*dede eeffd Zd!ej*d"ej*d#ej8dej*fd$Zdej*d"ej*d#ej8dej*fd%Zd&Zycc}}w)*CogVideoXDDIMSchedulera@ `DDIMScheduler` extends the denoising procedure introduced in denoising diffusion probabilistic models (DDPMs) with non-Markovian guidance. 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`, `scaled_linear`, or `squaredcos_cap_v2`. trained_betas (`np.ndarray`, *optional*): Pass an array of betas directly to the constructor to bypass `beta_start` and `beta_end`. clip_sample (`bool`, defaults to `True`): Clip the predicted sample for numerical stability. clip_sample_range (`float`, defaults to 1.0): The maximum magnitude for sample clipping. Valid only when `clip_sample=True`. set_alpha_to_one (`bool`, defaults to `True`): Each diffusion step uses the alphas product value at that step and at the previous one. For the final step there is no previous alpha. When this option is `True` the previous alpha product is fixed to `1`, otherwise it uses the alpha value at step 0. steps_offset (`int`, defaults to 0): An offset added to the inference steps, as required by some model families. 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). thresholding (`bool`, defaults to `False`): Whether to use the "dynamic thresholding" method. This is unsuitable for latent-space diffusion models such as Stable Diffusion. dynamic_thresholding_ratio (`float`, defaults to 0.995): The ratio for the dynamic thresholding method. Valid only when `thresholding=True`. sample_max_value (`float`, defaults to 1.0): The threshold value for dynamic thresholding. Valid only when `thresholding=True`. timestep_spacing (`str`, defaults to `"leading"`): 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. rescale_betas_zero_snr (`bool`, defaults to `False`): Whether to rescale the betas to have zero terminal SNR. This enables the model to generate very bright and dark samples instead of limiting it to samples with medium brightness. Loosely related to [`--offset_noise`](https://github.com/huggingface/diffusers/blob/74fd735eb073eb1d774b1ab4154a0876eb82f055/examples/dreambooth/train_dreambooth.py#L506). r Nnum_train_timesteps beta_startbeta_end beta_schedule trained_betas clip_sampleset_alpha_to_one steps_offsetprediction_typeclip_sample_rangesample_max_valuetimestep_spacingrescale_betas_zero_snrsnr_shift_scalec|+tj|tj|_n|dk(r-tj|||tj|_nk|dk(r6tj|dz|dz|tj dz|_n0|dk(rt ||_nt|d|jd|jz |_ tj|jd |_ |j|d |z |jzzz |_ | rt|j|_ |rtjdn|jd |_ d|_d|_tj t#j$d |ddd j'j)t"j*|_y) Nr(linear scaled_linear?rsquaredcos_cap_v2z is not implemented for ?r)dimr r9)rr.r/r3linspacefloat64r7NotImplementedError __class__alphascumprodr<rAfinal_alpha_cumprodinit_noise_sigmanum_inference_steps from_numpynparangecopyastypeint64 timesteps)selfrDrErFrGrHrIrJrKrLrMrNrOrPrQs r__init__zCogVideoXDDIMScheduler.__init__s$  $m5==IDJ h & H>QY^YfYfgDJ o - C3H[chcpcpquvvDJ 1 1,-@ADJ%7OPTP^P^O_&`a aDJJ& #mmDKKQ?#11_OH[_c_r_rGr5rs "";DZ^ZrZr,& 00${2q?KK3345 KK;;>BDWX2!  [[ ) )Y 688DZ^ZrZr,& ;; & &) 3$*[S-AL-P$PT`ehTi#i [[ ( (H 4#/ [[ ( (N :$0#$5#?;PSCSWcBc#c +DKK,G,G+HI""  %%!l*:;C$|S'83'>>FlS+?%?? $  #{Qeffroriginal_samplesnoiserhc|jj|j|_|jj|j}|j|j}||dz}|j }t |j t |j kr=|jd}t |j t |j kr=d||z dz}|j }t |j t |j kr=|jd}t |j t |j kr=||z||zz}|SN)rxr(rUr9r r<r~rxr)flattenlenshape unsqueeze)rirrrhr<sqrt_alpha_prodsqrt_one_minus_alpha_prod noisy_sampless r add_noisez CogVideoXDDIMScheduler.add_noisesa#1144>qAFF>L E$(# ,BF !%(#&"% )',!$3k 3k3k 3k  3k  bjj$u+&= >? 3k3k3k3k3k!3k 3k3k!%3k3k3kj  Y^YeYe"+@+@eCDU>V+@d).15 agllagag ag  ag #' ag!.agag "E) *agH,,||??   65<< QVQ`Q`ejeqeq&/g?sErC)g+?r)r dataclassesrtypingrrrrnumpyrcrconfiguration_utilsr r utilsr scheduling_utilsr rrr7rArCrrrrs`& !// AG 8*8 8(!)4X>F/^[F/r