wi_^ddlmZmZmZmZddlZddlmZddlm Z m Z Gdde Z dS) )ListOptionalTupleUnionN) randn_tensor)DiffusionPipelineImagePipelineOutputceZdZdZdZfdZej dd ed e e ej e ej fd ed e e d ede eeff dZxZS) DDPMPipelinea1 Pipeline for image generation. This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods implemented for all pipelines (downloading, saving, running on a particular device, etc.). Parameters: unet ([`UNet2DModel`]): A `UNet2DModel` to denoise the encoded image latents. scheduler ([`SchedulerMixin`]): A scheduler to be used in combination with `unet` to denoise the encoded image. Can be one of [`DDPMScheduler`], or [`DDIMScheduler`]. unetcvt|||dS)N)r scheduler)super__init__register_modules)selfrr __class__s v/root/.openclaw/workspace/chatterbox_venv_py311/lib/python3.11/site-packages/diffusers/pipelines/ddpm/pipeline_ddpm.pyrzDDPMPipeline.__init__)s8  49=====NpilT batch_size generatornum_inference_steps output_type return_dictreturnct|jjjtr4||jjj|jjj|jjjf}n%||jjjg|jjjR}|jjdkr,t||}| |j}nt|||j}|j || |j j D]@}|||j} |j | |||j}A|dz dzdd}|dddd}|d kr||}|s|fSt-| S) aa The call function to the pipeline for generation. Args: batch_size (`int`, *optional*, defaults to 1): The number of images to generate. generator (`torch.Generator`, *optional*): A [`torch.Generator`](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make generation deterministic. num_inference_steps (`int`, *optional*, defaults to 1000): The number of denoising steps. More denoising steps usually lead to a higher quality image at the expense of slower inference. output_type (`str`, *optional*, defaults to `"pil"`): The output format of the generated image. Choose between `PIL.Image` or `np.array`. return_dict (`bool`, *optional*, defaults to `True`): Whether or not to return a [`~pipelines.ImagePipelineOutput`] instead of a plain tuple. Example: ```py >>> from diffusers import DDPMPipeline >>> # load model and scheduler >>> pipe = DDPMPipeline.from_pretrained("google/ddpm-cat-256") >>> # run pipeline in inference (sample random noise and denoise) >>> image = pipe().images[0] >>> # save image >>> image.save("ddpm_generated_image.png") ``` Returns: [`~pipelines.ImagePipelineOutput`] or `tuple`: If `return_dict` is `True`, [`~pipelines.ImagePipelineOutput`] is returned, otherwise a `tuple` is returned where the first element is a list with the generated images mps)r)rdevicer g?rrrr)images) isinstancerconfig sample_sizeint in_channelsr#typertor set_timesteps progress_bar timestepssamplestep prev_sampleclampcpupermutenumpy numpy_to_pilr ) rrrrrr image_shapeimaget model_outputs r__call__zDDPMPipeline.__call__-s^ di&2C 8 8 d  ,  ,  , KK&ty'7'CcdiFVFbccK ; u $ $  BBBEHHT[))EE  $+VVVE $$%8999""4>#;<< a aA99UA..5LN'' a)'TT`EES''1-- ##Aq!Q//5577 %  %%e,,E 8O"%0000r)rNrrT)__name__ __module__ __qualname____doc__model_cpu_offload_seqrtorchno_gradr(rr Generatorrstrboolr rr; __classcell__)rs@rr r s  #>>>>>U]__MQ#'%* Q1Q1Q1E%/43H"HIJQ1! Q1 c] Q1  Q1 "E) *Q1Q1Q1_Q1Q1Q1Q1Q1rr ) typingrrrrrAutils.torch_utilsrpipeline_utilsr r r rrrKs 0/////////// ------CCCCCCCCg1g1g1g1g1$g1g1g1g1g1r