+h(SSKJrJrJrJrJr SSKrSSKJrJ r SSK J r SSK J r SSKJr SS KJrJr \ "5(a SSKJs Jr S rOS r"S S \5rg))DictListOptionalTupleUnionN) AutoencoderKLDiTTransformer2DModel)KarrasDiffusionSchedulers)is_torch_xla_available) randn_tensor)DiffusionPipelineImagePipelineOutputTFc2^\rSrSrSrSrSS\S\S\S\ \ \ \ 44U4Sjjjr S \\ \\ 4S \\ 4S jr\R$"5SS \\ S \S\ \\R(\\R(4S\ S\ \ S\S \\\44Sjj5rSrU=r$) DiTPipeline(a Pipeline for image generation based on a Transformer backbone instead of a UNet. 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: transformer ([`DiTTransformer2DModel`]): A class conditioned `DiTTransformer2DModel` to denoise the encoded image latents. Initially published as [`Transformer2DModel`](https://huggingface.co/facebook/DiT-XL-2-256/blob/main/transformer/config.json#L2) in the config, but the mismatch can be ignored. vae ([`AutoencoderKL`]): Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations. scheduler ([`DDIMScheduler`]): A scheduler to be used in combination with `transformer` to denoise the encoded image latents. ztransformer->vae transformervae schedulerid2labelc>[TU]5 URXUS9 0UlUbUR 5HQupVUR S5H7n[ U5URUR5R5'M9 MS [[URR 555Ulgg)N)rrr,) super__init__register_moduleslabelsitemssplitintlstriprstripdictsorted) selfrrrrkeyvaluelabel __class__s ^/home/james-whalen/.local/lib/python3.13/site-packages/diffusers/pipelines/dit/pipeline_dit.pyrDiTPipeline.__init__<s  +)T  &nn. "[[-E;>s8DKK  5 5 78./vdkk&7&7&9:;DK r(returnc[U[5(d [U5nUH-nX R;dM[USURS35e UVs/sHo RUPM sn$s snf)z Map label strings from ImageNet to corresponding class ids. Parameters: label (`str` or `dict` of `str`): Label strings to be mapped to class ids. Returns: `list` of `int`: Class ids to be processed by pipeline. zK does not exist. Please make sure to select one of the following labels: .) isinstancelistr ValueError)r%r(ls r* get_label_idsDiTPipeline.get_label_idsNsu%&&KEA # cefjfqfqerrst )..1 A...sA3 class_labelsguidance_scale generatornum_inference_steps output_type return_dictc [U5nURRRnURRRn [ XyX4UUR URRS9n US:a[R"U /S-5OU n [R"XR S9RS5n[R"S/U-UR S9n US:a[R"X/S5OUn URRU5 URURR5GHnUS:a'U S[U 5S-n[R"X/SS 9n URR!X5n Un[R""U5(dU R$R&S :HnU R$R&S :Hn[)U[*5(a/U(dU(a[R,O[R.nO.U(dU(a[R0O[R2n[R"U/UU R$S 9nO7[UR45S:XaUSR7U R$5nUR9U R4S5nURU UU S 9R:nUS:atUSS2SU 24USS2U S24nn[R<"U[U5S-SS 9unnUUUU- --n[R"UU/SS 9n[R"UU/SS 9nURRR>S-U :Xa[R<"UU SS 9unnOUnURRAUX5RBn [D(dGM[FRH"5 GM US:aU RKSSS 9un nOU n SURLRRN- U -n URLRQU 5R:nUS- S-RSSS5nURU5RWSSSS5R+5RY5nUS:XaUR[U5nUR]5 U(dU4$[_US9$)a The call function to the pipeline for generation. Args: class_labels (List[int]): List of ImageNet class labels for the images to be generated. guidance_scale (`float`, *optional*, defaults to 4.0): A higher guidance scale value encourages the model to generate images closely linked to the text `prompt` at the expense of lower image quality. Guidance scale is enabled when `guidance_scale > 1`. 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 250): 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 [`ImagePipelineOutput`] instead of a plain tuple. Examples: ```py >>> from diffusers import DiTPipeline, DPMSolverMultistepScheduler >>> import torch >>> pipe = DiTPipeline.from_pretrained("facebook/DiT-XL-2-256", torch_dtype=torch.float16) >>> pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config) >>> pipe = pipe.to("cuda") >>> # pick words from Imagenet class labels >>> pipe.labels # to print all available words >>> # pick words that exist in ImageNet >>> words = ["white shark", "umbrella"] >>> class_ids = pipe.get_label_ids(words) >>> generator = torch.manual_seed(33) >>> output = pipe(class_labels=class_ids, num_inference_steps=25, generator=generator) >>> image = output.images[0] # label 'white shark' ``` 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 )shaper8devicedtyper)r>irN)dimmpsnpu)r?r>)timestepr6g?rpil)images)0lenrconfig sample_size in_channelsr _execution_devicer?torchcattensorreshaper set_timesteps progress_bar timestepsscale_model_input is_tensorr>typer0floatfloat32float64int32int64r=toexpandsampler out_channelsstep prev_sample XLA_AVAILABLExm mark_stepchunkrscaling_factordecodeclampcpupermutenumpy numpy_to_pilmaybe_free_model_hooksr)r%r6r7r8r9r:r; batch_size latent_sizelatent_channelslatentslatent_model_input class_nullclass_labels_inputthalfrSis_mpsis_npur? noise_predepsrestcond_eps uncond_epshalf_eps model_output_sampless r*__call__DiTPipeline.__call__gs=x& &&--99 **11== I))""((   :H!9KUYYy1}5QX||L9O9OPXXY[\ \\4&:"5d>T>TU IWZ[I[UYY 'A1Eam $$%89""4>>#;#;((//'*11Q;$++Aq1++-''1a399;AAC % ''0G ##%: "'22r,)r)N)g@N2rFT)__name__ __module__ __qualname____firstlineno____doc__model_cpu_offload_seqr r r rrr strrrrr4rMno_gradrW Generatorboolrrr__static_attributes__ __classcell__)r)s@r*rr(s"/.2 <*<<- < 4S>* <<$/5d3i#8/T#Y/2 ]]_!$MQ#%%* S33iS3S3E%//43H"HIJ S3 ! S3 c] S3S3 "E) *S3S3r,r)typingrrrrrrMmodelsr r schedulersr utilsr utils.torch_utilsr pipeline_utilsrrtorch_xla.core.xla_modelcore xla_modelrcrbrr,r*rsI*65 :3+-C))MMS3#S3r,