force inpaint

modules/processing_diffusers.py

@@ -214,10 +214,8 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
                 p.mask = TF.to_pil_image(torch.ones_like(TF.to_tensor(p.init_images[0]))).convert("L")
             width = 8 * math.ceil(p.init_images[0].width / 8)
             height = 8 * math.ceil(p.init_images[0].height / 8)
-
             # option-1: use images as inputs
             task_args = {"image": p.init_images, "mask_image": p.mask, "strength": p.denoising_strength, "height": height, "width": width}
-
             """ # option-2: preprocess images into latents using diffusers
             vae_scale_factor = 2 ** (len(model.vae.config.block_out_channels) - 1)
             image_processor = diffusers.image_processor.VaeImageProcessor(vae_scale_factor=vae_scale_factor)
@@ -226,7 +224,6 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
             mask_image = mask_processor.preprocess(p.mask, width=width, height=height)
             task_args = {"image": p.init_images, "mask_image": p.mask, "strength": p.denoising_strength, "height": height, "width": width}
             """
-
             """ # option-2: manually assemble masked image latents
             masked_image_latents = []
             mask_image = TF.to_tensor(p.mask)
@@ -237,7 +234,6 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
             masked_image_latents = torch.stack(masked_image_latents, dim=0).to(shared.device)
             task_args = {"image": p.init_images, "mask_image": mask_image, "masked_image_latents": masked_image_latents, "strength": p.denoising_strength, "height": height, "width": width}
             """
-
         if model.__class__.__name__ == 'LatentConsistencyModelPipeline' and hasattr(p, 'init_images') and len(p.init_images) > 0:
             init_latents = [vae_encode(image, model=shared.sd_model, full_quality=p.full_quality).squeeze(dim=0) for image in p.init_images]
             init_latent = torch.stack(init_latents, dim=0).to(shared.device)

modules/sd_models.py

@@ -1010,10 +1010,6 @@ def set_diffuser_pipe(pipe, new_pipe_type):
     )
     diffusers.pipelines.auto_pipeline.AUTO_TEXT2IMAGE_PIPELINES_MAPPING = AUTO_TEXT2IMAGE_PIPELINES_MAPPING
     """
-
-    if shared.opts.diffusers_force_inpaint:
-        if new_pipe_type == DiffusersTaskType.IMAGE_2_IMAGE:
-            new_pipe_type = DiffusersTaskType.INPAINTING # sdxl may work better with init mask
     try:
         if new_pipe_type == DiffusersTaskType.TEXT_2_IMAGE:
             new_pipe = diffusers.AutoPipelineForText2Image.from_pipe(pipe)

modules/shared.py

@@ -368,7 +368,6 @@ options_templates.update(options_section(('diffusers', "Diffusers Settings"), {
     "diffusers_eval": OptionInfo(True, "Force model eval"),
     "diffusers_force_zeros": OptionInfo(True, "Force zeros for prompts when empty"),
     "diffusers_aesthetics_score": OptionInfo(False, "Require aesthetics score"),
-    "diffusers_force_inpaint": OptionInfo(False, 'Diffusers force inpaint pipeline'),
     "diffusers_pooled": OptionInfo("default", "Diffusers SDXL pooled embeds (experimental)", gr.Radio, {"choices": ['default', 'weighted']}),
 }))
 
@@ -531,9 +530,9 @@ options_templates.update(options_section(('sampler-params', "Sampler Settings"),
     'uni_pc_variant': OptionInfo("bh1", "UniPC variant", gr.Radio, {"choices": ["bh1", "bh2", "vary_coeff"]}),
     'uni_pc_skip_type': OptionInfo("time_uniform", "UniPC skip type", gr.Radio, {"choices": ["time_uniform", "time_quadratic", "logSNR"]}),
     "ddim_discretize": OptionInfo('uniform', "DDIM discretize img2img", gr.Radio, {"choices": ['uniform', 'quad']}),
-    # TODO pad_cond_uncond implementation missing
+    # TODO pad_cond_uncond implementation missing for original backend
     "pad_cond_uncond": OptionInfo(True, "Pad prompt and negative prompt to be same length", gr.Checkbox, {"visible": False}),
-    # TODO batch_cond-uncond implementation missing
+    # TODO batch_cond-uncond implementation missing for original backend
     "batch_cond_uncond": OptionInfo(True, "Do conditional and unconditional denoising in one batch", gr.Checkbox, {"visible": False}),
 }))