Use pipelines without vae (#10441)

* Use pipelines without vae

* getattr

* vqvae

---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

src/diffusers/pipelines/aura_flow/pipeline_aura_flow.py

@@ -146,9 +146,7 @@ class AuraFlowPipeline(DiffusionPipeline):
             tokenizer=tokenizer, text_encoder=text_encoder, vae=vae, transformer=transformer, scheduler=scheduler
         )
 
-        self.vae_scale_factor = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
-        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
 
     def check_inputs(

src/diffusers/pipelines/cogvideo/pipeline_cogvideox.py

@@ -183,14 +183,12 @@ class CogVideoXPipeline(DiffusionPipeline, CogVideoXLoraLoaderMixin):
             tokenizer=tokenizer, text_encoder=text_encoder, vae=vae, transformer=transformer, scheduler=scheduler
         )
         self.vae_scale_factor_spatial = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
+            2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         )
         self.vae_scale_factor_temporal = (
-            self.vae.config.temporal_compression_ratio if hasattr(self, "vae") and self.vae is not None else 4
-        )
-        self.vae_scaling_factor_image = (
-            self.vae.config.scaling_factor if hasattr(self, "vae") and self.vae is not None else 0.7
+            self.vae.config.temporal_compression_ratio if getattr(self, "vae", None) else 4
         )
+        self.vae_scaling_factor_image = self.vae.config.scaling_factor if getattr(self, "vae", None) else 0.7
 
         self.video_processor = VideoProcessor(vae_scale_factor=self.vae_scale_factor_spatial)
 

src/diffusers/pipelines/cogvideo/pipeline_cogvideox_fun_control.py

@@ -190,14 +190,12 @@ class CogVideoXFunControlPipeline(DiffusionPipeline, CogVideoXLoraLoaderMixin):
             tokenizer=tokenizer, text_encoder=text_encoder, vae=vae, transformer=transformer, scheduler=scheduler
         )
         self.vae_scale_factor_spatial = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
+            2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         )
         self.vae_scale_factor_temporal = (
-            self.vae.config.temporal_compression_ratio if hasattr(self, "vae") and self.vae is not None else 4
-        )
-        self.vae_scaling_factor_image = (
-            self.vae.config.scaling_factor if hasattr(self, "vae") and self.vae is not None else 0.7
+            self.vae.config.temporal_compression_ratio if getattr(self, "vae", None) else 4
         )
+        self.vae_scaling_factor_image = self.vae.config.scaling_factor if getattr(self, "vae", None) else 0.7
 
         self.video_processor = VideoProcessor(vae_scale_factor=self.vae_scale_factor_spatial)
 

src/diffusers/pipelines/cogvideo/pipeline_cogvideox_image2video.py

@@ -203,14 +203,12 @@ class CogVideoXImageToVideoPipeline(DiffusionPipeline, CogVideoXLoraLoaderMixin)
             scheduler=scheduler,
         )
         self.vae_scale_factor_spatial = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
+            2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         )
         self.vae_scale_factor_temporal = (
-            self.vae.config.temporal_compression_ratio if hasattr(self, "vae") and self.vae is not None else 4
-        )
-        self.vae_scaling_factor_image = (
-            self.vae.config.scaling_factor if hasattr(self, "vae") and self.vae is not None else 0.7
+            self.vae.config.temporal_compression_ratio if getattr(self, "vae", None) else 4
         )
+        self.vae_scaling_factor_image = self.vae.config.scaling_factor if getattr(self, "vae", None) else 0.7
 
         self.video_processor = VideoProcessor(vae_scale_factor=self.vae_scale_factor_spatial)
 

src/diffusers/pipelines/cogvideo/pipeline_cogvideox_video2video.py

@@ -206,14 +206,12 @@ class CogVideoXVideoToVideoPipeline(DiffusionPipeline, CogVideoXLoraLoaderMixin)
         )
 
         self.vae_scale_factor_spatial = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
+            2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         )
         self.vae_scale_factor_temporal = (
-            self.vae.config.temporal_compression_ratio if hasattr(self, "vae") and self.vae is not None else 4
-        )
-        self.vae_scaling_factor_image = (
-            self.vae.config.scaling_factor if hasattr(self, "vae") and self.vae is not None else 0.7
+            self.vae.config.temporal_compression_ratio if getattr(self, "vae", None) else 4
         )
+        self.vae_scaling_factor_image = self.vae.config.scaling_factor if getattr(self, "vae", None) else 0.7
 
         self.video_processor = VideoProcessor(vae_scale_factor=self.vae_scale_factor_spatial)
 

src/diffusers/pipelines/cogview3/pipeline_cogview3plus.py

@@ -153,9 +153,7 @@ class CogView3PlusPipeline(DiffusionPipeline):
         self.register_modules(
             tokenizer=tokenizer, text_encoder=text_encoder, vae=vae, transformer=transformer, scheduler=scheduler
         )
-        self.vae_scale_factor = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
-        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
 
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
 

src/diffusers/pipelines/controlnet/pipeline_controlnet.py

@@ -254,7 +254,7 @@ class StableDiffusionControlNetPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False

src/diffusers/pipelines/controlnet/pipeline_controlnet_img2img.py

@@ -224,7 +224,7 @@ class StableDiffusionControlNetImg2ImgPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False

src/diffusers/pipelines/controlnet/pipeline_controlnet_inpaint.py

@@ -223,7 +223,7 @@ class StableDiffusionControlNetInpaintPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
         self.mask_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_normalize=False, do_binarize=True, do_convert_grayscale=True

src/diffusers/pipelines/controlnet/pipeline_controlnet_inpaint_sd_xl.py

@@ -264,7 +264,7 @@ class StableDiffusionXLControlNetInpaintPipeline(
         )
         self.register_to_config(force_zeros_for_empty_prompt=force_zeros_for_empty_prompt)
         self.register_to_config(requires_aesthetics_score=requires_aesthetics_score)
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
         self.mask_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_normalize=False, do_binarize=True, do_convert_grayscale=True

src/diffusers/pipelines/controlnet/pipeline_controlnet_sd_xl.py

@@ -275,7 +275,7 @@ class StableDiffusionXLControlNetPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False

src/diffusers/pipelines/controlnet/pipeline_controlnet_sd_xl_img2img.py

@@ -267,7 +267,7 @@ class StableDiffusionXLControlNetImg2ImgPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False

src/diffusers/pipelines/controlnet/pipeline_controlnet_union_inpaint_sd_xl.py

@@ -246,7 +246,7 @@ class StableDiffusionXLControlNetUnionInpaintPipeline(
         )
         self.register_to_config(force_zeros_for_empty_prompt=force_zeros_for_empty_prompt)
         self.register_to_config(requires_aesthetics_score=requires_aesthetics_score)
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
         self.mask_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_normalize=False, do_binarize=True, do_convert_grayscale=True

src/diffusers/pipelines/controlnet/pipeline_controlnet_union_sd_xl.py

@@ -257,7 +257,7 @@ class StableDiffusionXLControlNetUnionPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False

src/diffusers/pipelines/controlnet/pipeline_controlnet_union_sd_xl_img2img.py

@@ -281,7 +281,7 @@ class StableDiffusionXLControlNetUnionImg2ImgPipeline(
             feature_extractor=feature_extractor,
             image_encoder=image_encoder,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False

src/diffusers/pipelines/controlnet/pipeline_flax_controlnet.py

@@ -178,7 +178,7 @@ class FlaxStableDiffusionControlNetPipeline(FlaxDiffusionPipeline):
             safety_checker=safety_checker,
             feature_extractor=feature_extractor,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
 
     def prepare_text_inputs(self, prompt: Union[str, List[str]]):
         if not isinstance(prompt, (str, list)):

src/diffusers/pipelines/controlnet_hunyuandit/pipeline_hunyuandit_controlnet.py

@@ -269,9 +269,7 @@ class HunyuanDiTControlNetPipeline(DiffusionPipeline):
                 " checker. If you do not want to use the safety checker, you can pass `'safety_checker=None'` instead."
             )
 
-        self.vae_scale_factor = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
-        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
         self.register_to_config(requires_safety_checker=requires_safety_checker)
         self.default_sample_size = (

src/diffusers/pipelines/controlnet_sd3/pipeline_stable_diffusion_3_controlnet.py

@@ -236,9 +236,7 @@ class StableDiffusion3ControlNetPipeline(
             image_encoder=image_encoder,
             feature_extractor=feature_extractor,
         )
-        self.vae_scale_factor = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
-        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
         self.tokenizer_max_length = (
             self.tokenizer.model_max_length if hasattr(self, "tokenizer") and self.tokenizer is not None else 77

src/diffusers/pipelines/controlnet_sd3/pipeline_stable_diffusion_3_controlnet_inpainting.py

@@ -230,9 +230,7 @@ class StableDiffusion3ControlNetInpaintingPipeline(DiffusionPipeline, SD3LoraLoa
             scheduler=scheduler,
             controlnet=controlnet,
         )
-        self.vae_scale_factor = (
-            2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
-        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_resize=True, do_convert_rgb=True, do_normalize=True
         )

src/diffusers/pipelines/controlnet_xs/pipeline_controlnet_xs.py

@@ -178,7 +178,7 @@ class StableDiffusionControlNetXSPipeline(
             safety_checker=safety_checker,
             feature_extractor=feature_extractor,
         )
-        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
         self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
         self.control_image_processor = VaeImageProcessor(
             vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False