Cast height, width to int inside prepare latents (#7691)

update

examples/community/composable_stable_diffusion.py

@@ -321,7 +321,12 @@ def check_inputs(self, prompt, height, width, callback_steps):
             )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if latents is None:
             if device.type == "mps":
                 # randn does not work reproducibly on mps

examples/community/gluegen.py

@@ -500,7 +500,12 @@ def check_inputs(
                 )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/instaflow_one_step.py

@@ -468,7 +468,12 @@ def check_inputs(
                 )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/ip_adapter_face_id.py

@@ -753,7 +753,12 @@ def check_inputs(
                 )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/latent_consistency_img2img.py

@@ -177,7 +177,12 @@ def prepare_latents(
         latents=None,
         generator=None,
     ):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
 
         if not isinstance(image, (torch.Tensor, PIL.Image.Image, list)):
             raise ValueError(

examples/community/latent_consistency_interpolate.py

@@ -472,7 +472,12 @@ def run_safety_checker(self, image, device, dtype):
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/latent_consistency_txt2img.py

@@ -163,7 +163,12 @@ def run_safety_checker(self, image, device, dtype):
         return image, has_nsfw_concept
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if latents is None:
             latents = torch.randn(shape, dtype=dtype).to(device)
         else:

examples/community/lpw_stable_diffusion.py

@@ -726,7 +726,12 @@ def prepare_latents(
     ):
         if image is None:
             batch_size = batch_size * num_images_per_prompt
-            shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+            shape = (
+                batch_size,
+                num_channels_latents,
+                int(height) // self.vae_scale_factor,
+                int(width) // self.vae_scale_factor,
+            )
             if isinstance(generator, list) and len(generator) != batch_size:
                 raise ValueError(
                     f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/lpw_stable_diffusion_xl.py

@@ -1060,7 +1060,12 @@ def prepare_latents(
         batch_size *= num_images_per_prompt
 
         if image is None:
-            shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+            shape = (
+                batch_size,
+                num_channels_latents,
+                int(height) // self.vae_scale_factor,
+                int(width) // self.vae_scale_factor,
+            )
             if isinstance(generator, list) and len(generator) != batch_size:
                 raise ValueError(
                     f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"
@@ -1140,7 +1145,12 @@ def prepare_latents(
             return latents
 
         else:
-            shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+            shape = (
+                batch_size,
+                num_channels_latents,
+                int(height) // self.vae_scale_factor,
+                int(width) // self.vae_scale_factor,
+            )
             if isinstance(generator, list) and len(generator) != batch_size:
                 raise ValueError(
                     f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/pipeline_demofusion_sdxl.py

@@ -477,7 +477,12 @@ def check_inputs(
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/pipeline_sdxl_style_aligned.py

@@ -919,7 +919,12 @@ def prepare_latents(
         batch_size *= num_images_per_prompt
 
         if image is None:
-            shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+            shape = (
+                batch_size,
+                num_channels_latents,
+                int(height) // self.vae_scale_factor,
+                int(width) // self.vae_scale_factor,
+            )
             if isinstance(generator, list) and len(generator) != batch_size:
                 raise ValueError(
                     f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"
@@ -999,7 +1004,12 @@ def prepare_latents(
             return latents
 
         else:
-            shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+            shape = (
+                batch_size,
+                num_channels_latents,
+                int(height) // self.vae_scale_factor,
+                int(width) // self.vae_scale_factor,
+            )
             if isinstance(generator, list) and len(generator) != batch_size:
                 raise ValueError(
                     f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/pipeline_stable_diffusion_pag.py

@@ -857,7 +857,12 @@ def check_inputs(
             )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/pipeline_stable_diffusion_xl_controlnet_adapter.py

@@ -751,7 +751,12 @@ def check_conditions(
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/pipeline_stable_diffusion_xl_ipex.py

@@ -614,7 +614,12 @@ def check_inputs(
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/pipeline_zero1to3.py

@@ -497,7 +497,12 @@ def check_inputs(self, image, height, width, callback_steps):
             )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/stable_diffusion_controlnet_inpaint.py

@@ -635,7 +635,12 @@ def check_inputs(
             )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/stable_diffusion_ipex.py

@@ -533,7 +533,12 @@ def check_inputs(
                 )
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/community/stable_diffusion_reference.py

@@ -609,7 +609,12 @@ def prepare_latents(
         Returns:
             torch.Tensor: The prepared latent vectors.
         """
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/research_projects/promptdiffusion/pipeline_prompt_diffusion.py

@@ -789,7 +789,12 @@ def prepare_image(
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

examples/research_projects/rdm/pipeline_rdm.py

@@ -123,7 +123,12 @@ def _encode_image(self, retrieved_images, batch_size):
         return image_embeddings
 
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

src/diffusers/pipelines/audioldm/pipeline_audioldm.py

@@ -330,8 +330,8 @@ def prepare_latents(self, batch_size, num_channels_latents, height, dtype, devic
         shape = (
             batch_size,
             num_channels_latents,
-            height // self.vae_scale_factor,
-            self.vocoder.config.model_in_dim // self.vae_scale_factor,
+            int(height) // self.vae_scale_factor,
+            int(self.vocoder.config.model_in_dim) // self.vae_scale_factor,
         )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(

src/diffusers/pipelines/audioldm2/pipeline_audioldm2.py

@@ -790,8 +790,8 @@ def prepare_latents(self, batch_size, num_channels_latents, height, dtype, devic
         shape = (
             batch_size,
             num_channels_latents,
-            height // self.vae_scale_factor,
-            self.vocoder.config.model_in_dim // self.vae_scale_factor,
+            int(height) // self.vae_scale_factor,
+            int(self.vocoder.config.model_in_dim) // self.vae_scale_factor,
         )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(

src/diffusers/pipelines/controlnet/pipeline_controlnet.py

@@ -807,7 +807,12 @@ def prepare_image(
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

src/diffusers/pipelines/controlnet/pipeline_controlnet_inpaint.py

@@ -972,7 +972,12 @@ def prepare_latents(
         return_noise=False,
         return_image_latents=False,
     ):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

src/diffusers/pipelines/controlnet/pipeline_controlnet_inpaint_sd_xl.py

@@ -880,7 +880,12 @@ def prepare_latents(
         return_noise=False,
         return_image_latents=False,
     ):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"

src/diffusers/pipelines/controlnet/pipeline_controlnet_sd_xl.py

@@ -813,7 +813,12 @@ def prepare_image(
 
     # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
     def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
-        shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor)
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
         if isinstance(generator, list) and len(generator) != batch_size:
             raise ValueError(
                 f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"