Merge remote-tracking branch 'upstream/main' into cogview3-plus

Author: zRzRzRzRzRzRzR

src/diffusers/models/autoencoders/autoencoder_kl_cogvideox.py

@@ -1234,7 +1234,7 @@ def _decode(self, z: torch.Tensor, return_dict: bool = True) -> Union[DecoderOut
             return self.tiled_decode(z, return_dict=return_dict)
 
         frame_batch_size = self.num_latent_frames_batch_size
-        num_batches = num_frames // frame_batch_size
+        num_batches = max(num_frames // frame_batch_size, 1)
         conv_cache = None
         dec = []
 

src/diffusers/pipelines/controlnet_sd3/pipeline_stable_diffusion_3_controlnet_inpainting.py

@@ -251,6 +251,9 @@ def __init__(
             if hasattr(self, "transformer") and self.transformer is not None
             else 128
         )
+        self.patch_size = (
+            self.transformer.config.patch_size if hasattr(self, "transformer") and self.transformer is not None else 2
+        )
 
     # Copied from diffusers.pipelines.stable_diffusion_3.pipeline_stable_diffusion_3.StableDiffusion3Pipeline._get_t5_prompt_embeds
     def _get_t5_prompt_embeds(
@@ -577,8 +580,14 @@ def check_inputs(
         callback_on_step_end_tensor_inputs=None,
         max_sequence_length=None,
     ):
-        if height % 8 != 0 or width % 8 != 0:
-            raise ValueError(f"`height` and `width` have to be divisible by 8 but are {height} and {width}.")
+        if (
+            height % (self.vae_scale_factor * self.patch_size) != 0
+            or width % (self.vae_scale_factor * self.patch_size) != 0
+        ):
+            raise ValueError(
+                f"`height` and `width` have to be divisible by {self.vae_scale_factor * self.patch_size} but are {height} and {width}."
+                f"You can use height {height - height % (self.vae_scale_factor * self.patch_size)} and width {width - width % (self.vae_scale_factor * self.patch_size)}."
+            )
 
         if callback_on_step_end_tensor_inputs is not None and not all(
             k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs

src/diffusers/pipelines/pag/pipeline_pag_sd_3.py

@@ -212,6 +212,9 @@ def __init__(
             if hasattr(self, "transformer") and self.transformer is not None
             else 128
         )
+        self.patch_size = (
+            self.transformer.config.patch_size if hasattr(self, "transformer") and self.transformer is not None else 2
+        )
 
         self.set_pag_applied_layers(
             pag_applied_layers, pag_attn_processors=(PAGCFGJointAttnProcessor2_0(), PAGJointAttnProcessor2_0())
@@ -542,8 +545,14 @@ def check_inputs(
         callback_on_step_end_tensor_inputs=None,
         max_sequence_length=None,
     ):
-        if height % 8 != 0 or width % 8 != 0:
-            raise ValueError(f"`height` and `width` have to be divisible by 8 but are {height} and {width}.")
+        if (
+            height % (self.vae_scale_factor * self.patch_size) != 0
+            or width % (self.vae_scale_factor * self.patch_size) != 0
+        ):
+            raise ValueError(
+                f"`height` and `width` have to be divisible by {self.vae_scale_factor * self.patch_size} but are {height} and {width}."
+                f"You can use height {height - height % (self.vae_scale_factor * self.patch_size)} and width {width - width % (self.vae_scale_factor * self.patch_size)}."
+            )
 
         if callback_on_step_end_tensor_inputs is not None and not all(
             k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs

src/diffusers/pipelines/stable_diffusion_3/pipeline_stable_diffusion_3.py

@@ -203,6 +203,9 @@ def __init__(
             if hasattr(self, "transformer") and self.transformer is not None
             else 128
         )
+        self.patch_size = (
+            self.transformer.config.patch_size if hasattr(self, "transformer") and self.transformer is not None else 2
+        )
 
     def _get_t5_prompt_embeds(
         self,
@@ -525,8 +528,14 @@ def check_inputs(
         callback_on_step_end_tensor_inputs=None,
         max_sequence_length=None,
     ):
-        if height % 8 != 0 or width % 8 != 0:
-            raise ValueError(f"`height` and `width` have to be divisible by 8 but are {height} and {width}.")
+        if (
+            height % (self.vae_scale_factor * self.patch_size) != 0
+            or width % (self.vae_scale_factor * self.patch_size) != 0
+        ):
+            raise ValueError(
+                f"`height` and `width` have to be divisible by {self.vae_scale_factor * self.patch_size} but are {height} and {width}."
+                f"You can use height {height - height % (self.vae_scale_factor * self.patch_size)} and width {width - width % (self.vae_scale_factor * self.patch_size)}."
+            )
 
         if callback_on_step_end_tensor_inputs is not None and not all(
             k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs