refactor: address dg845 second-round review comments

  - Fix vae_image_processor -> image_processor in docs
  - Remove empty __init__ from AttnProcessor
  - Restore # Copied from on retrieve_timesteps and sync with source
  - Add check_inputs method for basic input validation
  - Remove unnecessary test skip for test_from_save_pretrained_dtype_inference

Author: tangyanfei.8

docs/source/en/api/pipelines/joyimage_edit_plus.md

@@ -35,7 +35,7 @@ images = [
     Image.open("reference_1.png").convert("RGB"),
 ]
 
-target_h, target_w = pipeline.vae_image_processor.get_default_height_width(images[-1])
+target_h, target_w = pipeline.image_processor.get_default_height_width(images[-1])
 
 output = pipeline(
     images=images,

src/diffusers/models/transformers/transformer_joyimage_edit_plus.py

@@ -80,9 +80,6 @@ class JoyImageEditPlusAttnProcessor:
     _attention_backend = None
     _parallel_config = None
 
-    def __init__(self):
-        pass
-
     def __call__(
         self,
         attn: "JoyImageEditPlusAttention",

src/diffusers/pipelines/joyimage/pipeline_joyimage_edit_plus.py

@@ -66,6 +66,7 @@
 """
 
 
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.retrieve_timesteps
 def retrieve_timesteps(
     scheduler,
     num_inference_steps: int | None = None,
@@ -98,24 +99,30 @@ def retrieve_timesteps(
         second element is the number of inference steps.
     """
     if timesteps is not None and sigmas is not None:
-        raise ValueError("Only one of `timesteps` or `sigmas` can be passed.")
-
+        raise ValueError("Only one of `timesteps` or `sigmas` can be passed. Please choose one to set custom values")
     if timesteps is not None:
-        if "timesteps" not in set(inspect.signature(scheduler.set_timesteps).parameters.keys()):
-            raise ValueError(f"{scheduler.__class__} does not support custom timesteps.")
+        accepts_timesteps = "timesteps" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accepts_timesteps:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" timestep schedules. Please check whether you are using the correct scheduler."
+            )
         scheduler.set_timesteps(timesteps=timesteps, device=device, **kwargs)
         timesteps = scheduler.timesteps
         num_inference_steps = len(timesteps)
     elif sigmas is not None:
-        if "sigmas" not in set(inspect.signature(scheduler.set_timesteps).parameters.keys()):
-            raise ValueError(f"{scheduler.__class__} does not support custom sigmas.")
+        accept_sigmas = "sigmas" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accept_sigmas:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" sigmas schedules. Please check whether you are using the correct scheduler."
+            )
         scheduler.set_timesteps(sigmas=sigmas, device=device, **kwargs)
         timesteps = scheduler.timesteps
         num_inference_steps = len(timesteps)
     else:
         scheduler.set_timesteps(num_inference_steps, device=device, **kwargs)
         timesteps = scheduler.timesteps
-
     return timesteps, num_inference_steps
 
 
@@ -390,6 +397,47 @@ def interrupt(self) -> bool:
     # Forward pass
     # ------------------------------------------------------------------
 
+    def check_inputs(
+        self,
+        prompt,
+        height,
+        width,
+        negative_prompt=None,
+        prompt_embeds=None,
+        negative_prompt_embeds=None,
+        callback_on_step_end_tensor_inputs=None,
+    ):
+        if height is not None and height % self.vae_scale_factor_spatial != 0:
+            raise ValueError(f"`height` must be divisible by {self.vae_scale_factor_spatial} but is {height}.")
+        if width is not None and width % self.vae_scale_factor_spatial != 0:
+            raise ValueError(f"`width` must be divisible by {self.vae_scale_factor_spatial} but is {width}.")
+
+        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
+        ):
+            raise ValueError(
+                f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found"
+                f" {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}"
+            )
+
+        if prompt is not None and prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `prompt`: {prompt} and `prompt_embeds`: {prompt_embeds}. Please make sure to"
+                " only forward one of the two."
+            )
+        elif prompt is None and prompt_embeds is None:
+            raise ValueError(
+                "Provide either `prompt` or `prompt_embeds`. Cannot leave both `prompt` and `prompt_embeds` undefined."
+            )
+        elif prompt is not None and (not isinstance(prompt, str) and not isinstance(prompt, list)):
+            raise ValueError(f"`prompt` has to be of type `str` or `list` but is {type(prompt)}")
+
+        if negative_prompt is not None and negative_prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `negative_prompt`: {negative_prompt} and `negative_prompt_embeds`:"
+                f" {negative_prompt_embeds}. Please make sure to only forward one of the two."
+            )
+
     @torch.no_grad()
     @replace_example_docstring(EXAMPLE_DOC_STRING)
     def __call__(
@@ -483,6 +531,16 @@ def __call__(
             if isinstance(images[0], Image.Image):
                 images = [images]  # single sample
 
+        self.check_inputs(
+            prompt=prompt,
+            height=height,
+            width=width,
+            negative_prompt=negative_prompt,
+            prompt_embeds=prompt_embeds,
+            negative_prompt_embeds=negative_prompt_embeds,
+            callback_on_step_end_tensor_inputs=callback_on_step_end_tensor_inputs,
+        )
+
         self._guidance_scale = guidance_scale
         self._interrupt = False
 

tests/models/transformers/test_models_transformer_joyimage_edit_plus.py

@@ -91,9 +91,7 @@ def get_dummy_inputs(self) -> dict[str, torch.Tensor]:
 
 
 class TestJoyImageEditPlusTransformer(JoyImageEditPlusTransformerTesterConfig, ModelTesterMixin):
-    @pytest.mark.parametrize("dtype", [torch.float16, torch.bfloat16], ids=["fp16", "bf16"])
-    def test_from_save_pretrained_dtype_inference(self, tmp_path, dtype):
-        pytest.skip("Tolerance requirements too high for meaningful test")
+    pass
 
 
 class TestJoyImageEditPlusTransformerMemory(JoyImageEditPlusTransformerTesterConfig, MemoryTesterMixin):