[`Tests`] Update a deprecated parameter in test files and fix several typos (#7277)

* Add properties and `IPAdapterTesterMixin` tests for `StableDiffusionPanoramaPipeline`

* Fix variable name typo and update comments

* Update deprecated `output_type="numpy"` to "np" in test files

* Discard changes to src/diffusers/pipelines/stable_diffusion_panorama/pipeline_stable_diffusion_panorama.py

* Update test_stable_diffusion_panorama.py

* Update numbers in README.md

* Update get_guidance_scale_embedding method to use timesteps instead of w

* Update number of checkpoints in README.md

* Add type hints and fix var name

* Fix PyTorch's convention for inplace functions

* Fix a typo

* Revert "Fix PyTorch's convention for inplace functions"

This reverts commit 74350cf65b2c9aa77f08bec7937d7a8b13edb509.

* Fix typos

* Indent

* Refactor get_guidance_scale_embedding method in LEditsPPPipelineStableDiffusionXL class

src/diffusers/pipelines/stable_diffusion_sag/pipeline_stable_diffusion_sag.py

@@ -632,7 +632,7 @@ class StableDiffusionSAGPipeline(DiffusionPipeline, StableDiffusionMixin, Textua
         # corresponds to doing no classifier free guidance.
         do_classifier_free_guidance = guidance_scale > 1.0
         # and `sag_scale` is` `s` of equation (16)
-        # of the self-attentnion guidance paper: https://arxiv.org/pdf/2210.00939.pdf
+        # of the self-attention guidance paper: https://arxiv.org/pdf/2210.00939.pdf
         # `sag_scale = 0` means no self-attention guidance
         do_self_attention_guidance = sag_scale > 0.0
 
@@ -667,7 +667,7 @@ class StableDiffusionSAGPipeline(DiffusionPipeline, StableDiffusionMixin, Textua
 
         if timesteps.dtype not in [torch.int16, torch.int32, torch.int64]:
             raise ValueError(
-                f"{self.__class__.__name__} does not support using a scheduler of type {self.scheduler.__class__.__name__}. Please make sure to use one of 'DDIMScheduler, PNDMScheduler, DDPMScheduler, DEISMultistepScheduler, UniPCMultistepScheduler, DPMSolverMultistepScheduler, DPMSolverSinlgestepScheduler'."
+                f"{self.__class__.__name__} does not support using a scheduler of type {self.scheduler.__class__.__name__}. Please make sure to use one of 'DDIMScheduler, PNDMScheduler, DDPMScheduler, DEISMultistepScheduler, UniPCMultistepScheduler, DPMSolverMultistepScheduler, DPMSolverSinglestepScheduler'."
             )
 
         # 5. Prepare latent variables
@@ -723,7 +723,7 @@ class StableDiffusionSAGPipeline(DiffusionPipeline, StableDiffusionMixin, Textua
                         noise_pred_uncond, noise_pred_text = noise_pred.chunk(2)
                         noise_pred = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond)
 
-                    # perform self-attention guidance with the stored self-attentnion map
+                    # perform self-attention guidance with the stored self-attention map
                     if do_self_attention_guidance:
                         # classifier-free guidance produces two chunks of attention map
                         # and we only use unconditional one according to equation (25)

src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl.py

@@ -740,20 +740,22 @@ class StableDiffusionXLPipeline(
             self.vae.decoder.mid_block.to(dtype)
 
     # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding
-    def get_guidance_scale_embedding(self, w, embedding_dim=512, dtype=torch.float32):
+    def get_guidance_scale_embedding(
+        self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32
+    ) -> torch.FloatTensor:
         """
         See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
 
         Args:
-            timesteps (`torch.Tensor`):
-                generate embedding vectors at these timesteps
+            w (`torch.Tensor`):
+                Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings.
             embedding_dim (`int`, *optional*, defaults to 512):
-                dimension of the embeddings to generate
-            dtype:
-                data type of the generated embeddings
+                Dimension of the embeddings to generate.
+            dtype (`torch.dtype`, *optional*, defaults to `torch.float32`):
+                Data type of the generated embeddings.
 
         Returns:
-            `torch.FloatTensor`: Embedding vectors with shape `(len(timesteps), embedding_dim)`
+            `torch.FloatTensor`: Embedding vectors with shape `(len(w), embedding_dim)`.
         """
         assert len(w.shape) == 1
         w = w * 1000.0

src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_img2img.py

@@ -874,20 +874,22 @@ class StableDiffusionXLImg2ImgPipeline(
             self.vae.decoder.mid_block.to(dtype)
 
     # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding
-    def get_guidance_scale_embedding(self, w, embedding_dim=512, dtype=torch.float32):
+    def get_guidance_scale_embedding(
+        self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32
+    ) -> torch.FloatTensor:
         """
         See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
 
         Args:
-            timesteps (`torch.Tensor`):
-                generate embedding vectors at these timesteps
+            w (`torch.Tensor`):
+                Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings.
             embedding_dim (`int`, *optional*, defaults to 512):
-                dimension of the embeddings to generate
-            dtype:
-                data type of the generated embeddings
+                Dimension of the embeddings to generate.
+            dtype (`torch.dtype`, *optional*, defaults to `torch.float32`):
+                Data type of the generated embeddings.
 
         Returns:
-            `torch.FloatTensor`: Embedding vectors with shape `(len(timesteps), embedding_dim)`
+            `torch.FloatTensor`: Embedding vectors with shape `(len(w), embedding_dim)`.
         """
         assert len(w.shape) == 1
         w = w * 1000.0

src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_inpaint.py

@@ -1110,20 +1110,22 @@ class StableDiffusionXLInpaintPipeline(
             self.vae.decoder.mid_block.to(dtype)
 
     # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding
-    def get_guidance_scale_embedding(self, w, embedding_dim=512, dtype=torch.float32):
+    def get_guidance_scale_embedding(
+        self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32
+    ) -> torch.FloatTensor:
         """
         See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
 
         Args:
-            timesteps (`torch.Tensor`):
-                generate embedding vectors at these timesteps
+            w (`torch.Tensor`):
+                Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings.
             embedding_dim (`int`, *optional*, defaults to 512):
-                dimension of the embeddings to generate
-            dtype:
-                data type of the generated embeddings
+                Dimension of the embeddings to generate.
+            dtype (`torch.dtype`, *optional*, defaults to `torch.float32`):
+                Data type of the generated embeddings.
 
         Returns:
-            `torch.FloatTensor`: Embedding vectors with shape `(len(timesteps), embedding_dim)`
+            `torch.FloatTensor`: Embedding vectors with shape `(len(w), embedding_dim)`.
         """
         assert len(w.shape) == 1
         w = w * 1000.0

src/diffusers/pipelines/t2i_adapter/pipeline_stable_diffusion_adapter.py

@@ -613,20 +613,22 @@ class StableDiffusionAdapterPipeline(DiffusionPipeline, StableDiffusionMixin):
         return height, width
 
     # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding
-    def get_guidance_scale_embedding(self, w, embedding_dim=512, dtype=torch.float32):
+    def get_guidance_scale_embedding(
+        self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32
+    ) -> torch.FloatTensor:
         """
         See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
 
         Args:
-            timesteps (`torch.Tensor`):
-                generate embedding vectors at these timesteps
+            w (`torch.Tensor`):
+                Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings.
             embedding_dim (`int`, *optional*, defaults to 512):
-                dimension of the embeddings to generate
-            dtype:
-                data type of the generated embeddings
+                Dimension of the embeddings to generate.
+            dtype (`torch.dtype`, *optional*, defaults to `torch.float32`):
+                Data type of the generated embeddings.
 
         Returns:
-            `torch.FloatTensor`: Embedding vectors with shape `(len(timesteps), embedding_dim)`
+            `torch.FloatTensor`: Embedding vectors with shape `(len(w), embedding_dim)`.
         """
         assert len(w.shape) == 1
         w = w * 1000.0

src/diffusers/pipelines/t2i_adapter/pipeline_stable_diffusion_xl_adapter.py

@@ -784,20 +784,22 @@ class StableDiffusionXLAdapterPipeline(
         return height, width
 
     # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding
-    def get_guidance_scale_embedding(self, w, embedding_dim=512, dtype=torch.float32):
+    def get_guidance_scale_embedding(
+        self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32
+    ) -> torch.FloatTensor:
         """
         See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
 
         Args:
-            timesteps (`torch.Tensor`):
-                generate embedding vectors at these timesteps
+            w (`torch.Tensor`):
+                Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings.
             embedding_dim (`int`, *optional*, defaults to 512):
-                dimension of the embeddings to generate
-            dtype:
-                data type of the generated embeddings
+                Dimension of the embeddings to generate.
+            dtype (`torch.dtype`, *optional*, defaults to `torch.float32`):
+                Data type of the generated embeddings.
 
         Returns:
-            `torch.FloatTensor`: Embedding vectors with shape `(len(timesteps), embedding_dim)`
+            `torch.FloatTensor`: Embedding vectors with shape `(len(w), embedding_dim)`.
         """
         assert len(w.shape) == 1
         w = w * 1000.0

src/diffusers/pipelines/text_to_video_synthesis/pipeline_text_to_video_zero.py

@@ -575,8 +575,8 @@ class TextToVideoZeroPipeline(DiffusionPipeline, StableDiffusionMixin, TextualIn
                 Pre-generated noisy latents sampled from a Gaussian distribution, to be used as inputs for video
                 generation. Can be used to tweak the same generation with different prompts. If not provided, a latents
                 tensor is generated by sampling using the supplied random `generator`.
-            output_type (`str`, *optional*, defaults to `"numpy"`):
-                The output format of the generated video. Choose between `"latent"` and `"numpy"`.
+            output_type (`str`, *optional*, defaults to `"np"`):
+                The output format of the generated video. Choose between `"latent"` and `"np"`.
             return_dict (`bool`, *optional*, defaults to `True`):
                 Whether or not to return a
                 [`~pipelines.text_to_video_synthesis.pipeline_text_to_video_zero.TextToVideoPipelineOutput`] instead of

tests/pipelines/controlnet/test_controlnet.py

@@ -211,7 +211,7 @@ class ControlNetPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": image,
         }
 
@@ -402,7 +402,7 @@ class StableDiffusionMultiControlNetPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": images,
         }
 
@@ -602,7 +602,7 @@ class StableDiffusionMultiControlNetOneModelPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": images,
         }
 

tests/pipelines/controlnet/test_controlnet_img2img.py

@@ -164,7 +164,7 @@ class ControlNetImg2ImgPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": image,
             "control_image": control_image,
         }
@@ -313,7 +313,7 @@ class StableDiffusionMultiControlNetPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": image,
             "control_image": control_image,
         }

tests/pipelines/controlnet/test_controlnet_inpaint.py

@@ -155,7 +155,7 @@ class ControlNetInpaintPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": image,
             "mask_image": mask_image,
             "control_image": control_image,
@@ -375,7 +375,7 @@ class MultiControlNetInpaintPipelineFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": image,
             "mask_image": mask_image,
             "control_image": control_image,

tests/pipelines/controlnet/test_controlnet_inpaint_sdxl.py

@@ -172,7 +172,7 @@ class ControlNetPipelineSDXLFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": init_image,
             "mask_image": mask_image,
             "control_image": control_image,

tests/pipelines/controlnet/test_controlnet_sdxl_img2img.py

@@ -163,7 +163,7 @@ class ControlNetPipelineSDXLImg2ImgFastTests(
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,
-            "output_type": "numpy",
+            "output_type": "np",
             "image": image,
             "control_image": image,
         }

tests/pipelines/ddim/test_ddim.py

@@ -63,7 +63,7 @@ class DDIMPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
             "batch_size": 1,
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
         return inputs
 
@@ -113,7 +113,7 @@ class DDIMPipelineIntegrationTests(unittest.TestCase):
         ddim.set_progress_bar_config(disable=None)
 
         generator = torch.manual_seed(0)
-        image = ddim(generator=generator, eta=0.0, output_type="numpy").images
+        image = ddim(generator=generator, eta=0.0, output_type="np").images
 
         image_slice = image[0, -3:, -3:, -1]
 
@@ -133,7 +133,7 @@ class DDIMPipelineIntegrationTests(unittest.TestCase):
         ddpm.set_progress_bar_config(disable=None)
 
         generator = torch.manual_seed(0)
-        image = ddpm(generator=generator, output_type="numpy").images
+        image = ddpm(generator=generator, output_type="np").images
 
         image_slice = image[0, -3:, -3:, -1]
 

tests/pipelines/ddpm/test_ddpm.py

@@ -50,10 +50,10 @@ class DDPMPipelineFastTests(unittest.TestCase):
         ddpm.set_progress_bar_config(disable=None)
 
         generator = torch.Generator(device=device).manual_seed(0)
-        image = ddpm(generator=generator, num_inference_steps=2, output_type="numpy").images
+        image = ddpm(generator=generator, num_inference_steps=2, output_type="np").images
 
         generator = torch.Generator(device=device).manual_seed(0)
-        image_from_tuple = ddpm(generator=generator, num_inference_steps=2, output_type="numpy", return_dict=False)[0]
+        image_from_tuple = ddpm(generator=generator, num_inference_steps=2, output_type="np", return_dict=False)[0]
 
         image_slice = image[0, -3:, -3:, -1]
         image_from_tuple_slice = image_from_tuple[0, -3:, -3:, -1]
@@ -75,10 +75,10 @@ class DDPMPipelineFastTests(unittest.TestCase):
         ddpm.set_progress_bar_config(disable=None)
 
         generator = torch.manual_seed(0)
-        image = ddpm(generator=generator, num_inference_steps=2, output_type="numpy").images
+        image = ddpm(generator=generator, num_inference_steps=2, output_type="np").images
 
         generator = torch.manual_seed(0)
-        image_eps = ddpm(generator=generator, num_inference_steps=2, output_type="numpy")[0]
+        image_eps = ddpm(generator=generator, num_inference_steps=2, output_type="np")[0]
 
         image_slice = image[0, -3:, -3:, -1]
         image_eps_slice = image_eps[0, -3:, -3:, -1]
@@ -102,7 +102,7 @@ class DDPMPipelineIntegrationTests(unittest.TestCase):
         ddpm.set_progress_bar_config(disable=None)
 
         generator = torch.manual_seed(0)
-        image = ddpm(generator=generator, output_type="numpy").images
+        image = ddpm(generator=generator, output_type="np").images
 
         image_slice = image[0, -3:, -3:, -1]
 

tests/pipelines/deepfloyd_if/test_if.py

@@ -50,7 +50,7 @@ class IFPipelineFastTests(PipelineTesterMixin, IFPipelineTesterMixin, unittest.T
             "prompt": "A painting of a squirrel eating a burger",
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
 
         return inputs

tests/pipelines/deepfloyd_if/test_if_img2img.py

@@ -55,7 +55,7 @@ class IFImg2ImgPipelineFastTests(PipelineTesterMixin, IFPipelineTesterMixin, uni
             "image": image,
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
 
         return inputs

tests/pipelines/deepfloyd_if/test_if_img2img_superresolution.py

@@ -57,7 +57,7 @@ class IFImg2ImgSuperResolutionPipelineFastTests(PipelineTesterMixin, IFPipelineT
             "original_image": original_image,
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
 
         return inputs

tests/pipelines/deepfloyd_if/test_if_inpainting.py

@@ -57,7 +57,7 @@ class IFInpaintingPipelineFastTests(PipelineTesterMixin, IFPipelineTesterMixin,
             "mask_image": mask_image,
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
 
         return inputs

tests/pipelines/deepfloyd_if/test_if_inpainting_superresolution.py

@@ -59,7 +59,7 @@ class IFInpaintingSuperResolutionPipelineFastTests(PipelineTesterMixin, IFPipeli
             "mask_image": mask_image,
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
 
         return inputs

tests/pipelines/deepfloyd_if/test_if_superresolution.py

@@ -52,7 +52,7 @@ class IFSuperResolutionPipelineFastTests(PipelineTesterMixin, IFPipelineTesterMi
             "image": image,
             "generator": generator,
             "num_inference_steps": 2,
-            "output_type": "numpy",
+            "output_type": "np",
         }
 
         return inputs