remove changes from pipeline_stable_diffusion as part of imagic pipeline

Author: MarkRich

examples/community/imagic_stable_diffusion.py

@@ -2,9 +2,9 @@
     modeled after the textual_inversion.py / train_dreambooth.py and the work
     of justinpinkney here: https://github.com/justinpinkney/stable-diffusion/blob/main/notebooks/imagic.ipynb
 """
+import inspect
 import warnings
 from typing import List, Optional, Union
-import inspect
 
 import numpy as np
 import torch
@@ -18,9 +18,10 @@
 from diffusers.pipelines.stable_diffusion import StableDiffusionPipelineOutput
 from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
 from diffusers.schedulers import DDIMScheduler, LMSDiscreteScheduler, PNDMScheduler
+from diffusers.utils import logging
 from tqdm.auto import tqdm
 from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
-from diffusers.utils import logging
+
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
 
@@ -119,7 +120,6 @@ def disable_attention_slicing(self):
         # set slice_size = `None` to disable `attention slicing`
         self.enable_attention_slicing(None)
 
-
     def train(
         self,
         prompt: Union[str, List[str]],
@@ -379,14 +379,13 @@ def __call__(
         """
         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 self.text_embeddings is None: 
+        if self.text_embeddings is None:
             raise ValueError("Please run the pipe.train() before trying to generate an image.")
         if self.text_embeddings_orig is None:
             raise ValueError("Please run the pipe.train() before trying to generate an image.")
 
         text_embeddings = alpha * self.text_embeddings_orig + (1 - alpha) * self.text_embeddings
 
-
         # here `guidance_scale` is defined analog to the guidance weight `w` of equation (2)
         # of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1`
         # corresponds to doing no classifier free guidance.
@@ -487,4 +486,4 @@ def __call__(
         if not return_dict:
             return (image, has_nsfw_concept)
 
-        return StableDiffusionPipelineOutput(images=image, nsfw_content_detected=has_nsfw_concept)
+        return StableDiffusionPipelineOutput(images=image, nsfw_content_detected=has_nsfw_concept)

src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py

@@ -174,7 +174,6 @@ def __call__(
         return_dict: bool = True,
         callback: Optional[Callable[[int, int, torch.FloatTensor], None]] = None,
         callback_steps: Optional[int] = 1,
-        text_embeddings: Optional[torch.FloatTensor] = None,
         **kwargs,
     ):
         r"""
@@ -266,9 +265,6 @@ def __call__(
             )
             text_input_ids = text_input_ids[:, : self.tokenizer.model_max_length]
 
-        if text_embeddings is None:
-            text_embeddings = self.text_encoder(text_input_ids.to(self.device))[0]
-
         # duplicate text embeddings for each generation per prompt, using mps friendly method
         bs_embed, seq_len, _ = text_embeddings.shape
         text_embeddings = text_embeddings.repeat(1, num_images_per_prompt, 1)