[Tests] Lower required memory for clip guided and fix super edge-case git pipeline module bug (#754)

* [Tests] Lower required memory

* fix

* up

* uP

Author: patrickvonplaten

src/diffusers/dynamic_modules_utils.py

@@ -259,7 +259,8 @@ def get_cached_module_file(
                 local_files_only=local_files_only,
                 use_auth_token=False,
             )
-            submodule = "local"
+            submodule = "git"
+            module_file = pretrained_model_name_or_path + ".py"
         except EnvironmentError:
             logger.error(f"Could not locate the {module_file} inside {pretrained_model_name_or_path}.")
             raise
@@ -288,7 +289,7 @@ def get_cached_module_file(
     full_submodule = DIFFUSERS_DYNAMIC_MODULE_NAME + os.path.sep + submodule
     create_dynamic_module(full_submodule)
     submodule_path = Path(HF_MODULES_CACHE) / full_submodule
-    if submodule == "local":
+    if submodule == "local" or submodule == "git":
         # We always copy local files (we could hash the file to see if there was a change, and give them the name of
         # that hash, to only copy when there is a modification but it seems overkill for now).
         # The only reason we do the copy is to avoid putting too many folders in sys.path.

tests/test_pipelines.py

@@ -112,18 +112,22 @@ def test_local_custom_pipeline(self):
         assert output_str == "This is a local test"
 
     @slow
+    @unittest.skipIf(torch_device == "cpu", "Stable diffusion is supposed to run on GPU")
     def test_load_pipeline_from_git(self):
         clip_model_id = "laion/CLIP-ViT-B-32-laion2B-s34B-b79K"
 
         feature_extractor = CLIPFeatureExtractor.from_pretrained(clip_model_id)
-        clip_model = CLIPModel.from_pretrained(clip_model_id)
+        clip_model = CLIPModel.from_pretrained(clip_model_id, torch_dtype=torch.float16)
 
         pipeline = DiffusionPipeline.from_pretrained(
             "CompVis/stable-diffusion-v1-4",
             custom_pipeline="clip_guided_stable_diffusion",
             clip_model=clip_model,
             feature_extractor=feature_extractor,
+            torch_dtype=torch.float16,
+            revision="fp16",
         )
+        pipeline.enable_attention_slicing()
         pipeline = pipeline.to(torch_device)
 
         # NOTE that `"CLIPGuidedStableDiffusion"` is not a class that is defined in the pypi package of th e library, but solely on the community examples folder of GitHub under: