add pipeline_class_name argument to Stable Diffusion conversion script (#4461)

* add pipeline class

* Apply suggestions from code review

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* style

---------

Co-authored-by: yiyixuxu <yixu310@gmail,com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

scripts/convert_original_stable_diffusion_to_diffusers.py

@@ -15,6 +15,7 @@
 """ Conversion script for the LDM checkpoints. """
 
 import argparse
+import importlib
 
 import torch
 
@@ -133,8 +134,22 @@ if __name__ == "__main__":
         required=False,
         help="Set to a path, hub id to an already converted vae to not convert it again.",
     )
+    parser.add_argument(
+        "--pipeline_class_name",
+        type=str,
+        default=None,
+        required=False,
+        help="Specify the pipeline class name",
+    )
+
     args = parser.parse_args()
 
+    if args.pipeline_class_name is not None:
+        library = importlib.import_module("diffusers")
+        class_obj = getattr(library, args.pipeline_class_name)
+    else:
+        pipeline_class = None
+
     pipe = download_from_original_stable_diffusion_ckpt(
         checkpoint_path=args.checkpoint_path,
         original_config_file=args.original_config_file,
@@ -152,6 +167,7 @@ if __name__ == "__main__":
         clip_stats_path=args.clip_stats_path,
         controlnet=args.controlnet,
         vae_path=args.vae_path,
+        pipeline_class=pipeline_class,
     )
 
     if args.half: