[advanced dreambooth lora sdxl training script] load pipeline for inference...

[advanced dreambooth lora sdxl training script] load pipeline for inference only if validation prompt is used (#6171)

* load pipeline for inference only if validation prompt is used

* move things outside

* load pipeline for inference only if validation prompt is used

* fix readme when validation prompt is used

---------
Co-authored-by: linoytsaban <linoy@huggingface.co>
Co-authored-by: apolinário <joaopaulo.passos@gmail.com>

examples/advanced_diffusion_training/train_dreambooth_lora_sdxl_advanced.py

@@ -112,7 +112,7 @@ def save_model_card(
     repo_folder=None,
     vae_path=None,
 ):
-    img_str = "widget:\n" if images else ""
+    img_str = "widget:\n"
     for i, image in enumerate(images):
         image.save(os.path.join(repo_folder, f"image_{i}.png"))
         img_str += f"""
@@ -121,6 +121,10 @@ def save_model_card(
             url:
                 "image_{i}.png"
         """
+    if not images:
+        img_str += f"""
+        - text: '{instance_prompt}'
+        """
 
     trigger_str = f"You should use {instance_prompt} to trigger the image generation."
     diffusers_imports_pivotal = ""
@@ -157,8 +161,6 @@ tags:
 base_model: {base_model}
 instance_prompt: {instance_prompt}
 license: openrail++
-widget:
-    - text: '{validation_prompt if validation_prompt else instance_prompt}'
 ---
 """
 
@@ -2010,7 +2012,8 @@ def main(args):
             text_encoder_lora_layers=text_encoder_lora_layers,
             text_encoder_2_lora_layers=text_encoder_2_lora_layers,
         )
-
+        images = []
+        if args.validation_prompt and args.num_validation_images > 0:
             # Final inference
             # Load previous pipeline
             vae = AutoencoderKL.from_pretrained(
@@ -2045,8 +2048,6 @@ def main(args):
             pipeline.load_lora_weights(args.output_dir)
 
             # run inference
-        images = []
-        if args.validation_prompt and args.num_validation_images > 0:
             pipeline = pipeline.to(accelerator.device)
             generator = torch.Generator(device=accelerator.device).manual_seed(args.seed) if args.seed else None
             images = [