fix img2img interrogate

modules/interrogate.py

@@ -80,12 +80,13 @@ class InterrogateModels:
         self.create_fake_fairscale()
         import models.blip # pylint: disable=no-name-in-module
         model_path = os.path.join(paths.models_path, "BLIP")
-        shared.log.debug(f'Loading interrogate model: type=BLIP folder={model_path}')
+        download_name='model_base_caption_capfilt_large.pth',
+        shared.log.debug(f'Model interrogate load: type=BLiP model={download_name} path={model_path}')
         files = modelloader.load_models(
             model_path=model_path,
             model_url='https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_base_caption_capfilt_large.pth',
             ext_filter=[".pth"],
-            download_name='model_base_caption_capfilt_large.pth',
+            download_name=download_name,
         )
         blip_model = models.blip.blip_decoder(pretrained=files[0], image_size=blip_image_eval_size, vit='base', med_config=os.path.join(paths.paths["BLIP"], "configs", "med_config.json")) # pylint: disable=c-extension-no-member
         blip_model.eval()
@@ -93,6 +94,7 @@ class InterrogateModels:
         return blip_model
 
     def load_clip_model(self):
+        shared.log.debug(f'Model interrogate load: type=CLiP model={clip_model_name} path={shared.opts.clip_models_path}')
         import clip
         if self.running_on_cpu:
             model, preprocess = clip.load(clip_model_name, device="cpu", download_root=shared.opts.clip_models_path)

modules/processing_diffusers.py

@@ -217,14 +217,14 @@ def process_diffusers(p: StableDiffusionProcessing):
         if 'prompt' in possible:
             if hasattr(model, 'text_encoder') and 'prompt_embeds' in possible and len(p.prompt_embeds) > 0 and p.prompt_embeds[0] is not None:
                 args['prompt_embeds'] = p.prompt_embeds[0]
-                if 'XL' in model.__class__.__name__:
-                    args['pooled_prompt_embeds'] = p.positive_pooleds[0]
+                if 'XL' in model.__class__.__name__ and len(getattr(p, 'negative_pooleds', [])) > 0:
+                    args['pooled_prompt_embeds'] = p.negative_pooleds[0]
             else:
                 args['prompt'] = prompts
         if 'negative_prompt' in possible:
             if hasattr(model, 'text_encoder') and 'negative_prompt_embeds' in possible and len(p.negative_embeds) > 0 and p.negative_embeds[0] is not None:
                 args['negative_prompt_embeds'] = p.negative_embeds[0]
-                if 'XL' in model.__class__.__name__:
+                if 'XL' in model.__class__.__name__ and len(getattr(p, 'negative_pooleds', [])) > 0:
                     args['negative_pooled_prompt_embeds'] = p.negative_pooleds[0]
             else:
                 args['negative_prompt'] = negative_prompts

modules/ui.py

@@ -111,6 +111,7 @@ def parse_style(styles):
 
 
 def process_interrogate(interrogation_function, mode, ii_input_files, ii_input_dir, ii_output_dir, *ii_singles):
+    mode = int(mode)
     if mode in {0, 1, 3, 4}:
         return [interrogation_function(ii_singles[mode]), None]
     if mode == 2: