update cap

multimodal/open_flamingo/chat/conversation.py

@@ -425,6 +425,7 @@ class Chat:
         if radio in ["Cap"]:
             output_text, out_image = captioner(self.model, self.tokenizer, image_ori, batch_images, input_ids,
                                                attention_mask, image_start_index_list, image_nums, added_bbox_list)
+
         else:
             with torch.inference_mode():
                 text_outputs = self.model.generate(
@@ -477,6 +478,7 @@ class Chat:
             print(output_text)
             output_text = re.findall(r'Assistant:(.+)', output_text)[-1]
             print(output_text)
+        print(output_text)
 
         return output_text, out_image