Merge pull request #122 from Reboot-Video/main

Caption to Phrase Grounding: return parsed data
2025-03-11 14:01:12 +02:00
parent 77230ddf79 841a4ae736
commit f38eeb8997
1 changed files with 4 additions and 1 deletions
@@ -454,7 +454,10 @@ class Florence2Run:
                out_tensor = annotated_image_tensor[:3, :, :].unsqueeze(0).permute(0, 2, 3, 1).cpu().float()
                out.append(out_tensor)
               
-                out_data.append(bboxes)
+                if task == 'caption_to_phrase_grounding':
+                    out_data.append(parsed_answer[task_prompt])
+                else:
+                    out_data.append(bboxes)

                
                pbar.update(1)