display image

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+example_images/1.jpg filter=lfs diff=lfs merge=lfs -text

app.py

@@ -250,8 +250,10 @@ def model_inference(input_dict, history):
             print(raw_result)
             proc_img = raw_result
             all_images += [proc_img]
-            # complete_assistant_response_for_gradio += [(proc_img, "Visual Operation Result")]
-            # yield complete_assistant_response_for_gradio # Update Gradio display
+            proc_img.save("tmp.png")
+            display = [dict(text="", files=["tmp.png"])]
+            complete_assistant_response_for_gradio = complete_assistant_response_for_gradio + display
+            yield complete_assistant_response_for_gradio # Update Gradio display
             
             new_piece = dict(role='user', content=[
                                     dict(type='text', text="\nHere is the cropped image (Image Size: {}x{}):".format(proc_img.size[0], proc_img.size[1])),
@@ -275,9 +277,12 @@ with gr.Blocks() as demo:
 
     gr.HTML(html_header)
     
+    # image_op_display = gr.Image(label="Visual Operation Result", type="pil", height=480, show_download_button=True, interactive=False)
+
     gr.ChatInterface(
         fn=model_inference,
         description="# **Pixel Reasoner**",
+        chatbot=gr.Chatbot(label="Conversation", layout="bubble", bubble_full_width=False, show_copy_button=True, height=600),
         examples=examples,
         fill_height=True,
         textbox=gr.MultimodalTextbox(label="Query Input", file_types=["image"], file_count="multiple"),
@@ -290,4 +295,4 @@ with gr.Blocks() as demo:
     gr.Markdown(learn_more_markdown)
     gr.Markdown(bibtext)
 
-demo.launch(debug=True, share=True)
+demo.launch(debug=True, share=False)