Update app.py

app.py

@@ -140,11 +140,13 @@ def response_audio(state:AppState, progress=gr.Progress(track_tqdm=True)):
             return state, synthesize_speech("Đã chuyển sang chế độ " + ("vẽ" if state.painting else "nói chuyện"))
         if state.painting is True:
             promptx = prompt_hugingface(textin,"Hugging Face","Qwen/Qwen2.5-72B-Instruct","Medium")
+            imgtxt=""
             if state.image_in:
                 img=resize(state.image_in)
+                imgtxt = creator_prompt.img2text(img)
             else:
                 img=None
-            state.image_out = render.generate_images(promptx,img,progress)
+            state.image_out = render.generate_images(imgtxt+promptx,img,progress)
             audio_bytes = synthesize_speech("Bạn thấy tôi vẽ "+textin+" có đẹp không")
             return state, audio_bytes
         else:
@@ -170,11 +172,13 @@ def response_text(state:AppState,textin,image:Image, prompt,progress=gr.Progress
         if state.painting is True:
             state.conversation.append({"role": "user", "content":"Bạn: " + textin})
             #state.image_out = generate_image(textin, image, streng, ckpt,guidance)
+            imgtxt=""
             if image:
                 img=resize(image)
+                imgtxt = creator_prompt.img2text(img)
             else:
                 img=None
-            image_out = render.generate_images(textin,img,progress)
+            image_out = render.generate_images(imgtxt+textin,img,progress)
             state.image_out = image_out
             audio_bytes = synthesize_speech("Bạn thấy tôi vẽ "+prompt+" có đẹp không")
             return state, audio_bytes