Update app.py

app.py

@@ -53,22 +53,39 @@ def delay_execution(agent):
     time.sleep(5)
 
 # Agent stuff starts here
+
 class BasicAgent:
     def __init__(self):
+        # Initialize a VLM for Image understanding
+        vlmodel = InferenceClientModel(model_id='Qwen/Qwen2.5-VL-72B-Instruct')
+        
+        # Create the agent
+        self.vlm_agent = CodeAgent(
+            tools=[],
+            model=vlmodel,
+            additional_authorized_imports=[],
+            step_callbacks=[],
+            max_steps=5,
+            verbosity_level=2,
+            name='image_understanding_agent',
+            description='Gives textual descriptions of images.'
+        )
+
         # If the agent does not answer, the model is overloaded, please use another model or the following Hugging Face Endpoint that also contains qwen2.5 coder:
         # model_id='https://pflgm2locj2t89co.us-east-1.aws.endpoints.huggingface.cloud' 
         model = InferenceClientModel(
-            #model_id='Qwen/Qwen2.5-VL-72B-Instruct',
-            #provider='nebius',
+            model_id='deepseek-ai/DeepSeek-R1-0528',
+            provider='nebius',
             token=os.getenv('HF_Token'))
         
         self.agent = CodeAgent(
             model=model,
             tools=[FinalAnswerTool(), BroadfieldWebsearch(), VisitWebpageTool()], ## add your tools here (don't remove final answer),
-            additional_authorized_imports=['pandas', 'requests', 'markdownify'],
+            additional_authorized_imports=['pandas', 'requests', 'markdownify', 'openpyxl'],
             max_steps=10,
             verbosity_level=1,
-            step_callbacks=[delay_execution]
+            step_callbacks=[delay_execution],
+            managed_agents=[self.vlm_agent]
         )