Fix errors

app.py

@@ -1,5 +1,6 @@
 import os
 import tempfile
+from base64 import b64encode
 from contextlib import suppress
 from io import BytesIO
 from pprint import pprint
@@ -31,6 +32,8 @@ from langchain_tavily import TavilySearch
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
+model = ChatOpenAI(model="gpt-4o", temperature=0)
+
 
 class State(MessagesState):
     question: str
@@ -48,8 +51,7 @@ class BasicAgent:
         ]
 
         # Chat model with tool support
-        self.model = ChatOpenAI(model="gpt-4o", temperature=0)
-        self.model_with_tools = self.model.bind_tools(self.tools, parallel_tool_calls=False)
+        self.model_with_tools = model.bind_tools(self.tools, parallel_tool_calls=False)
 
         # LangGraph
         self.graph = StateGraph(State)
@@ -72,7 +74,8 @@ class BasicAgent:
         1. **Thought:** Briefly outline your reasoning step.  
         2. **Reflect:** Check “Did I use all observations? Did my tool call succeed?”  
         3. **Action:** Either call a tool (with arguments) or prepare your final answer.  
-        4. **Final Answer:** Provide only the bare result (no labels, no extra text).
+        4. **Final Answer:** Provide only the bare result (no labels, no extra text, no thoughts, no reflection, no "Final Answer" string in the result). For question that contain phrases like `what is the number` or 
+        `what is the highest number` return just the number, e.g., 2.
         
         **Answer Format Rules**  
         - If the answer is a number, output digits only (no commas, no units, no strings like “one”, “twenty three”).  
@@ -109,7 +112,7 @@ class BasicAgent:
             question=question, 
             messages=[sys_msg, HumanMessage(content=question)]
         )
-        config = RunnableConfig(recursion_limit=10)
+        config = RunnableConfig(recursion_limit=15)
         result = self.compiled_graph.invoke(state, config)
         final_answer = result["messages"][-1].content
         print(f"\nFinal Answer: {final_answer}")
@@ -175,13 +178,23 @@ class BasicAgent:
             return f"Error: file not found at {path}"
 
         print("File metadata:", os.stat(path))
-        img = Image.open(path)
-        img_bytes = BytesIO()
-        img.save(img_bytes, format=img.format)
-        img_bytes.seek(0)
+            
+        # img = Image.open(path)
+        # img_bytes = BytesIO()
+        # img.save(img_bytes, format=img.format)
+        # img_bytes.seek(0)
+
+        with open("photo.png","rb") as f:
+            b64 = b64encode(f.read()).decode()
     
         vision = ChatOpenAI(model="gpt-4o-vision", temperature=0)
-        result = vision.analyze_image(img_bytes, question)
+        msg = HumanMessage(content=[
+            {"type":"text", "text": question},
+            {"type":"image_base64", "image_base64": {"data": b64}}
+        ])
+        response = vision.invoke([SystemMessage(content="Analyze the image."), msg])
+        result = response.content
+        print("Result:", result)
         return result
 
     @staticmethod
@@ -216,7 +229,7 @@ class BasicAgent:
 
         Please answer briefly based on this transcript, and give only the answer."""
         
-        response = self.model(completion_kwargs={"max_tokens": 200})(prompt)
+        response = model(completion_kwargs={"max_tokens": 200})(prompt)
         answer = response.choices[0].text.strip()
     
         return answer[:max_chars]