Modified files

README.md

@@ -1,5 +1,5 @@
 ---
-title: Agent Agents Course
+title: Agents Course
 emoji: 📚
 colorFrom: green
 colorTo: yellow

agent.py

@@ -66,6 +66,7 @@ def initialize_agent():
             model=model,
             managed_agents=[web_agent],
             additional_authorized_imports=["time", "numpy", "pandas"],
+            system_prompt="""You are a general AI assistant. I will ask you a question. Report your thoughts, and finish your answer with the following template: FINAL ANSWER: [YOUR FINAL ANSWER]. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings. If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise. If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string."""
         )
         return manager_agent
     else:

app.py

@@ -1,19 +1,21 @@
+
 import gradio as gr
 import requests
+import re
 from agent import initialize_agent
 
 # 1. Initialize the agent
 manager_agent = initialize_agent()
 
 # 2. Define the API URL
-API_URL = "https://gaia-benchmark-dev.hf.space"
+API_URL = "https://agents-course-unit4-scoring.hf.space"
 
 # 3. Gradio interface
 def run_evaluation(hf_username, space_url):
     if not manager_agent:
-        return "Agent could not be initialized. Please check the model and token."
+        return {"error": "Agent could not be initialized. Please check the model and token."}
     if not hf_username or not space_url:
-        return "Please provide your Hugging Face username and Space URL."
+        return {"error": "Please provide your Hugging Face username and Space URL."}
 
     try:
         # Get questions
@@ -29,12 +31,19 @@ def run_evaluation(hf_username, space_url):
             prompt = question["question"]
             print(f"Answering question {i+1}/{len(questions)}: {prompt}")
             try:
-                answer = manager_agent.run(prompt)
-                answers.append({"question_id": question["id"], "answer": answer})
+                raw_answer = manager_agent.run(prompt)
+                # Extract the final answer
+                match = re.search(r"FINAL ANSWER: (.*)", raw_answer, re.IGNORECASE)
+                if match:
+                    answer = match.group(1).strip()
+                else:
+                    answer = raw_answer # Fallback to the raw answer if the pattern is not found
+                
+                answers.append({"task_id": question["id"], "submitted_answer": answer})
                 print(f"Got answer: {answer}")
             except Exception as e:
                 print(f"Error running agent on question {question['id']}: {e}")
-                answers.append({"question_id": question["id"], "answer": "Error: Could not generate an answer."})
+                answers.append({"task_id": question["id"], "submitted_answer": "Error: Could not generate an answer."})
 
         # Submit answers
         print("Submitting answers...")
@@ -43,7 +52,7 @@ def run_evaluation(hf_username, space_url):
             json={
                 "hf_username": hf_username,
                 "answers": answers,
-                "space_url": space_url
+                "agent_code": space_url
             }
         )
         submission_response.raise_for_status()
@@ -51,10 +60,10 @@ def run_evaluation(hf_username, space_url):
         return submission_response.json()
     except requests.exceptions.RequestException as e:
         print(f"An error occurred: {e}")
-        return f"An error occurred: {e}"
+        return {"error": f"An error occurred: {e}"}
     except Exception as e:
         print(f"An unexpected error occurred: {e}")
-        return f"An unexpected error occurred: {e}"
+        return {"error": f"An unexpected error occurred: {e}"}
 
 with gr.Blocks() as demo:
     gr.Markdown("# GAIA Benchmark Evaluation with smolagent")
@@ -72,4 +81,4 @@ with gr.Blocks() as demo:
     )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()