Update app.py

app.py

@@ -6,7 +6,7 @@ from datetime import datetime
 from smolagents import Tool, ToolCallingAgent
 from smolagents.models import InferenceClientModel
 
-# --- Constants ---
+# Constants
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
 # --- Custom Tools ---
@@ -16,13 +16,13 @@ class CalculatorTool(Tool):
     input_schema = {
         "expression": {
             "type": "string",
-            "description": "Mathematical expression to evaluate (e.g., '2+2')"
+            "description": "Math expression to evaluate (e.g. '2+2')"
         }
     }
     output_schema = {
         "result": {
             "type": "string",
-            "description": "The calculated result of the expression"
+            "description": "The result of the expression"
         }
     }
 
@@ -30,26 +30,26 @@ class CalculatorTool(Tool):
         try:
             return {"result": str(eval(expression))}
         except Exception as e:
-            return {"result": f"Error: {str(e)}"}
+            return {"result": f"Error: {e}"}
+
 
 class TimeTool(Tool):
     name = "current_time"
     description = "Gets current UTC time"
-    input_schema = {}  # No input
+    input_schema = {}
     output_schema = {
         "time": {
             "type": "string",
-            "description": "Current UTC time (YYYY-MM-DD HH:MM:SS)"
+            "description": "Current time in UTC"
         }
     }
 
     def use(self) -> dict:
         return {"time": datetime.utcnow().strftime("%Y-%m-%d %H:%M:%S UTC")}
 
-# --- Agent Setup ---
+# --- Agent ---
 class LocalAgent:
     def __init__(self):
-        print("Initializing agent...")
         self.tools = [CalculatorTool(), TimeTool()]
         self.agent = ToolCallingAgent(
             tools=self.tools,
@@ -61,17 +61,16 @@ class LocalAgent:
 
     def __call__(self, question: str) -> str:
         question_lower = question.lower()
-        if any(op in question_lower for op in ["calculate", "what is", "+", "-", "*", "/"]):
+        if any(op in question_lower for op in ["calculate", "+", "-", "*", "/", "what is"]):
             return CalculatorTool().use(question.replace("?", ""))["result"]
         if "time" in question_lower:
             return TimeTool().use()["time"]
-
         try:
             return str(self.agent.run(question))
         except Exception as e:
             return f"Error: {e}"
 
-# --- Evaluation Logic ---
+# --- Evaluation Function ---
 def run_and_submit_all(profile: gr.OAuthProfile | None):
     if not profile:
         return "Please login first.", None
@@ -84,22 +83,22 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
         questions = requests.get(f"{api_url}/questions", timeout=15).json()
 
         answers = []
-        results = []
+        logs = []
 
         for q in questions:
             try:
-                answer = agent(q["question"])
+                ans = agent(q["question"])
                 answers.append({
                     "task_id": q["task_id"],
-                    "submitted_answer": answer
+                    "submitted_answer": ans
                 })
-                results.append({
+                logs.append({
                     "Task ID": q["task_id"],
                     "Question": q["question"],
-                    "Answer": answer
+                    "Answer": ans
                 })
             except Exception as e:
-                results.append({
+                logs.append({
                     "Task ID": q["task_id"],
                     "Question": q["question"],
                     "Answer": f"Error: {e}"
@@ -116,29 +115,31 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
         return (
             f"✅ Score: {result.get('score', 'N/A')}%\n"
             f"Correct: {result.get('correct_count', '?')}/{result.get('total_attempted', '?')}",
-            pd.DataFrame(results)
+            pd.DataFrame(logs)
         )
 
     except Exception as e:
-        return f"Evaluation failed: {str(e)}", pd.DataFrame([])
+        return f"Evaluation failed: {e}", pd.DataFrame([])
 
 # --- Gradio UI ---
-with gr.Blocks(title="Agent Evaluation Runner") as app:
+with gr.Blocks(title="Agent Evaluation") as app:
     gr.Markdown("## 🤖 Agent Evaluation with smolagents")
-    gr.Markdown("Login, then run the evaluation to test your agent.")
+    gr.Markdown("Login, then click 'Run Evaluation' to test your agent.")
 
-    gr.LoginButton()  # OAuth Login UI
+    gr.LoginButton()               # Login button visible
+    profile = gr.OAuthProfile()   # Profile input (not visible)
 
     run_btn = gr.Button("🚀 Run Evaluation")
-    output = gr.Textbox(label="Result Summary")
-    results_table = gr.DataFrame(label="Answers and Logs")
+    output = gr.Textbox(label="Evaluation Result")
+    results_table = gr.DataFrame(label="Answer Log")
 
     run_btn.click(
         fn=run_and_submit_all,
-        inputs=[],  # No need to pass profile manually
+        inputs=[profile],          # ✅ Must include this
         outputs=[output, results_table]
     )
 
 if __name__ == "__main__":
     app.launch()
 
+