Update app.py

app.py

@@ -4,26 +4,20 @@ import os
 import gradio as gr
 import requests
 import pandas as pd
+
 import google.generativeai as genai
 from smolagents import CodeAgent, DuckDuckGoSearchTool
+from smolagents.model.base import ModelOutput  # import ModelOutput if available
 
-# System prompt for Gemini
+# System prompt used by the agent
 SYSTEM_PROMPT = """You are a general AI assistant. I will ask you a question.
-Report your thoughts, and finish your answer with just the answer — no prefixes like \"FINAL ANSWER:\".
+Report your thoughts, and finish your answer with just the answer — no prefixes like "FINAL ANSWER:".
 Your answer should be a number OR as few words as possible OR a comma-separated list of numbers and/or strings.
 If you're asked for a number, don’t use commas or units like $ or %, unless specified.
 If you're asked for a string, don’t use articles or abbreviations (e.g. for cities), and write digits in plain text unless told otherwise."""
 
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# Generation result wrapper
-class GenerationResult:
-    def __init__(self, content, token_usage=None, input_tokens=0, output_tokens=0):
-        self.content = content
-        self.token_usage = token_usage or {}
-        self.input_tokens = input_tokens
-        self.output_tokens = output_tokens
-
 # Gemini model wrapper
 class GeminiFlashModel:
     def __init__(self, model_id="gemini-1.5-flash", api_key=None):
@@ -46,32 +40,42 @@ class GeminiFlashModel:
 
         try:
             response = self.model.generate_content(prompt)
-            return GenerationResult(
+            return ModelOutput(
                 content=response.text.strip(),
-                token_usage={},
                 input_tokens=0,
-                output_tokens=0
+                output_tokens=0,
+                token_usage={}
             )
         except Exception as e:
-            return GenerationResult(
+            return ModelOutput(
                 content=f"GENERATION ERROR: {e}",
-                token_usage={},
                 input_tokens=0,
-                output_tokens=0
+                output_tokens=0,
+                token_usage={}
             )
 
-# Agent class
+# Agent wrapper
 class MyAgent:
     def __init__(self):
-        self.model = GeminiFlashModel()
+        self.model = GeminiFlashModel(model_id="gemini-1.5-flash")
         self.agent = CodeAgent(tools=[DuckDuckGoSearchTool()], model=self.model)
 
-    def __call__(self, question: str):
-        return self.agent.run(question)
+    def __call__(self, question: str) -> str:
+        result = self.agent.run(question)
+        print(f"[DEBUG] Agent run result type: {type(result)}; value: {result}")
 
-# Evaluation function
+        # Return string content only
+        if hasattr(result, "content"):
+            return result.content
+        elif isinstance(result, dict):
+            return result.get("content", str(result))
+        else:
+            return str(result)
 
+# Main evaluation function
 def run_and_submit_all(profile: gr.OAuthProfile | None):
+    print("Starting run_and_submit_all...")
+
     space_id = os.getenv("SPACE_ID")
 
     if profile:
@@ -105,32 +109,19 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
         if not task_id or question_text is None:
             continue
 
+        print(f"Running agent on question: {question_text}")
         try:
-            result = agent(question_text)
-
-            if hasattr(result, "content"):
-                submitted_answer = result.content
-            elif isinstance(result, dict) and "content" in result:
-                submitted_answer = result["content"]
-            else:
-                submitted_answer = str(result)
-
-            answers_payload.append({
-                "task_id": task_id,
-                "submitted_answer": submitted_answer
-            })
-
-            results_log.append({
-                "Task ID": task_id,
-                "Question": question_text,
-                "Submitted Answer": submitted_answer
-            })
-
+            submitted_answer = agent(question_text)
+            print(f"Agent answer: {submitted_answer} (type: {type(submitted_answer)})")
+            answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
+            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})
         except Exception as e:
+            error_msg = f"AGENT ERROR: {e}"
+            print(error_msg)
             results_log.append({
                 "Task ID": task_id,
                 "Question": question_text,
-                "Submitted Answer": f"AGENT ERROR: {e}"
+                "Submitted Answer": error_msg
             })
 
     if not answers_payload:
@@ -157,7 +148,7 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
     except Exception as e:
         return f"Submission failed: {e}", pd.DataFrame(results_log)
 
-# UI setup
+# Gradio UI setup
 with gr.Blocks() as demo:
     gr.Markdown("# Basic Agent Evaluation Runner")
     gr.Markdown("""
@@ -183,3 +174,4 @@ if __name__ == "__main__":
 
 
 
+