Update app.py

app.py

@@ -3,42 +3,30 @@ import gradio as gr
 import requests
 import pandas as pd
 import json
-import torch
-from transformers import AutoTokenizer, AutoModelForCausalLM
+from transformers import pipeline
 
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# --- Smart Agent using Mistral-7B ---
-class SmartLLMAgent:
+# --- Hugging Face Agent using Transformers Pipeline ---
+class HFTransformersAgent:
     def __init__(self):
-        print("🔁 Loading SmartLLMAgent with Mistral-7B...")
-        self.tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
-        self.model = AutoModelForCausalLM.from_pretrained(
-            "HuggingFaceH4/zephyr-7b-beta",
-            torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
-            device_map="auto"
+        print("🤖 Loading text-generation agent...")
+        self.generator = pipeline(
+            "text-generation",
+            model="tiiuae/falcon-7b-instruct",
+            max_new_tokens=100,
+            do_sample=True,
+            temperature=0.7,
+            top_p=0.9
         )
 
     def __call__(self, question: str) -> str:
-        prompt = f"Answer the following question clearly and precisely:\n\nQuestion: {question}\n\nAnswer:"
-        inputs = self.tokenizer(prompt, return_tensors="pt", truncation=True, max_length=2048)
-        inputs = {k: v.to(self.model.device) for k, v in inputs.items()}
-
-        with torch.no_grad():
-            outputs = self.model.generate(
-                **inputs,
-                max_new_tokens=150,
-                do_sample=True,
-                top_p=0.9,
-                temperature=0.7,
-                pad_token_id=self.tokenizer.eos_token_id
-            )
-
-        decoded = self.tokenizer.decode(outputs[0], skip_special_tokens=True)
-        final_answer = decoded.split("Answer:")[-1].strip()
-        print(f"✅ Generated answer: {final_answer}")
-        return final_answer or "Not Enough Information"
+        prompt = f"Answer the following question:\n\n{question}\n\nAnswer:"
+        result = self.generator(prompt)[0]["generated_text"]
+        answer = result.split("Answer:")[-1].strip()
+        print(f"✅ Generated answer: {answer}")
+        return answer if answer else "Not Enough Information"
 
 # --- Run Agent & Submit ---
 def run_and_submit_all(profile: gr.OAuthProfile | None):
@@ -46,39 +34,35 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
 
     if profile:
         username = f"{profile.username}"
-        print(f"👤 User logged in: {username}")
+        print(f"👤 Logged in: {username}")
     else:
-        return "Please Login to Hugging Face with the button.", None
+        return "Please log in to Hugging Face using the button.", None
 
+    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
     questions_url = f"{DEFAULT_API_URL}/questions"
     submit_url = f"{DEFAULT_API_URL}/submit"
 
     try:
-        agent = SmartLLMAgent()
+        agent = HFTransformersAgent()
     except Exception as e:
-        return f"❌ Error loading agent: {e}", None
-
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(f"🔗 Agent Code: {agent_code}")
+        return f"❌ Error initializing agent: {e}", None
 
     try:
         response = requests.get(questions_url, timeout=15)
         response.raise_for_status()
-        questions = response.json()
-        print(f"📥 Fetched {len(questions)} questions.")
+        questions_data = response.json()
+        print(f"📥 Fetched {len(questions_data)} questions.")
     except Exception as e:
-        return f"❌ Failed to fetch questions: {e}", None
+        return f"❌ Error fetching questions: {e}", None
 
-    answers_payload = []
     results_log = []
+    answers_payload = []
 
-    for item in questions:
+    for item in questions_data:
         task_id = item.get("task_id")
         question_text = item.get("question")
-
         if not task_id or not question_text:
             continue
-
         try:
             answer = agent(question_text)
             answers_payload.append({"task_id": task_id, "submitted_answer": answer})
@@ -87,7 +71,7 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"ERROR: {e}"})
 
     if not answers_payload:
-        return "❌ No answers produced.", pd.DataFrame(results_log)
+        return "❌ Agent did not produce any answers.", pd.DataFrame(results_log)
 
     submission_data = {
         "username": username.strip(),
@@ -95,7 +79,7 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
         "answers": answers_payload
     }
 
-    print("📤 Submission Payload:")
+    print("📤 Submission preview:")
     print(json.dumps(submission_data, indent=2)[:1000])
 
     try:
@@ -115,13 +99,13 @@ def run_and_submit_all(profile: gr.OAuthProfile | None):
 
 # --- Gradio UI ---
 with gr.Blocks() as demo:
-    gr.Markdown("# 🧠 Smart Agent Evaluation Runner (Mistral-7B)")
+    gr.Markdown("# 🤖 Smart Agent Runner (HF Transformers)")
     gr.Markdown(
         """
         ✅ **Instructions:**
-        1. Login to your Hugging Face account.
-        2. Click **"Run Evaluation & Submit All Answers"**.
-        3. Wait 1–2 mins. You need at least **30% correct** to pass Level 1 and get the certificate.
+        1. Log in to your Hugging Face account.
+        2. Click **“Run Evaluation & Submit All Answers”**.
+        3. Wait ~1–2 mins. You need at least **30% correct** to unlock the certificate.
         """
     )
 
@@ -133,5 +117,5 @@ with gr.Blocks() as demo:
     run_button.click(fn=run_and_submit_all, outputs=[status_output, results_table])
 
 if __name__ == "__main__":
-    print("🚀 Launching Smart Agent with Mistral-7B...")
+    print("🚀 Launching agent...")
     demo.launch(debug=True, share=False)