Update app.py

app.py

@@ -1,158 +1,66 @@
-import os
-import requests
-import gradio as gr
-import pandas as pd
-from smolagents import SmolAgent
-from smolagents.models import InferenceClientModel
-from smolagents.tools import Tool
-from duckduckgo_search import DDGS
-
-# --- Constants ---
-DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
-
-# --- Custom Search Tool ---
-class CustomSearchTool(Tool):
-    name = "duckduckgo_search"
-    description = "A simple tool that performs web search using DuckDuckGo."
-    inputs = {"query": str}
-    outputs = str
-
-    def run(self, query: str) -> str:
-        print(f"[SearchTool] Searching DuckDuckGo for: {query}")
-        try:
-            with DDGS() as ddgs:
-                results = ddgs.text(query, max_results=3)
-                output = "\n".join([r["body"] for r in results if "body" in r])
-                return output.strip() or "No results found."
-        except Exception as e:
-            print(f"[SearchTool] Search failed: {e}")
-            return f"Search error: {e}"
-
-# --- BasicAgent Class ---
-class BasicAgent(SmolAgent):
-    def __init__(self):
-        model = InferenceClientModel("mistralai/Mixtral-8x7B-Instruct-v0.1")
-        tools = [CustomSearchTool()]
-        super().__init__(model=model, tools=tools)
-
-# --- Submission + Evaluation ---
-def run_and_submit_all(profile: gr.OAuthProfile | None):
-    space_id = os.getenv("SPACE_ID")
-
-    if profile:
-        username = f"{profile.username}"
-        print(f"User logged in: {username}")
-    else:
-        print("User not logged in.")
-        return "Please Login to Hugging Face with the button.", None
-
-    api_url = DEFAULT_API_URL
-    questions_url = f"{api_url}/questions"
-    submit_url = f"{api_url}/submit"
-
-    try:
-        agent = BasicAgent()
-    except Exception as e:
-        print(f"Error instantiating agent: {e}")
-        return f"Error initializing agent: {e}", None
-
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(f"Agent code: {agent_code}")
-
-    # Fetch questions
-    print(f"Fetching questions from: {questions_url}")
-    try:
-        response = requests.get(questions_url, timeout=15)
-        response.raise_for_status()
-        questions_data = response.json()
-        if not questions_data:
-            return "Fetched questions list is empty or invalid format.", None
-    except Exception as e:
-        return f"Error fetching questions: {e}", None
-
-    results_log = []
-    answers_payload = []
-
-    print(f"Running agent on {len(questions_data)} questions...")
-    for item in questions_data:
-        task_id = item.get("task_id")
-        question_text = item.get("question")
-        if not task_id or question_text is None:
-            continue
-        try:
-            submitted_answer = agent(question_text)
-            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:
-            results_log.append({
-                "Task ID": task_id,
-                "Question": question_text,
-                "Submitted Answer": f"AGENT ERROR: {e}"
-            })
+import datasets
+from langchain.docstore.document import Document
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.retrievers import BM25Retriever
+from smolagents import Tool, CodeAgent, InferenceClientModel
+
+# Define RetrieverTool exactly as in your example
+class RetrieverTool(Tool):
+    name = "retriever"
+    description = (
+        "Uses lexical search to retrieve the parts of transformers documentation most relevant to answer your query."
+    )
+    inputs = {
+        "query": {
+            "type": "string",
+            "description": "The query to perform. Use affirmative form rather than a question.",
+        }
+    }
+    output_type = "string"
 
-    if not answers_payload:
-        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
+    def __init__(self, docs, **kwargs):
+        super().__init__(**kwargs)
+        self.retriever = BM25Retriever.from_documents(docs, k=10)
 
-    # Submit
-    submission_data = {
-        "username": username.strip(),
-        "agent_code": agent_code,
-        "answers": answers_payload
-    }
-    try:
-        response = requests.post(submit_url, json=submission_data, timeout=60)
-        response.raise_for_status()
-        result_data = response.json()
-        final_status = (
-            f"Submission Successful!\n"
-            f"User: {result_data.get('username')}\n"
-            f"Overall Score: {result_data.get('score', 'N/A')}% "
-            f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
-            f"Message: {result_data.get('message', 'No message received.')}"
+    def forward(self, query: str) -> str:
+        assert isinstance(query, str), "Query must be a string"
+        docs = self.retriever.invoke(query)
+        return "\nRetrieved documents:\n" + "".join(
+            [f"\n\n===== Document {i} =====\n{doc.page_content}" for i, doc in enumerate(docs)]
         )
-        return final_status, pd.DataFrame(results_log)
-    except Exception as e:
-        return f"Submission Failed: {e}", pd.DataFrame(results_log)
 
-# --- Gradio UI ---
-with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
-    gr.Markdown(
-        """
-        **Instructions:**
-        1. Clone this space, then modify your agent logic, tools, and setup.
-        2. Log in to your Hugging Face account below.
-        3. Click "Run Evaluation & Submit All Answers" to start.
-        ---
-        **Note:** This is intentionally basic. Improve and optimize your agent for better scores.
-        """
+# Replace agent creation inside try block
+try:
+    # Load and prepare docs once here
+    knowledge_base = datasets.load_dataset("m-ric/huggingface_doc", split="train")
+    knowledge_base = knowledge_base.filter(lambda row: row["source"].startswith("huggingface/transformers"))
+
+    source_docs = [
+        Document(page_content=doc["text"], metadata={"source": doc["source"].split("/")[1]})
+        for doc in knowledge_base
+    ]
+
+    text_splitter = RecursiveCharacterTextSplitter(
+        chunk_size=500,
+        chunk_overlap=50,
+        add_start_index=True,
+        strip_whitespace=True,
+        separators=["\n\n", "\n", ".", " ", ""],
+    )
+    docs_processed = text_splitter.split_documents(source_docs)
+
+    # Instantiate RetrieverTool with processed docs
+    retriever_tool = RetrieverTool(docs_processed)
+
+    # Instantiate the smolagents CodeAgent with model
+    agent = CodeAgent(
+        tools=[retriever_tool],
+        model=InferenceClientModel(model_id="Qwen/Qwen2.5-Coder-32B-Instruct"),
+        max_steps=4,
+        verbosity_level=2,
+        stream_outputs=True,
     )
 
-    gr.LoginButton()
-
-    run_button = gr.Button("Run Evaluation & Submit All Answers")
-    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
-    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
-
-    run_button.click(fn=run_and_submit_all, outputs=[status_output, results_table])
-
-# --- Launch ---
-if __name__ == "__main__":
-    print("\n" + "-" * 30 + " App Starting " + "-" * 30)
-    space_host = os.getenv("SPACE_HOST")
-    space_id = os.getenv("SPACE_ID")
-
-    if space_host:
-        print(f"✅ SPACE_HOST found: {space_host}")
-        print(f"   Runtime URL: https://{space_host}.hf.space")
-    if space_id:
-        print(f"✅ SPACE_ID found: {space_id}")
-        print(f"   Repo URL: https://huggingface.co/spaces/{space_id}")
-    print("-" * 70 + "\n")
-
-    print("Launching Gradio Interface...")
-    demo.launch(debug=True, share=False)
+except Exception as e:
+    print(f"Error instantiating agent: {e}")
+    return f"Error initializing agent: {e}", None