Update app.py

app.py

@@ -1,240 +1,42 @@
-import requests
 import os
 import gradio as gr
-import inspect
+import requests
 import pandas as pd
 import time
 import re
 
-# === TOOLS SECTION (scalone z helper.py) ===
-
-from langchain_experimental.utilities import PythonREPL
-from langchain.tools import Tool
-
-# 1. Python REPL Tool
-python_repl = PythonREPL()
-repl_tool = Tool(
-    name="python_repl",
-    description="""
-    A Python REPL (Read-Eval-Print Loop) for executing Python code.
-    Use this tool for:
-    - Performing accurate calculations (arithmetic, complex math).
-    - Manipulating and analyzing data (e.g., lists, numbers).
-    - Executing small, self-contained Python scripts.
-    Input MUST be valid Python code, and all outputs must be printed.
-    """,
-    func=python_repl.run,
-)
-
-# 2. File Saver Tool
-def download_and_save_file(args: dict) -> str:
-    """
-    Downloads a file from a given URL and saves it to a specified local filename.
-    Input: JSON string with 'url' and 'local_filename' keys.
-    Example: {"url": "https://example.com/data.xlsx", "local_filename": "data.xlsx"}
-    """
-    try:
-        if isinstance(args, str):
-            import json
-            args = json.loads(args)
-        url = args.get("url")
-        local_filename = args.get("local_filename")
-        if not url or not local_filename:
-            return "Error: Both 'url' and 'local_filename' must be provided."
-        response = requests.get(url, stream=True, timeout=30)
-        response.raise_for_status()
-        os.makedirs(os.path.dirname(local_filename) or '.', exist_ok=True)
-        with open(local_filename, 'wb') as f:
-            for chunk in response.iter_content(chunk_size=8192):
-                f.write(chunk)
-        return f"File downloaded successfully to {local_filename}"
-    except Exception as e:
-        return f"An unexpected error occurred: {e}"
-
-file_saver_tool = Tool(
-    name="file_saver",
-    description="Downloads a file from a URL and saves it to a specified local filename. Input: JSON with 'url' and 'local_filename'.",
-    func=download_and_save_file,
-)
-
-# 3. Audio Transcriber Tool
-import speech_recognition as sr
-from pydub import AudioSegment
-
-def transcribe_audio_from_path(local_audio_path: str, language: str = "en-US") -> str:
-    """
-    Transcribes audio content from a local file path to text.
-    Only local file paths. Converts to WAV if needed.
-    """
-    r = sr.Recognizer()
-    temp_wav_path = "temp_audio_to_transcribe.wav"
-    transcribed_text = ""
-    try:
-        if local_audio_path.startswith("http://") or local_audio_path.startswith("https://"):
-            return "Error: This tool only accepts local file paths, not URLs. Please use 'file_saver' first."
-        if not os.path.exists(local_audio_path):
-            return f"Error: Local audio file not found at '{local_audio_path}'."
-        audio = AudioSegment.from_file(local_audio_path)
-        audio.export(temp_wav_path, format="wav")
-        with sr.AudioFile(temp_wav_path) as source:
-            audio_listened = r.record(source)
-            try:
-                transcribed_text = r.recognize_google(audio_listened, language=language)
-            except sr.UnknownValueError:
-                return "Could not understand audio (speech not clear or too short)."
-            except sr.RequestError as e:
-                return f"Could not request results from Google Speech Recognition service; {e}"
-    except Exception as e:
-        return f"An unexpected error occurred during audio processing or transcription: {e}"
-    finally:
-        if os.path.exists(temp_wav_path):
-            os.remove(temp_wav_path)
-    return transcribed_text.strip()
-
-audio_transcriber_tool = Tool(
-    name="audio_transcriber_tool",
-    description=(
-        "Transcribes audio content from a **local file path** to a text transcript. "
-        "Use for extracting spoken information from audio recordings downloaded using 'file_saver'."
-    ),
-    func=transcribe_audio_from_path,
-)
-
-# 4. Gemini Multimodal Tool (for images)
-import base64
-from langchain.tools import Tool
-from langchain_google_genai import ChatGoogleGenerativeAI
-from langchain_core.messages import HumanMessage
-
-def analyze_image_with_gemini(args: dict) -> str:
-    """
-    Analyzes an image using Gemini Multimodal LLM to answer a given question.
-    Input: JSON with 'image_path' and 'question'.
-    """
-    try:
-        if isinstance(args, str):
-            import json
-            args = json.loads(args)
-        image_path = args.get("image_path")
-        question = args.get("question")
-        if not image_path or not question:
-            return "Error: Both 'image_path' and 'question' must be provided."
-        if not os.path.exists(image_path):
-            return f"Error: Local image file not found at '{image_path}'."
-        google_api_key = os.getenv("GOOGLE_API_KEY")
-        if not google_api_key:
-            return "Error: GOOGLE_API_KEY not found in environment variables for multimodal tool."
-        llm = ChatGoogleGenerativeAI(
-            model="gemini-2.0-flash",
-            google_api_key=google_api_key,
-            temperature=0.0
-        )
-        with open(image_path, "rb") as f:
-            image_bytes = f.read()
-            image_base64 = base64.b64encode(image_bytes).decode('utf-8')
-        message = HumanMessage(
-            content=[
-                {"type": "text", "text": question},
-                {"type": "image_url", "image_url": {"url": f"data:image/png;base64,{image_base64}"}},
-            ]
-        )
-        response = llm.invoke([message])
-        return response.content
-    except Exception as e:
-        return f"Error in gemini_multimodal_tool: {e}"
-
-gemini_multimodal_tool = Tool(
-    name="gemini_multimodal_tool",
-    description="Analyze an image with Gemini LLM. Input: JSON with 'image_path' and 'question'.",
-    func=analyze_image_with_gemini,
-)
-
-# 5. Wikipedia Search Tool
-from langchain_community.document_loaders import WikipediaLoader
-
-def wiki_search(query: str) -> str:
-    """Search Wikipedia for a query and return up to 2 results."""
-    search_docs = WikipediaLoader(query=query, load_max_docs=2).load()
-    formatted_search_docs = "\n\n---\n\n".join(
-        [
-            f'<Document source="{doc.metadata.get("source", "")}" page="{doc.metadata.get("page", "")}"/>\n{doc.page_content}\n</Document>'
-            for doc in search_docs
-        ])
-    return formatted_search_docs
-
-wikipedia_search_tool2 = Tool(
-    name="wikipedia_search_tool2",
-    description="Search Wikipedia for a query and return up to 2 results.",
-    func=wiki_search,
-)
-
-# ========== END TOOLS SECTION ==========
-
-# --- AGENT SECTION ---
-from langchain_google_genai import ChatGoogleGenerativeAI
-from langchain.memory import ConversationSummaryMemory
+from langchain_openai import ChatOpenAI
 from langchain.prompts import PromptTemplate
 from langchain.agents import AgentExecutor, create_react_agent
+from langchain.memory import ConversationSummaryMemory
 from typing import List, Optional
 
-DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
+# === TOOL IMPORTS ===
+from helper import repl_tool, file_saver_tool, audio_transcriber_tool, gemini_multimodal_tool, wikipedia_search_tool2
 
-# Set up LLM (Google Gemini - requires GOOGLE_API_KEY env variable)
-google_api_key = os.getenv("GOOGLE_API_KEY")
-if not google_api_key:
-    raise RuntimeError("GOOGLE_API_KEY not found in environment. Please set it.")
-
-gemini_model = "gemini-2.0-flash"
-llm_client = ChatGoogleGenerativeAI(
-    model=gemini_model,
-    google_api_key=google_api_key,
-    temperature=0,
-)
-summary_memory = ConversationSummaryMemory(llm=llm_client, memory_key="chat_history")
+# Constants
+DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# Prompt
+# --- Prompt ---
 prompt = PromptTemplate(
     input_variables=["input", "agent_scratchpad", "chat_history", "tool_names"],
     template="""
-You are a smart and helpful AI Agent/Assistant that excels at fact-based reasoning. You are allowed and encouraged to use one or more tools as needed to answer complex questions and perform tasks.
-Your FINAL ANSWER must be one of these formats and ONLY the answer itself (no intro phrases):
-- A number (e.g., '26', '1977', '519')
-- As few words as possible (e.g., 'Paris', 'down', 'LUX')
-- A comma-separated list of numbers and/or strings (e.g., '10,20,30', 'apple,banana,orange')
----
-Previous conversation history:
-{chat_history}
-New input: {input}
----
-{agent_scratchpad}
-"""
-)
-
-tools = [repl_tool, file_saver_tool, audio_transcriber_tool, gemini_multimodal_tool, wikipedia_search_tool2]
-
-summary_llm = ChatGoogleGenerativeAI(
-    model=gemini_model,
-    google_api_key=google_api_key,
-    temperature=0,
-    streaming=True
-)
-
-summary_react_agent = create_react_agent(
-    llm=summary_llm,
-    tools=tools,
-    prompt=prompt
+    You are a smart and helpful AI Agent/Assistant that excels at fact-based reasoning. You are allowed and encouraged to use one or more tools as needed to answer complex questions and perform tasks.
+    [ ...cut for brevity: insert your strict format rules and examples here ... ]
+    {chat_history}
+    New input: {input}
+    ---
+    {agent_scratchpad}
+    """
 )
 
+# === AGENT DEFINITION ===
 class BasicAgent:
     def __init__(
         self,
-        agent,
-        tools: List,
-        verbose: bool = False,
-        handle_parsing_errors: bool = True,
-        max_iterations: int = 9,
-        memory: Optional[ConversationSummaryMemory] = None
-    ) -> None:
+        agent, tools: List, verbose: bool = False, handle_parsing_errors: bool = True,
+        max_iterations: int = 9, memory: Optional[ConversationSummaryMemory] = None
+    ):
         self.agent = agent
         self.tools = tools
         self.verbose = verbose
@@ -249,6 +51,7 @@ class BasicAgent:
             max_iterations=self.max_iterations,
             memory=self.memory
         )
+
     def __call__(self, question: str) -> str:
         result = self.agent_obj.invoke(
             {"input": question},
@@ -256,10 +59,11 @@ class BasicAgent:
         )
         return result['output']
 
-def run_and_submit_all( profile: gr.OAuthProfile | None):
+def run_and_submit_all(profile: gr.OAuthProfile | None):
     space_id = os.getenv("SPACE_ID")
+
     if profile:
-        username= f"{profile.username}"
+        username = f"{profile.username}"
         print(f"User logged in: {username}")
     else:
         print("User not logged in.")
@@ -269,10 +73,42 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
     questions_url = f"{api_url}/questions"
     submit_url = f"{api_url}/submit"
 
-    agent = BasicAgent(summary_react_agent, tools, True, True, 30, summary_memory)
+    # OpenAI API key only!
+    openai_api_key = os.getenv("OPENAI_API_KEY")
+    if not openai_api_key:
+        print("OpenAI API key not found in environment variables.")
+        return "OpenAI API key not found. Please set OPENAI_API_KEY environment variable.", None
+
+    # Use GPT-4o (or another allowed OpenAI model)
+    llm_client = ChatOpenAI(model='gpt-4o', temperature=0, api_key=openai_api_key)
+
+    # Tools: only offline/tools not requiring other APIs
+    tools = [
+        repl_tool,
+        file_saver_tool,
+        audio_transcriber_tool,
+        gemini_multimodal_tool,   # If this is purely local or adapted for OpenAI images, otherwise remove!
+        wikipedia_search_tool2
+    ]
+
+    summary_memory = ConversationSummaryMemory(llm=llm_client, memory_key="chat_history")
+
+    summary_react_agent = create_react_agent(
+        llm=llm_client,
+        tools=tools,
+        prompt=prompt
+    )
+
+    # 1. Instantiate Agent
+    try:
+        agent = BasicAgent(summary_react_agent, tools, True, True, 30, summary_memory)
+    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(agent_code)
 
+    # 2. Fetch Questions
     print(f"Fetching questions from: {questions_url}")
     try:
         response = requests.get(questions_url, timeout=15)
@@ -283,8 +119,10 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
              return "Fetched questions list is empty or invalid format.", None
         print(f"Fetched {len(questions_data)} questions.")
     except Exception as e:
+        print(f"Error fetching questions: {e}")
         return f"Error fetching questions: {e}", None
 
+    # 3. Run your Agent
     results_log = []
     answers_payload = []
     print(f"Running agent on {len(questions_data)} questions...")
@@ -294,22 +132,26 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
         file_name = item.get("file_name")
         full_question_for_agent = question_text
         if file_name:
-            attachment_url = f"{api_url}/files/{task_id}"
+            attachment_url = f"{DEFAULT_API_URL}/files/{task_id}"
             full_question_for_agent += f"\n\nAttachment '{file_name}' available at EXACT URL: {attachment_url}"
-            print(f"Running agent on task {task_id}: {full_question_for_agent}",flush=True)
+            print(f"Running agent on task {task_id}: {full_question_for_agent}", flush=True)
         try:
             submitted_answer = agent(full_question_for_agent)
             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})
-            time.sleep(5)  # for demo, zmień na 60 przy real eval!
+            time.sleep(2)  # Decrease or remove if not rate-limited!
         except Exception as e:
             print(f"Error running agent on task {task_id}: {e}")
             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})
 
     if not answers_payload:
+        print("Agent did not produce any answers to submit.")
         return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
 
     submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
+    status_update = f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..."
+    print(status_update)
+
     print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
     try:
         response = requests.post(submit_url, json=submission_data, timeout=60)
@@ -322,29 +164,27 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
             f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
             f"Message: {result_data.get('message', 'No message received.')}"
         )
-        cleaned_final_status = re.sub(r'[^\x20-\x7E\n\r\t]+', '', final_status)
-        cleaned_final_status = cleaned_final_status.strip()
+        cleaned_final_status = re.sub(r'[^\x20-\x7E\n\r\t]+', '', final_status).strip()
         results_df = pd.DataFrame(results_log)
         return cleaned_final_status, results_df
     except Exception as e:
-        status_message = f"An unexpected error occurred during submission: {e}"
+        print(f"Error submitting answers: {e}")
         results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return f"Submission Failed: {e}", results_df
 
+# --- Build Gradio Interface using Blocks ---
 with gr.Blocks() as demo:
     gr.Markdown("# Basic Agent Evaluation Runner")
     gr.Markdown(
         """
         **Instructions:**
-        1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
-        2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
-        3.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
+        1. Log in to your Hugging Face account using the button below.
+        2. Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
         ---
-        **Disclaimers:**
-        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
-        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution.
+        **Note:** Only OpenAI API key is needed!
         """
     )
+
     gr.LoginButton()
     run_button = gr.Button("Run Evaluation & Submit All Answers")
     status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)