[CHANGES]

app-withshortcut.py

@@ -0,0 +1,234 @@
+import os
+import gradio as gr
+import requests
+import inspect
+import pandas as pd
+from dotenv import load_dotenv
+
+from myagent import BasicAgent  # Import your agent class from myagent.py
+from multiagents import MultiAgent
+
+from phoenix.otel import register
+from openinference.instrumentation.smolagents import SmolagentsInstrumentor
+
+# use space_host var to determine if running in HF space or locally, if so register local instrumentation
+space_host_startup = os.getenv("SPACE_HOST")
+if not space_host_startup:
+    register()
+    SmolagentsInstrumentor().instrument()
+
+# (Keep Constants as is)
+# --- Constants ---
+DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
+load_dotenv()
+
+max_questions = 20
+
+# known answer, already solved, to avoid computation cost
+known_answers = {
+    "f918266a-b3e0-4914-865d-4faa564f1aef": "0",
+    "a1e91b78-d3d8-4675-bb8d-62741b4b68a6": "3",
+    "2d83110e-a098-4ebb-9987-066c06fa42d0": "right",
+    "8e867cd7-cff9-4e6c-867a-ff5ddc2550be": "3",
+    "9d191bce-651d-4746-be2d-7ef8ecadb9c2": "extremely",
+    # Add more known answers as needed
+}
+
+
+
+def run_and_submit_all(nb_questions: int, profile: gr.OAuthProfile | None): 
+    """
+    Fetches all questions, runs my Agent on them, submits all answers,
+    and displays the results.
+    """
+    # --- Determine HF Space Runtime URL and Repo URL ---
+    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code   
+
+    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"
+    file_url = f"{api_url}/files"    
+    submit_url = f"{api_url}/submit"
+
+    # 1. Instantiate Agent 
+    try:
+        # agent = BasicAgent()
+        agent = MultiAgent()
+    except Exception as e:
+        print(f"Error instantiating agent: {e}")
+        return f"Error initializing agent: {e}", None
+    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
+    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)
+        response.raise_for_status()
+        questions_data = response.json()
+        if not questions_data:
+             print("Fetched questions list is empty.")
+             return "Fetched questions list is empty or invalid format.", None
+        print(f"Fetched {len(questions_data)} questions.")
+    except requests.exceptions.RequestException as e:
+        print(f"Error fetching questions: {e}")
+        return f"Error fetching questions: {e}", None
+    except requests.exceptions.JSONDecodeError as e:
+         print(f"Error decoding JSON response from questions endpoint: {e}")
+         print(f"Response text: {response.text[:500]}")
+         return f"Error decoding server response for questions: {e}", None
+    except Exception as e:
+        print(f"An unexpected error occurred fetching questions: {e}")
+        return f"An unexpected error occurred fetching questions: {e}", None
+
+    # 3. Run your Agent
+    results_log = []
+    answers_payload = []
+
+    # for testing keep only some questions
+    questions_data = questions_data[:nb_questions]
+
+    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")
+        file_name = item.get("file_name")
+        file_question_url = None
+        if file_name:
+            file_question_url = f"{file_url}/{task_id}"
+        if not task_id or question_text is None:
+            print(f"Skipping item with missing task_id or question: {item}")
+            continue
+        try:
+            agent_question = question_text
+            if file_question_url:
+                agent_question += f"\n\nFile URL: {file_question_url}"
+
+            shortcut = known_answers.get(task_id)
+            if shortcut:
+                submitted_answer = shortcut
+            else:
+                submitted_answer = agent(agent_question)
+            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:
+             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)
+
+    # 4. Prepare Submission 
+    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)
+
+    # 5. Submit
+    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
+    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.')}"
+        )
+        print("Submission successful.")
+        results_df = pd.DataFrame(results_log)
+        return final_status, results_df
+    except requests.exceptions.HTTPError as e:
+        error_detail = f"Server responded with status {e.response.status_code}."
+        try:
+            error_json = e.response.json()
+            error_detail += f" Detail: {error_json.get('detail', e.response.text)}"
+        except requests.exceptions.JSONDecodeError:
+            error_detail += f" Response: {e.response.text[:500]}"
+        status_message = f"Submission Failed: {error_detail}"
+        print(status_message)
+        results_df = pd.DataFrame(results_log)
+        return status_message, results_df
+    except requests.exceptions.Timeout:
+        status_message = "Submission Failed: The request timed out."
+        print(status_message)
+        results_df = pd.DataFrame(results_log)
+        return status_message, results_df
+    except requests.exceptions.RequestException as e:
+        status_message = f"Submission Failed: Network error - {e}"
+        print(status_message)
+        results_df = pd.DataFrame(results_log)
+        return status_message, results_df
+    except Exception as e:
+        status_message = f"An unexpected error occurred during submission: {e}"
+        print(status_message)
+        results_df = pd.DataFrame(results_log)
+        return status_message, 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.
+
+        ---
+        **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. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
+        """
+    )
+
+    gr.LoginButton()
+
+    nb_questions = gr.Number(value=20)
+
+    run_button = gr.Button("Run Evaluation & Submit All Answers")
+
+    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
+    # Removed max_rows=10 from DataFrame constructor
+    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
+
+    run_button.click(
+        fn=run_and_submit_all,
+        inputs=[nb_questions],
+        outputs=[status_output, results_table]
+    )
+
+if __name__ == "__main__":
+    print("\n" + "-"*30 + " App Starting " + "-"*30)
+    # Check for SPACE_HOST and SPACE_ID at startup for information
+    space_host_startup = os.getenv("SPACE_HOST")
+    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup
+
+    if space_host_startup:
+        print(f"✅ SPACE_HOST found: {space_host_startup}")
+        print(f"   Runtime URL should be: https://{space_host_startup}.hf.space")
+    else:
+        print("ℹ️  SPACE_HOST environment variable not found (running locally?).")
+
+    if space_id_startup: # Print repo URLs if SPACE_ID is found
+        print(f"✅ SPACE_ID found: {space_id_startup}")
+        print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
+        print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
+    else:
+        print("ℹ️  SPACE_ID environment variable not found (running locally?). Repo URL cannot be determined.")
+
+    print("-"*(60 + len(" App Starting ")) + "\n")
+
+    print("Launching Gradio Interface for Basic Agent Evaluation...")
+    demo.launch(debug=True, share=False)

common/mylogger.py

@@ -0,0 +1,41 @@
+import time
+
+def save_file_with_timestamp(content: str, file_name: str, extension: str) -> str:
+    """
+    Save content to a file with a timestamp.
+    Args:
+        content (str): The content to save.
+        file_name (str): The base name of the file.
+    Returns:
+        str: The path to the saved file.
+    """
+    try:
+        # save content to a file in test folder before returning
+        # compute filepath with correct extension based on convert_to_markdown and add a timestamp for unicity
+        
+        unicity_suffix = str(int(time.time()))
+        
+        file_path = f"test/{file_name}_{unicity_suffix}.{extension}"
+        with open(file_name, "w", encoding="utf-8") as f:
+            f.write(content)
+    except Exception as e:
+        print(f"Error saving content to file: {e}")
+    return file_name
+
+
+def mylog(agent_name: str, message: str, depth: int = 0) -> None:
+    """
+    Log a message with indentation based on the depth.
+    Args:
+        agent_name (str): The name of the agent.
+        message (str): The message to log.
+        depth (int): The depth of the log message.
+    """
+    indent = " " * (depth * 4)
+    try:            
+        # log agent call in file
+        with open("logs/agent_calls.log", "a") as log_file:
+            log_file.write(f"{indent}{agent_name}: {message}\n")
+    except Exception as e:
+            print(f"Error logging agent call: {e}")            
+    

env.example

@@ -0,0 +1,3 @@
+HF_TOKEN=<<your hf token>>
+SPACE_ID=<<your hf sapce id>>
+OPENAI_API_KEY=<<your open ai key>>

multiagents.py

@@ -0,0 +1,137 @@
+
+# a multi agent proposal to solve HF agent course final assignment
+import os
+import dotenv
+from smolagents import CodeAgent
+from smolagents import OpenAIServerModel
+from tools.fetch import fetch_webpage, search_web
+from smolagents import PythonInterpreterTool
+from tools.yttranscript import get_youtube_transcript, get_youtube_title_description
+from tools.stt import get_text_transcript_from_audio_file
+from tools.image import analyze_image
+from common.mylogger import mylog
+import myprompts
+
+dotenv.load_dotenv()
+
+gemini_model = OpenAIServerModel(
+    model_id="gemini-2.0-flash",
+    api_key=os.environ["GEMINI_API_KEY"],
+    # Google Gemini OpenAI-compatible API base URL
+    api_base="https://generativelanguage.googleapis.com/v1beta/openai/",
+)
+
+vllm_model = OpenAIServerModel(
+    model_id="Qwen/Qwen2.5-1.5B-Instruct",
+    api_base="http://192.168.1.39:18000/v1",  
+    api_key="token-abc123",  
+)
+
+openai_41nano_model = OpenAIServerModel(
+    model_id="gpt-4.1-nano",
+    api_base="https://api.openai.com/v1",
+    api_key=os.environ["OPENAI_API_KEY"],
+)
+
+openai_41mini_model = OpenAIServerModel(
+    model_id="gpt-4.1-mini",
+    api_base="https://api.openai.com/v1",
+    api_key=os.environ["OPENAI_API_KEY"],
+)
+
+
+def check_final_answer(final_answer, agent_memory)  -> bool:
+    """
+    Check if the final answer is correct.
+    basic check on the length of the answer.
+    """
+    mylog("check_final_answer", final_answer)
+    # if return answer is more than 200 characters, we will assume it is not correct    
+    if len(str(final_answer)) > 200:
+        return False
+    else:
+        return True
+
+
+web_agent = CodeAgent(
+    model=openai_41nano_model,
+    tools=[
+        search_web,
+        fetch_webpage,                
+    ],
+    name="web_agent",
+    description="Use search engine to find webpages related to a subject and get the page content",
+    additional_authorized_imports=["pandas", "numpy","bs4"],
+    verbosity_level=1,    
+    max_steps=7,
+)
+
+audiovideo_agent = CodeAgent(
+    model=openai_41nano_model,
+    tools=[
+        get_youtube_transcript,
+        get_youtube_title_description,
+        get_text_transcript_from_audio_file,
+        analyze_image
+    ],
+    name="audiovideo_agent",
+    description="Extracts information from image, video or audio files from the web",
+    additional_authorized_imports=["pandas", "numpy","bs4", "requests"],
+    verbosity_level=1,
+    max_steps=7,
+)
+
+
+
+manager_agent = CodeAgent(
+    model=openai_41mini_model,
+    tools=[ PythonInterpreterTool()],
+    managed_agents=[web_agent, audiovideo_agent],    
+    additional_authorized_imports=["pandas", "numpy","bs4"],
+    planning_interval=5,
+    verbosity_level=2,
+    final_answer_checks=[check_final_answer],
+    max_steps=15,
+    name="manager_agent",
+    description="A manager agent that coordinates the work of other agents to answer questions.",
+)
+
+class MultiAgent:
+    def __init__(self):
+        print("BasicAgent initialized.")
+
+    def __call__(self, question: str) -> str:
+        mylog(self.__class__.__name__, question)        
+
+        try:
+            prefix = """You are the top agent of a multi-agent system that can answer questions by coordinating the work of other agents.
+            You will receive a question and you will decide which agent to use to answer it.
+            You can use the web_agent to search the web for information and for fetching the content of a web page, or the audiovideo_agent to extract information from video or audio files.
+            You can also use your own knowledge to answer the question.
+            You need to respect the output format that is given to you.
+            Finding the correct answer to the question need reasoning and plannig, read the question carrefully, think step by step and do not skip any steps.
+            """
+
+            question = prefix + "\nTHE QUESTION:\n" + question + '\n' + myprompts.output_format
+
+            fixed_answer = ""
+
+            fixed_answer = manager_agent.run(question)
+            
+            return fixed_answer
+        except Exception as e:
+            error = f"An error occurred while processing the question: {e}"
+            print(error)
+            return error
+
+
+if __name__ == "__main__":
+    # Example usage
+   
+    question = """
+What was the actual enrollment of the Malko competition in 2023?
+"""
+    agent = MultiAgent()
+    answer = agent(question)
+    print(f"Answer: {answer}")
+    

myprompts.py

@@ -0,0 +1,14 @@
+review_prompt = """You are a reviewer agent. You will be given a question. 
+Your task it to assert if a LLM agent with access to web content can answer the question or not, or if a coding agent and more tools is needed to answer the question.
+If the question is too complex for a LLM agent, you should return "code" as the answer else you should return "model".
+"""
+
+model_prompt = """You are a LLM agent. You will be given a question.
+Your task it to answer the question using the tools you have access to.
+take time to analyse the steps to answer the question.
+if a filename is given in the question you can infer the url 
+"""
+
+output_format = """OUTPUT FORMAT should be a number OR as few words as possible OR a comma separated list of numbers and/or strings.
+ If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise, use digits only.
+ If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string."""

requirements.txt

@@ -1,7 +1,14 @@
 gradio
+gradio[oauth]
 requests
-pandas
-smolagents
+python-dotenv
+smolagents 
 duckduckgo-search
-openai
-instructor>=1.2.3
+smolagents[openai]
+markdownify
+beautifulsoup4
+transformers
+smolagents[transformers]
+smolagents[audio]
+smolagents[telemetry,toolkit]
+youtube_transcript_api

tools/fetch.py

@@ -0,0 +1,100 @@
+# this is  asmolagent too to fetch html content from a url
+from smolagents import tool
+import requests
+from markdownify import markdownify as md
+from bs4 import BeautifulSoup
+from common.mylogger import save_file_with_timestamp, mylog
+
+@tool
+def fetch_webpage(url: str, convert_to_markdown: bool = True) -> str:
+    """
+    Fetches the HTML content of a given URL. 
+    if markdown conversion is enabled, it will remove script and style and return the text content as markdown else return raw unfiltered HTML
+    Args:
+        url (str): The URL to fetch.
+        convert_to_markdown (bool): If True, convert the HTML content to Markdown format. else return the raw HTML.
+    Returns:
+        str: The HTML content of the URL.
+    """    
+    content = None
+    response = requests.get(url, timeout=30)
+    if (convert_to_markdown):
+        soup = BeautifulSoup(response.text, "html.parser")
+        # remove script and style tags
+        for script in soup(["script", "style"]):
+            script.extract()
+
+        # for wikipedia only keep the main content
+        if "wikipedia.org" in url:
+            main_content = soup.find("main",{"id":"content"})
+            if main_content:
+                content = md(str(main_content),strip=['script', 'style'], heading_style="ATX").strip()
+            else:
+                content = md(response.text,strip=['script', 'style'], heading_style="ATX").strip()
+    else:
+        content = response.text
+    
+    save_file_with_timestamp(content, "webpage", ".md" if convert_to_markdown else ".html")
+           
+    return content
+
+@tool
+# this tool allow web search on a local SearXNG instance
+def search_web(query: str, num_results: int = 5) -> list:
+    """
+    Perform a web search using local SearXNG instance.
+    Args:
+        query (str): The search query.
+        num_results (int): The number of results to return.
+    Returns:
+        list: A list of search results sorted by score with {url, title, content, score} for each result.
+    """
+    # local metaserach engine searxng, run on localhost:8888
+    searxng_url = "http://localhost:8888/search"
+    params = {"q": query, "format": 'json'}
+    response = requests.get(searxng_url, params=params)
+    if response.status_code == 200:
+        ret = response.json()
+        # keep only the response'results' array
+        results = ret.get("results", [])
+        # keep only the first num_results   
+        results = results[:num_results]
+        # for each result keep only the url, title and content ans score
+        results = [
+            {
+                "url": result.get("url"),
+                "title": result.get("title"),
+                "content": result.get("content"),
+                "score": result.get("score"),
+            }
+            for result in results
+        ]
+        
+        return results
+        
+    else:
+        print(f"Error: {response.status_code}")
+        return []
+
+if __name__ == "__main__":
+   
+    try:
+        # Test the function
+        query = "What is the capital of France?"
+        results = search_web(query,3)
+        print(results)
+    except Exception as e:
+        print(f"An error occurred: {e}")
+
+    try:
+        # Test the function
+        video_id = "L1vXCYZAYYM"  # Replace with your YouTube video ID
+        video_url = "https://www.youtube.com/watch?v=" + video_id
+        url = "https://en.wikipedia.org/wiki/Malko_Competition"
+        # page_content = fetch_webpage(video_url)
+        page_content = fetch_webpage(url, convert_to_markdown=True)
+        print(page_content.encode("utf-8"))
+    except Exception as e:
+        print(f"An error occurred: {e}")
+ 
+   

tools/image.py

@@ -0,0 +1,40 @@
+# image analyzer using open ai model
+
+from smolagents import tool
+from openai import OpenAI
+import dotenv
+dotenv.load_dotenv()
+
+@tool
+def analyze_image(question: str, image_url: str) -> str:
+    """
+    Analyze an image using OpenAI's API.
+    Args:
+        question (str): The question to ask about the image. eg. "What is in this image?"
+        image_url (str): The URL of the image to analyze.
+    """
+    client = OpenAI()
+
+    response = client.responses.create(
+        model="gpt-4o-mini",
+        input=[
+            {
+                "role": "user",
+                "content": [
+                    { "type": "input_text", "text": f"{question}" },
+                    {
+                        "type": "input_image",
+                        "image_url": f"{image_url}",
+                    }
+                ]
+            }
+        ]
+    )
+
+    return response
+
+if __name__ == "__main__":
+    question = "What is the main subject of this image?"
+    image_url = "https://agents-course-unit4-scoring.hf.space/files/cca530fc-4052-43b2-b130-b30968d8aa44"
+    answer = analyze_image(question, image_url)
+    print(f"Answer: {answer}")

tools/stt.py

@@ -0,0 +1,72 @@
+import io
+import os  
+import shutil
+import subprocess
+import requests
+import uuid
+from smolagents import tool
+import dotenv
+dotenv.load_dotenv()
+
+@tool
+def get_text_transcript_from_audio_file(file_url: str, language: str = "en-US") -> str:
+    """
+    Convert speech to text using local whisper model.
+    This function downloads an audio file from a given URL, converts it to WAV format if necessary,
+    then use whisper model to transcribe the audio to text.
+    
+    Args:
+        file_url (str): The URL of the audio file to transcribe.        
+        language (str): The language code for the transcription. Default is "en-US".
+        
+    Returns:
+        str: The transcribed text.
+    """
+
+    file_name = uuid.uuid4().hex +".mp3"
+
+    dest_folder = os.getenv("STT_FOLDER")
+    if not dest_folder:
+        dest_folder = '.'
+    file_path = os.path.join(dest_folder + "\\tmp", file_name)
+    # 1. download the file from url (in pure python without wget or curl)
+    if not os.path.exists(file_name):            
+        response = requests.get(file_url)
+        if response.status_code == 200:
+            with open(file_path, "wb") as f:
+                f.write(response.content)
+        else:
+            raise Exception(f"Error downloading file: {response.status_code}")
+    
+    # 2. if it is a mp3 convert to wav with ffmpeg exec
+    if file_name.endswith(".mp3"):
+        cmd = f"ffmpeg -i {file_path} -ac 1 -ar 16000 -c:a pcm_s16le {file_path[:-4]}.wav"
+        cmd_as_list = cmd.split()
+        subprocess.run(cmd_as_list, cwd=dest_folder, check=True)
+        file_path = file_path[:-4] + ".wav"
+        file_name = file_name[:-4] + ".wav"
+
+    # 3. copy file to data folder
+    shutil.copy2(file_path, os.path.join(dest_folder, "testdata/"))
+    
+
+    # 4. call docker run  command
+    docker_command = f"""
+        docker run 
+        -v {dest_folder}/models:/app/models 
+        -v {dest_folder}/testdata:/app/testdata 
+        ghcr.io/appleboy/go-whisper:latest 
+        --model /app/models/ggml-small.bin 
+        --audio-path /app/testdata/{file_name}
+    """
+
+    subprocess.run(docker_command.split(), cwd=dest_folder, check=True)
+    # 5. cat the output file an return it   
+    output_filepath = os.path.join(dest_folder, "testdata", f"{file_name[:-4]}.txt")
+    with open(output_filepath, "r") as f:
+        text = f.read()
+    return text
+
+if __name__ == "__main__":
+    transcript = get_text_transcript_from_audio_file("https://agents-course-unit4-scoring.hf.space/files/99c9cc74-fdc8-46c6-8f8d-3ce2d3bfeea3", )
+    print(transcript)   

tools/yttranscript.py

@@ -0,0 +1,72 @@
+
+from smolagents import tool
+from youtube_transcript_api import YouTubeTranscriptApi
+from bs4 import BeautifulSoup
+from tools.fetch import fetch_webpage
+
+@tool
+# a function to get youtube transcript from video id
+def get_youtube_transcript(video_id: str) -> str:
+    """
+    Fetches the transcript of a YouTube video given its video ID.    
+    Args:
+        video_id (str): The ID of the YouTube video. Pass in the video ID, NOT the video URL. For a video with the URL https://www.youtube.com/watch?v=12345 the ID is 12345.
+    Returns:
+        str: The transcript of the YouTube video. as a single string with each line separated by a newline character.
+    """
+    # Initialize the YouTubeTranscriptApi
+    ytt_api = YouTubeTranscriptApi()
+    fetched_transcript = ytt_api.fetch(video_id)
+    raw_data = fetched_transcript.to_raw_data()
+    # raw data is in the form of [{        'text': 'Hey there',        'start': 0.0,        'duration': 1.54    },    {        'text': 'how are you',,        'start': 1.54,        'duration': 4.16    },   ... ] we will return ony the text element as lines
+    transcript = "\n".join([item['text'] for item in raw_data])
+    return transcript
+
+
+@tool
+# a function to get video title and description from video url
+def get_youtube_title_description(video_url: str) -> str:
+    """
+    Fetches the title and description of a YouTube video given its video ID.    
+    Args:
+        video_url (str): The url of the YouTube video.
+    Returns:
+        str: The title and description of the YouTube video.
+    """
+    # Initialize the YouTube object
+    soup = BeautifulSoup(fetch_webpage(video_url, convert_to_markdown=False), "html.parser")
+    # Extract the title by looking at the meta tag with name="title" and getting the content
+    metatitle = soup.find("meta", {"name": "title"})
+    if metatitle is not None:
+        title = metatitle["content"]
+    else:
+        title = "No title found"
+
+    # same for description
+    metadescription = soup.find("meta", {"name": "description"})    
+    if metadescription is not None:
+        description = metadescription["content"]
+    else:
+        description = "No description found"    
+    
+    return f"Title: {title}\nDescription: {description}"
+
+
+if __name__ == "__main__":
+    from dotenv import load_dotenv
+    load_dotenv
+    # Test the function
+    video_id = "1htKBjuUWec"  # Replace with your YouTube video ID
+    video_url = "https://www.youtube.com/watch?v=" + video_id
+    # Get the title and description
+    try:
+        title_description = get_youtube_title_description(video_url)
+        print(title_description)
+    except Exception as e:
+        print(f"Error fetching title and description: {e}")
+    
+    try:
+        transcript = get_youtube_transcript(video_id)
+    except Exception as e:
+        print(f"Error fetching transcript: {e}")
+    print(transcript)

vllm_asopenai_test.py

@@ -0,0 +1,17 @@
+import os
+from smolagents import OpenAIServerModel, CodeAgent
+
+# test file in cas I need to run entirely locally
+
+model = OpenAIServerModel(
+    model_id="Qwen/Qwen2.5-1.5B-Instruct",
+    api_base="http://192.168.1.39:18000/v1",  
+    api_key="token-abc123",  
+)
+
+myagent = CodeAgent(
+    model=model,
+    tools=[])
+
+result = myagent.run("Hello who are you?")  # Replace with your question
+print(result)

vllm_test.py

@@ -0,0 +1,18 @@
+from openai import OpenAI
+
+# test file in cas I need to run entirely locally
+# this test comptaibility with open ai
+
+client = OpenAI(
+    base_url="http://192.168.1.39:18000/v1",
+    api_key="token-abc123",
+)
+
+completion = client.chat.completions.create(
+  model="Qwen/Qwen2.5-1.5B-Instruct",
+  messages=[
+    {"role": "user", "content": "Hello!"}
+  ]
+)
+
+print(completion.choices[0].message)