app.py

from smolagents import CodeAgent,DuckDuckGoSearchTool, HfApiModel,load_tool,tool
import datetime
import requests
import pytz
import yaml
from tools.final_answer import FinalAnswerTool

from Gradio_UI import GradioUI
verbose = True

# Below is an example of a tool that does nothing. Amaze us with your creativity !
@tool
def tell_me_a_joke(topic:int)-> str: #it's import to specify the return type
    """A tool that tells a joke for a given topic
    Args:
        topic: topic for the joke
        arg2: the second argument
    """
    return "Who's there? Bohoo!"

@tool
def get_current_time_in_timezone(timezone: str) -> str:
    """A tool that fetches the current local time in a specified timezone.
    Args:
        timezone: A string representing a valid timezone (e.g., 'America/New_York').
    """
    try:
        # Create timezone object
        tz = pytz.timezone(timezone)
        # Get current time in that timezone
        local_time = datetime.datetime.now(tz).strftime("%Y-%m-%d %H:%M:%S")
        return f"The current local time in {timezone} is: {local_time}"
    except Exception as e:
        return f"Error fetching time for timezone '{timezone}': {str(e)}"


MODEL_IDS = [
    #'https://wxknx1kg971u7k1n.us-east-1.aws.endpoints.huggingface.cloud/',
    #'https://jc26mwg228mkj8dw.us-east-1.aws.endpoints.huggingface.cloud/', 
    #'meta-llama/Llama-3.2-1B-Instruct', ## Does a poor job of interpreting my questions and matching them to the tools
    'Qwen/Qwen2.5-Coder-32B-Instruct',
    'Qwen/Qwen2.5-Coder-14B-Instruct',
    'Qwen/Qwen2.5-Coder-7B-Instruct',
    'Qwen/Qwen2.5-Coder-3B-Instruct',
    'Qwen/Qwen2.5-Coder-1.5B-Instruct'
    # Add here wherever model is working for you
]

def is_model_overloaded(model_url):
    """Verify if the model is overloaded doing a test call."""
    try:
        response = requests.post(model_url, json={"inputs": "Test"})
        if verbose: 
            print(response.status_code)
        if response.status_code == 503:  # 503 Service Unavailable = Overloaded
            return True
        if response.status_code == 404:  # 404 Client Error: Not Found 
            return True
        if response.status_code == 424:  # 424 Client Error: Failed Dependency for url:
            return True
        return False
    except requests.RequestException:
        return True  # if there are an error is overloaded

def get_available_model():
    """Select the first model available from the list."""
    for model_url in MODEL_IDS:
        print("trying",model_url)
        if not is_model_overloaded(model_url):
            return model_url
    return MODEL_IDS[0]  # if all are failing, use the first model by dfault

if verbose: print("Checking available models.")

selected_model_id = get_available_model()

final_answer = FinalAnswerTool()
model = HfApiModel(
max_tokens=2096,
temperature=0.5,
model_id= selected_model_id,#'https://wxknx1kg971u7k1n.us-east-1.aws.endpoints.huggingface.cloud',# it is possible that this model may be overloaded
custom_role_conversions=None,
)


# Import tool from Hub
# image_generation_tool = load_tool("agents-course/text-to-image", trust_remote_code=True)

with open("prompts.yaml", 'r') as stream:
    prompt_templates = yaml.safe_load(stream)
    
agent = CodeAgent(
    model=model,
    tools=[get_current_time_in_timezone, final_answer, tell_me_a_joke], ## add your tools here (don't remove final answer)
    max_steps=6,
    verbosity_level=1,
    grammar=None,
    planning_interval=None,
    name=None,
    description=None,
    prompt_templates=prompt_templates
)


GradioUI(agent).launch()