app.py

import gradio as gr
from huggingface_hub import InferenceClient
import os
from typing import Optional

from datasets import load_dataset
from langchain.schema import Document
from tools.guestinforetriever import GuestInfoRetrieverTool
from tools.search_tool import SearchTool
from tools.weather_tool import WeatherTool


# Load dataset and initialize tools once at module import
try:
    ds = load_dataset("agents-course/unit3-invitees")
    docs = []
    for split in ds.keys():
        for item in ds[split]:
            # attempt to use common text fields, fallback to stringified item
            text = None
            for key in ("text", "content", "body", "description", "name"):
                if key in item and item[key]:
                    text = item[key]
                    break
            if text is None:
                text = str(item)
            docs.append(Document(page_content=str(text), metadata={"source": f"{split}"}))

    guest_tool = GuestInfoRetrieverTool(docs)
    search_tool = SearchTool(docs)
except Exception:
    # dataset load failed; provide empty fallback tools
    docs = []
    guest_tool = None
    search_tool = None

weather_tool = WeatherTool()


def respond(
    message,
    history: list[dict[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    hf_token: Optional[gr.OAuthToken],
):
    """
    For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
    """
    # simple command routing for tools
    text = message.strip()
    if text.lower().startswith("/guest "):
        query = text[len("/guest "):].strip()
        if guest_tool:
            yield guest_tool.forward(query)
        else:
            yield "Guest retriever not available (dataset failed to load)."
        return

    if text.lower().startswith("/search "):
        query = text[len("/search "):].strip()
        if search_tool:
            yield search_tool.forward(query)
        else:
            yield "Search tool not available (dataset failed to load)."
        return

    if text.lower().startswith("/weather "):
        location = text[len("/weather "):].strip()
        yield weather_tool.forward(location)
        return

    # Default: call the HF chat model
    # Prefer the Gradio OAuth token, fall back to env var `HUGGINGFACEHUB_API_TOKEN`.
    hf_token_value = None
    if hf_token and getattr(hf_token, "token", None):
        hf_token_value = hf_token.token
    else:
        hf_token_value = os.environ.get("HUGGINGFACEHUB_API_TOKEN")

    if not hf_token_value:
        yield (
            "Missing Hugging Face API token. Please run `huggingface-cli login` or set the"
            " environment variable `HUGGINGFACEHUB_API_TOKEN` with a valid token (starts with 'hf_')."
        )
        return

    try:
        client = InferenceClient(token=hf_token_value, model="openai/gpt-oss-20b")
    except Exception as e:
        yield f"Failed to initialize Hugging Face InferenceClient: {e}"
        return

    messages = [{"role": "system", "content": system_message}]
    messages.extend(history)
    messages.append({"role": "user", "content": message})

    response = ""
    for message in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        choices = message.choices
        token = ""
        if len(choices) and choices[0].delta.content:
            token = choices[0].delta.content

        response += token
        yield response


"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
chatbot = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)

with gr.Blocks() as demo:
    with gr.Sidebar():
        gr.LoginButton()
    chatbot.render()


if __name__ == "__main__":
    demo.launch()