app.py

# Integrates all components into an agent
# import gradio as gr
import asyncio
from llama_index.core.workflow import Context
from llama_index.core.agent.workflow import AgentWorkflow, ToolCallResult, AgentStream
from llama_index.llms.huggingface_api import HuggingFaceInferenceAPI
from retriever import guest_info_retriever
from tools import get_weather_info, get_hub_stats, google_search
from dotenv import load_dotenv

llm = HuggingFaceInferenceAPI(model_name="Qwen/Qwen2.5-Coder-32B-Instruct")

alfred = AgentWorkflow.from_tools_or_functions(
    [guest_info_retriever, get_weather_info, get_hub_stats, google_search], llm=llm
)

ctx = Context(alfred)


async def main():
    handler = alfred.run(
        "Tell me about Lady Ada Lovelace.",
        ctx=ctx,
    )
    async for ev in handler.stream_events():
        if isinstance(ev, ToolCallResult):
            print("")
            print("Called tool: ", ev.tool_name, ev.tool_kwargs, "=>", ev.tool_output)
        elif isinstance(ev, AgentStream):  # showing the thought process
            print(ev.delta, end="", flush=True)
    print("🎩 Alfred's Response:")
    response = await handler
    print(response)

    handler2 = alfred.run("What projects is she currently working on?", ctx=ctx)
    async for ev in handler2.stream_events():
        if isinstance(ev, ToolCallResult):
            print("")
            print("Called tool: ", ev.tool_name, ev.tool_kwargs, "=>", ev.tool_output)
        elif isinstance(ev, AgentStream):  # showing the thought process
            print(ev.delta, end="", flush=True)
    print("🎩 Alfred's Second Response:")
    response2 = await handler2
    print(response2)


# demo = gr.Interface(fn=greet, inputs="text", outputs="text")
# demo.launch()

if __name__ == "__main__":
    load_dotenv()
    asyncio.run(main())