Hugging Face smolagents

How to use the SmolagentsInstrumentor to trace smolagents by Hugging Face

smolagents is a minimalist AI agent framework developed by Hugging Face, designed to simplify the creation and deployment of powerful agents with just a few lines of code. It focuses on simplicity and efficiency, making it easy for developers to leverage large language models (LLMs) for various applications.

Install

pip install openinference-instrumentation-smolagents smolagents arize-otel

Setup

Add your HF_TOKEN as an environment variable:

os.environ["HF_TOKEN"] = "<your_hf_token_value>"

Next, register a tracer provider to send traces to the Arize platform.

from arize.otel import register

# Setup OTel via our convenience function
tracer_provider = register(
    space_id = "your-space-id", # in app space settings page
    api_key = "your-api-key", # in app space settings page
    project_name = "your-project-name", # name this to whatever you would like
)

Finally, use the smolagents autoinstrumentor to automatically track and visualize every step and call made by your agent.

from openinference.instrumentation.smolagents import SmolagentsInstrumentor

SmolagentsInstrumentor().instrument(tracer_provider=tracer_provider)

Create & Run an Agent

from smolagents import (
   CodeAgent,
   DuckDuckGoSearchTool,
   VisitWebpageTool,
   HfApiModel,
)

hf_model = HfApiModel()

agent = CodeAgent(
    tools=[DuckDuckGoSearchTool(), VisitWebpageTool()],
    model=hf_model,
    add_base_tools=True
)

agent.run("fetch the share price of google from 2020 to 2024, and create a line graph from it?")

Observe

Now that you have tracing setup, all invocations and steps of your Agent will be sent to the Arize platform for observability and evaluation.