hegel 0.0.3

Supercharge your LLM with a frictionless wrapper

hegel

Welcome to hegel, the python client library for Hegel AI. This library adds the full functionality of our Dialectic product to your LLM, including observability, computing metrics, and tracking customer feedback.

Monitoring your LLM

To use hegel, you'll wrap your current call to an LLM API with the HegelScribe object. HegelScribe helps record the input and response of a LLM call with minimal overhead. The call to the LLM's API is not routed through our server; only the input and response are captured and written to the database through a separate thread without blocking the main thread.

import openai
from hegel.scribe import HegelScribe

scribe = HegelScribe(completion_fn=openai.ChatCompletion.create)
response = scribe(
    model= "gpt-3.5-turbo",
    messages = [  
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Who won the world series in 2020?"},
    ])

Setting your API key

An OpenAI API key is needed if you would like to use OpenAI's API (we do not store or track your API key in anyway). A separate Hegel API key is needed to use our service. You can get a key by contacting us.

To set your API keys:

# Starting from this `hegel` directory
cp .env.example .env

Edit the .env file to include your keys. Avoid sharing that file.

Viewing Logs and Metrics

Scribe collects metrics, monitors failures, and helps gather customer feedback. You can view aggregate metrics, drill into logs, and more on the Hegel Dashboard.

Contributing

We welcome PRs and suggestions! Don't hesitate to open a PR/issue or to reach out to us via email.

Usage and Feedback

We will be delighted to work with early adopters to shape our designs. Please reach out to us via email if you're interested in using this tooling for your project or have any feedback.

License

We will be gradually releasing more components to the open-source community. The current license can be found in the LICENSE file. If there is any concern, please contact us and we will be happy to work with you.