publang is a tool for semantic search and information retrieval using large lange models over biomedical documents. Currently, we are focused on using OpenAI's API, but we have plans to expand to locally available and domain specific LLMs.
This tool facilitates extracting and evaluating structured annotations over corpora of biomedical articles, such as demographic information, or
git clone https://github.com/adelavega/publang.git
cd publang
pip install -e .
To use LLM APIs, you'll need to configure your API keys.
PubLang uses OpenAI's client library for inference. Set your API keys using the OPENAI_API_KEY
environment variable.
Note that many third-party LLM inference APIs support using OpenAI's library, and thus are compatible.
To use other API, simply set the OPENAI_API_BASE
variable.
For example, to use Fireworks AI:
export OPENAI_API_BASE="https://api.fireworks.ai/inference/v1"
export OPENAI_API_KEY="<YOUR_FIREWORKS_API_KEY>"
Alternatively, you can pass an initialized client
object to extraction functions.