restful backends for GPTs build with ChatGPTBuilder and also for huggingface UI
micromamba create -f environment.yaml
micromamba activate longevity_gpts
There is a package with its own REST API per each module as well as index.py where everything is connected together
python rest_endpoint.py
You should fill in .env file to parametrize your host, port and other configuration
You could use tools with openai like api. It is usefull when you use LLM chats like huggingface chatua or librechat.
python openai_api_endpoint.py
This endpoint uses its endpoint_options.yaml file for configurations. It takes the model as a key and then applies all the parameters in litellm call. The required parameters are model and system_prompt. system_prompt is the name of the file in the prompts folder with system prompt. key_getter is the file name for a list of keys.
If it complains on non-finding local packages do:
pip install -e .
It will also dynamically react to all your code updates in submodules.
gene_lookup will not work without disgenet_2020.sqlite database. You should download it from https://www.disgenet.org/downloads and put it in genetics/data folder.