Design, conduct and analyze results of AI-powered surveys and experiments. Simulate social science and market research with large numbers of AI agents and LLMs.
Currently, edsl covers mainstream models. However, researchers may be interested in using some customized model services. I just explored a way to add support for any model services. Maybe some examples to be included in the documentation can help other scholars customize it more easily.
This customization includes:
Create a new CustomService class
Modify the API URL, API token/key, request header, body, and response parser
Store API token/key in .env
Modify enum.py and registry.py to cover the customized model
I can create that documentation and PR if this is a good idea.
Currently,
edsl
covers mainstream models. However, researchers may be interested in using some customized model services. I just explored a way to add support for any model services. Maybe some examples to be included in the documentation can help other scholars customize it more easily.This customization includes:
CustomService
class.env
enum.py
andregistry.py
to cover the customized modelI can create that documentation and PR if this is a good idea.