The documented setup process is pretty involved between the GitHub repo, choosing the datasets to use, downloading those from huggingface, setting up the config file to use that dataset, etc etc..
There should be an app that users run when setting up Vigil that automates the process through menu selections.
Menu prompts user to input embedding model to use or 'openai'
If openai, prompt for API key
Set key in config file
Set app file paths in config file
Clone huggingface repo for selected embedding model into data/datasets
Load the data into the vector db with the config
Maybe perform some kind of simple test to verify everything is working as expected
Display commands to user can copy-paste to start the server / cli
The documented setup process is pretty involved between the GitHub repo, choosing the datasets to use, downloading those from huggingface, setting up the config file to use that dataset, etc etc..
There should be an app that users run when setting up Vigil that automates the process through menu selections.
data/datasets