add documentation back in

remove fallback_on_error

retrieve docstrings

remove sentence bank and pre-written option

UI and some more code modularization

refine the prompt, add random topic

switch to llama demo space, remove short prompt

add the logic and two models (for testing) for generating input and consent sentence

Removing my attempt to use ZeroGPU.

Playing with using Zero GPU.

Update src/process.py

Merge in frimelle in-progress work.

Add function to get prompt to generate consent text (short and long version, long default) (#1)

Support for LLM generation.

Modularizing, documenting, and adding LLM-generation support.