Reason-Act SQLite Demo

This is a demonstration of how to use reason and act with llama.cpp and a LLM to pose plain english queries to a SQLite database using one of two strategies:

1. Actions that mimic interaction with a frontend like Datasette. Actions: list tables, list table columns, facet, filter
2. Let the LLM use SQLite queries directly. Actions: list tables, list table schema, execute sql

The things you'll need to do are:

1. Provide a SQLite database (named example.db or you need to change the name in the Python files)
2. Change the prompts in both Python scripts (the prompt string inside the execute functions) to be specific to your data and problems. You'll also want to date the DATA_HELP table and column descriptions in run-sql-queries.py.
3. Download a GGUF model for use. The default is to look for dolphin-2.2.1-mistral-7b.Q5_K_M.gguf in the current dir. If you want to use a different model, edit the script you're running.

There are some dependencies for this project that you need, first. You can install with using pip:

pip install -r requirements.txt

Once you have everything installed and configured, you can kick off a session by coming up with a question and asking it on the command line:

python run_interface.py "What kind of data do I have available?"
python llm_sql_queries.py "What are some interesting records in the database?"

The model output will be printed to stdout.