This is a template repo that is meant to be used whenever you are trying to use LangChain to automate some task with LLMs. This example takes about three minutes to run.
First, you should define what it is you are trying to have the LLM accomplish in the form of a simple step by step guide.
An example is in how_to_make_good_polls.md.
This will be your guide for structuring the tool you are trying to make.
For each step you can try:
- Asking chatgpt outright to do it in the form of a prompt, be sure to ask for its reasoning, encourage it, and include few shot prompting/examples,
- Iterate to make a part of it better
- Rank output and cherry pick the best option
- Python 3.11+
- Poetry
- For installation, see their official docs.
poetry install --no-root
cp example.env .env # put the proper env vars herepoetry run python main.py generate_poll --topic "programming"
poetry run python main.py generate_question --topic "programming"
poetry run python main.py generate_answers --question "which programming language is easiest to use"
poetry run python main.py rank_poll --topic "programming" --target_audience "software developers" # generates and ranks a poll