2024-03-08 07:33:01 +00:00
# DEPRECATED: Try Mentat instead! https://github.com/AbanteAI/mentat
2023-03-18 22:16:47 +00:00
# Wolverine
## About
Give your python scripts regenerative healing abilities!
Run your scripts with Wolverine and when they crash, GPT-4 edits them and explains what went wrong. Even if you have many bugs it will repeatedly rerun until it's fixed.
For a quick demonstration see my [demo video on twitter ](https://twitter.com/bio_bootloader/status/1636880208304431104 ).
## Setup
python3 -m venv venv
source venv/bin/activate
2023-04-08 20:10:04 +00:00
pip install -r requirements.txt
2023-04-15 05:22:57 +00:00
cp .env.sample .env
2023-03-18 22:16:47 +00:00
2023-04-15 05:22:57 +00:00
Add your openAI api key to `.env`
_warning!_ By default wolverine uses GPT-4 and may make many repeated calls to the api.
2023-03-18 22:16:47 +00:00
## Example Usage
2023-04-08 19:38:14 +00:00
To run with gpt-4 (the default, tested option):
2023-04-24 16:21:14 +00:00
python -m wolverine examples/buggy_script.py "subtract" 20 3
2023-03-18 22:16:47 +00:00
2023-04-08 19:38:14 +00:00
You can also run with other models, but be warned they may not adhere to the edit format as well:
2023-04-24 16:21:14 +00:00
python -m wolverine --model=gpt-3.5-turbo examples/buggy_script.py "subtract" 20 3
2023-04-08 19:38:14 +00:00
2023-04-15 05:22:57 +00:00
If you want to use GPT-3.5 by default instead of GPT-4 uncomment the default model line in `.env` :
DEFAULT_MODEL=gpt-3.5-turbo
2023-04-15 05:28:13 +00:00
You can also use flag `--confirm=True` which will ask you `yes or no` before making changes to the file. If flag is not used then it will apply the changes to the file
2023-04-24 16:21:14 +00:00
python -m wolverine examples/buggy_script.py "subtract" 20 3 --confirm=True
2023-04-15 05:28:13 +00:00
2023-04-27 00:33:36 +00:00
## Environment variables
| env name | description | default value |
| ------------------- | ----------------------------------------------------------------- | ------------- |
| OPENAI_API_KEY | OpenAI API key | None |
| DEFAULT_MODEL | GPT model to use | "gpt-4" |
| VALIDATE_JSON_RETRY | Number of retries when requesting OpenAI API (-1 means unlimites) | -1 |
2023-04-25 21:15:08 +00:00
2023-03-18 22:16:47 +00:00
## Future Plans
This is just a quick prototype I threw together in a few hours. There are many possible extensions and contributions are welcome:
2023-04-08 19:38:14 +00:00
- add flags to customize usage, such as asking for user confirmation before running changed code
2023-03-18 22:16:47 +00:00
- further iterations on the edit format that GPT responds in. Currently it struggles a bit with indentation, but I'm sure that can be improved
2023-04-15 05:22:57 +00:00
- a suite of example buggy files that we can test prompts on to ensure reliability and measure improvement
2023-03-18 22:16:47 +00:00
- multiple files / codebases: send GPT everything that appears in the stacktrace
- graceful handling of large files - should we just send GPT relevant classes / functions?
- extension to languages other than python
2023-04-15 05:22:57 +00:00
## Star History
[![Star History Chart ](https://api.star-history.com/svg?repos=biobootloader/wolverine&type=Date )](https://star-history.com/#biobootloader/wolverine)