Warning The classification API is deprecated. Need to replace it with https://platform.openai.com/docs/guides/fine-tuning
This GitHub action labels issues using OpenAI's Classification API powered by GPT-3 models! We are using curie as our completion model and ada as the search model.
It uses your existing labels and past GitHub issues to train a model that can predict the labels of new issues. When someone opens a new issue this action will automatically label it.
You will need to sign up for the OpenAI API and get an OpenAI API Key.
Add this OpenAI API Key as a secret called OPENAI_API_KEY. See Creating encrypted secrets for a repository.
Create a workflow (eg: .github/workflows/labeler.yml). See Creating a Workflow file.
name: "OpenAI Issue Labeler"
on:
issues:
types: [opened, edited]
jobs:
triage:
runs-on: ubuntu-latest
steps:
- uses: austenstone/openai-issue-labeler@v2
with:
openai-api-key: "${{ secrets.OPENAI_API_KEY }}"That's it. Your issues are now labeled by powerful AI models! 🧠
Various inputs are defined in action.yml to let you configure the labeler. See the Classifications API reference for more information.
| Name | Description | Default |
|---|---|---|
| openai‑api‑key | The OpenAI API key secret | N/A |
| token | Token to use to authorize label changes. | ${{ github.token }} |
| temperature | Higher values mean the model will take more risks. | 0 |
| model | ID of the engine to use for completion. You can select one of ada, babbage, curie, or davinci. | curie |
| search‑model | ID of the engine to use for Search. You can select one of ada, babbage, curie, or davinci. | ada |
The model uses Labeled data to classify the new piece of data.
- All repository labels
- Examples of past label usage
- Label description and label
- Issue title and label
- Issue body and label
- A query to find a label for
- New Issue title
- New Issue body
The label! It will be automatically added to your issue.
This workflow runs on this repo. Check out the issues and feel free to open a new one.