ailabel 0.4.0

A tool for creating and managing labeled datasets for AI training

AILabel

A tool for creating and managing labeled datasets for AI training.

Features

• Create and manage topics (categories for classification)
• Label text payloads within topics
• Predict labels for new data using AI (Google Gemini)
• Fast, Unix-style CLI with streaming and batch processing support

Installation

From PyPI

# Install from PyPI using uv
uv pip install ailabel

# For development, install test dependencies
uv pip install "ailabel[test]"

From Source

# Clone the repository
git clone https://github.com/yourusername/ailabel.git
cd ailabel

# Install the package using uv
uv pip install -e .

# For development, install test dependencies
uv pip install -e ".[test]"

Usage

# Create a new topic
label topics new sentiment

# List all topics
label topics list

# Get information about a topic
label topics info sentiment --labels

# Label a payload
label label "This product is amazing!" --topic=sentiment --as=positive

# Label from stdin
echo "This product is amazing!" | label label - --topic=sentiment --as=positive

# Interactive labeling
label label --topic=sentiment --interactive

# JSON output format
label label "Product was great" --topic=sentiment --as=positive --json

# Predict a label for a new payload
label predict "I love this product" --topic=sentiment

# Predict from stdin and get JSON output
echo "I love this product" | label predict - --topic=sentiment --json

# Process multiple items in batch mode
cat items.txt | label predict - --topic=lang-or-animal --batch

# Show debug information
label --debug

Environment Variables

Create a .env.secret file with the following variables:

GEMINI_API_KEY=your_gemini_api_key

Development

Running Tests

# Run all tests
pytest

# Run tests with coverage
pytest --cov=ailabel

License

MIT

uvx --no-cache --from . label -t school math