llm-jina

LLM plugin for using Jina AI Search Foundation APIs

Installation

Install this plugin in the same environment as LLM.

llm install llm-jina

Usage

This plugin adds a new jina-engineer command to LLM. Before using the command, make sure to set your Jina AI API key as an environment variable:

export JINA_API_KEY=your_api_key_here

You can get your Jina AI API key for free at https://jina.ai/?sui=apikey

The jina-engineer command supports various operations:

1. Web scraping:

llm jina-engineer --url https://example.com "Extract main content"

2. Web search:

llm jina-engineer --search "Latest news about AI"

3. Text classification:

llm jina-engineer --classify --labels "positive,negative,neutral" "This product is amazing!"

4. Generate embeddings:

llm jina-engineer --embed "Embed this text"

5. Rerank search results:

llm jina-engineer --search --rerank "Best restaurants in New York"

6. Verify factual accuracy:

llm jina-engineer --ground "The Earth is flat"

7. Segment text:

llm jina-engineer --segment "This is a long text that needs to be segmented into smaller chunks."

You can combine multiple operations in a single command. For example:

llm jina-engineer --search --rerank --embed "AI advancements in 2023"

Options

• --url: Specify a URL for web scraping
• --search: Perform a web search
• --classify: Classify text
• --embed: Generate embeddings
• --rerank: Rerank search results (requires --search)
• --ground: Verify factual accuracy
• --segment: Segment text
• --model: Specify the model to use for embeddings or classification (default: jina-embeddings-v3)
• --labels: Provide labels for classification (comma-separated)

Development

To set up this plugin locally, first checkout the code. Then create a new virtual environment:

cd llm-jina
python3 -m venv venv
source venv/bin/activate

Now install the dependencies and test dependencies:

pip install -e '.[test]'

To run the tests:

pytest

Contributing

Contributions to this plugin are welcome! Please follow these steps:

1. Fork the repository
2. Create a new branch for your feature
3. Make your changes and add tests
4. Run the tests to ensure everything is working
5. Submit a pull request

License

This plugin is released under the Apache 2.0 License.