retrieve-sentence-embeddings 0.1.0a0

A simple function to retrieve sentence embeddings from an OpenAI Embeddings-compatible API.

retrieve-sentence-embeddings

A simple function to retrieve sentence embeddings from an OpenAI Embeddings-compatible API. That includes:

• OpenAI's embedding models (base_url='https://api.openai.com/v1)
• Google Gemini (base_url='https://generativelanguage.googleapis.com/v1beta/openai')
• ollama (base_url='http://localhost:11434/v1').

with compatible models.

Features

• Compatibility: Unlike openai, works on Python 2+ with zero non-Python dependencies.

Installation

pip install retrieve-sentence-embeddings

Usage

# coding=utf-8
from retrieve_sentence_embeddings import retrieve_sentence_embeddings

api_key = u'OPENAI_API_KEY'
base_url = u'https://api.openai.com/v1'
model = u'text-embedding-ada-002'
sentences = [
    u'Hello, world!',
    u'This is a test sentence.',
    u'Embeddings are useful!'
]

sentence_embeddings = retrieve_sentence_embeddings(api_key, base_url, model, sentences)

sentence_embeddings is a List[List[float]], e.g.,

[
    [0.123, -0.982, ...],
    [0.222, 0.088, ...],
    ...
]

Error Handling

If a network error or malformed response is encountered, the exception is propagated.

Contributing

Contributions are welcome! Please submit pull requests or open issues on the GitHub repository.

License

This project is licensed under the MIT License.