FLock SDK is a powerful software dev kit that enables data and computation resources owners to collaboratively train models using any source data.

Key Features

  1. Federated learning:

    Distributed approach that allows data owners to train models collaboratively while keeping their data locally.

  2. Collaborative training:

    By sharing source data, FLock enables multiple participants to contribute.

  3. Rewards and smart contracts:

    Participants are incentivised with rewards and penalties specified by pre-defined smart contracts, ensuring fair and transparent compensation.

  4. Secure and privacy-preserving:

    Data owners retain control over their sensitive information.

  5. Flexible integration:

    Designed to be easily integrated into your existing workflows and systems.

Example Uses

  1. Flock LLM:

    flock_llm directory - example use of Flock SDK with the Flock Large Language Model(LLM). This demonstrates how to fine-tune a Vicuna-7B and train a LoRA adaptor.

  2. Credit Card Fraud Detection:

    credit_card_fraud_detection directory - example use of Flock SDK to train a fraud detection model.

  3. MobileNet Example:

    mobilenet_example directory - example implementation of Flock SDK to train a MobileNet model for image classification.

Last updated