Open Collective
Open Collective
Loading

FlowerTune LLM Labs

PROJECT
Fiscal Host: Ethicalabs.ai

FlowerTune LLM federates fine-tuning of LLMs with LoRA, quantization, and adaptive training, enabling decentralized, efficient, and privacy-preserving AI

About


FlowerTune LLM Labs is a federated fine-tuning initiative evaluating large language models (LLMs) using techniques like LoRA, quantization, and adaptive training.

It enables decentralized, efficient AI adaptation, allowing collaborative model improvement without sharing raw data. This approach is particularly beneficial for domain-specific applications, such as healthcare, where data privacy is paramount.

By leveraging the Flower framework, FlowerTune LLM facilitates seamless integration of federated learning into existing workflows, promoting privacy-preserving, scalable AI solutions.

The project also emphasizes resource efficiency, making it accessible for participants with varying computational capabilities.โ€‹

Ethicalabs aims to push leaderboard experiments to the next level by leveraging DevOps automation for instance provisioning, using sustainable VMs to optimize resource usage.

Your support helps sustain research into privacy-preserving, decentralized AI, enabling open experimentation, scalable model fine-tuning, and responsible AI development.

Our team

Contribute


Become a financial contributor.

Financial Contributions

Custom contribution
Donation
Make a custom one-time or recurring contribution.