Riva

Fine-tune open source AI models in minutes.

Riva is building the fastest way to turn foundation models into production-ready, task-specific AI. Pick a model, connect your data, launch a fine-tune, and ship.

This organization is home to the models, adapters, datasets, experiments, and demos we publish along the way.

What you'll find here

• 🚀 Fine-tuned open source models
• 🧩 LoRA adapters
• 📚 Curated datasets
• 📈 Benchmarks & evaluations
• 🧪 Experiments
• 🔬 Reproducible training examples

Every repository is designed to be explored, remixed, and improved.

Our Philosophy

General models are only the starting point.

The real value comes from models trained on the right data for the right task.

We believe fine-tuning should be as simple as choosing a model, choosing a dataset, and clicking Train.

Built with Riva

Everything published under this organization was trained, evaluated, or managed using Riva.

As we grow, this page will become a living collection of hundreds of open models across every domain, from coding and finance to healthcare, education, law, customer support, and beyond.

Follow Our Work

We're constantly training new models, publishing new datasets, and experimenting with the latest open source releases.

If a new model is worth fine-tuning, you'll probably find it here.

---

Website: https://riva.fly.dev/

Train faster. Experiment more. Build with open models.