Open-R1: a fully open reproduction of DeepSeek-R1

From the article:

DeepSeek-R1 release leaves open several questions about:

• Data collection: How were the reasoning-specific datasets curated?
• Model training: No training code was released by DeepSeek, so it is unknown which hyperparameters work best and how they differ across different model families and scales.
• Scaling laws: What are the compute and data trade-offs in training reasoning models?

These questions prompted us to launch the Open-R1 project, an initiative to systematically reconstruct DeepSeek-R1’s data and training pipeline, validate its claims, and push the boundaries of open reasoning models. By building Open-R1, we aim to provide transparency on how reinforcement learning can enhance reasoning, share reproducible insights with the open-source community, and create a foundation for future models to leverage these techniques.

In this blog post we take a look at key ingredients behind DeepSeek-R1, which parts we plan to replicate, and how to contribute to the Open-R1 project

that's why gen ai models are not "open source", ever. If they were, this group would't have to "try", they could just run the build script.

Of course, the training data and software is not available. The weights are just a binary blob. It's not the source, but merely the "compiled binary"