I would think it's the fact that there's no secret sauce except data and compute to get to a decent model along with a good team. It's a Nemtron 3 Nano architecture as far as I understand, and of course there seems to be some special data curation for German.
The main constraint was licensing rather than regulations like GDPR. To ensure compliance, we restricted ourselves to datasets with permissive licenses.
The one exception was the Genios dataset (a high-quality German corpus), which we used during pretraining under a separate licensing agreement. Beyond that, the growing availability of high-quality permissive datasets meant we could assemble sufficient training data without compromising on quality.
Impressive timeline, but I’m curious whether the real advantage is model quality or simply owning the full stack and data.
I would think it's the fact that there's no secret sauce except data and compute to get to a decent model along with a good team. It's a Nemtron 3 Nano architecture as far as I understand, and of course there seems to be some special data curation for German.
one of the authors here, AMA.
What sort of regulations did you run into while gathering training data?
The main constraint was licensing rather than regulations like GDPR. To ensure compliance, we restricted ourselves to datasets with permissive licenses.
The one exception was the Genios dataset (a high-quality German corpus), which we used during pretraining under a separate licensing agreement. Beyond that, the growing availability of high-quality permissive datasets meant we could assemble sufficient training data without compromising on quality.