Advances in language modeling have been driven by scaling pretraining on ever more data. Yet, the training data is still largely curated on the model's behalf. A more general approach to pretraining would let the model…
0 comments
No comments yet.
Related stories
- Self-Play Pretraining with Zero Dataarxiv.orgHacker News · 2 points · 1 day ago
- Hacker News · 6 points · about 12 hours ago
- Hacker News · 1 points · about 13 hours ago
- Physical Self-Playskild.aiHacker News · 2 points · 7 days ago
- Hacker News · 377 points · 16 days ago
- Ars Technica · 0 points · 7 days ago