Imanol Schlag
@ImanolSchlagAI Researcher training LLMs in Switzerland, for Switzerland.
Imanol Schlag Reposted
MoEUT: Mixture-of-Experts Universal Transformers Their UT model, for the first time, slightly outperforms standard Transformers on LM tasks such as BLiMP and PIQA, while using significantly less compute and memory repo: github.com/robertcsordas/… abs: arxiv.org/abs/2405.16039
United States Trends
- 1. Mike 1,81 Mn posts
- 2. Serrano 240 B posts
- 3. Canelo 16,6 B posts
- 4. #NetflixFight 72,3 B posts
- 5. Father Time 10,8 B posts
- 6. #netflixcrash 15,8 B posts
- 7. Logan 79 B posts
- 8. He's 58 26 B posts
- 9. Rosie Perez 14,9 B posts
- 10. ROBBED 102 B posts
- 11. Boxing 300 B posts
- 12. Shaq 16,2 B posts
- 13. #buffering 10,9 B posts
- 14. My Netflix 83,2 B posts
- 15. Roy Jones 7.181 posts
- 16. Tori Kelly 5.249 posts
- 17. Ramos 69,9 B posts
- 18. Muhammad Ali 18,5 B posts
- 19. Cedric 22 B posts
- 20. #netfilx 5.062 posts
Loading...
Something went wrong.
Something went wrong.