Stanford's Marin 535B-A23B Open Training Run Hits 13%, Funded by Jensen Huang's Foundation
stanfordnlp · x · 2026-09-05
Stanford's Percy Liang announced the fully open Marin project is 13% through training a 535B-A23B model, with compute on CoreWeave funded by the Jen-Hsun and Lori Huang Foundation. Commenters note Marin pairs with the cs336 LM-from-scratch course, offering an open window into frontier-scale training.
More from Models
- OpenAI Devs launches 24-hour GPT-6 Astra shipping challenge — OpenAIDevs · 2026-09-05
- Anthropic reportedly prepping Fable 5.2 to answer GPT-6 Astra; Grok 4.7 due next week — kimmonismus · 2026-09-05
- OpenAI quietly changed GPT-6 Astra benchmark results around launch, Fortune reports — mark_k · 2026-09-05
- Robot control idea: local high-frequency model consuming latent predictions from a larger model — chris_j_paxton · 2026-09-05
- Astra beats Sol but not SOTA on hard wet-lab biology, citing scarce public data — nlarusstone · 2026-09-05
- User Cancels Fable 5.1 After One Prompt Burns 1% of Weekly Quota — BLUECOW009 · 2026-09-05