Nous Research releases Hermes 4.3 36B
Released Hermes 4.3 36B — trained the same model both centrally (FSDP+AdamW) and on Psyche (TP+DisTrO) and compared them. The Psyche run completed at an average of 144k tok/s on 24 nodes and beat the centralized run on downstream performance
Why it matters
The first controlled comparison showing distributed training is not just possible but equal or better
Source — Primary sources first; secondary reports are labelled “Press”
Officialnousresearch.comhttps://nousresearch.com/introducing-hermes-4-3/Added in Week 35, 2026 · See that week