Every story and comment from the Hacker News API, in a layout borrowed from 2001.
A comment on Arcee Trinity Mini: US-Trained Moe Model.
Trinity Mini: 26B parameter MoE (3B active), fully post-trained reasoning model
They did pretraining on their own and are still training the large version on 2048 B300 GPUs
There are no comments on this comment yet.
Trinity Mini: 26B parameter MoE (3B active), fully post-trained reasoning model
They did pretraining on their own and are still training the large version on 2048 B300 GPUs