24 hours on a 4070. Seems quite doable. https://github.com/sapientinc/HRM | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		PJones2000 7 months ago \| parent \| context \| favorite \| on: Hierarchical Reasoning Model 24 hours on a 4070. Seems quite doable. https://github.com/sapientinc/HRM

suspended_state 7 months ago [–]

> Experiments below assume an 8-GPU setup.

Careful with those numbers.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact