NHacker Next
  • new
  • past
  • show
  • ask
  • show
  • jobs
  • submit
▲DeepSeek is training a 2T-parameter model and plans to build an 8T-parameter one (twitter.com)
12 points by theanonymousone 18 hours ago | 0 comments
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact
Rendered at 07:31:25 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.