Clayfoot5d ago
Hacker News 100

Run a 1T parameter model on a 32gb Mac by streaming tensors from NVMe

Link: https://github.com/t8/hypura
Discussion: https://news.ycombinator.com/item?id=47504695

GitHub - t8/hypura: Run models too big for your Mac's memory

Run models too big for your Mac's memory. Contribute to t8/hypura development by creating an account on GitHub.

GitHub
Mar 24 at 6:15pmWeb