logoalt Hacker News

gpm • today at 5:14 PM • 2 replies • view on HN

Huh, so the metadata says 1.1 trillion parameters, each 32 or 16 bits.

But the files are only roughly 640GB in size (~10GB * 64 files, slightly less in fact). Shouldn't they be closer to 2.2TB?

Replies

coder543 • today at 5:34 PM

The description specifically says:

"Kimi-K2.6 adopts the same native int4 quantization method as Kimi-K2-Thinking."

johndough • today at 5:21 PM

The bulk of Kimi-K2.6's parameters are stored with 4 bits per weight, not 16 or 32. There are a few parameters that are stored with higher precision, but they make up only a fraction of the total parameters.

➕ show 1 reply