More

killingtime74 · 2026-05-29T12:23:42 1780057422

Are you next going to say YouTube rankings don't take into account videos that aren't on YouTube and Spotify rankings don't take into account songs that aren't on Spotify?

killingtime74 · 2026-05-29T12:11:07 1780056667

What's a good site to browse when on the toilet

killingtime74 · 2026-05-27T10:30:16 1779877816

You mean they were slaves right

red75prime · 2026-05-27T18:25:19 1779906319

Nah. It was welfare for Roman citizens: cura annonae.

killingtime74 · 2026-05-25T18:55:08 1779735308

By 2025 we will have AGI and software developers don't be necessary. Also next year we will have self driving.

killingtime74 · 2026-05-24T10:26:54 1779618414

There are definitely admired companies in the world. Costco comes to mind.

killingtime74 · 2026-05-20T03:22:25 1779247345

I was wondering who this was. figured out eventually it's the author of Openclaw

killingtime74 · 2026-05-20T01:52:13 1779241933

Time to switch to Gitlab, Bitbucket or self-hosted

killingtime74 · 2026-05-18T22:02:52 1779141772

There's no contract. these people are drafted in Korea

killingtime74 · 2026-05-15T19:25:38 1778873138

You don't need a machine. You need a rack of them. 1.34TB VRAM https://wavespeed.ai/blog/posts/deepseek-v4-gpu-vram-require...

fgonzag · 2026-05-16T02:38:05 1778899085

Nobody is serving models in BF16 precision, not even commercial providers. Especially with newer quant methods (like nv4)

The article states you can fit Q4 in 4 x 4090 and it works reasonably well.

I'd personally fo for deepseek V4 flash at Q8, hardware prices need to come down though. Once an NV4 version get released it'll be easier to run on commodity hardware.

sterlind · 2026-05-16T00:07:11 1778890031

less if you quantize. apparently Q8 and Q4 do pretty well.

killingtime74 · 2026-05-13T22:04:42 1778709882

Probably confession