"e235b8529acd7ebaa882a3cc4e16223aeb20d929b2e4a8a864304a281e0c8d654342a8e64ae757230f76e0a1404a6156f8f691a2d8a0bfb1b94148faa9b727bc"

Last Notes

o hey dere @npub1wcx…jz9t i see we share some follows. do i know u or do we just have some shared pals

2026-05-13 19:56:57 CEST

yes hello i am nostring

2026-05-11 18:14:32 CEST

- reply

honestly not sure they're worth it, unless u rly want to run the biggest models. inference speed is quite slow and in terms of actually USING it i think thats an important metric. altho if you are more focused on training and stuff they are decent i believe. 5090 blows it away for inference on any model that can fit on it and the rtx 6000 same, and that goes up to 96gb of vram (altho more exepnsive, closer to $10k)

2026-05-11 18:12:34 CEST

gm nostr i am receiving of a steam cleaner today and will hence be commencing with many cleanings

2026-05-11 00:18:33 CEST

@npub1c5z…v99c o hello my darling

2026-05-10 23:10:27 CEST

- reply

i for one appreciate the ephemeral nature of stories. its what ppl are doing now, today. i rarely care about someone’s historical tweets

2026-05-10 16:34:42 CEST

- reply

twitter was popular for a reason, its a formula people enjoy. brief updates from their pals. that same basic formula powers instagram and now tiktok/reels/shorts/etc. while underlying tech might change, we haven’t really moved away from the core idea

2026-05-10 16:32:11 CEST

- reply

i have a 4090 btw

2026-05-10 14:55:00 CEST

what up gamers #gamers #gaming

2026-05-10 14:53:21 CEST

- reply

ya this just looks like NMS

2026-05-10 14:52:38 CEST

- reply

is this a JOI?

2026-05-10 02:25:38 CEST