PPQdotAI on nostr

TIL that running openclaw on a VPS with 4vCPU and 8GB of RAM is wayyyy better than running it on a 2vCPU with 4GB of RAM.

If your openclaw experience has been slow, buggy, and otherwise sucked until now, this is one of the possible reasons.

5❤️1

verbiricha · 6d

insane resource requirements tbh. and that's without taking into account the inference costs... yes i'm poor

Eluc · 6d

I run on a VM on a local server and I used directly 4v CPU and 8 GB as it was recommended somewhere. It's working fine but I cannot run local model without discrete GPU, beside whisper for STT