10-11-2026, 02:42 AM
Thanks, that helps. The pattern around prefill/decode split has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.
|
multi-GPU offload advice
|
|
« Next Oldest | Next Newest »
|
| Messages In This Thread |
|
multi-GPU offload advice - by vectorport - 10-10-2026, 01:28 AM
Re: multi-GPU offload advice - by signalpeak - 10-11-2026, 01:58 AM
Re: multi-GPU offload advice - by northcraft90 - 10-11-2026, 02:42 AM
|