working config for 32k context for Llama

Thread Rating:

0 Vote(s) - 0 Average
1
2
3
4
5

Linear Mode

working config for 32k context for Llama

rocmhouse90 Offline

Offline

Member

Posts: 63
Threads: 1
Joined: Sep 2026
Reputation: 0

#11

08-30-2026, 09:48 PM

Quote:Appreciate the detail. That matches what I saw on 7900 XTX.

I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher. The main thing that moved was latency, not quality.

still benchmarking everything

Find

Reply

« Next Oldest | Next Newest »

Messages In This Thread

working config for 32k context for Llama - by mlxgarden - 08-24-2026, 11:23 PM

Re: working config for 32k context for Llama - by rocmriley - 08-25-2026, 11:34 AM

Re: working config for 32k context for Llama - by mangoquill - 08-25-2026, 01:02 PM

Re: working config for 32k context for Llama - by rocmhouse90 - 08-25-2026, 01:58 PM

Re: working config for 32k context for Llama - by vectorport - 08-25-2026, 02:43 PM

Re: working config for 32k context for Llama - by relayforge - 08-31-2026, 09:53 PM

Re: working config for 32k context for Llama - by rocmhouse90 - 08-25-2026, 10:52 PM

Re: working config for 32k context for Llama - by audioriver - 08-26-2026, 11:39 PM

Re: working config for 32k context for Llama - by rocmcrate26 - 08-28-2026, 11:46 AM

Re: working config for 32k context for Llama - by rocmhouse90 - 08-28-2026, 08:33 PM

Re: working config for 32k context for Llama - by rasterbench - 08-30-2026, 09:15 AM

Re: working config for 32k context for Llama - by rocmhouse90 - 08-30-2026, 09:48 PM

Re: working config for 32k context for Llama - by cozysmith - 09-02-2026, 09:54 AM

Re: working config for 32k context for Llama - by llmpeak - 09-05-2026, 10:42 AM

Re: working config for 32k context for Llama - by bytehall - 09-05-2026, 11:33 PM

View a Printable Version

Forum Jump:

Users browsing this thread: 1 Guest(s)