Inference › Optimization & Performance › Context / KV Cache v
paged attention observations

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Linear Mode
paged attention observations
cacheline Offline
Senior Member
****
Posts: 341
Threads: 51
Joined: Sep 2026
Reputation: 0
#13
10-02-2026, 02:51 AM
A couple things I'd check first. If you're on TensorRT-LLM, I'd confirm the version first and then reduce variables until the behavior is obvious. Once you know the plain path works, then add the faster path back in piece by piece.
measure twice, benchmark three times
Find
Reply


Messages In This Thread
paged attention observations - by cacheline - 09-23-2026, 12:46 AM
Re: paged attention observations - by cacheline - 09-23-2026, 02:27 AM
Re: paged attention observations - by cacheline - 09-26-2026, 03:29 PM
Re: paged attention observations - by cacheline - 09-27-2026, 04:37 AM
Re: paged attention observations - by edgecoast - 09-27-2026, 09:15 AM
Re: paged attention observations - by cacheline - 09-28-2026, 09:34 AM
Re: paged attention observations - by cacheline - 09-28-2026, 11:19 AM
Re: paged attention observations - by mangohouse - 09-28-2026, 11:49 PM
Re: paged attention observations - by cacheline - 09-29-2026, 02:20 AM
Re: paged attention observations - by sparrowhall - 10-02-2026, 02:29 AM
Re: paged attention observations - by cacheline - 10-02-2026, 02:51 AM

Forum Jump:


Users browsing this thread: 1 Guest(s)