Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
serving on Blackwell
#11
What worked for me was pretty simple. On B200, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.
Mostly CUDA, sometimes wrong, usually measured
Reply
#12
Maybe, but I had the opposite result. I've seen people blame batch size when the real issue was much more mundane.
Reply
#13
Quote:I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with Phi and compare from there.

A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at prefix caching.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)