Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
best small quant that still feels smart on 7900 XTX
#1
I figured I'd ask here before changing everything.

I'm mostly working with DeepSeek in SGLang on Threadripper box. The part I'm trying to understand is prefill/decode split.

Would appreciate practical advice rather than theory.
post logs or it didn't happen
Reply


Messages In This Thread
best small quant that still feels smart on 7900 XTX - by benchplease - 10-12-2026, 04:01 AM

Forum Jump:


Users browsing this thread: 1 Guest(s)