Inference › Optimization & Performance › Quantization v
quant choices for 70B local

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Linear Mode
quant choices for 70B local
relaynest82 Offline
Junior Member
**
Posts: 34
Threads: 6
Joined: Sep 2026
Reputation: 0
#14
10-18-2026, 09:24 PM
Did a small comparison this afternoon. On M3 Max with llama.cpp, I saw basically the same behavior. The main thing that moved was latency, not quality.
Find
Reply


Messages In This Thread
quant choices for 70B local - by southden - 10-11-2026, 01:50 AM
Re: quant choices for 70B local - by vramscope - 10-11-2026, 01:50 PM
Re: quant choices for 70B local - by benchplease - 10-12-2026, 02:24 AM
Re: quant choices for 70B local - by northquill - 10-12-2026, 10:42 AM
Re: quant choices for 70B local - by southtinker - 10-12-2026, 10:59 AM
Re: quant choices for 70B local - by torchshed - 10-15-2026, 11:45 AM
Re: quant choices for 70B local - by fusedtinker - 10-18-2026, 12:08 PM
Re: quant choices for 70B local - by benchplease - 10-18-2026, 01:21 PM
Re: quant choices for 70B local - by vrambandit - 10-18-2026, 04:51 PM
Re: quant choices for 70B local - by turboscope - 10-18-2026, 08:08 PM
Re: quant choices for 70B local - by cacheworks - 10-18-2026, 09:02 PM
Re: quant choices for 70B local - by relaynest82 - 10-18-2026, 09:24 PM
Re: quant choices for 70B local - by homelabhenry - 10-18-2026, 06:07 PM
Re: quant choices for 70B local - by cacheworks - 10-19-2026, 02:50 AM
Re: quant choices for 70B local - by vrambandit - 10-19-2026, 05:23 AM

Forum Jump:


Users browsing this thread: 1 Guest(s)