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

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Linear Mode
quant choices for 70B local for Llama
vrambandit Offline
Member
***
Posts: 168
Threads: 14
Joined: Sep 2026
Reputation: 0
#2
10-13-2026, 07:14 PM
Good point. The pattern around speculative decoding has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
2x3090 • llama.cpp • too much VRAM is still not enough
Find
Reply


Messages In This Thread
Re: quant choices for 70B local for Llama - by vrambandit - 10-13-2026, 07:14 PM

Forum Jump:


Users browsing this thread: 1 Guest(s)