speculative decoding in llama.cpp

[quantsupply]

Tried to reproduce it just now. The issue went away once I changed one setting and restarted the stack cleanly.

[thetagrid15]

What worked for me was pretty simple. On GB200, the easiest win is often to back off the aggressive settings and see whether the problem survives. Not glamorous advice, but repeatable baselines save a lot of time.

[phasescope]

I think the bottleneck is somewhere else. I've seen people blame tool calling when the real issue was much more mundane.

[quietarch]

My first guess would be this: Before rebuilding anything, I'd test one clean baseline with DeepSeek and compare from there.

[northcraft90]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Llama and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.