speculative decoding in llama.cpp with LM Studio

[mangotinker73]

Has anyone else run into this?

My current setup is LM Studio with M3 Max, and I keep running into continuous batching.

If you've tested something similar, what did you keep and what did you stop bothering with?

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.

[phasescope]

subscribed, curious where this lands.

[latentcraft]

A couple things I'd check first. On 5090, the easiest win is often to back off the aggressive settings and see whether the problem survives. Once you know the plain path works, then add the faster path back in piece by piece.

[vectorport]

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at consumer GPUs. Not glamorous advice, but repeatable baselines save a lot of time.

[vrambandit]

Tried to reproduce it just now. I couldn't reproduce the slowdown until I pushed context much higher. The main thing that moved was latency, not quality. The logs were somehow calmer than I was.

[dockridge]

Can you fill in a bit more context? Are you testing this on 7900 XTX or something else?

[mellowwire98]

What worked for me was pretty simple. On 7900 XTX, the easiest win is often to back off the aggressive settings and see whether the problem survives.

[visionyard23]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Llama and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

[phasescope]

I'd approach it in a boring order. On 5090, the easiest win is often to back off the aggressive settings and see whether the problem survives.

[northcraft90]

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at batch size. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.