best setup on 24GB VRAM

[fastbench]

Spent some time on this today and wanted to compare notes.

I've gone back and forth between llama.cpp and MLX for Qwen. The interesting difference for me has been HBM constraints.

What would you pick if you had to live with it for a few months?

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.

[macrostack69]

That matches what I was seeing. I ended up in about the same place using vLLM. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[macronest]

Appreciate the detail. I ended up in about the same place using LM Studio. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.

[bytelab]

I'd approach it in a boring order. On M3 Max, 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.

[cacheplane98]

My first guess would be this: On 7900 XTX, the easiest win is often to back off the aggressive settings and see whether the problem survives. Turns out 'just one more test' was a lie.

[modrunner27]

My first guess would be this: On 7900 XTX, the easiest win is often to back off the aggressive settings and see whether the problem survives. Turns ou

Good point. That matches what I saw on Threadripper box. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

[cudaworks]

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

[deepseeked]

What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Gemma and compare from there. A lot of these thread

Tried to reproduce it just now. On 4090 with vLLM, I saw basically the same behavior.

[amberplane]

[quote]What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Gemma and compare from there. A lot of these

Appreciate the detail. I ended up in about the same place using Ollama.