Search / Retrieval / Reranking discussion

[deepseeked]

This keeps coming up in conversation so I figured I'd start a thread.

One thing I keep noticing is that people underestimate reranking until they actually try to run this stuff for more than one afternoon.

Curious whether that lines up with what other people are seeing.

[cudascope]

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.

[turboarch95]

A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at NVFP4. Not glamorous advice, but repeatable baselines save a lot of time.

[cobaltlab]

Maybe, but I had the opposite result. The reported numbers can look compelling, but in ordinary usage I didn't see the same benefit.

[latentpeak22]

following this because I'm in the same boat.

[thetaplane]

following this because I'm in the same boat.

My first guess would be this: On A6000, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

[northhall88]

What exact setup are you running? Is this local-only or are you routing anything externally?

[deltastation]

I'd start with the simpler stack. Turns out 'just one more test' was a lie.

[fusedtinker]

Did a small comparison this afternoon. On 5090 with MLX, I saw basically the same behavior. I expected the model choice to matter more than it did.

I'm happy to test one more variant if you want a second data point.

[torchkite]

Tried to reproduce it just now. On A6000 with Ollama, I saw basically the same behavior. The surprise was that memory pressure showed up before throughput collapsed.

There are a few different conversations hiding inside this thread, so being explicit about the goal helps.