TensorRT-LLM worth the setup time with TensorRT-LLM

[syshouse]

Did a bit of testing and thought I'd post the result here.

After a few runs on 7900 XTX, I ended up with decent performance, but the setup only really made sense once I adjusted around consumer GPUs.

Curious if anyone else landed in the same place.

[syshouse]

that fixed it for me too.

[vectorport]

Good point. The pattern around structured output has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.

[rasterbench]

Good point. The pattern around structured output has been pretty similar for me. I got better results once I stopped chasing one metric and just looke

I'd approach it in a boring order. If you're on SGLang, I'd confirm the version first and then reduce variables until the behavior is obvious.

[syshouse]

[quote]Good point. The pattern around structured output has been pretty similar for me. I got better results once I stopped chasing one metric and jus

Did a small comparison this afternoon. My result was mixed: first run looked bad, second run looked normal.

[batchcraft]

I hit almost the same issue last night.

[plainscope50]

Can you fill in a bit more context? What model, what runtime, and roughly what context size?

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

[vectorshed43]

Can you fill in a bit more context? What model, what runtime, and roughly what context size?

[kernelpeak]

Can you fill in a bit more context? What model, what runtime, and roughly what context size?

I tested this tonight. On A6000 with vLLM, I saw basically the same behavior.

[batchcraft]

A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at HBM constraints. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.