+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Inference Engines (https://www.datasculpting.ai/inference/forumdisplay.php?fid=9)
+--- Forum: TensorRT-LLM (https://www.datasculpting.ai/inference/forumdisplay.php?fid=37)
+--- Thread: quant support gaps (/showthread.php?tid=914)
Pages:
1
2
quant support gaps - sysstation69 - 10-09-2026
Thought I'd post a small recipe because a lot of threads ask similar things.
Just sharing a setup that ended up working reasonably well: Mistral, LM Studio, A6000, with a lot of attention on INT4.
Feel free to tear it apart if you see something dumb.
I care more about stability than squeezing the last possible tokens/sec out of it.
If I had to summarize the lesson so far, it would be that clean setup and observability beat clever guesses.
Re: quant support gaps - routeforge64 - 10-09-2026
That matches what I was seeing. I ended up in about the same place using llama.cpp.
Re: quant support gaps - kernelgrid - 10-09-2026
I don't think that's the whole story. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.
Re: quant support gaps - batchcraft - 10-10-2026
A couple questions before I guess too hard. Is this local-only or are you routing anything externally?
Re: quant support gaps - routecrate - 10-11-2026
Interesting, I hadn't tried that. I don't think your conclusion is far off.
Re: quant support gaps - torchtrack24 - 10-11-2026
I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher.
Re: quant support gaps - rocmriley - 10-11-2026
Quote:I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher.
Thanks, that helps. I ended up in about the same place using PyTorch. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
Re: quant support gaps - batchscope72 - 10-11-2026
Quote:Quote:I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher.
I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with FLUX and compare from there.
Re: quant support gaps - sysstation69 - 10-11-2026
following this because I'm in the same boat.
Re: quant support gaps - ggufvault - 10-13-2026
following this because I'm in the same boat.