+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Optimization & Performance (https://www.datasculpting.ai/inference/forumdisplay.php?fid=10)
+--- Forum: Benchmarks (https://www.datasculpting.ai/inference/forumdisplay.php?fid=43)
+--- Thread: QLoRA resource reality check on Threadripper box (/showthread.php?tid=742)
QLoRA resource reality check on Threadripper box - kernelpeak - 10-05-2026
Posting this in case it saves someone else a little time.
If you were starting from scratch today for this use case, what would you choose first and what would you ignore?
I'm especially interested in mistakes worth avoiding.
Re: QLoRA resource reality check on Threadripper box - batchsize - 10-05-2026
Quote:Posting this in case it saves someone else a little time.
Thanks, that helps. I don't think your conclusion is far off. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.