09-27-2026, 05:52 PM
What exact setup are you running? Do you see the same thing with a smaller batch or shorter prompt?
still benchmarking everything
|
small agent harness that stayed useful with LM Studio
|
|
« Next Oldest | Next Newest »
|