10-18-2026, 10:44 PM
I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher.
|
what changed after the inference shift for Qwen
|
|
« Next Oldest | Next Newest »
|