TY  - RPRT
TI  - Inference First, Training Underneath: Layer-Granular LoRA Co-Serving with LayerLoom
AU  - Zongpu Zhang
AU  - Yishu Yin
AU  - Xuehai Qian
PY  - 2026
UR  - https://arxiv.org/abs/2611.00652
ID  - 2611.00652
ER  -