TY  - RPRT
TI  - SqueezeAttention: 2D Management of KV-Cache in LLM Inference via Layer-wise Optimal Budget
AU  - Zihao Wang
AU  - Bin Cui
AU  - Shaoduo Gan
PY  - 2024
UR  - https://arxiv.org/abs/2404.04793
ID  - 2404.04793
ER  -