vllm/benchmarks/kernels at f1f6cc10c77d6ee40b3ce769124cb2760428dc48 - vllm

mirror of https://github.com/wassname/vllm.git synced 2026-07-10 08:05:39 +08:00

Files

T

wbn dacaf5a400 Replace head_mapping params with num_kv_heads to attention kernel. (#1997 )

Co-authored-by: wangguoya <wangguoya@baidu.com>
Co-authored-by: Yang Zhao <zhaoyangstar@foxmail.com>

2023-12-10 10:12:53 -08:00

benchmark_paged_attention.py

2023-12-10 10:12:53 -08:00