[PD] Refactor parallel sizes and add pp support for mooncake (#8571)

Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
This commit is contained in:
Shangming Cai
2025-08-05 11:18:11 +08:00
committed by GitHub
parent 08f8f49016
commit d98a4913ea
4 changed files with 257 additions and 139 deletions

View File

@@ -25,10 +25,13 @@ class KVArgs:
gpu_id: int
# for different tp
decode_tp_size: int
# for pp prefill
prefill_pp_size: int
kv_head_num: int
page_size: int
# for pp prefill
prefill_pp_size: int
pp_rank: int
# for system dp
system_dp_rank: int
class KVPoll: