GitLab

新增环境变量 VLLM_USE_MARLIN_W16A16_MOE，用于显式启用 Marlin W16A16 MoE experts
在 fused_moe 中当开关开启且实现可用时，调用 fused_experts_impl_w16a16_marlin
增加 Marlin W16A16 MoE 实现与 reduce 路径

replace the fp8_mqa_logits and fp8_paged_mqa_logits interfaces in deepgemm with mqa_logits and paged_mqa_logits from lightop

This reverts commit 3b121add.

add VLLM_USE_OPT_RESHAPE_AND_CACHE、VLLM_USE_FUSE_SILU_AND_MUL and VLLM_USE_TOPK_RENORM for qwen3-30b

update VLLM_USE_PD_SPLIT=0 (for dspk)and  VLLM_USE_PD_SPLIT=1 (for others)

add VLLM_USE_PP_SYNC to use pp sync
update qwen3 of rmsnorm

set default_max_num_batched_tokens = 10240
update qwen3_moe of layernorm
off lightop of moe_fused_gate

add VLLM_USE_PD_SPLIT to split prefill and decode
replace triton_ of rms and act_and_mul

add deps of pd-pp