GitLab

[qwen3-480b] MoE(TN) configs for nmz TP=8
[opt] 优化deepep相关代码
[fix] 修复deepseek moe模型的awq量化推理bug和精度问题, 修复awq模型的VLLM_USE_LIGHTOP_MOE_SUM_MUL_ADD设置位置, update_state,优化性能，去除冗余操作
pcie 解决custom cudagraph模式需要拷贝的问题，需要配合dtk进行使用
[feat] Switch default w8a8 gemm impl to blaslt. Support w8a8-fp8 GEMM backend.MoE 路由抓取：新增 router_capture 工具链与 envs 统一配置
[envs] set VLLM_CUSTOM_CACHE=1、VLLM_USE_FUSED_RMS_ROPE=1、VLLM_USE_FUSED_FILL_RMS_CAT=1、VLLM_USE_FLASH_ATTN_FP8=1、VLLM_USE_FLASH_MLA_FP8=1、update VLLM_USE_TOPK_RENORM

feat: support moe sum when topk==9

bugfix: Fix mtp model load error when eable shared experts fusion.

vllm：export VLLM_CUSTOM_CACHE=1
dtk：export HIP_KERNEL_EVENT_SYSTENFENCE=1

This reverts commit d6ba8ed5.

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

Signed-off-by: Vadim Gimpelson <vadim.gimpelson@centml.ai>

Signed-off-by: Duncan Moss <djm.moss@gmail.com>
Co-authored-by: Duncan Moss <dmoss@nvidia.com>

Signed-off-by: yewentao256 <zhyanwentao@126.com>

Signed-off-by: kaln27 <liaojuncheng123@foxmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>

Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>

Signed-off-by: jiang1.li <jiang1.li@intel.com>

Co-authored-by: mgoin <mgoin64@gmail.com>

Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>

Signed-off-by: Hosang Yoon <hosang.yoon@amd.com>

Signed-off-by: ilmarkov <imarkov@redhat.com>
Signed-off-by: Haoyang Li <Haoyang.Li@amd.com>
Co-authored-by: ilmarkov <imarkov@redhat.com>

Signed-off-by: mgoin <mgoin64@gmail.com>

Signed-off-by: ilmarkov <imarkov@redhat.com>
Co-authored-by: ilmarkov <imarkov@redhat.com>

Signed-off-by: jiang1.li <jiang1.li@intel.com>

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Co-authored-by: mgoin <mgoin64@gmail.com>

Signed-off-by: SzymonOzog <szymon.ozog@gmail.com>

Signed-off-by: Lu Fang <lufang@fb.com>