EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

[fix] fix determine_num_fused_shared_experts (#7180)

Browse Source
This commit is contained in:
JieXin Liang
2025-06-15 08:41:22 +08:00
committed by GitHub
parent 4473320380
commit 55561e2553
1 changed files with 29 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files

46
python/sglang/srt/models/deepseek_v2.py
View File

@@ -1709,53 +1709,35 @@ class DeepseekV2ForCausalLM(nn.Module):
def determine_num_fused_shared_experts( def determine_num_fused_shared_experts(
self, architecture: str = "DeepseekV3ForCausalLM" self, architecture: str = "DeepseekV3ForCausalLM"
): ):
self.num_fused_shared_experts = ( self.num_fused_shared_experts = 0
0 if global_server_args_dict["disable_shared_experts_fusion"]:
if global_server_args_dict["disable_shared_experts_fusion"] return
else self.config.n_shared_experts
)
if self.num_fused_shared_experts > 0:
# Only Deepseek V3/R1 can use shared experts fusion optimization now. # Only Deepseek V3/R1 can use shared experts fusion optimization now.
disable_reason = None
if ( if (
not _is_cuda not _is_cuda
or torch.cuda.get_device_capability("cuda") < (9, 0)
or self.config.architectures[0] != architecture or self.config.architectures[0] != architecture
or self.config.n_routed_experts != 256 or self.config.n_routed_experts != 256
or self.config.n_shared_experts != 1
): ):
self.num_fused_shared_experts = 0 disable_reason = "Only Deepseek V3/R1 on NV-platform with capability >= 90 can use shared experts fusion optimization."
global_server_args_dict["disable_shared_experts_fusion"] = True
log_info_on_rank0(
logger,
"Only Deepseek V3/R1 on NV-platform can use shared experts fusion optimization. Shared experts fusion optimization is disabled.",
)
elif ( elif (
global_server_args_dict["enable_deepep_moe"] global_server_args_dict["enable_deepep_moe"]
or global_server_args_dict["enable_ep_moe"] or global_server_args_dict["enable_ep_moe"]
): ):
self.num_fused_shared_experts = 0 disable_reason = "Deepseek V3/R1 can not use shared experts fusion optimization when in deepep_moe or ep_moe mode."
if disable_reason is not None:
global_server_args_dict["disable_shared_experts_fusion"] = True global_server_args_dict["disable_shared_experts_fusion"] = True
log_info_on_rank0( log_info_on_rank0(
logger, logger,
"Deepseek V3/R1 can not use shared experts fusion optimization when in deepep_moe or ep_moe mode. Shared experts fusion optimization is disabled.", f"{disable_reason} Shared experts fusion optimization is disabled.",
) )
elif self.num_fused_shared_experts == 0: return
if (
_is_cuda
and torch.cuda.get_device_capability("cuda") >= (9, 0)
and self.config.architectures[0] == architecture
and self.config.n_routed_experts == 256
and (
not (
global_server_args_dict["enable_deepep_moe"]
or global_server_args_dict["enable_ep_moe"]
)
)
):
self.num_fused_shared_experts = self.config.n_shared_experts self.num_fused_shared_experts = self.config.n_shared_experts
global_server_args_dict["disable_shared_experts_fusion"] = False
log_info_on_rank0(
logger,
"Deepseek V3/R1 with fp8/fp4 can use shared experts fusion optimization when SM version >=90. Shared experts fusion optimization is enabled.",
)
def get_input_embeddings(self) -> nn.Embedding: def get_input_embeddings(self) -> nn.Embedding:
return self.model.embed_tokens return self.model.embed_tokens