xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-10 03:54:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Misc] Print FusedMoE detail info (#13974)

Browse Source
This commit is contained in:
Jee Jee Li 2025-02-28 07:53:13 +08:00 committed by GitHub
parent 9804145cac
commit 67fc426845
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 20 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
vllm/model_executor/layers/fused_moe/layer.py
View File

@ -737,3 +737,23 @@ class FusedMoE(torch.nn.Module):
# If we are in the row parallel case (down_proj)
else:
param_data[expert_id] = loaded_weight
def extra_repr(self) -> str:
s = (
f"global_num_experts={self.global_num_experts}, "
f"local_num_experts={self.local_num_experts}, "
f"top_k={self.top_k}, "
f"intermediate_size_per_partition={self.intermediate_size_per_partition}, " # noqa: E501
f"tp_size={self.tp_size},\n"
f"ep_size={self.ep_size}, "
f"reduce_results={self.reduce_results}, "
f"renormalize={self.renormalize}, "
f"use_grouped_topk={self.use_grouped_topk}")
if self.use_grouped_topk:
s += f", num_expert_group={self.num_expert_group}, topk_group={self.topk_group}" # noqa: E501
s += f", scoring_func='{self.scoring_func}', activation='{self.activation}'" # noqa: E501
return s