EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

[BugFix]Fix eplb problems when using dynamic eplb. (#3364)

Browse Source 
### What this PR does / why we need it?
When using dynamic eplb,it will be blocking by nz tensor.We fix these
prolems by clone src tensor and recv tensor.

### Does this PR introduce any user-facing change?

### How was this patch tested?
Qwen3_moe in A3.

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: offline0806 <3337230449@qq.com>
Co-authored-by: offline0806 <3337230449@qq.com>
This commit is contained in:
offline893
2025-10-11 14:04:02 +08:00
committed by GitHub
parent ca05f7d632
commit 82b6c846ca
8 changed files with 58 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
vllm_ascend/ops/common_fused_moe.py
View File

@@ -23,6 +23,7 @@ from vllm.config import CompilationLevel, get_current_vllm_config
from vllm.distributed import (get_dp_group, get_ep_group, get_tp_group,
tensor_model_parallel_all_reduce)
from vllm.forward_context import get_forward_context
from vllm.logger import logger
from vllm.model_executor.layers.fused_moe.config import FusedMoEConfig
from vllm.model_executor.layers.fused_moe.layer import (
FusedMoE, UnquantizedFusedMoEMethod, determine_expert_map)
@@ -185,13 +186,23 @@ class AscendFusedMoE(FusedMoE):
os.R_OK):
self.expert_load_balancer = ExpertLoadBalancer(
self.expert_map_path, self.global_num_experts)
self.local_num_experts, self.expert_map = (
self.expert_load_balancer.get_rank_placement_map(
self.moe_instance_id, self.ep_rank))
self.log2phy = self.expert_load_balancer.get_rank_log2phy_map(
self.moe_instance_id, self.ep_rank).npu()
self.global_redundant_expert_num = (
self.expert_load_balancer.get_global_redundant_expert_num())
try:
self.local_num_experts, self.expert_map = (
self.expert_load_balancer.get_rank_placement_map(
self.moe_instance_id, self.ep_rank))
self.log2phy = self.expert_load_balancer.get_rank_log2phy_map(
self.moe_instance_id, self.ep_rank).npu()
except Exception as e:
logger.warning(
f"Init expert map of mtp/eagle when using sample.{e}")
self.local_num_experts, self.expert_map = determine_default_expert_map(
self.global_num_experts, self.ep_size, self.ep_rank,
self.global_redundant_expert_num)
self.log2phy = determine_default_log2phy_map(
self.global_num_experts, self.ep_size, self.ep_rank,
self.global_redundant_expert_num).npu()
else:
# init moe.
self.local_num_experts, self.expert_map = determine_expert_map(
@@ -227,6 +238,7 @@ class AscendFusedMoE(FusedMoE):
if (self.quant_method.__class__.__name__
in ("GPTQMarlinMoEMethod", "CompressedTensorsWNA16MoEMethod")):
moe_quant_params["intermediate_size_full"] = intermediate_size
self.quant_method.create_weights(layer=self, **moe_quant_params)
self.enable_shared_expert_dp = ascend_config.enable_shared_expert_dp