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

8
tests/ut/eplb/core/test_eplb_device_transfer_loader.py
View File

@@ -48,13 +48,7 @@ def test_generate_task_and_state_flow(mock_adaptor):
loader_obj.generate_expert_d2d_transfer_task([], [], {}, 0)
assert loader_obj.comm_op_list is None
updated_map = {20: torch.tensor(0)}
loader_obj.generate_expert_d2d_transfer_task([(1, 10)], [(2, 20)],
updated_map, 0)
assert loader_obj.state == loader.ExpertWeightUpdateState.READY
assert loader_obj.comm_op_list
assert loader_obj.recv_expert_list
assert loader_obj.state == loader.ExpertWeightUpdateState.WAITING
def test_asyn_transfer_and_update(mock_adaptor):