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

[Test]add ut for deepseek_v2. (#1964)

Browse Source 
What this PR does / why we need it?
Add uts for deepseek_v2

Does this PR introduce any user-facing change?
No

How was this patch tested?
- vLLM version: v0.9.2
- vLLM main:
f3137cdd81

---------

Signed-off-by: 张帮政 <zhangbangzheng@huawei.com>
This commit is contained in:
Zac
2025-07-24 10:27:50 +08:00
committed by GitHub
parent 846555cdb5
commit 2ffe051859
2 changed files with 309 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
tests/ut/ops/test_fused_ops.py
View File

@@ -188,7 +188,6 @@ class TestAscendFusedMoe:
assert layer.top_k == default_moe_config['top_k']
assert hasattr(layer, 'w13_weight')
assert hasattr(layer, 'w2_weight')
assert layer.moe_instance_id == 0
# check group_topk
with pytest.raises(AssertionError):