xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-06-07 00:49:09 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(tests): Ensure reliable CUDA cache clearing in MoE test (#23416)

Browse Source 
Signed-off-by: AzizCode92 <azizbenothman76@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
This commit is contained in:
Aziz 2025-08-22 19:20:59 +02:00 committed by GitHub
parent 424fb7a5d2
commit 341923b982
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tests/kernels/moe/test_moe.py
View File

@ -429,11 +429,11 @@ def test_mixtral_moe(dtype: torch.dtype, padding: bool, use_rocm_aiter: bool,
vllm_moe.experts.w13_weight, (0, 128), "constant", 0)[..., vllm_moe.experts.w13_weight, (0, 128), "constant", 0)[...,
0:-128], 0:-128],
requires_grad=False) requires_grad=False)
torch.cuda.empty_cache()
vllm_moe.experts.w2_weight = Parameter(F.pad( vllm_moe.experts.w2_weight = Parameter(F.pad(
vllm_moe.experts.w2_weight, (0, 128), "constant", 0)[..., vllm_moe.experts.w2_weight, (0, 128), "constant", 0)[...,
0:-128], 0:-128],
requires_grad=False) requires_grad=False)
torch.cuda.synchronize()
torch.cuda.empty_cache() torch.cuda.empty_cache()
# Run forward passes for both MoE blocks # Run forward passes for both MoE blocks