biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[CI] Prune back the number of tests in tests/kernels/* (#9932)

Browse Source 
Signed-off-by: mgoin <michael@neuralmagic.com>
This commit is contained in:
Michael Goin
2024-11-05 16:02:32 -05:00
committed by GitHub
parent 02462465ea
commit 235366fe2e
10 changed files with 60 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
tests/kernels/test_awq_marlin.py
View File

@@ -14,13 +14,17 @@ from vllm.model_executor.layers.quantization.utils.marlin_utils_test import (
awq_marlin_quantize)
from vllm.scalar_type import scalar_types
NUM_EXPERTS = [8, 64]
TOP_KS = [2, 6]
GROUP_SIZES = [-1, 32, 128]
@pytest.mark.parametrize("m", [64, 512, 222, 33, 1])
@pytest.mark.parametrize("n", [128, 2048, 256, 1024])
@pytest.mark.parametrize("k", [128, 1024, 512])
@pytest.mark.parametrize("e", [8, 64])
@pytest.mark.parametrize("topk", [2, 6])
@pytest.mark.parametrize("group_size", [-1, 32, 64, 128])
@pytest.mark.parametrize("m", [1, 33, 64, 222])
@pytest.mark.parametrize("n", [128, 2048])
@pytest.mark.parametrize("k", [128, 1024])
@pytest.mark.parametrize("e", NUM_EXPERTS)
@pytest.mark.parametrize("topk", TOP_KS)
@pytest.mark.parametrize("group_size", GROUP_SIZES)
@pytest.mark.skipif(not (ops.supports_moe_ops
and hasattr(torch.ops._moe_C, "marlin_gemm_moe")),
reason="Marlin is not supported on this GPU type.")