A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 23:30 UTC
[Bugfix/CI] Turn test_compressed_tensors_2of4_sparse back on (#13250)
Tyler Michael Smith committed
Feb 13, 2025 at 23:19 UTC
09545c0a94800468dc6ef3f01d22a6ce7a70cb30
1 file changed
-1
tests/quantization/test_compressed_tensors.py
-1
@@ -495,7 +495,6 @@ def test_compressed_tensors_2of4_quant_int8(vllm_runner, args_2of4):
495
assert output
496
497
498
-@pytest.mark.skip(reason="2of4 sparse w16a16 CUTLASS produces bad output.")
498
@pytest.mark.skipif(
499
not sparse_cutlass_supported(),
500
reason="2of4 Sparse is not yet supported on this GPU type.",