[CI] Drop duplicate bf16 skinny GEMM test from Kimi K3 B200 job (#59850)
Signed-off-by: khluu <khluu000@gmail.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 5.5
parent
c28834df00
commit
1a001d5842
@@ -124,10 +124,9 @@ steps:
|
||||
- csrc/libtorch_stable/kimi_k3/
|
||||
- tests/models/kimi_k3/
|
||||
- tests/kernels/attention/test_kimi_k3_mla_fused_epilogue.py
|
||||
- tests/kernels/test_bf16_skinny_gemm.py
|
||||
commands:
|
||||
# The native NVIDIA Kimi K3 kernels require the SM100 family.
|
||||
- pytest -v -s models/kimi_k3 --ignore=models/kimi_k3/test_prefix_cache.py kernels/attention/test_kimi_k3_mla_fused_epilogue.py kernels/test_bf16_skinny_gemm.py
|
||||
- pytest -v -s models/kimi_k3 --ignore=models/kimi_k3/test_prefix_cache.py kernels/attention/test_kimi_k3_mla_fused_epilogue.py
|
||||
mirror:
|
||||
amd:
|
||||
label: ":amd: (MI355) Kimi K3"
|
||||
@@ -142,8 +141,8 @@ steps:
|
||||
- vllm/platforms/rocm.py
|
||||
commands:
|
||||
# The ROCm KDA/AttnRes kernels are gated on gfx950, so this has to be MI355.
|
||||
# The two kernels/ files the NVIDIA job adds are CUDA-only (they import
|
||||
# cuda.bindings.driver), so they are deliberately left out here.
|
||||
# The kernels/ file the NVIDIA job adds is CUDA-only (it imports
|
||||
# cuda.bindings.driver), so it is deliberately left out here.
|
||||
- pytest -v -s models/kimi_k3
|
||||
|
||||
- label: ":nvidia: (4xB200) Kimi K3 Prefix Cache Shard %N"
|
||||
|
||||
Reference in New Issue
Block a user