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 5, 14:45 UTC
Bump helion dependency from 0.3.2 to 0.3.3 (#38062)
Signed-off-by: Yanan Cao <gmagogsfm@gmail.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Yanan Cao committed
Apr 2, 2026 at 10:59 UTC
ecd5443dbcd1de360fff067547c68a1128dcfb1b
3 files changed
+6
-3
.buildkite/test-amd.yaml
+1
-1
@@ -790,7 +790,7 @@ steps:
790
- tests/kernels/helion/
791
- vllm/platforms/rocm.py
792
commands:
793
- - pip install helion
793
+ - pip install helion==0.3.3
794
- pytest -v -s kernels/helion/
795
796
.buildkite/test_areas/kernels.yaml
+1
-1
@@ -139,7 +139,7 @@ steps:
139
- vllm/utils/import_utils.py
140
- tests/kernels/helion/
141
commands:
142
- - pip install helion
142
+ - pip install helion==0.3.3
143
- pytest -v -s kernels/helion/
144
145
setup.py
+4
-1
@@ -1063,7 +1063,10 @@ setup(
1063
# Optional deps for AMD FP4 quantization support
1064
"petit-kernel": ["petit-kernel"],
1065
# Optional deps for Helion kernel development
1066
- "helion": ["helion==0.3.2"],
1066
+ # NOTE: When updating helion version, also update CI files:
1067
+ # - .buildkite/test_areas/kernels.yaml
1068
+ # - .buildkite/test-amd.yaml
1069
+ "helion": ["helion==0.3.3"],
1070
# Optional deps for gRPC server (vllm serve --grpc)
1071
"grpc": ["smg-grpc-servicer[vllm] >= 0.5.0"],
1072
# Optional deps for OpenTelemetry tracing