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 3, 19:00 UTC
[CI] Use uv with Python 3.12 for PyPI wheel upload (#42470)
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Kevin H. Luu committed
May 13, 2026 at 02:12 UTC
f6e868fbdf23eb5fcab4a91b450e96837b72bda8
1 file changed
+10
-4
.buildkite/scripts/upload-release-wheels-pypi.sh
+10
-4
@@ -39,11 +39,17 @@ fi
39
40
set -x # avoid printing secrets above
41
42
-# install twine and sdist build prerequisites from pypi
43
-python3 -m venv /tmp/vllm-release-env
42
+# install uv if not already available
43
+if ! command -v uv &> /dev/null; then
44
+ curl -LsSf https://astral.sh/uv/install.sh | UV_VERSION=0.11.14 sh
45
+ export PATH="$HOME/.local/bin:$PATH"
46
+fi
47
+
48
+# install twine and sdist build prerequisites using uv with Python 3.12
49
+uv venv --python 3.12 /tmp/vllm-release-env
50
source /tmp/vllm-release-env/bin/activate
45
-pip install twine
46
-pip install -r requirements/build/cuda.txt
51
+uv pip install twine
52
+uv pip install -r requirements/build/cuda.txt
53
python3 -m twine --version
54
55
# copy release wheels to local directory