# vLLM 0.20+ ships an aarch64 manylinux wheel on PyPI whose Requires-Dist pins
# torch==2.11.0 / torchvision==0.26.0 / torchaudio==2.11.0, locking an ABI-
# consistent set with the cu130 torch wheel installed above.
#
# Pinned to match the cublas13 build (requirements-cublas13-after.txt) and to
# stay deterministic on GB10 / DGX Spark: 0.24.0 carries vllm-project/vllm#45179
# ("release cached device memory under pressure on UMA GPUs during weight
# loading"), without which cold model loads on the Grace Blackwell unified-
# memory architecture crash deterministically with an empty "Engine core init
# failed" set (mudler/LocalAI#10722). Leaving this unpinned let the L4T image
# drift onto whatever wheel was latest at build time.
vllm==0.24.0