Latest vllm custom build on CUDA 12.8.1, Torch 2.9.1, Vllm 0.9.1 Target to support Nvidia blackwell architecture (RTX 5000 series, RTX PRO series, B200)
Content type
Image
Digest
sha256:0f25ba95d…
Size
14.5 GB
Last updated
about 1 year ago
docker pull utarn/vllm