Sign inSign up

clustermax/cmax-inference:0.2.0

Index digest

sha256:a9fbd980f90da689d96a932a928d4812bad6cedb0d83d84951e02064cf91595d

OS/ARCH

linux/arm64

Compressed size

8.71 GB

Last pushed

3 months by clustermax

Type

Image

Manifest digest

sha256:7efca7b2f6df19793169cdf6b37d706e5f86edd2c1ddc1d3accf9ca340e05674

Image Layers

Image Layers

1ARG RELEASE0 B
2ARG LAUNCHPAD_BUILD_ARCH0 B
3LABEL org.opencontainers.image.ref.name=ubuntu0 B
4LABEL org.opencontainers.image.version=24.040 B
5ADD file ... in / 28.55 MB
6CMD ["/bin/bash"]0 B
7ENV NVARCH=sbsa0 B
8ENV NVIDIA_REQUIRE_CUDA=cuda>=13.0 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<5360 B
9ENV NV_CUDA_CUDART_VERSION=13.0.96-10 B
10ARG TARGETARCH0 B
11LABEL maintainer=NVIDIA CORPORATION <[email protected]>0 B
12RUN |1 TARGETARCH=arm64 /bin/sh -c4.32 MB
13ENV CUDA_VERSION=13.0.20 B
14RUN |1 TARGETARCH=arm64 /bin/sh -c95.11 MB
15RUN |1 TARGETARCH=arm64 /bin/sh -c186 B
16ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin0 B
17ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
18COPY NGC-DL-CONTAINER-LICENSE / # buildkit6.72 KB
19ENV NVIDIA_VISIBLE_DEVICES=all0 B
20ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility0 B
21ARG CUDA_VERSION0 B
22ARG PYTHON_VERSION0 B
23ARG DEADSNAKES_MIRROR_URL0 B
24ARG DEADSNAKES_GPGKEY_URL0 B
25ARG GET_PIP_URL0 B
26ENV DEBIAN_FRONTEND=noninteractive0 B
27WORKDIR /vllm-workspace104 B
28RUN |5 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=231 B
29RUN |5 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=216.06 MB
30RUN |5 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=1.21 GB
31RUN |5 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=45.55 MB
32ENV UV_HTTP_TIMEOUT=5000 B
33ENV UV_INDEX_STRATEGY=unsafe-best-match0 B
34ENV UV_LINK_MODE=copy0 B
35ENV VLLM_ENABLE_CUDA_COMPATIBILITY=00 B
36ARG PYTORCH_CUDA_INDEX_BASE_URL0 B
37COPY requirements/common.txt /tmp/common.txt # buildkit1.52 KB
38COPY requirements/cuda.txt /tmp/requirements-cuda.txt # buildkit628 B
39RUN |6 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=3.29 GB
40ARG FLASHINFER_VERSION=0.6.8.post10 B
41RUN |7 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=2.2 GB
42ARG GDRCOPY_CUDA_VERSION=12.80 B
43ARG GDRCOPY_OS_VERSION=Ubuntu22_040 B
44ARG TARGETPLATFORM0 B
45COPY tools/install_gdrcopy.sh /tmp/install_gdrcopy.sh # buildkit897 B
46RUN |10 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=518.41 KB
47ARG BITSANDBYTES_VERSION_X86=0.46.10 B
48ARG BITSANDBYTES_VERSION_ARM64=0.42.00 B
49ARG TIMM_VERSION=>=1.0.170 B
50ARG RUNAI_MODEL_STREAMER_VERSION=>=0.15.70 B
51RUN |14 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=281.07 MB
52ARG PIP_INDEX_URL UV_INDEX_URL0 B
53ARG PIP_EXTRA_INDEX_URL UV_EXTRA_INDEX_URL0 B
54ARG PYTORCH_CUDA_INDEX_BASE_URL0 B
55ARG PIP_KEYRING_PROVIDER UV_KEYRING_PROVIDER0 B
56ARG PYTORCH_NIGHTLY0 B
57COPY /workspace/torch_lib_versions.txt torch_lib_versions.txt # buildkit200 B
58RUN |22 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=433.1 MB
59RUN |22 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=124 B
60ENV LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
61RUN |22 CUDA_VERSION=13.0.2 PYTHON_VERSION=3.12 DEADSNAKES_MIRROR_URL=10.86 MB
62ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
63COPY examples examples # buildkit241.11 KB
64COPY benchmarks benchmarks # buildkit223.6 KB
65COPY ./vllm/collect_env.py . # buildkit9.62 KB
66ARG TARGETPLATFORM0 B
67ARG INSTALL_KV_CONNECTORS=false0 B
68ARG CUDA_VERSION0 B
69ARG PIP_INDEX_URL UV_INDEX_URL0 B
70ARG PIP_EXTRA_INDEX_URL UV_EXTRA_INDEX_URL0 B
71ENV UV_HTTP_TIMEOUT=5000 B
72ARG torch_cuda_arch_list=7.5 8.0 8.6 8.90 B
73ENV TORCH_CUDA_ARCH_LIST=8.0 8.7 8.9 9.00 B
74RUN |8 TARGETPLATFORM=linux/arm64 INSTALL_KV_CONNECTORS=true CUDA_VERSION=13.0.2863.93 MB
75ENV VLLM_USAGE_SOURCE=production-docker-image0 B
76ENTRYPOINT ["vllm" "serve"]0 B
77LABEL ai.clustermax.image=cmax-inference ai.clustermax.base=vllm/vllm-openai:v0.20.0-aarch64-cu130-ubuntu2404 ai.clustermax.base.digest=sha256:f81415d5682a03566d92dc46e883d73a6ac001e6ef1fc4293512269f1372b8b9 ai.clustermax.vllm.version=0.20.00 B
78RUN /bin/sh -c pip install78.38 MB

Command

 ARG RELEASE