Sign inSign up

docker/model-runner:v1.2.8-vllm-cuda

Multi-platform
Index digest

sha256:eaa0654180a88d6ebb6524f6f339e6a9cb32ae0f2441a5cc14a1573bc25e201d

OS/ARCH

Compressed size

5.08 GB

Last pushed

about 1 month by docker

Type

Image

Manifest digest

sha256:d74aaf61216261180e8008a7f5521dc47fceb683b1b7b8725770c234cc97a718

Image Layers

Image Layers

1ARG RELEASE0 B
2ARG LAUNCHPAD_BUILD_ARCH0 B
3LABEL org.opencontainers.image.version=24.040 B
4ADD file ... in / 28.36 MB
5CMD ["/bin/bash"]0 B
6ENV NVARCH=x86_640 B
7ENV NVIDIA_REQUIRE_CUDA=cuda>=13.3 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<5360 B
8ENV NV_CUDA_CUDART_VERSION=13.3.29-10 B
9ARG TARGETARCH0 B
10LABEL maintainer=NVIDIA CORPORATION <[email protected]>0 B
11ARG IMAGE_UPGRADE_ID=00 B
12RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260527 /bin/sh6.42 MB
13ENV CUDA_VERSION=13.3.00 B
14RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260527 /bin/sh378.73 KB
15RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260527 /bin/sh151.54 MB
16RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260527 /bin/sh32 B
17RUN |2 TARGETARCH=amd64 IMAGE_UPGRADE_ID=20260527 /bin/sh184 B
18ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin0 B
19ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
20COPY NGC-DL-CONTAINER-LICENSE / # buildkit6.72 KB
21ENV NVIDIA_VISIBLE_DEVICES=all0 B
22ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility0 B
23ENV NV_CUDA_LIB_VERSION=13.3.0-10 B
24ENV NV_NVTX_VERSION=13.3.29-10 B
25ENV NV_LIBNPP_VERSION=13.1.2.48-10 B
26ENV NV_LIBNPP_PACKAGE=libnpp-13-3=13.1.2.48-10 B
27ENV NV_LIBCUSPARSE_VERSION=12.8.1.7-10 B
28ENV NV_LIBCUBLAS_PACKAGE_NAME=libcublas-13-30 B
29ENV NV_LIBCUBLAS_VERSION=13.5.1.27-10 B
30ENV NV_LIBCUBLAS_PACKAGE=libcublas-13-3=13.5.1.27-10 B
31ARG TARGETARCH0 B
32LABEL maintainer=NVIDIA CORPORATION <[email protected]>0 B
33RUN |1 TARGETARCH=amd64 /bin/sh -c1.19 GB
34RUN |1 TARGETARCH=amd64 /bin/sh -c58.43 KB
35COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit1.64 KB
36COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit1.49 KB
37ENV NVIDIA_PRODUCT_NAME=CUDA0 B
38ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]0 B
39ARG BUILD_DATE=2026-07-06T05:42:04Z0 B
40ARG APP_VERSION=b98790 B
41ARG APP_REVISION=72874f559c598b8f89fbb24864868337cf5afb4c0 B
42ARG IMAGE_URL=https://github.com/ggml-org/llama.cpp0 B
43ARG IMAGE_SOURCE=https://github.com/ggml-org/llama.cpp0 B
44LABEL org.opencontainers.image.created=2026-07-06T05:42:04Z org.opencontainers.image.version=b9879 org.opencontainers.image.revision=72874f559c598b8f89fbb24864868337cf5afb4c org.opencontainers.image.title=llama.cpp0 B
45RUN |5 BUILD_DATE=2026-07-06T05:42:04Z APP_VERSION=b9879 APP_REVISION=72874f559c598b8f89fbb24864868337cf5afb4c247.29 MB
46COPY /app/lib/ /app # buildkit153.33 MB
47ENV LLAMA_ARG_HOST=0.0.0.00 B
48COPY /app/full/llama /app/full/llama-server /app #22.7 KB
49WORKDIR /app32 B
50HEALTHCHECK {Test:[CMD curl -f http://localhost:8080/health]0 B
51ENTRYPOINT ["/app/llama-server"]0 B
52ARG LLAMA_SERVER_VARIANT=cuda0 B
53RUN |1 LLAMA_SERVER_VARIANT=cuda /bin/sh -c3.54 KB
54COPY scripts/ /scripts/ # buildkit10.91 KB
55RUN |1 LLAMA_SERVER_VARIANT=cuda /bin/sh -c504.82 KB
56WORKDIR /app32 B
57RUN |1 LLAMA_SERVER_VARIANT=cuda /bin/sh -c153.36 MB
58USER modelrunner0 B
59ENV MODEL_RUNNER_SOCK=/var/run/model-runner/model-runner.sock0 B
60ENV MODEL_RUNNER_PORT=124340 B
61ENV LLAMA_SERVER_PATH=/app0 B
62ENV LD_LIBRARY_PATH=/app0 B
63ENV HOME=/home/modelrunner0 B
64ENV MODELS_PATH=/models0 B
65LABEL com.docker.desktop.service=model-runner0 B
66EXPOSE [12434/tcp]0 B
67HEALTHCHECK {Test:[CMD-SHELL curl -f "http://localhost:${MODEL_RUNNER_PORT}/engines/status"0 B
68ENTRYPOINT ["/app/model-runner"]0 B
69ARG VLLM_VERSION=0.19.10 B
70ARG VLLM_CUDA_VERSION=cu1300 B
71ARG VLLM_PYTHON_TAG=cp38-abi30 B
72ARG TARGETARCH=amd640 B
73USER root0 B
74RUN |5 LLAMA_SERVER_VARIANT=cuda VLLM_VERSION=0.19.1 VLLM_CUDA_VERSION=cu130130.54 MB
75RUN |5 LLAMA_SERVER_VARIANT=cuda VLLM_VERSION=0.19.1 VLLM_CUDA_VERSION=cu130126 B
76USER modelrunner0 B
77RUN |5 LLAMA_SERVER_VARIANT=cuda VLLM_VERSION=0.19.1 VLLM_CUDA_VERSION=cu1303.01 GB
78RUN |5 LLAMA_SERVER_VARIANT=cuda VLLM_VERSION=0.19.1 VLLM_CUDA_VERSION=cu13020.73 MB
79COPY /app/model-runner /app/model-runner # buildkit5 MB

Command

 ARG RELEASE