Sign inSign up

runpod/worker-v1-vllm:dev-refs-pull-356-merge

Index digest

sha256:b3a495823b460dd369992a954d29fc22d62de48fed1e5b64973f9cdb4c790407

OS/ARCH

linux/amd64

Compressed size

8.68 GB

Last pushed

8 days by runpod

Type

Image

Manifest digest

sha256:eef52dbf942c327867c8412e0c2df2295ede5a600a4dfd52e6dc7f2792077bb0

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 / 29.18 MB
6CMD ["/bin/bash"]0 B
7ENV NVARCH=x86_640 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=amd64 /bin/sh -c4.34 MB
13ENV CUDA_VERSION=13.0.20 B
14RUN |1 TARGETARCH=amd64 /bin/sh -c100.29 MB
15RUN |1 TARGETARCH=amd64 /bin/sh -c182 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 INSTALL_RUBIN_PRERELEASE0 B
22ARG TARGETPLATFORM0 B
23ARG max_jobs=20 B
24ARG CUDA_VERSION0 B
25ARG PYTHON_VERSION0 B
26ARG NCCL_VERSION0 B
27ARG DEADSNAKES_MIRROR_URL0 B
28ARG DEADSNAKES_GPGKEY_URL0 B
29ARG GET_PIP_URL0 B
30ENV CUDA_VERSION=13.0.20 B
31ENV DEBIAN_FRONTEND=noninteractive0 B
32WORKDIR /vllm-workspace104 B
33RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=16230 B
34RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=16222.66 MB
35RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=161.35 GB
36ENV CUDA_HOME=/usr/local/cuda0 B
37ENV PATH=/usr/local/cuda/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin0 B
38ENV LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
39RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=1632 B
40COPY tools/install_snapshot_runtime.sh /usr/local/bin/install_snapshot_runtime.sh # buildkit1.62 KB
41RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=168.91 MB
42ENV CRIU_CUDA_PLUGIN_DIR=/usr/local/lib/criu0 B
43RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=1639.87 MB
44ENV UV_HTTP_TIMEOUT=5000 B
45ENV UV_INDEX_STRATEGY=unsafe-best-match0 B
46ENV UV_LINK_MODE=copy0 B
47ENV UV_PYTHON_INSTALL_DIR=/opt/uv/python0 B
48ENV UV_CACHE_DIR=/opt/uv/cache0 B
49RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=16150 B
50RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=16162 B
51ENV UV_OVERRIDE=/etc/uv-overrides.txt0 B
52RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=163.64 KB
53COPY docker/entrypoints/vllm-nonroot-entrypoint.sh /usr/local/bin/vllm-nonroot-entrypoint.sh # buildkit2.03 KB
54RUN |9 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=162.03 KB
55ENV VLLM_ENABLE_CUDA_COMPATIBILITY=00 B
56ARG PYTORCH_CUDA_INDEX_BASE_URL0 B
57COPY requirements/common.txt /tmp/common.txt # buildkit1.67 KB
58COPY requirements/cuda.txt /tmp/requirements-cuda.txt # buildkit870 B
59RUN |10 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=164.8 GB
60ARG FLASHINFER_VERSION=0.7.0.post10 B
61RUN |11 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=161.18 GB
62ARG GDRCOPY_CUDA_VERSION=12.80 B
63ARG GDRCOPY_OS_VERSION=Ubuntu24_040 B
64ARG TARGETPLATFORM0 B
65COPY tools/install_gdrcopy.sh /tmp/install_gdrcopy.sh # buildkit897 B
66RUN |14 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=16539.32 KB
67ARG TIMM_VERSION=>=1.0.170 B
68ARG RUNAI_MODEL_STREAMER_VERSION=>=0.15.70 B
69RUN |16 INSTALL_RUBIN_PRERELEASE=false TARGETPLATFORM=linux/amd64 max_jobs=1659.14 MB
70ARG PIP_INDEX_URL UV_INDEX_URL0 B
71ARG PIP_EXTRA_INDEX_URL UV_EXTRA_INDEX_URL0 B
72ARG PYTORCH_CUDA_INDEX_BASE_URL0 B
73ARG PIP_KEYRING_PROVIDER UV_KEYRING_PROVIDER0 B
74ARG PYTORCH_NIGHTLY0 B
75COPY /workspace/torch_lib_versions.txt torch_lib_versions.txt # buildkit198 B
76COPY /workspace/dist/wheel.sha256 /tmp/vllm-wheel.sha256 # buildkit226 B
77RUN |8 PIP_INDEX_URL= UV_INDEX_URL= PIP_EXTRA_INDEX_URL=300.88 MB
78RUN |8 PIP_INDEX_URL= UV_INDEX_URL= PIP_EXTRA_INDEX_URL=5.22 KB
79RUN |8 PIP_INDEX_URL= UV_INDEX_URL= PIP_EXTRA_INDEX_URL=32 B
80ENV LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
81COPY /tmp/ep_kernels_workspace/dist/wheels.sha256 /tmp/ep-kernels-wheels.sha256 # buildkit251 B
82RUN |8 PIP_INDEX_URL= UV_INDEX_URL= PIP_EXTRA_INDEX_URL=13.81 MB
83ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/cuda/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib640 B
84COPY examples examples # buildkit298.11 KB
85COPY benchmarks benchmarks # buildkit297.15 KB
86COPY ./vllm/collect_env.py . # buildkit9.61 KB
87ARG TARGETPLATFORM0 B
88ARG INSTALL_KV_CONNECTORS=false0 B
89ARG INSTALL_RUBIN_PRERELEASE0 B
90ARG TRITON_INSTALL_FROM_SOURCE_REPO0 B
91ARG CUDA_VERSION0 B
92ARG VLLM_BUILD_COMMIT0 B
93ARG VLLM_BUILD_PIPELINE0 B
94ARG VLLM_BUILD_URL0 B
95ARG VLLM_IMAGE_TAG0 B
96ARG PIP_INDEX_URL UV_INDEX_URL0 B
97ARG PIP_EXTRA_INDEX_URL UV_EXTRA_INDEX_URL0 B
98ENV UV_HTTP_TIMEOUT=5000 B
99ARG torch_cuda_arch_list=7.5 8.0 8.6 8.90 B
100ENV TORCH_CUDA_ARCH_LIST=7.5 8.0 8.6 8.90 B
101RUN |14 TARGETPLATFORM=linux/amd64 INSTALL_KV_CONNECTORS=true INSTALL_RUBIN_PRERELEASE=false329.6 MB
102RUN |14 TARGETPLATFORM=linux/amd64 INSTALL_KV_CONNECTORS=true INSTALL_RUBIN_PRERELEASE=false92 B
103COPY /tmp/triton_from_source_workspace/dist/wheel.sha256 /tmp/triton-source-wheel.sha256 # buildkit212 B
104COPY tools/install_triton_from_source.sh /tmp/install_triton_from_source.sh # buildkit824 B
105RUN |14 TARGETPLATFORM=linux/amd64 INSTALL_KV_CONNECTORS=true INSTALL_RUBIN_PRERELEASE=false183 B
106ENV VLLM_USAGE_SOURCE=production-docker-image0 B
107ENV VLLM_BUILD_COMMIT=db9527a46873454610df6dbedf79a36d6bf1a7f6 VLLM_BUILD_PIPELINE=019d130e-464e-4ff7-b84b-492992c0c06b VLLM_BUILD_URL=https://buildkite.com/vllm/release-v2/builds/7746 VLLM_IMAGE_TAG=vllm/vllm-openai:v0.31.00 B
108LABEL org.opencontainers.image.source=https://github.com/vllm-project/vllm org.opencontainers.image.revision=db9527a46873454610df6dbedf79a36d6bf1a7f6 org.opencontainers.image.version=vllm/vllm-openai:v0.31.0 org.opencontainers.image.url=https://buildkite.com/vllm/release-v2/builds/77460 B
109ENTRYPOINT ["vllm" "serve"]0 B
110ARG VLLM_VERSION=v0.31.00 B
111COPY builder/requirements.txt /requirements.txt # buildkit451 B
112RUN |1 VLLM_VERSION=v0.31.0 /bin/sh -c247.46 MB
113RUN |1 VLLM_VERSION=v0.31.0 /bin/sh -c53.65 KB
114RUN |1 VLLM_VERSION=v0.31.0 /bin/sh -c21.68 MB
115ARG MODEL_NAME=0 B
116ARG MODEL_REVISION=0 B
117ARG TOKENIZER_NAME=0 B
118ARG TOKENIZER_REVISION=0 B
119ARG QUANTIZATION=0 B
120ARG BASE_PATH=/runpod-volume0 B
121ENV MODEL_NAME= MODEL_REVISION= TOKENIZER_NAME= TOKENIZER_REVISION=0 B
122COPY src /src # buildkit17.82 KB
123RUN |7 VLLM_VERSION=v0.31.0 MODEL_NAME= MODEL_REVISION=32 B
124ENTRYPOINT ["python3" "/src/main.py"]0 B

Command

 ARG RELEASE