Sign inSign up

sparksnail/prism-infer:v0.3.0-qwen3-8b

Manifest digest

sha256:f434a607e2d87ff3fe0c9dded2dbf2ec4123f67adb21878f4ab82c91cda9f058

OS/ARCH

linux/amd64

Compressed size

15.44 GB

Last pushed

about 1 month by sparksnail

Type

Image

Manifest digest

sha256:f434a607e2d87ff3fe0c9dded2dbf2ec4123f67adb21878f4ab82c91cda9f058

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=22.040 B
5ADD file ... in / 29.03 MB
6CMD ["/bin/bash"]0 B
7ARG PYTORCH_VERSION=2.6.00 B
8ARG TRITON_VERSION=0 B
9ARG TARGETPLATFORM=linux/amd640 B
10ARG CUDA_VERSION=12.4.10 B
11LABEL com.nvidia.volumes.needed=nvidia_driver0 B
12RUN |4 PYTORCH_VERSION=2.6.0 TRITON_VERSION= TARGETPLATFORM=linux/amd646.87 MB
13COPY /opt/conda /opt/conda # buildkit3.05 GB
14RUN |4 PYTORCH_VERSION=2.6.0 TRITON_VERSION= TARGETPLATFORM=linux/amd6432 B
15ENV PATH=/opt/conda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin0 B
16ENV NVIDIA_VISIBLE_DEVICES=all0 B
17ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility0 B
18ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib640 B
19ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/opt/conda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin0 B
20ENV PYTORCH_VERSION=2.6.00 B
21WORKDIR /workspace99 B
22ENV PYTHONDONTWRITEBYTECODE=1 PYTHONUNBUFFERED=1 HF_HUB_DISABLE_TELEMETRY=10 B
23WORKDIR /app92 B
24RUN /bin/sh -c apt-get update86.15 MB
25RUN /bin/sh -c python -m230.84 MB
26ENV PRISM_IMAGE_VARIANT=performance PRISM_MODEL_PROFILE=qwen3-8b-bf16-tp1 PRISM_MODEL=/opt/models/Qwen3-8B PRISM_MODEL_ID=Qwen/Qwen3-8B0 B
27COPY /opt/models/Qwen3-8B/.prism-model-revision /opt/models/Qwen3-8B/config.json /opt/models/Qwen3-8B/generation_config.json /opt/models/Qwen3-8B/merges.txt3.75 MB
28COPY /opt/models/Qwen3-8B/model-00001-of-00005.safetensors /opt/models/Qwen3-8B/ # buildkit2.95 GB
29COPY /opt/models/Qwen3-8B/model-00002-of-00005.safetensors /opt/models/Qwen3-8B/ # buildkit2.92 GB
30COPY /opt/models/Qwen3-8B/model-00003-of-00005.safetensors /opt/models/Qwen3-8B/ # buildkit2.9 GB
31COPY /opt/models/Qwen3-8B/model-00004-of-00005.safetensors /opt/models/Qwen3-8B/ # buildkit2.33 GB
32COPY /opt/models/Qwen3-8B/model-00005-of-00005.safetensors /opt/models/Qwen3-8B/ # buildkit926.33 MB
33COPY pyproject.toml README.md LICENSE ./4.28 KB
34COPY prism_infer ./prism_infer # buildkit84.01 KB
35ARG GIT_SHA0 B
36ARG SOURCE_URL=https://github.com/SparkSnail/prism-infer0 B
37ENV PRISM_IMAGE_GIT_SHA=dfe0cecf57fc9838841ac7b360fac7b9c68044160 B
38RUN |2 GIT_SHA=dfe0cecf57fc9838841ac7b360fac7b9c6804416 SOURCE_URL=https://github.com/SparkSnail/prism-infer /bin/sh446.06 KB
39ENV HF_HUB_OFFLINE=1 TRANSFORMERS_OFFLINE=1 PIP_NO_CACHE_DIR=10 B
40LABEL org.opencontainers.image.source=https://github.com/SparkSnail/prism-infer org.opencontainers.image.revision=dfe0cecf57fc9838841ac7b360fac7b9c6804416 org.opencontainers.image.title=prism-infer 2P2D0 B
41EXPOSE map[29500/tcp:{} 8001/tcp:{}]0 B
42STOPSIGNAL SIGTERM0 B
43ENTRYPOINT ["prism-infer-worker"]0 B
44ARG GIT_SHA=af553c2aff0ebcce8113df7aac57ed4c4e99bc8a0 B
45ENV PRISM_IMAGE_GIT_SHA=af553c2aff0ebcce8113df7aac57ed4c4e99bc8a0 B
46LABEL org.opencontainers.image.revision=af553c2aff0ebcce8113df7aac57ed4c4e99bc8a0 B
47ARG GIT_SHA=9224c2981eba8246789bca09daf753568095471b0 B
48ARG PRISM_RELEASE=true0 B
49ARG PRISM_IMAGE_VARIANT=performance0 B
50ARG SOURCE_URL=https://github.com/SparkSnail/prism-infer.git0 B
51WORKDIR /app32 B
52COPY pyproject.toml README.md LICENSE ./6.19 KB
53COPY prism_infer ./prism_infer # buildkit382.39 KB
54ENV PRISM_IMAGE_GIT_SHA=9224c2981eba8246789bca09daf753568095471b PRISM_IMAGE_VARIANT=performance0 B
55RUN |4 GIT_SHA=9224c2981eba8246789bca09daf753568095471b PRISM_RELEASE=true PRISM_IMAGE_VARIANT=performance337.41 KB
56ENV HF_HUB_OFFLINE=1 TRANSFORMERS_OFFLINE=1 HOME=/var/run/prism TRITON_HOME=/var/run/prism/triton0 B
57LABEL org.opencontainers.image.source=https://github.com/SparkSnail/prism-infer.git org.opencontainers.image.revision=9224c2981eba8246789bca09daf753568095471b org.opencontainers.image.title=prism-infer 2P2D0 B
58USER prism0 B
59EXPOSE [29500/tcp 8001/tcp]0 B
60STOPSIGNAL SIGTERM0 B
61ENTRYPOINT ["prism-infer-worker"]0 B
62USER root0 B
63COPY /verified-model/ /tmp/verified-model/ # buildkit3.75 MB
64RUN /bin/sh -c python -3.75 MB
65USER root0 B
66ARG INFER_GIT_SHA0 B
67ARG PRISM_RELEASE=false0 B
68ARG INFER_SOURCE_URL=https://github.com/SparkSnail/prism-infer.git0 B
69WORKDIR /app32 B
70RUN |3 INFER_GIT_SHA=9224c2981eba8246789bca09daf753568095471b PRISM_RELEASE=true INFER_SOURCE_URL=https://github.com/SparkSnail/prism-infer.git286 B
71COPY /opt/conda/lib/python3.11/site-packages/numpy /opt/conda/lib/python3.11/site-packages/numpy # buildkit12.01 MB
72COPY /opt/conda/lib/python3.11/site-packages/numpy.libs /opt/conda/lib/python3.11/site-packages/numpy.libs # buildkit8.09 MB
73COPY /opt/conda/lib/python3.11/site-packages/numpy-2.2.6.dist-info /opt/conda/lib/python3.11/site-packages/numpy-2.2.6.dist-info # buildkit80.03 KB
74COPY /opt/conda/lib/python3.11/site-packages/safetensors /opt/conda/lib/python3.11/site-packages/safetensors # buildkit478.79 KB
75COPY /opt/conda/lib/python3.11/site-packages/safetensors-0.5.3.dist-info /opt/conda/lib/python3.11/site-packages/safetensors-0.5.3.dist-info # buildkit2.21 KB
76COPY /pyproject.toml /README.md /LICENSE ./6.19 KB
77COPY /prism_infer ./prism_infer # buildkit382.38 KB
78ENV PRISM_IMAGE_GIT_SHA=9224c2981eba8246789bca09daf753568095471b0 B
79RUN |3 INFER_GIT_SHA=9224c2981eba8246789bca09daf753568095471b PRISM_RELEASE=true INFER_SOURCE_URL=https://github.com/SparkSnail/prism-infer.git326.41 KB
80ENV HF_HUB_OFFLINE=1 TRANSFORMERS_OFFLINE=1 HOME=/var/run/prism TRITON_HOME=/var/run/prism/triton0 B
81LABEL org.opencontainers.image.source=https://github.com/SparkSnail/prism-infer.git org.opencontainers.image.revision=9224c2981eba8246789bca09daf753568095471b org.opencontainers.image.title=prism-infer 2P2D0 B
82USER prism0 B
83EXPOSE map[29500/tcp:{} 8001/tcp:{}]0 B
84STOPSIGNAL SIGTERM0 B
85ENTRYPOINT ["prism-infer-worker"]0 B

Command

 ARG RELEASE