Container registry for ML (Machine Learning)

Repositories

1

Projects Packages

2

mnn (sha256:0097ea1ca19abb435e81fcab7b688bc52edba4b260c1087724bb89ebad8ef366)

Published 2026-01-05 02:03:36 +09:00 by enbay

Installation

Pull the image from the command line:

docker pull git.mattereal.org/ml_container/mnn@sha256:0097ea1ca19abb435e81fcab7b688bc52edba4b260c1087724bb89ebad8ef366

Digest

sha256:0097ea1ca19abb435e81fcab7b688bc52edba4b260c1087724bb89ebad8ef366

For more information on the Container registry, see the documentation.

Image layers

ARG RELEASE

ARG LAUNCHPAD_BUILD_ARCH

LABEL org.opencontainers.image.ref.name=ubuntu

LABEL org.opencontainers.image.version=22.04

ADD file:d025507456f1d7d19195885b1c02a346454d60c9348cbd3be92431f2d7e2666e in /

CMD ["/bin/bash"]

RUN /bin/sh -c apt update # buildkit

RUN /bin/sh -c apt install python3.10 pip gcc g++ clang cmake rustc cargo git -y # buildkit

WORKDIR /opt

RUN /bin/sh -c git clone https://git.mattereal.org/ai-model-converter/MNN # buildkit

WORKDIR /opt/MNN

RUN /bin/sh -c mkdir build # buildkit

WORKDIR /opt/MNN/build

RUN /bin/sh -c cmake .. -DMNN_BUILD_CONVERTER=true -DMNN_LOW_MEMORY=true -DMNN_CPU_WEIGHT_DEQUANT_GEMM=true -DMNN_BUILD_LLM=true -DMNN_SUPPORT_TRANSFORMER_FUSE=true # buildkit

RUN /bin/sh -c make -j # buildkit

RUN /bin/sh -c make install # buildkit

WORKDIR /opt/MNN/transformers/llm/export

RUN /bin/sh -c pip install --no-cache-dir torch --index-url https://download.pytorch.org/whl/cpu # buildkit

RUN /bin/sh -c pip install --no-cache-dir -r requirements.txt # buildkit

ENV LD_LIBRARY_PATH=/opt/MNN/build

ENV PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/MNN/transformers/llm/export:/opt/MNN/build

Labels

Key	Value
org.opencontainers.image.ref.name	ubuntu
org.opencontainers.image.version	22.04

Details

Container

2026-01-05 02:03:36 +09:00

0

OCI / Docker

linux/amd64

1.7 GiB

Versions (1) View all

ubuntu-22.04 2026-01-05