environments acft multimodal gpu

acft-multimodal-gpu

Overview

Environment used by Multimodal classification Finetune components

Version: 64

Tags

Preview

View in Studio: https://ml.azure.com/registries/azureml/environments/acft-multimodal-gpu/version/64

Docker image: mcr.microsoft.com/azureml/curated/acft-multimodal-gpu:64

Docker build context

Dockerfile

# PTCA image
FROM mcr.microsoft.com/aifx/acpt/stable-ubuntu2204-cu118-py310-torch271:biweekly.202509.2

USER root
RUN apt-get -y update && apt-get install -y expat libc-bin libc-bin libc-dev locales libc6 dpkg-dev dpkg libdpkg-perl libssl-dev libssl3 openssl


# Install required packages from pypi
COPY requirements.txt .
RUN pip install --upgrade "pip>=25.3"
RUN pip install -r requirements.txt --no-cache-dir
# Downgraded azureml-metrics, to prevent bug related to logger in azureml-metrics. Should be upgraded in next pypi release of acft-multimodal
# Removing it from requirements.txt to avoid conflict with scikit-learn version
RUN pip install azureml-metrics==0.0.33

# Vulnerability fix
RUN pip install --upgrade "pip>=25.3" "starlette>=0.49.1" && \
    pip install datasets==2.15.0 \
    pyarrow==14.0.1 \
    cryptography==44.0.1 \
    gunicorn==23.0.0 \
    onnxruntime==1.17.3 \
    onnx==1.17.0 \
    scikit-learn==1.5.0
RUN pip install azureml-acft-common-components==0.0.83
RUN pip install azureml-acft-accelerator==0.0.83
# Upgrade requests, pip, starlette in the system Python (3.13) for fixing vulnerability
RUN /opt/conda/bin/python3.13 -m pip install --upgrade pip requests urllib3 starlette || true