Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

creating new pymarlin docker #73

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 13 additions & 12 deletions Dockerfile
Original file line number Diff line number Diff line change
@@ -1,17 +1,18 @@
FROM mcr.microsoft.com/azureml/base:openmpi3.1.2-ubuntu18.04
RUN apt-get update
FROM mcr.microsoft.com/azureml/openmpi4.1.0-cuda11.1-cudnn8-ubuntu18.04

# install torch latest
RUN conda install pytorch=1.9.1 cudatoolkit=11.1 -c pytorch -c nvidia

# create conda environment
RUN conda update -n base -c defaults conda -y
RUN conda create -n marlin python=3.8 -y
RUN echo ". /opt/miniconda/etc/profile.d/conda.sh" >> ~/.bashrc
# install latest pymarlin published to pip with extra plugin dependencies like transformers (check setup.py)
RUN pip install pymarlin[plugins] --ignore-installed

#install torch latest
# Cuda toolkit other than 1.2 makes GPUs invisible. Base image issue
RUN conda install pytorch cudatoolkit=10.2 -c pytorch -y -n marlin
# install opacus
RUN pip install opacus

ADD . /workdir
WORKDIR /workdir
# install deepspeed
RUN pip install deepspeed

RUN /opt/miniconda/envs/marlin/bin/pip install -U -e .
# install ORT
RUN pip install onnxruntime-training -f https://download.onnxruntime.ai/onnxruntime_stable_cu111.html
RUN pip install torch-ort --ignore-installed
RUN python -m torch_ort.configure