Git Product home page Git Product logo

maple-ngc's Introduction

Docker

View the published images here: https://hub.docker.com/r/kastanday/maple_ngc_scratch_pip/tags

Regular docker usage

# 1. Must be x64
export DOCKER_DEFAULT_PLATFORM=linux/amd64
# 2. Build image
docker build -t kastanday/ngc_tf1:latest .
# 3. Run image to check installed packages work
docker run -it kastanday/ngc_tf1:latest /bin/bash

Set custom Docker Cache location

Docs on docker buildx

# first enable the buildx driver
docker buildx create --use --driver=docker-container

# build a container with custom platform and cache
sudo docker buildx build \
  --push -t kastanday/pytorch_falcon_aarch64 \
  --platform linux/aarch64 \  # or linux/amd64
  --cache-to type=local,dest=/mnt/teton/utils/docker_cache \
  --cache-from type=local,src=/mnt/teton/utils/docker_cache \
  .

Apptainer (formerly Singularity)

To convert docker -> singularity/apptainer:

First, push the Docker image to the Docker Hub.

sudo docker push kastanday/ngc_tf1:latest

Then pull and convert that image to a local .sif apptainer/singulairty file.

apptainer pull docker://kastanday/ngc_tf1:latest
# WARNING: Fast internet recommended. This takes a long time and uses significant disk space.

Finally, manually transfer this file to Delta. I use Globus for this.

Note: Delta does NOT support building images (like in this command). Note: By default, only the $HOME dir is mounted in the container. You can specify flags to have other dirs mounted too.

Documentation

Nvidia NGC docs: https://docs.nvidia.com/ngc/pdf/NGC-User-Guide.pdf

Delta Apptainer/Singularity guide: https://wiki.ncsa.illinois.edu/display/DSC/Delta+User+Guide#DeltaUserGuide-Apptainer(formerlySingularity)

Delta Supercomputer

Standard run command for single Node. (Multi-node is a bit harder.)

Note: Delta does NOT support building images (yet, so they say). Images must be built elsewhere then transferred to Delta.

export MY_ACCOUNT=<👉YOUR_GPU_ALLOCATION_HERE👈> && \
export CONTAINER=<👉PATH TO LOCAL IMAGE FILE maple_ngc_scratch_pip_mini.sif👈> && \
srun \
 --mem=32g \
 --nodes=1 \
 --ntasks-per-node=1 \
 --cpus-per-task=1 \
 --partition=gpuA100x4 \
 --account=${MY_ACCOUNT} \
 --time=00:30:00 \
 --gpus-per-node=1 \
 --gpus-per-task=1 \
 --gpu-bind=verbose,per_task:1 \
 --pty \
 singularity run --nv ${CONTAINER} /bin/bash

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.