submit_job_config.cfg

protocolVersion: 2
name: bitDelta37
type: job
jobRetryCount: 0
prerequisites:
  - type: dockerimage
    uri: 'nvcr.io/nvidia/pytorch:23.10-py3'
    name: docker_image_0
taskRoles:
  taskrole:
    instances: 1
    completion:
      minFailedInstances: 1
    taskRetryCount: 0
    dockerImage: docker_image_0
    resourcePerInstance:
      gpu: 8
      cpu: 96
      memoryMB: 960000
      ports:
        SynPort: 1
    commands:
      - 'echo ''root:password1234'' |chpasswd'
      - ln -s /mnt/glusterfs-pvc/usr/yibo.han /root/
      - ln -s /mnt/glusterfs-pvc/usr/yibo.han/huggingface /root/.cache/
      - apt update
      - apt upgrade -y
      - apt install -y tmux
      - apt install -y libgl1-mesa-glx 
      - cp ~/yibo.han/Anaconda3-2023.09-0-Linux-x86_64.sh ./
      - >-
        sh Anaconda3-2023.09-0-Linux-x86_64.sh -b -p ~/anaconda &&
        ~/anaconda/bin/conda init
      - echo '' >> ~/.bashrc
      - echo '' >> ~/.bashrc
      - >-
        echo 'alias proxyon="export http_proxy=http://127.0.0.1:10080;export
        https_proxy=http://127.0.0.1:10080"' >> ~/.bashrc
      - echo 'alias proxyoff="unset http_proxy;unset https_proxy"' >> ~/.bashrc
      - >-
        sed -ri 's/^#?PermitRootLogin\s+.*/PermitRootLogin yes/'
        /etc/ssh/sshd_config /usr/sbin/sshd -D &
      - sleep 365d
defaults:
  virtualCluster: default
extras:
  com.microsoft.pai.runtimeplugin:
    - plugin: ssh
      parameters:
        jobssh: true
        userssh:
          type: custom
          value: ''
    - plugin: teamwise_storage
      parameters:
        storageConfigNames:
          - pai-ssd-pvc
          - glusterfs-pvc
  hivedScheduler:
    taskRoles:
      taskrole:
        skuNum: 8
        skuType: gpu-machine