---
apiVersion: kubeflow.org/v1alpha1
kind: MPIJob
metadata:
  labels:
    ksonnet.io/component: tf-resnet50-horovod-job-fsx
  name: tf-resnet50-horovod-job-fsx
  namespace: kubeflow
spec:
  replicas: 2
  template:
    spec:
      containers:
      - command:
        - mpirun
        - -mca
        - btl_tcp_if_exclude
        - lo
        - -mca
        - pml
        - ob1
        - -mca
        - btl
        - ^openib
        - --bind-to
        - none
        - -map-by
        - slot
        - -x
        - LD_LIBRARY_PATH
        - -x
        - PATH
        - -x
        - NCCL_DEBUG=INFO
        - python
        - models/resnet/tensorflow/train_imagenet_resnet_hvd.py
        - --batch_size=256
        - --model=resnet50
        - --num_batches=300
        - --fp16
        - --display_every=50
        - --lr_decay_mode=poly
        - --data_dir=/data/imagenet/train
        image: seedjeffwan/eks-dl-benchmark:cuda10-tf1.13.1-hvd0.16.0-py3.5
        name: tf-resnet50-horovod-job
        resources:
          limits:
            nvidia.com/gpu: 4
        volumeMounts:
        - name: persistent-storage
          mountPath: /data/
      volumes:
      - name: persistent-storage
        persistentVolumeClaim:
          claimName: fsx-static-storage