Usage Examples

Note

Make sure you use TensorFlow on the Intel® Gaudi® software version 1.14.0 or below before proceeding.

hl-smi Example

1. Create a hl-smi.yaml file:

   apiVersion: v1
   kind: Pod
   metadata:
       name: habanalabs-gaudi-demo
   spec:
       containers:
       - name: habana-ai-base-container
       image: vault.habana.ai/gaudi-docker/1.14.0/ubuntu22.04/habanalabs/tensorflow-installer-tf-cpu-2.15.0:latest
       workingDir: /root
       command: ["hl-smi"]
       args: [""]
       resources:
           limits:
           habana.ai/gaudi: 1
   

2. Apply the .yaml file:

   oc apply -f hl-smi.yaml -n habana-system
   

3. Check the results:

   oc logs habanalabs-gaudi-demo -n habana-system
   

mnist Example

1. Create a Dockerfile_model file:

   FROM vault.habana.ai/gaudi-docker/1.14.0/ubuntu22.04/habanalabs/tensorflow-installer-tf-cpu-2.15.0:latest
   git clone -b 1.14.0 https://github.com/HabanaAI/Model-References.git
   

2. Build and run the Docker container. Make sure the Docker image is built on the same operating system where the Dockerfile was created.

   docker build -f Dockerfile_model -t mnist-example:rhel8.6 .
   

3. Once the build is finished, upload the image to your local Docker registry:

   docker login <your docker registry URL>
   docker push  <your docker registry URL/path (where your docker is stored) images/mnist-example:rhel8.6>
   

4. Create pod-mnist.yaml file:

   apiVersion: v1
   kind: Pod
   metadata:
       name: habanalabs-gaudi-demo2
   spec:
       containers:
       - name: habana-ai-base-container
       image: <PATH to the image in step a>
       workingDir: /
       command: ["python3"]
       args: ["/Model-References/TensorFlow/examples/hello_world/example.py"]
       resources:
           limits:
           habana.ai/gaudi: 1
   

5. Apply the pod-mnist.yaml file:

   oc apply -f pod-mnist.yaml -n habana-system
   

6. Check the results:

   oc logs habanalabs-gaudi-demo2 -n habana-system