Run Cloud TPU applications in a Docker container

Docker containers make configuring applications easier by combining your code and all needed dependencies in one distributable package. You can run Docker containers within TPU VMs to simplify configuring and sharing your Cloud TPU applications. This document describes how to set up a Docker container for each ML framework supported by Cloud TPU.

Important: You can run a maximum of 256 Docker pods on a TPU VM.

Train a TensorFlow model in a Docker container

TPU device

1. Create a file named Dockerfile in your current directory and paste the following text

   FROM python:3.8
   RUN pip install https://storage.googleapis.com/cloud-tpu-tpuvm-artifacts/tensorflow/tf-2.12.0/tensorflow-2.12.0-cp38-cp38-linux_x86_64.whl
   RUN curl -L https://storage.googleapis.com/cloud-tpu-tpuvm-artifacts/libtpu/1.6.0/libtpu.so -o /lib/libtpu.so
   RUN git clone https://github.com/tensorflow/models.git
   WORKDIR ./models
   RUN pip install -r official/requirements.txt
   ENV PYTHONPATH=/models

2. Create Cloud Storage bucket

   gcloud storage buckets create gs://your-bucket-name --location=europe-west4

3. Create a TPU VM

   gcloud compute tpus tpu-vm create your-tpu-name \
   --zone=europe-west4-a \
   --accelerator-type=v2-8 \
   --version=tpu-vm-tf-2.18.0-pjrt

4. Copy the Dockerfile to your TPU VM

   gcloud compute tpus tpu-vm scp ./Dockerfile your-tpu-name:

5. SSH into the TPU VM

   gcloud compute tpus tpu-vm ssh your-tpu-name \
   --zone=europe-west4-a

6. Build the Docker image

   sudo docker build -t your-image-name . 

7. Start the Docker container

   sudo docker run -ti --rm --net=host --name your-container-name --privileged your-image-name bash

8. Set environment variables

   export STORAGE_BUCKET=gs://your-bucket-name
   export DATA_DIR=gs://cloud-tpu-test-datasets/fake_imagenet
   export MODEL_DIR=${STORAGE_BUCKET}/resnet-2x

9. Train ResNet

   python3 official/vision/train.py \
   --tpu=local \
   --experiment=resnet_imagenet \
   --mode=train_and_eval \
   --config_file=official/vision/configs/experiments/image_classification/imagenet_resnet50_tpu.yaml \
   --model_dir=${MODEL_DIR} \
   --params_override="task.train_data.input_path=${DATA_DIR}/train*, task.validation_data.input_path=${DATA_DIR}/validation*,trainer.train_steps=100"

When the training script completes, make sure you clean up the resources.

1. Type exit to exit from the Docker container
2. Type exit to exit from the TPU VM
3. Delete the TPU VM

    $ gcloud compute tpus tpu-vm delete your-tpu-name --zone=europe-west4-a

TPU Pod

1. Create a file named Dockerfile in your current directory and paste the following text

   FROM python:3.8
   RUN pip install https://storage.googleapis.com/cloud-tpu-tpuvm-artifacts/tensorflow/tf-2.12.0/tensorflow-2.12.0-cp38-cp38-linux_x86_64.whl
   RUN curl -L https://storage.googleapis.com/cloud-tpu-tpuvm-artifacts/libtpu/1.6.0/libtpu.so -o /lib/libtpu.so
   RUN git clone https://github.com/tensorflow/models.git
   WORKDIR ./models
   RUN pip install -r official/requirements.txt
   ENV PYTHONPATH=/models

2. Create a TPU VM

   gcloud compute tpus tpu-vm create your-tpu-name \
   --zone=europe-west4-a \
   --accelerator-type=v3-32 \
   --version=tpu-vm-tf-2.18.0-pod-pjrt

3. Copy the Dockerfile to your TPU VM

   gcloud compute tpus tpu-vm scp ./Dockerfile your-tpu-name:

4. SSH into the TPU VM

   gcloud compute tpus tpu-vm ssh your-tpu-name \
   --zone=europe-west4-a

5. Build the Docker image

   sudo docker build -t your-image-name . 

6. Start a Docker container

   sudo docker run -ti --rm --net=host --name your-container-name --privileged your-image-name bash

7. Train ResNet

   python3 official/vision/train.py \
   --tpu=local \
   --experiment=resnet_imagenet \
   --mode=train_and_eval \
   --config_file=official/vision/configs/experiments/image_classification/imagenet_resnet50_tpu.yaml \
   --model_dir=${MODEL_DIR} \
   --params_override="task.train_data.input_path=${DATA_DIR}/train*, task.validation_data.input_path=${DATA_DIR}/validation*,task.train_data.global_batch_size=2048,task.validation_data.global_batch_size=2048,trainer.train_steps=100"

When the training script completes, make sure you clean up the resources.

1. Type exit to exit from the Docker container
2. Type exit to exit from the TPU VM
3. Delete the TPU VM

     $ gcloud compute tpus tpu-vm delete your-tpu-name --zone=europe-west4-a

Train a PyTorch model in a Docker container

TPU device

1. Create Cloud TPU VM

   gcloud compute tpus tpu-vm create your-tpu-name \
   --zone=europe-west4-a \
   --accelerator-type=v2-8 \
   --version=tpu-ubuntu2204-base

2. SSH into TPU VM

   gcloud compute tpus tpu-vm ssh your-tpu-name \
   --zone=europe-west4-a

3. Start a container in the TPU VM using the nightly PyTorch/XLA image.

   sudo docker run -ti --rm --name your-container-name --privileged gcr.io/tpu-pytorch/xla:r2.0_3.8_tpuvm bash

   Note: After running this command your command prompt will change indicating the terminal is connected to the running container.

4. Configure TPU runtime

   There are two PyTorch/XLA runtime options: PJRT and XRT. We recommend you use PJRT unless you have a reason to use XRT. To learn more about the different runtime configurations, see you have a reason to use XRT. To learn more about the different runtime configurations, see the PJRT runtime documentation.

   PJRT

   export PJRT_DEVICE=TPU

   XRT

   export XRT_TPU_CONFIG="localservice;0;localhost:51011"

5. Clone the PyTorch XLA repo

   git clone --recursive https://github.com/pytorch/xla.git

6. Train ResNet50

   python3 xla/test/test_train_mp_imagenet.py --fake_data --model=resnet50 --num_epochs=1

When the training script completes, make sure you clean up the resources.

1. Type exit to exit from the Docker container
2. Type exit to exit from the TPU VM
3. Delete the TPU VM

    $ gcloud compute tpus tpu-vm delete your-tpu-name --zone=europe-west4-a

TPU Pod

When you run PyTorch code on a TPU Pod, you must run your code on all TPU workers at the same time. One way to do this is to use the gcloud compute tpus tpu-vm ssh command with the --worker=all and --command flags. The following procedure shows you how create a Docker image to make setting up each TPU worker easier.

1. Create a TPU VM

   gcloud compute tpus tpu-vm create your-tpu-name \
   --zone=us-central2-b \
   --accelerator-type=v4-32 \
   --version=tpu-ubuntu2204-base

2. Add the current user to the docker group

   gcloud compute tpus tpu-vm ssh your-tpu-name \
   --zone=us-central2-b \
   --worker=all \
   --command="sudo usermod -a -G docker $USER"

3. Run the training script in a container on all TPU workers.

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   --zone=us-central2-b \
   --command="docker run --rm --privileged --net=host  -e PJRT_DEVICE=TPU gcr.io/tpu-pytorch/xla:r2.0_3.8_tpuvm python /pytorch/xla/test/test_train_mp_imagenet.py --fake_data --model=resnet50 --num_epochs=1"

   Docker command flags:

   • --rm remove the container after its process terminates.
   • --privileged exposes the TPU device to the container.
   • --net=host binds all of the container's ports to the TPU VM to allow communication between the hosts in the Pod.
   • -e set environment variables.

When the training script completes, make sure you clean up the resources.

Delete the TPU VM using the following command:

$ gcloud compute tpus tpu-vm delete your-tpu-name \
  --zone=us-central2-b

Train a JAX model in a Docker container

TPU Device

1. Create the TPU VM

   gcloud compute tpus tpu-vm create your-tpu-name \
   --zone=europe-west4-a \
   --accelerator-type=v2-8 \
   --version=tpu-ubuntu2204-base

2. SSH into TPU VM

   gcloud compute tpus tpu-vm ssh your-tpu-name  --zone=europe-west4-a

3. Start Docker daemon in TPU VM

   sudo systemctl start docker

4. Start Docker container

   sudo docker run -ti --rm --name your-container-name --privileged --network=host python:3.8 bash

5. Install JAX

   pip install jax[tpu] -f https://storage.googleapis.com/jax-releases/libtpu_releases.html

6. Install FLAX

   pip install --upgrade clu
   git clone https://github.com/google/flax.git
   pip install --user -e flax

7. Run the FLAX MNIST training script

   cd flax/examples/mnist
   python3 main.py --workdir=/tmp/mnist \
   --config=configs/default.py \
   --config.learning_rate=0.05 \
   --config.num_epochs=5

When the training script completes, make sure you clean up the resources.

1. Type exit to exit from the Docker container
2. Type exit to exit from the TPU VM

3. Delete the TPU VM

   $ gcloud compute tpus tpu-vm delete your-tpu-name --zone=europe-west4-a

TPU Pod

When you run JAX code on a TPU Pod, you must run your JAX code on all TPU workers at the same time. One way to do this is to use the gcloud compute tpus tpu-vm ssh command with the --worker=all and --command flags. The following procedure shows you how create a Docker image to make setting up each TPU worker easier.

1. Create a file named Dockerfile in your current directory and paste the following text

   FROM python:3.8
   RUN pip install "jax[tpu]" -f https://storage.googleapis.com/jax-releases/libtpu_releases.html
   RUN pip install --upgrade clu
   RUN git clone https://github.com/google/flax.git
   RUN pip install --user -e flax
   WORKDIR ./flax/examples/mnist

2. Build the Docker image

   docker build -t your-image-name .

3. Add a tag to your Docker image before pushing it to the Artifact Registry. For more information on working with Artifact Registry, see Work with container images.

   docker tag your-image-name europe-west-docker.pkg.dev/your-project/your-repo/your-image-name:your-tag

4. Push your Docker image to the Artifact Registry

   docker push europe-west4-docker.pkg.dev/your-project/your-repo/your-image-name:your-tag

5. Create a TPU VM

   gcloud compute tpus tpu-vm create your-tpu-name \
   --zone=europe-west4-a \
   --accelerator-type==v2-8 \
   --version=tpu-ubuntu2204-base

6. Pull the Docker image from the Artifact Registry on all TPU workers.

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   --zone=europe-west4-a \
   --command="sudo usermod -a -G docker ${USER}"

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   --zone=europe-west4-a \
   --command="gcloud auth configure-docker europe-west4-docker.pkg.dev --quiet"

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   --zone=europe-west4-a \
   --command="docker pull europe-west4-docker.pkg.dev/your-project/your-repo/your-image-name:your-tag"

7. Run the container on all TPU workers.

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   zone=europe-west4-a \
   --command="docker run -ti -d --privileged --net=host --name your-container-name europe-west4-docker.pkg.dev/your-project/your-repo/your-image:your-tag bash"

8. Run the training script on all TPU workers:

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   --zone=europe-west4-a \
   --command="docker exec --privileged your-container-name python3 main.py --workdir=/tmp/mnist \
   --config=configs/default.py \
   --config.learning_rate=0.05 \
   --config.num_epochs=5"

When the training script completes, make sure you clean up the resources.

1. Shut down the container on all workers:

   gcloud compute tpus tpu-vm ssh your-tpu-name --worker=all \
   --zone=europe-west4-a \
   --command="docker kill your-container-name"

2. Delete the TPU VM using the following command:

   $ gcloud compute tpus tpu-vm delete your-tpu-name \
   --zone=europe-west4-a

What's next

• Cloud TPU Tutorials
• Manage TPUs
• Cloud TPU System Architecture
• Run TensorFlow code on TPU Pod slices
• Run JAX code on TPU Pod slices