Covalent Google Cloud Platform (GCP) Batch Plugin

Covalent is a Pythonic workflow tool used to execute tasks on advanced computing hardware. This executor plugin interfaces Covalent with GCP Batch.

1. Installation

To use this plugin with Covalent, install it using pip:

pip install covalent-gcpbatch-plugin

2. Usage Example

This is an example of how a workflow can be constructed to use the GCP Batch executor. In the example, we train a Support Vector Machine (SVM) and use an instance of the executor to execute the train_svm electron. Note that we also require DepsPip which will be required to execute the electrons.

from numpy.random import permutation
from sklearn import svm, datasets
import covalent as ct


deps_pip = ct.DepsPip(
    packages=["numpy==1.22.4", "scikit-learn==1.1.2"]
)

executor = ct.executor.GCPBatchExecutor(
    project_id='covalent_gcp_batch',
    region='us-east1',
    bucket_name='covalent-storage-bucket',
    container_image_uri='us-east1-docker.pkg.dev/covalent_gcp_batch_/covalent/covalent-gcpbatch-executor',
    service_account_email='covalentsaaccount@covalenttesting.iam.gserviceaccount.com',
    vcpus = 2,  # Number of vCPUs to allocate
    memory = 512,  # Memory in MB to allocate
    time_limit = 300,  # Time limit of job in seconds
    poll_freq = 3,  # Number of seconds to pause before polling for the job's status
  )


# Use executor plugin to train our SVM model
@ct.electron(
    executor=executor,
    deps_pip=deps_pip
)
def train_svm(data, C, gamma):
    X, y = data
    clf = svm.SVC(C=C, gamma=gamma)
    clf.fit(X[90:], y[90:])
    return clf

@ct.electron
def load_data():
    iris = datasets.load_iris()
    perm = permutation(iris.target.size)
    iris.data = iris.data[perm]
    iris.target = iris.target[perm]
    return iris.data, iris.target

@ct.electron
def score_svm(data, clf):
    X_test, y_test = data
    return clf.score(
    	X_test[:90],y_test[:90]
    )

@ct.lattice
def run_experiment(C=1.0, gamma=0.7):
    data = load_data()
    clf = train_svm(
    	data=data,
	    C=C,
	    gamma=gamma
    )
    return score_svm(
    	data=data,
	    clf=clf
    )

# Dispatch the workflow.
dispatch_id = ct.dispatch(run_experiment)(
        C=1.0,
        gamma=0.7
)

# Wait for our result and get result value
result = ct.get_result(dispatch_id, wait=True).result

print(result)

During the execution of the workflow, one can navigate to the UI to see the status of the workflow. Once completed, the above script should also output a value with the score of our model.

0.8666666666666667

In order for the above workflow to run successfully, one has to provision the required cloud resources as mentioned in the section Required GCP Batch Resources.

3. Configuration

There are many configuration options that can be passed in to the class ct.executor.GCPBatchExecutor or by modifying the covalent config file under the section [executors.gcpbatch].

For more information about all of the possible configuration values visit our read the docs (RTD) guide for this plugin.

4. Required GCP Resources

In order to run your workflows with covalent there are a few notable GCP resources that need to be provisioned first. The required resources are Google storage bucket, docker artifact registry and service account.

For more information regarding which cloud resources need to be provisioned visit our read the docs (RTD) guide for this plugin.

Getting Started with Covalent

For more information on how to get started with Covalent, check out the project homepage and the official documentation.

Release Notes

Release notes are available in the Changelog.

Citation

Please use the following citation in any publications:

W. J. Cunningham, S. K. Radha, F. Hasan, J. Kanem, S. W. Neagle, and S. Sanand. Covalent. Zenodo, 2022. https://doi.org/10.5281/zenodo.5903364

License

Covalent is licensed under the Apache License 2.0. See the LICENSE file or contact the support team for more details.