SparkSubmitOperator deploy-mode

I'm using SparkSubmitOperator but I don't see any equivalent of Spark's deploy-mode. How can I specify whether it's client or cluster?

This is the signature of the SparkSubmitOperator's ctor:

@apply_defaults
def __init__(self,
             application='',
             conf=None,
             conn_id='spark_default',
             files=None,
             py_files=None,
             archives=None,
             driver_class_path=None,
             jars=None,
             java_class=None,
             packages=None,
             exclude_packages=None,
             repositories=None,
             total_executor_cores=None,
             executor_cores=None,
             executor_memory=None,
             driver_memory=None,
             keytab=None,
             principal=None,
             proxy_user=None,
             name='airflow-spark',
             num_executors=None,
             status_poll_interval=1,
             application_args=None,
             env_vars=None,
             verbose=False,
             spark_binary="spark-submit",
             *args,
             **kwargs):

I searched in Google for hours but didn't find anything.