Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

is there a way to achieve the async execution of `launch_other_task`?

Have you seen the FileSensor? Would this solve your problem? <https://docs.flyte.org/en/latest/flytesnacks/examples/sensor/file_sensor_example.html>

maybe but not in an ideal way. I am actually using `tf.train.checkpoint_iterator` which is a tensorflow function to poll for new checkpoints

so the logic of detecting a checkpoint is handled there

you could define a sensor to wait for the file to become available, similar to the file sensor: <https://github.com/flyteorg/flytekit/blob/master/flytekit/sensor/file_sensor.py>

problem is I dont know what the file is called

you could define a new sensor to handle this. have you tried triggering a task from within your dynamic workflow?