Distribute the training across TPUs

to_xla(object)

Arguments

object

learner / model

Value

None