tf.raw_ops.BatchCholeskyGrad

lA Tensor. Must be one of the following types: float32, float64.
gradA Tensor. Must have the same type as l.
nameA name for the operation (optional).

A Tensor. Has the same type as l.