tf.raw_ops.BatchCholeskyGrad

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

A Tensor. Has the same type as l.