'LookupError: gradient registry has no entry for: SparseReduceMax in TensorFlow
I have an extremely large sparse matrix, say A, and it is stored using SparseTensor in TensorFlow to save memory usage. In my code, I have tf.sparse.reduce_max(A, axis=1). When I try to obtain the gradient of loss, I got an error: LookupError: gradient registry has no entry for: SparseReduceMax. Any idea to avoid this error?
TensorFlow version: 2.8.0
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|