Closed fengjunhui closed 6 years ago
This is spurious warning, caused by the application of a tf.nn.embedding_lookup
operation on a non-Variable
object (namely, node_states_per_layer[-1]
). While this /can/ lead to massive performance degradation, it does not do so in this case, and so the warning can be safely ignored.
There is a UserWarning problem when I run
python ./chem_tensorflow_sparse.py
. This is the detail :UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory. "Converting sparse IndexedSlices to a dense Tensor of unknown shape. "