@Jacques_Pienaar Hi Jacques
Thank you for the reply. Trying to understand how your suggestion of a cast can be achieved. We can’t cast the DenseHashTable itself to int32. I did try that, even though logically it doesn’t make sense, and get this error:
TypeError: Failed to convert elements of <tensorflow.python.ops.lookup_ops.DenseHashTable object at 0x18b7b4b80> to Tensor. Consider casting elements to a supported type. See https://www.tensorflow.org/api_docs/python/tf/dtypes for supported TF dtypes.
I can’t see a way of restructuring the code without dropping tf.map_fn because the function needs to perform a lookup using the rail input.
Adrian