B.overlap-> tensor([[0.0000, 0.0000, 0.0000, ..., 0.6466, 0.7945, 0.5389]],
overlap_for_each_prior, object_for_each_prior = overlap.max(dim=0) # (8732)
this .max(dim=0) return the two returns.
is there equivalent method in tensorflow 2.0?
You might want to ask on the Tensorflow question board. More people there are familiar with it and will be able to help you I think.
But checking the tf doc, reduce_max looks like what you’re looking for no? https://www.tensorflow.org/api_docs/python/tf/math/reduce_max