A tool for assignment to a slice in TensorFlow.
In TensorFlow, as opposed to Pytorch, it is currently impossible to assign to
the slice of a tensor in a range of different settings.
To mitigate this issue, tf-slice-assign introduces a single function that
allows to do exactly this using tensor_scatter_nd_update.
from tf_slice_assign import slice_assign
new_tensor = slice_assign(old_tensor, assignment, *slice_args)You can find a relatively simple example here.
pip install tf-slice-assign
In the following table, I am trying to give the reasons as to why no mitigation for the current problem exists.
| Link | Status |
|---|---|
| SO | Current answer requires creating a tf.Variable for each slice assignment you make |
| GH | Question is about tf.Variable |
| SO | Answers for tf.Variable or using tensor_scatter_update in a non-adaptable way |
| GH | Suggestion to use tensor_scatter_nd_update |
| GH | An answer suggest creating a mask, but a mask can actually be as difficult to create as the indices for tensor_scatter_nd_update |