I had a query: Batch-Size cannot be extracted during eager-execution, however in window_partition and window_reverse you are extracting Batch-Size via tf.shape(), is this not going to throw an error? Alternatively, this code bears compatibility till what version of tensorflow?
Hi: Thanks for this really nice implementation!
I had a query: Batch-Size cannot be extracted during eager-execution, however in
window_partitionandwindow_reverseyou are extracting Batch-Size viatf.shape(), is this not going to throw an error? Alternatively, this code bears compatibility till what version of tensorflow?