col-12{-ms-flex:0 0 100%;flex:0 0 100%;max-width:100%}.order-first{-ms-flex-order:-1;order:-1}.order-last{-ms-flex-order:13;order:13}.order-0{-ms-flex-order:0;order:0}.order-1{-ms-flex-order:1;order:1}.order-2{-ms-flex-order:2;order:2}.order-3{-ms-flex-order:3;order:3}.order-4{-ms-flex-order:4;order:4}.order-5{-ms-flex-order:5;order:5}.order-6{-ms-flex-order:6;order:6}.order-7{-ms-flex-order:7;order:7}.order-8{-ms-flex-order:8;order:8}.order-9{-ms-flex-order:9;order:9}.order-10{-ms-flex-order:10...

.order-sm-5 {\n    -ms-flex-order: 5;\n    order: 5;\n  }\n  .order-sm-6 {\n    -ms-flex-order: 6;\n    order: 6;\n  }\n  .order-sm-7 {\n    -ms-flex-order: 7;\n    order: 7;\n  }\n  .order-sm-8 {\n    -ms-flex-order: 8;\n    order: 8;\n  }\n  .order-sm-9 {\n    -ms-flex-order: 9;\n    order: 9;\n  }\n  .order-sm-10 {\n    -ms-flex-order: 10;\n    order: 10;\n  }\n  .order-sm-11 {\n    -ms-flex-order: 11;\n    order: 11;\n  }\n  .order-sm-12 {\n    -ms-flex-order: 12;\n    order: 12;\n  }\n  .offset-sm-0...

.order-sm-5 {\n    -ms-flex-order: 5;\n    order: 5;\n  }\n  .order-sm-6 {\n    -ms-flex-order: 6;\n    order: 6;\n  }\n  .order-sm-7 {\n    -ms-flex-order: 7;\n    order: 7;\n  }\n  .order-sm-8 {\n    -ms-flex-order: 8;\n    order: 8;\n  }\n  .order-sm-9 {\n    -ms-flex-order: 9;\n    order: 9;\n  }\n  .order-sm-10 {\n    -ms-flex-order: 10;\n    order: 10;\n  }\n  .order-sm-11 {\n    -ms-flex-order: 11;\n    order: 11;\n  }\n  .order-sm-12 {\n    -ms-flex-order: 12;\n    order: 12;\n  }\n  .offset-sm-0...

semantics to match NumPy. * `tf.split` now takes arguments in a reversed order
and with different keywords. In particular, we now match NumPy order as
`tf.split(value, num_or_size_splits, axis)`. * `tf.sparse_split` now takes
arguments in reversed order and with different keywords. In particular we now
match NumPy order as `tf.sparse_split(sp_input, num_split, axis)`. NOTE: we have