ptrendx commented on a change in pull request #17977: Relaxing type
requirements for broadcast_like
URL: https://github.com/apache/incubator-mxnet/pull/17977#discussion_r405177613
##########
File path: src/operator/tensor/broadcast_reduce_op_value.cc
##########
@@ -138,7 +138,16 @@ NNVM_REGISTER_OP(broadcast_like)
[](const NodeAttrs& attrs) {
return std::vector<std::string>{"lhs", "rhs"};
})
-.set_attr<nnvm::FInferType>("FInferType", ElemwiseType<2, 1>)
+.set_attr<nnvm::FInferType>("FInferType", [](const nnvm::NodeAttrs& attrs,
+ std::vector<int> *in_attrs,
+ std::vector<int> *out_attrs) {
+ CHECK_EQ(in_attrs->size(), 2) << " in operator " << attrs.name;
+ std::vector<int> checked_in_attrs = { (*in_attrs)[0] };
+ bool ret = !type_is_none((*in_attrs)[1]) &&
Review comment:
Yes, it is necessary - what `FInferType` returns is whether it succeeded in
inferring all the types (so that if all operators return true we know that all
types are inferred). That is why it is important to not lie and return true
only if all types are really inferred (even if we do not actually do anything
with the other type).
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
