You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2022/06/18 14:08:25 UTC

[GitHub] [tvm-rfcs] yelite commented on pull request #74: [RFC] TUNIP: TVMScript Unified Printer

yelite commented on PR #74:
URL: https://github.com/apache/tvm-rfcs/pull/74#issuecomment-1159471608

   > what functionality does JSON provide that TVMScript doesn't do right now as it's designed? Is there a reason to have two formats to begin with? imo it is not worth having two formats here unless there's a clear need for structural analysis of TIR outside of TVM.
   
   A good analog is the ProtoBuf [text format](https://github.com/protocolbuffers/protobuf/blob/main/src/google/protobuf/text_format.cc) vs. wire format (binary format). People write proto in text format when they need to create proto manually, while the binary format is used in RPC payload for its portability and efficiency.
   
   Back to our own use case, TVMScript is the choice if user want to construct IR graph manually, but it's not portable because the TVMScript parser has a strong dependency on Python, especially after the Metaprogramming RFC (#79). On contrast, the JSON format will be portable because it only depends on the IR node definition in C++.


-- 
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.

To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org