@calvin886 ONNX was very different then it is today back in 2017 when we began work on Relay. Furthermore ONNX is designed as an interchange format and has very little of the utilities that come with compiler IRs, these days Relay has many more extensions then ONNX including closures, data structures, traditional control flow (not pseudo ops), JSON serialization, a textual format, a human readable and parseable interchange, deep TVM support, strong Python support and frameworks for doing program analysis and pattern matching. We also have made several different design decisions such as unifying functions and graphs, unifying parameters and inputs, and so on.
You can read the various Relay papers to learn more about the specific advantages. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-tvm/issues/1673#issuecomment-698537152
