Re: [dmlc/tvm] [RFC] Unifying Object Protocol in the Stack (#4116)

2019-10-13 Thread Yao Wang
Just curious, is this change also related to relay/tvm node system unification? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/dmlc/tvm/issues/4116#issuecomment-541469429

Re: [dmlc/tvm] [RFC] Unifying Object Protocol in the Stack (#4116)

2019-10-13 Thread Tianqi Chen
The change is mainly need for runtime object infrastructure it unifies relay vm's object with the tvm's AST node. The unification of relay.module and tvm.module would be in a separate topic, but this PR is a step towards that -- You are receiving this because you are subscribed to this thread.

Re: [dmlc/tvm] [RFC][RUNTIME] Introduce new object protocol. (#4115)

2019-10-13 Thread Tianqi Chen
@junrushao1994 @icemelon9 @yzh119 can you help review this PR? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/dmlc/tvm/pull/4115#issuecomment-541470010

[dmlc/tvm] [RFC] Dynamic Shape Support - Graph Dispatching (#4118)

2019-10-13 Thread Yao Wang
## Overview There are more and more deployment requirements regarding dynamic input graphs, such as dynamic batching CV models and dynamic BERT. While dynamic input graph is supported in eager mode(Pytorch, Tensorflow Eager, MXNet Gluon) for model developing, TVM still just support static shape