@junrushao1994 no. for tensorflow fold style dynamic batching, it seems like partial evaluation will give us a static computation graph, then u can just do static batching (CombineConv2d, etc) on it as usual. If it goes as plan it seems like we dont even need anymore pass (just intergrating these two pass will be enough).
-- 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/3874#issuecomment-527252522
