Yong Wu

Changed this topic for TensorArray/TensorList only. I moved control flow related discussion into an independent thread https://github.com/tlc-pack/relax/issues/93, feel free to leave comments regarding control flow there.

@hypercubestart sorry for my late response, forgot to check the notification. We plan to have a loop operator in the future, it is something like PyTorch prim::loop or tensorflow while_loop....

you can install tvm unity via "`pip install -I mlc_ai_nightly -f https://mlc.ai/wheels`".

@tvm-bot rerun

Thank you for the proposal! With the introduction of the SLM, we now are able to utilize the TVM nn.module for supporting models created with Torch, which has been working...

@chunit-quic thanks for the reply! I'm new to the Torch ExportedProgram, I noticed that the export API is still under development (https://github.com/pytorch/executorch/issues/290). If `ExportedProgram` is the preferred way to obtain...

The restriction is that return is not allowed in true/false branch of if. Not related to recursion. @slyubomirsky a workaround as below. ```python import tvm from tvm.script import relax as...

If we would like to support multiple return like the following example in Relax Function, one option is to introduce a tag member like `Bool return_body` in `SeqExprNode` , the...

Thanks TQ! I don't have a preference on if we need this syntax, a better err message should be good for me.

@tqchen @slyubomirsky I have updated the error message in https://github.com/apache/tvm/pull/14123. I agree it would be nice to have this syntactic sugar (more pythonic style). I can work on adding this...