Hi @tqchen ,
While working on the solution, I see two more problems that I'd like to discuss
here:
1. No Control over flattening of buffers: If a buffer is flattened via
`FlattenLowAxisSepDimensions` , then it shouldn't be flattened via TIR passes
like `FlattenBuffer` or `FlattenStorage`. Is
Thank you @tqchen!
I will try this approach and update here if I have further questions.
---
[Visit
Topic](https://discuss.tvm.apache.org/t/axis-separator-in-relax-tensor/15385/10)
to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails,
Thank you so much @tqchen @slyubomirsky @sanirudh for your detailed inputs on
this!
I have some thoughts and questions listed below. I also discussed them with
@sanirudh. It would be great if we can address them in the next unity meeting.
I'll add this topic to the agenda.
1. Axis separator
for the static memory planning, i think it is not a big issue, since that can
happens after we do `FlattenLowAxisSepDimensions `, so memory planner will be
able to capture the n-D needs
---
[Visit
Topic](https://discuss.tvm.apache.org/t/axis-separator-in-relax-tensor/15385/7)
to respond.
Thanks a lot for the detailed explanation @tqchen and @slyubomirsky
I understand the concern about modifying the StructInfo. My only thought was to
store the logical to physical memory mapping somewhere in Relax, so that when
we do static memory planning, we don't just assume 1d allocation, bu
