t-vi edited a comment on issue #6268:
URL: https://github.com/apache/incubator-tvm/issues/6268#issuecomment-690039542


   > `torch.result_type` is confused with one of the inputs being numpy scalar 
of type np.int64, and it returns float32 when both lhs and rhs are clearly int64
   
   Oh, indeed, I missed that at first! I tried to cover numpy scalars as good 
as I can, but I'll have to fix it before calling result_type. But at the same 
time, division is special w.r.t. result type.
   
   For the quantization: I would not hold my breath and try to cope with the 
representation of quantization as is. I'm looking at quantizing some models, so 
I might see how they fare in TVM.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to