Bugs in the code
Opened this issue ยท 0 comments
hsiehjackson commented
๐ Bug
x_scaling_factor
intransformer_sentence_encoder.py
:
I think the code here should modifiedscaling_factor
tox_scaling_factor
. Or do we use the same scaling factor for all transformer encoder layers?- freeze_model for
IntSoftmax
:
I think we should implement fix function inIntSoftmax
because we haveQuantAct
here. If we don't implement the fix function, we will skip fixing theQuantAct
inIntSoftmax
from here.
Problem
I am recently trying to implement IBERT on TVM an I found I should add FixedPointMul
and SymmetricQuantFunction
operators in TVM. Do you have any existing implementation? Thanks!