anijain2305 commented on a change in pull request #5286: [Requantize] Cleanup
and Optimize Lowering
URL: https://github.com/apache/incubator-tvm/pull/5286#discussion_r405910530
##########
File path: src/relay/qnn/op/requantize.cc
##########
@@ -132,36 +132,29 @@ Expr RequantizeLower(const Expr& input_tensor, const
Expr& input_scale,
const Expr& input_zero_point, const Expr& output_scale,
const Expr& output_zero_point, const RequantizeAttrs*
param,
const Array<IndexExpr>& input_shape, const DataType&
out_dtype) {
- DataType hp_dtype = DataType::Int(64);
- auto tensor = Cast(input_tensor, hp_dtype);
+ auto tensor = Cast(input_tensor, DataType::Int(32));
Review comment:
Might not be required.
----------------------------------------------------------------
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]
With regards,
Apache Git Services
