Skip to content

Conversation

@MingMingShangTian
Copy link
Contributor

@MingMingShangTian MingMingShangTian commented May 24, 2021

PR types

Others

PR changes

OPs

Describe

  • Replace complex64/128 with complex<float/double> in elementwise_add, elementwise_sub, elementwise_mul, elementwise_div ops.
  • Remove the useless codes in CudaShuffleDownSync and CudaShuffleXorSync functions which was remained when removing CUDA9.0 codes in the PR31883

@paddle-bot-old
Copy link

Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

#define PADDLE_CUDA_FP16
#include "paddle/fluid/platform/complex128.h"
#include "paddle/fluid/platform/complex64.h"
#include "paddle/fluid/platform/complex.h"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

why this need in #define PADDLE_CUDA_FP16?

Copy link
Contributor

@chenwhql chenwhql left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@MingMingShangTian MingMingShangTian merged commit dbc08d6 into PaddlePaddle:develop May 25, 2021
@MingMingShangTian MingMingShangTian deleted the template_elementwise_ops branch May 25, 2021 12:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants