[OpenCL] fuse conv prelu pass #5461
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ysh329
reviewed
Feb 3, 2021
| @@ -21,7 +21,8 @@ __kernel void conv2d_1x1_opt( | |||
| __private const int input_height, /* of one block */ | |||
| __private const int output_width, | |||
| __private const int output_height, | |||
| __private const int old_w) { | |||
| __private const int old_w, | |||
There was a problem hiding this comment.
这个是prelu的参数，是一个tensor。之前的act里面只要添加attribute就好了，prelu融合需要添加一个新的输入
| output0 = activation_type4(output0, alpha0); | ||
| output1 = activation_type4(output1, alpha1); | ||
| output2 = activation_type4(output2, alpha2); | ||
| output3 = activation_type4(output3, alpha3); |
There was a problem hiding this comment.
考虑能否把该值融合到activation_type和type4中，以增加默认参数的方式。这个与FUSE_SCALE_ACT还有些不同，后者FUSE_SCALE不仅在convkernel中有用到，在element kernel中也有用到
There was a problem hiding this comment.
我看原来的实现中有prelu的选项，如果有prelu定义时添加一个参数，如果没有就只有一个参数。你是说把第二个参数设置成默认就含有是么
Superjomn
reviewed
Mar 1, 2021
| output2 = activation_type4(output2); | ||
| output3 = activation_type4(output3); | ||
| CL_DTYPE4 alpha0,alpha1,alpha2,alpha3; | ||
| #ifdef PRELU_CH |
There was a problem hiding this comment.
全局宏比较多，可以增加一些注释标记
比如
#ifdef PRELU_CH // {
// }
#endif defined(PRELU_ELE) // {
// }等
daming5432
force-pushed
the
fuse_conv_prelu_pass
branch
from
ffdc200
to
8d1c252
Compare
…into fuse_conv_prelu_pass
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
mtcnn_det2 骁龙865融合前耗时2.28ms，融合后2.07ms
mtcnn_det3 骁龙865融合前耗时5.69ms，融合后4.72ms