You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2020/01/18 01:00:25 UTC
[GitHub] [incubator-mxnet] ptrendx opened a new pull request #17368: Faster
GPU frozen BatchNorm
ptrendx opened a new pull request #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368
## Description ##
This PR introduces an improved implementation of GPU BatchNorm when `use_global_stats` is `True`
Performance results (using V100 PCIe card, shape of data = `(208, 64, 112, 112)`)
dtype = `float32`
| FWD/BWD | axis | Old implementation time (ms) | New implementation time (ms) | speedup |
| - | - | - | - | - |
| FWD | 0 | 3.67 | 2.31 | 1.59x |
| BWD | 0 | 6.60 | 2.28 | 2.9x |
| FWD | 1 | 4.92 | 2.21 | 2.23x |
| BWD | 1 | 8.61 | 2.23 | 3.86x |
| FWD | 2 | 11.52 | 2.21 | 5.21x |
| BWD | 2 | 21.36 | 2.74 | 7.8x |
| FWD | 3 | 1010.55 | 2.24 | 451x |
| BWD | 3 | 1901.58 | 2.77 | 686.5x |
dtype = `float16`
| FWD/BWD | axis | Old implementation time (ms) | New implementation time (ms) | speedup |
| - | - | - | - | - |
| FWD | 0 | 3.1 | 0.92 | 3.37x |
| BWD | 0 | 6.56 | 1.04 | 6.31x |
| FWD | 1 | 4.34 | 0.92 | 4.72x |
| BWD | 1 | 9.21 | 1.34 | 6.87x |
| FWD | 2 | 10.23 | 0.92 | 11.12x |
| BWD | 2 | 22.61 | 1.60 | 14.13x |
| FWD | 3 | 948.1 | 1.01 | 939x |
| BWD | 3 | 2061.9 | 1.32 | 1562x |
@Kh4L @Jerryzcn FYI
## Checklist ##
### Essentials ###
Please feel free to remove inapplicable items for your PR.
- [x] Changes are complete (i.e. I finished coding on this PR)
- [x] All changes have test coverage:
- [x] To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-608647098
Reviving this PR. @apeforest I did change `my_id` to `thread_idx_in_warp`, but other cases of `my_*` variables are actually quite intuitive (and basically mean a private value of a given thing owned by a thread).
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-612156765
@apeforest Is this PR good to go now?
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Kh4L commented on a change in pull request
#17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Kh4L commented on a change in pull request #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#discussion_r368199696
##########
File path: src/operator/nn/batch_norm.cu
##########
@@ -44,9 +46,30 @@
using namespace mxnet;
+namespace {
+
/*! \brief inverse standard deviation <-> variance */
-#define VARIANCE_TO_INVSTD(__var$, __eps$) (1.0/sqrt((__var$) + DType(__eps$)))
-#define INVSTD_TO_VARIANCE(__invstd$, __eps$) ((1.0 / ((__invstd$) * (__invstd$))) - (__eps$))
+template <typename DType, typename AccReal>
+MSHADOW_XINLINE AccReal variance_to_invstd(DType var, AccReal eps) {
+ return rsqrtf(static_cast<AccReal>(var) + eps);
+}
+
+template <>
+MSHADOW_XINLINE double variance_to_invstd(double var, double eps) {
+ return rsqrt(var + eps);
+}
+
+template <typename AccReal>
+MSHADOW_XINLINE AccReal invstd_to_variance(AccReal invstd, AccReal eps) {
+ return 1.0f / (invstd * invstd) - eps;
Review comment:
What about:
```suggestion
return static_cast<AccReal>(1.0) / (invstd * invstd) - eps;
```
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-609940534
@mxnet-bot run ci [unix-gpu]
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on a change in pull request
#17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on a change in pull request #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#discussion_r369155050
##########
File path: src/operator/nn/batch_norm.cu
##########
@@ -44,9 +46,30 @@
using namespace mxnet;
+namespace {
+
/*! \brief inverse standard deviation <-> variance */
-#define VARIANCE_TO_INVSTD(__var$, __eps$) (1.0/sqrt((__var$) + DType(__eps$)))
-#define INVSTD_TO_VARIANCE(__invstd$, __eps$) ((1.0 / ((__invstd$) * (__invstd$))) - (__eps$))
+template <typename DType, typename AccReal>
+MSHADOW_XINLINE AccReal variance_to_invstd(DType var, AccReal eps) {
+ return rsqrtf(static_cast<AccReal>(var) + eps);
+}
+
+template <>
+MSHADOW_XINLINE double variance_to_invstd(double var, double eps) {
+ return rsqrt(var + eps);
+}
+
+template <typename AccReal>
+MSHADOW_XINLINE AccReal invstd_to_variance(AccReal invstd, AccReal eps) {
+ return 1.0f / (invstd * invstd) - eps;
Review comment:
Good suggestion.
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] mxnet-bot commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-608701123
Jenkins CI successfully triggered : [centos-cpu]
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Jerryzcn commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Jerryzcn commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-575855289
@haojin2 @sxjscience
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Jerryzcn commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Jerryzcn commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-585426547
Lgtm
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Jerryzcn commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Jerryzcn commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-585426958
BTW, were you able to find the bug that cause master MXNet to fail on mask rcnn?
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Kh4L commented on issue #17368: Faster GPU frozen
BatchNorm
Posted by GitBox <gi...@apache.org>.
Kh4L commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-585425575
@ptrendx @Jerryzcn sorry for the delay, please find it here:
[joblog.log](https://github.com/apache/incubator-mxnet/files/4195098/joblog.log)
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] apeforest commented on a change in pull request
#17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
apeforest commented on a change in pull request #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#discussion_r369401124
##########
File path: src/common/cuda_utils.h
##########
@@ -764,27 +764,86 @@ __device__ inline DType ldg(const DType* address) {
#endif
}
-template <typename OP, typename T>
+namespace mxnet {
+namespace common {
+/*! \brief common utils for cuda */
+namespace cuda {
+
+static constexpr const int warp_size = 32;
+
+/*! \brief Reduction inside a warp.
+ * Template parameters:
+ * NVALUES - number of values to reduce (defaults to warp_size).
+ * \param value - values to be reduced.
+ * \param redfun - function used to perform reduction.
+ */
+template <int NVALUES = warp_size, typename OP, typename T>
__device__ inline T warp_reduce(T value, OP redfun) {
- value = redfun(value, __shfl_down_sync(0xffffffff, value, 16));
- value = redfun(value, __shfl_down_sync(0xffffffff, value, 8));
- value = redfun(value, __shfl_down_sync(0xffffffff, value, 4));
- value = redfun(value, __shfl_down_sync(0xffffffff, value, 2));
- value = redfun(value, __shfl_down_sync(0xffffffff, value, 1));
+#pragma unroll
+ for (int i = warp_size / 2; i >= 1; i /= 2) {
+ if (NVALUES > i) value = redfun(value, __shfl_down_sync(0xffffffff, value, i));
+ }
return value;
}
-template <typename OP>
+template <int NValues = warp_size, typename OP>
__device__ inline mshadow::half::half_t warp_reduce(mshadow::half::half_t value, OP redfun) {
float v = static_cast<float>(value);
- v = redfun(v, __shfl_down_sync(0xffffffff, v, 16));
- v = redfun(v, __shfl_down_sync(0xffffffff, v, 8));
- v = redfun(v, __shfl_down_sync(0xffffffff, v, 4));
- v = redfun(v, __shfl_down_sync(0xffffffff, v, 2));
- v = redfun(v, __shfl_down_sync(0xffffffff, v, 1));
+#pragma unroll
+ for (int i = warp_size / 2; i >= 1; i /= 2) {
+ if (NValues > i) v = redfun(v, __shfl_down_sync(0xffffffff, v, i));
+ }
return mshadow::half::half_t(v);
}
+/*! \brief Reduction inside a block, requires all threads in a block to participate.
+ * It uses a 2 step approach:
+ * - all warps in a block perform intermediate reduction
+ * - first warp reduces the intermediate results.
+ * Template parameters:
+ * NTHREADS - number of threads in a block.
+ * all_reduce - whether all threads need the result of the reduction. If set to
+ * true, then all threads return with the same value. If set to
+ * false, then only thread 0 has the valid result. Defaults to true.
+ * \param value - value from each thread to be reduced
+ * \param redfun - function used to perform reduction
+ */
+template <int NTHREADS, bool all_reduce = true, typename OP, typename T>
+__device__ inline T reduce(const T& value, OP redfun) {
+ static_assert(NTHREADS <= warp_size * warp_size,
+ "Number of threads too large for reduction");
+ __shared__ T scratch[NTHREADS / warp_size];
+ const int my_id = threadIdx.x % warp_size;
Review comment:
Could we use a more informative names other than my_*?
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Kh4L commented on issue #17368: Faster GPU frozen
BatchNorm
Posted by GitBox <gi...@apache.org>.
Kh4L commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-585517849
@Jerryzcn no, we didn't really look into it, as we thought that you were taking over this task
There is an open issue here https://github.com/apache/incubator-mxnet/issues/17485
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] mxnet-bot commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-609940621
Jenkins CI successfully triggered : [unix-gpu]
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Jerryzcn commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Jerryzcn commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-575855412
Have you tried training MaskRCNN with this?
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-583141547
@Kh4L Could you post results of your runs?
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-614286699
@apeforest Gentle ping
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-608983596
@mxnet-bot run ci [centos-cpu, unix-gpu]
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-575855578
@Jerryzcn Not yet.
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Jerryzcn edited a comment on issue #17368: Faster
GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Jerryzcn edited a comment on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-575855289
@haojin2 @sxjscience Could u take a look as well since u both worked on norm layers before.
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] mxnet-bot commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-608983609
Jenkins CI successfully triggered : [centos-cpu, unix-gpu]
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Jerryzcn commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
Jerryzcn commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-577471171
Could we get some convergence result on fasterrcnn before merging thanks!
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] Kh4L commented on issue #17368: Faster GPU frozen
BatchNorm
Posted by GitBox <gi...@apache.org>.
Kh4L commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-577475334
@Jerryzcn it's in progress, I'll post results here
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] ptrendx commented on issue #17368: Faster GPU
frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on issue #17368: Faster GPU frozen BatchNorm
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-608701091
@mxnet-bot run ci [centos-cpu]
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #17368:
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-675227485
Jenkins CI successfully triggered : [miscellaneous, centos-cpu]
----------------------------------------------------------------
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:
users@infra.apache.org
[GitHub] [incubator-mxnet] szha commented on pull request #17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
szha commented on pull request #17368:
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-671670108
hmm not sure what's happening with the CI.
----------------------------------------------------------------
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:
users@infra.apache.org
[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #17368:
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-672126297
Jenkins CI successfully triggered : [centos-cpu, miscellaneous]
----------------------------------------------------------------
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:
users@infra.apache.org
[GitHub] [incubator-mxnet] szha merged pull request #17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
szha merged pull request #17368:
URL: https://github.com/apache/incubator-mxnet/pull/17368
----------------------------------------------------------------
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:
users@infra.apache.org
[GitHub] [incubator-mxnet] ptrendx commented on pull request #17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on pull request #17368:
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-675227459
@mxnet-bot run ci [centos-cpu, miscellaneous]
----------------------------------------------------------------
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:
users@infra.apache.org
[GitHub] [incubator-mxnet] ptrendx commented on pull request #17368: Faster GPU frozen BatchNorm
Posted by GitBox <gi...@apache.org>.
ptrendx commented on pull request #17368:
URL: https://github.com/apache/incubator-mxnet/pull/17368#issuecomment-672126249
@mxnet-bot run ci [centos-cpu, miscellaneous]
----------------------------------------------------------------
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:
users@infra.apache.org