Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Mkldnn #2

Closed
wants to merge 75 commits into from
Closed

Mkldnn #2

wants to merge 75 commits into from

Conversation

wentingj
Copy link

@wentingj wentingj commented Dec 5, 2017

Description

(Brief description on what this PR is about)

Checklist

Essentials

  • Passed code style checking (make lint)
  • Changes are complete (i.e. I finished coding on this PR)
  • All changes have test coverage
  • For user-facing API changes, API doc string has been updated. For new C++ functions in header files, their functionalities and arguments are well-documented.
  • To my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

  • Feature1, tests, (and when applicable, API doc)
  • Feature2, tests, (and when applicable, API doc)

Comments

  • If this change is a backward incompatible change, why must this change be made.
  • Interesting edge cases to note here

@zheng-da
Use NNVM for convolution.
a10bc72
@zheng-da
Fix bugs in CuDNN convolution.
cf6293f
@zheng-da
Use NNVM for activation.
cba26a4
@zheng-da
Rename the input macro for convolution.
f7a9e72
@zheng-da
Use NNVM for batch norm.
dcafbb3
@zheng-da
Use NNVM for FullyConnected.
94143ab
@zheng-da
Use NNVM for SoftmaxActivation.
ecb33ea
@zheng-da
Use NNVM for dropout.
d45e0d3
@zheng-da
Use NNVM for Pooling.
b73e762
@zheng-da
Use NNVM for Deconvolution and Upsampling.
df581ef
@zheng-da
Fix a bug in deconvolution.
8491f72
@zheng-da
Handle aux states in batch norm.
ac317cd
@zheng-da
Fix GPU versions of the operators.
672ce33
@zheng-da
Fix coding style.
a5cadd4
@zheng-da
Fix bugs in CuDNN convolution.
54f699e
@zheng-da
Fix bugs in other CuDNN operators.
9147a51 
Make these operators stateless. Every time forward or backward is invoked,
they will be reinitialized.
@zheng-da
enable depthwise convolution.
10ae5f0
@zheng-da
Move CuDNN code to src/operator/nn/cudnn
2279f93
@zheng-da
Fix a bug in convolution.
b713524
@zheng-da
Remove MKL code.
2a7d71e
@zheng-da
Update MXNet for MKLDNN.
cfd862b
@zheng-da
Enable MKLDNN Relu.
063b504
@zheng-da
Merge remote-tracking branch 'mxnet_master/master' into mkldnn
f1431e9
@zheng-da
Change Makefile for MKLDNN.
756ec14
@zheng-da
Temporarily disable part of dropout.
604300c
@zheng-da
Remove infer storage in convolution.
f1f27d1
@zheng-da
Update MXNet for MKLDNN.
df73da1
@zheng-da
Support MKLDNN storage type in python.
6da4528
@zheng-da
Update activation.
1b3e210
@zheng-da
Add MKLDNN base classes.
b2a8b60
zheng-da and others added 25 commits November 7, 2017 22:17
@zheng-da
Fix a bug in NDArray.
8335d27
@zheng-da
Reorder MKLDNN memory to default format in SetTBlob.
131a141
@zheng-da
Disable MKLDNN correctly.
61ac839
@zheng-da
Fix a bug in activation.
64cf57c
@zheng-da
Reshape of NDArray supports MKLDNN.
4a2a98b
@zheng-da
Fix a memory ref bug in NDArray.
8d5ad60
@zheng-da
Reshape NDArray in MKLDNN FullyConnected.
e83c9c0
@zheng-da
Fix data format conversion.
97a6910
@zheng-da
Create MKLDNN NDArray in python.
f87d8b9
@zheng-da
Support Slice for MKLDNN NDArray.
1b97bc7
@zheng-da
Reduce the overhead of summing the result to the output array.
3cad7c9
@zheng-da
Avoid unnecessary memory copy in NDArray.
0044a9a
@zheng-da
Fix a bug in data reordering.
1494a44
@zheng-da
Fix a bug in NDArray.
8f7da06
@zheng-da
Don't hard code MKLDNN type.
ac06afe
@zheng-da
Support dilation in MKLDNN convolution.
ca6b1f7
@zheng-da
Merge branch 'mkldnn' of https://github.com/zheng-da/incubator-mxnet …
39cc06a 
…into mkldnn
@zheng-da
Fix a bug in sum results.
ee28ebe
@zheng-da
Rewrite GetMKLDNNData.
70d5b75
@zheng-da
Add prepare_mkldnn.sh
4996db5
@zheng-da
Enable MKLDNN activation.
40841b6
@zheng-da
Fix a bug on FullyConnected.
10412f5
@zheng-da
Handle 3 dims for MKLDNN NDArray.
2819797
@wentingj
add mkldnn_concat.cc
2832750
@wentingj
declare concat func
94e4dcc
@wentingj wentingj closed this Dec 5, 2017
@wentingj wentingj deleted the mkldnn branch December 5, 2017 05:47
xinyu-intel pushed a commit to wentingj/incubator-mxnet-int8 that referenced this pull request May 6, 2018
@thomelane @indhub
[MXNET-340] Updated tutorials page. (apache#10621)
97511ba 
* Updated tutorials page.

* Combined tutorial links

Added "alternative" links.

* Corrected typo

* Force build.

* Force build zheng-da#2

* Force build zheng-da#3

* Force zheng-da#4
jinhuang415 pushed a commit to jinhuang415/incubator-mxnet that referenced this pull request May 29, 2018
@thomelane
[MXNET-340] Updated tutorials page. (apache#10621)
373e92b 
* Updated tutorials page.

* Combined tutorial links

Added "alternative" links.

* Corrected typo

* Force build.

* Force build zheng-da#2

* Force build zheng-da#3

* Force zheng-da#4
zheng-da pushed a commit that referenced this pull request Jun 28, 2018
@thomelane @zheng-da
[MXNET-340] Updated tutorials page. (apache#10621)
2fce148 
* Updated tutorials page.

* Combined tutorial links

Added "alternative" links.

* Corrected typo

* Force build.

* Force build #2

* Force build #3

* Force #4
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@wentingj @zheng-da