Ecosyste.ms: OpenCollective

An open API service for software projects hosted on Open Collective.

github.com/deeplearning4j/deeplearning4j

Suite of tools for deploying and training deep learning models using the JVM. Highlights include model import for keras, tensorflow, and onnx/pytorch, a modular and tiny c++ library for running math code and a java based math library on top of the core c++ library. Also includes samediff: a pytorch/tensorflow like library for running deep learn...
https://github.com/deeplearning4j/deeplearning4j

Add default interface method for aggregate conditions

e8d92d25341a4e6c9a83fc54a170f55abdadb704 authored over 3 years ago by Adam Gibson <[email protected]>
Improve docs

efbeccf3985d0cc1bfa0a012c16cae29e1075a76 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9401 from eclipse/ag_fix_9391

Fix mistake in ROCScoreCalculator

288073a259aa9549743ba2ee8e2fc96574226118 authored over 3 years ago by Adam Gibson <[email protected]>
Add graph optimizer framework

49495294ddf1ab0e1190b3d6c2ffe2dedc691cdd authored over 3 years ago by Adam Gibson <[email protected]>
Fix mistake in ROCScoreCalculator

1c4992f12eff6713f09db490bfe66acaea24a979 authored over 3 years ago by Adam Gibson <[email protected]>
Fixes https://github.com/eclipse/deeplearning4j/issues/9393

9a22ff679d6637769b9bea095a7fcead9c48a694 authored over 3 years ago by Adam Gibson <[email protected]>
Add deconvolution 3d for import

75899208427bb755f8629d6a17d594162a078fa4 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9396 from eclipse/ag_cuda_compile_fix

Fix compilation error with system properties in cuda

a7a123592854e34f039dcff40c2c0000b1dfb809 authored over 3 years ago by Adam Gibson <[email protected]>
Fix compilation error with system properties inc uda

00d88298e24b993a06829d4f639a04cb455eb463 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9392 from quickwritereader/qwr_cudnn_refactor

cudnn: refactor introducing thin cudnn descriptors to fix descriptor leaks

72a51359ef4746031caa7dc7c08105c12cfe71e8 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9395 from eclipse/ag_npy_create

Add more flexible handling of creating numpy arrays

46a4ff11d6a3f53b8d1367343496b9bdb541d159 authored over 3 years ago by Adam Gibson <[email protected]>
Update NumpyArray.java

2b4d930c35cb353a2ac245a8741d611d7c205ae4 authored over 3 years ago by Adam Gibson <[email protected]>
Add more flexible handling of creating numpy arrays

10f2808e0ce90bd1da3eaec2e72f7a9a25f67265 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9394 from eclipse/ag_disable_python_thread_management

Allow disabling of python thread management

6f75741a5a9beae2c11044c479aa4af564cf4d29 authored over 3 years ago by Adam Gibson <[email protected]>
Update naming suggestions

30887f937418aa27543aaba918a4d92419bcc00d authored over 3 years ago by Adam Gibson <[email protected]>
Refactor constants and system properties usage in to 1 class

892b6e5e26d72c0fd6d8900941556053e54b7793 authored over 3 years ago by Adam Gibson <[email protected]>
Allow disabling of python thread management

9469401fa020d81f288c566cd8a550bcf2ffd666 authored over 3 years ago by Adam Gibson <[email protected]>
minor fix: change error message

Signed-off-by: AbdelRauf <[email protected]>

bffca66ea1e553ca7e193671f9a26411e5e73868 authored over 3 years ago by AbdelRauf <[email protected]>
cudnn: refactor introducing thin cudnn descriptors to fix obvious descriptor leaks and also possible memory leaks when error is thrown

Signed-off-by: AbdelRauf <[email protected]>

0a12956c5fb47a06bffedf3820e3963b53bb75d6 authored over 3 years ago by AbdelRauf <[email protected]>
Merge pull request #9386 from eclipse/ag_python4j_init_change

Change PySetPath to PyAddPath, add more configurability for python path setting

9d83f7cf9758a9eaccee639f2b992dee28a33e95 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9384 from eclipse/ag_global_pooling_collapse_fix_2

Global Pooling Collapse Dimension change

89197956b635e44c16989a36f2b75e5a780d3b10 authored over 3 years ago by Adam Gibson <[email protected]>
Change python executioner to use add path rather than set path, add more configurability for calling py init.

1727ab2d76507c27646d8e22db658a01449c9225 authored over 3 years ago by Adam Gibson <[email protected]>
Fix batch norm for rnn and multiple batch sizes

6d93fd2905e3b26b3575c6ab97be621c42000e30 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9385 from quickwritereader/qwr_applyscalar_typo

fix: applyScalar typo

d6dafb0863f34dab630455f60d7f3d5663cd3d0a authored over 3 years ago by Adam Gibson <[email protected]>
fix: applyScalar typo

Signed-off-by: AbdelRauf <[email protected]>

d20fce7d34cbb60dad46a19710cb40e0c764f5d7 authored over 3 years ago by AbdelRauf <[email protected]>
Update KerasGlobalPooling.java

2407e0fc62a8769bb977a440785fc8f2f3082f6c authored over 3 years ago by Adam Gibson <[email protected]>
Update KerasGlobalPooling.java

a3e977df4494d84bfbd4f1867db7d94c6d0dd388 authored over 3 years ago by Adam Gibson <[email protected]>
Merge conflict fix

5634973a33189ad24353bc273e1eefb13021d0cd authored over 3 years ago by Adam Gibson <[email protected]>
Merge branch 'master' of https://github.com/eclipse/deeplearning4j into ag_global_pooling_collapse_fix_2

89ee36f741972447fe4450e05b406d194504281e authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9382 from eclipse/ag_compat_release

Update compat workflow for releases

d495603ea658950a2f1fd97b52a6541c629570ad authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9322 from eclipse/ag_helper_utils_conv_assertion_add

Add missing assertion test for conv helper

cb236a123624ab89f7313678f4ad2f63e29d5ba2 authored over 3 years ago by Adam Gibson <[email protected]>
Test changes for global pooling related tests

eeb1266bf00ea24b15cb6b222b1a4ae7d5e0682b authored over 3 years ago by Adam Gibson <[email protected]>
update collapse dimensions conventions for global pooling

9620e6be9a0fb4a0b7ecd6ef9f68764a605a5f2e authored over 3 years ago by Adam Gibson <[email protected]>
Update feeed forward types to be more flexible, always return recurrent for more consistent interactions with global pooling

2f1d33eba1e6f02c0b345028a95560054d47badd authored over 3 years ago by Adam Gibson <[email protected]>
Update build-deploy-linux-x86_64-compat.yml

383191327ef6d6e736f5fa689d6fb7a6568c106a authored over 3 years ago by Adam Gibson <[email protected]>
Update build-deploy-linux-x86_64-compat.yml

c0a149420429c60fc14e2b2dfcb9e222ec9170f8 authored over 3 years ago by Adam Gibson <[email protected]>
Update compat workflow for releases

abed4050cc6f4d17869aba0d2ccb58f0e4e15490 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9381 from eclipse/ag_remove_mkldnn_lstm

Remove MKLDNN LSTM Helper

8a428e01aee694ec3ca719fcc7c3932d4a438238 authored over 3 years ago by Adam Gibson <[email protected]>
Delete MKLDNNLSTMHelper.java

8b002ec5480cfdbc6e8a49096c00e52d4fe47266 authored over 3 years ago by Adam Gibson <[email protected]>
Remove MKLDNN Helper

7aa35cdce60ebae01f1f24de65eb71e694a95779 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9379 from quickwritereader/qwr_stdvarmean

optimized for Cpu "running Standard deviation and Variance."

c4e664b580c5bfae1dda851ec39c324229de2535 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9380 from quickwritereader/qwr_aggexpr

Added possibility to use different intermediate variable type in aggr…

743489e68201c0c2bfeef061189e85ff96743b01 authored over 3 years ago by Adam Gibson <[email protected]>
Add missing comment and ReadMe section for splitting template instantiations

Signed-off-by: AbdelRauf <[email protected]>

ee89872f95859f49b0eb3f1639a8802b064076af authored over 3 years ago by AbdelRauf <[email protected]>
Merge pull request #9378 from eclipse/ag_global_pooling_nwc_ncw_fix

Fix global pooling for NWC/NCW channel variants

3827a94d79960be0b44dcd31a3c6599fadf14e78 authored over 3 years ago by Adam Gibson <[email protected]>
Added possibility to use different intermediate variable type in aggregation expressions

Signed-off-by: AbdelRauf <[email protected]>

0ba0f933a95d2dceeff3651bc540d03b5f3b1631 authored over 3 years ago by AbdelRauf <[email protected]>
optimized stdev and variance. We use the same running parallel algorithm (used in summary stats) but this time only variables necessary for variance are calculated. Furthermore, we exploit the auto-vectorization and parallelization nature of the algorithm where it is possible.

Signed-off-by: AbdelRauf <[email protected]>

88ee9bfedea2def3e318c7479ad49e4606731a80 authored over 3 years ago by AbdelRauf <[email protected]>
Create KerasGlobalPoolingTest.java

80be09cd0e8e945c7bb623e56808b003588f90f4 authored over 3 years ago by Adam Gibson <[email protected]>
Fix global pooling for NWC/NCW channel variants

9ddaa1665cc075e91b07a4fa6fad77f68e608577 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9377 from yumg/fix#9266

Fix#9266

05419970d774749c446efb9f622e0a66876ee6a6 authored over 3 years ago by Adam Gibson <[email protected]>
fix eclipse/deeplearning4j#9266 and eclipse/deeplearning4j-examples#1029

Signed-off-by: yumg <[email protected]>

77e03670c87e7635aad0f572b39320e8860a0263 authored over 3 years ago by yumg <[email protected]>
Update docs

8f6f6f86167fd6daddaa9f7fa4341a234e04c155 authored over 3 years ago by Adam Gibson <[email protected]>
Cudnn Lstm:

Forward inference implemented using v6, and v8 (when version > 8.0.1) api calls.
As our Cuda Lstm...

c6c50b55a8d8a4897f856e363b431fe813b7e723 authored over 3 years ago by AbdelRauf <[email protected]>
Update Nd4jCpuPresets.java

d256f08d98c594efbb52625813a09b02141bceba authored over 3 years ago by Adam Gibson <[email protected]>
ctcLoss: refactored to use new scoped Descriptors in cudnn call

Signed-off-by: AbdelRauf <[email protected]>

10fb49fa6f4a1476d5aac83e303514a3453d5e3b authored over 3 years ago by AbdelRauf <[email protected]>
minor refactoring

Signed-off-by: AbdelRauf <[email protected]>

82edf877dbe7c71dfec78ef60e006706fbb6de76 authored over 3 years ago by AbdelRauf <[email protected]>
ctcBeam fix: run when NdArray allocated on device

Signed-off-by: AbdelRauf <[email protected]>

03516ddafe02c6dd2a803437a4091eb981a43195 authored over 3 years ago by AbdelRauf <[email protected]>
Allow feed forward as type in conv1d (with test)

a3b6b7a8f641de57753f7c849c6c136d17301c48 authored over 3 years ago by Adam Gibson <[email protected]>
fix: comment typos

Signed-off-by: AbdelRauf <[email protected]>

f56614e92d26b59fc5a6dcf13e46a5a7b262421d authored over 3 years ago by AbdelRauf <[email protected]>
Add centos compat build based on docker

707d5f1f165aa55600a10927fd0fee8981c9f007 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9372 from quickwritereader/qwr_cudnn

cudnn lstm forward inference and other fixes

886c52116378d87bcd7723ddc9fcd9abcea0cc84 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9365 from eclipse/ag_cnn1d_ff

Allow feed forward as type in conv1d (with test)

1f2657e518f1174a8452232c98930f73191c71a3 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9373 from eclipse/ag_compat_fix_2

Adds compat classifier fixes enabling builds on centos 7

50d29f3ada007caaa405684fb29063d7600e772c authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9374 from eclipse/ag_benchmark_suite_removal

Fix compilation error after removal of running benchmark suite

82e481d076fcd751f1b43fa366482cc53a30cbf7 authored over 3 years ago by Adam Gibson <[email protected]>
Update Nd4jCpuPresets.java

37f0b0274811df62aeec0354429b7232c82b2cbe authored over 3 years ago by Adam Gibson <[email protected]>
Fix compilation error after removal of running benchmark suite

c2f9c9ec2e4290bdeb591a5557fd3101ecac7cdb authored over 3 years ago by Adam Gibson <[email protected]>
Update docs

3ef99ef2b8715158dc6ed04e0183c6cd3763b26b authored over 3 years ago by Adam Gibson <[email protected]>
Add centos compat build based on docker

cfac543b5fc99cd2f7071f0fba9cc48060cdd3df authored over 3 years ago by Adam Gibson <[email protected]>
minor refactoring

Signed-off-by: AbdelRauf <[email protected]>

d370b23c1167c65d24873350a5d582d44b2d265d authored over 3 years ago by AbdelRauf <[email protected]>
Merge pull request #9364 from eclipse/ag_fix_visual_studio_version_cuda

Ensure older visual studio

3587b9a58305e2e0e2b36046eabec037dd01e1b7 authored over 3 years ago by Adam Gibson <[email protected]>
fix: comment typos

Signed-off-by: AbdelRauf <[email protected]>

abba89f4f1398295fe8da80612da4f3fbae2c283 authored over 3 years ago by AbdelRauf <[email protected]>
ctcLoss: refactored to use new scoped Descriptors in cudnn call

Signed-off-by: AbdelRauf <[email protected]>

603e33e92338c1d2f7825e3747d3a0ad0c320b8b authored over 3 years ago by AbdelRauf <[email protected]>
Cudnn Lstm:

Forward inference implemented using v6, and v8 (when version > 8.0.1) api calls.
As our Cuda Lstm...

0e796ae62cc2dedd86d6ae95412a2f7987ed0a96 authored over 3 years ago by AbdelRauf <[email protected]>
ctcBeam fix: run when NdArray allocated on device

Signed-off-by: AbdelRauf <[email protected]>

d3e66b81a05021c178d4e7d2cb3c2a52edd0d8fb authored over 3 years ago by AbdelRauf <[email protected]>
Remove benchmark suite from running definition

cba63a25a5c0fd297f9f141529603ade86103322 authored over 3 years ago by Adam Gibson <[email protected]>
More benchmark refactoring

75fc5edc94ef056540b54c3945eaa8c15cd3da6d authored over 3 years ago by Adam Gibson <[email protected]>
Move benchmark to contrib and unblock c++ 17 compilation

584c35a7c604d213e1681cfb4fc39280c5247c9e authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9368 from eclipse/ag_centos_6

Add new way of specifying OS/Version and enabling centos 6 builds

4f6a9f8f7df569613ee6ea1b56095c55b147227c authored over 3 years ago by Adam Gibson <[email protected]>
Update CMakeLists.txt

22af1ab97a52b02aeac3083800286b7ba5b757f8 authored over 3 years ago by Adam Gibson <[email protected]>
Remove setting cuda c++ 14 standard in 2 places

6ec4a8d22920b12e0b17bf22678c3b31171ec2d9 authored over 3 years ago by Adam Gibson <[email protected]>
Update buildnativeoperations.sh

93114c9048ba6442058fdf8d2e1c7ddba3cd0d5c authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

d1bb271850985d5c9482b24913d908454f4d7d99 authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

42ca9d2a6616cacfa9b0fd48df017bd6f0d1d2e4 authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

3642667286179d17995d54cdb1c682c0489172b8 authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

e7daaf86243f269162f0aeb0a3e594471b21112f authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

8dcea1efda7124729266a36bc22f9e93ca8f843b authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

690b97f704c605c054458a2df3577501863e5680 authored over 3 years ago by Adam Gibson <[email protected]>
Add missing dot for setup

268728bee64f0e2e0e95c3a74d7af0e455ca8b96 authored over 3 years ago by Adam Gibson <[email protected]>
Remove generator toolset env

4580997b578b8a8feca47bf26c6d39a5043bcbdd authored over 3 years ago by Adam Gibson <[email protected]>
Update action.yml

fe360d7a06b5436df551feee653e854bea0dca45 authored over 3 years ago by Adam Gibson <[email protected]>
Remove visual studio, instead just enable new standard

a13661908c1f692a18e58ca5b4280c79f2b3997d authored over 3 years ago by Adam Gibson <[email protected]>
Add new way of specifying OS/Version and enabling centos 6 builds

f87907e0d594390e7c727b24dec1e1148c67a082 authored over 3 years ago by Adam Gibson <[email protected]>
Update CMakeLists.txt

4c6bca0af4b28623cf2ca4428ef1265389a052ca authored over 3 years ago by Adam Gibson <[email protected]>
Update CMakeLists.txt

8e590a18506a467ea0ca105cd61a9a6637f6f470 authored over 3 years ago by Adam Gibson <[email protected]>
Update buildnativeoperations.sh

35e274e29d705a34662d3e1fdb90d520de677f20 authored over 3 years ago by Adam Gibson <[email protected]>
Attempt cuda windows workaround

2177836b28b3aa994c62516c14d34a0f65fb9c2a authored over 3 years ago by Adam Gibson <[email protected]>
Allow feed forward as type in conv1d (with test)

2d54eff854b93526d719425bbdab4605d0939503 authored over 3 years ago by Adam Gibson <[email protected]>
Update buildnativeoperations.sh

12baabf11dd9659428f7085ed6d8d46bb5281c6e authored over 3 years ago by Adam Gibson <[email protected]>
Set the cmake tools et version for cuda

b97505d06b478c108078212dace83e638e015ed1 authored over 3 years ago by Adam Gibson <[email protected]>
Ensure older visual studio

6f7ccd99a57eb4af6afff3f4662d7dd2dff6f0a2 authored over 3 years ago by Adam Gibson <[email protected]>
Merge pull request #9363 from eclipse/ag_logback_move

Fix compilation errors from logback dep removal

0591d628a12bcda0a3940142b9d0b9c9ff448079 authored over 3 years ago by Adam Gibson <[email protected]>