...
- Fix Nightly CI (#20019)
- correcting cuda 11.2 image name in CI and CD (#19960)
- CI fixes to make more stable and upgradable (#19895)
- Address CI failures with docker timeouts (v2) (#19890)
- Attempt to fix v1.x CI issues. (#19872)
- Update CI build scripts to install python 3.6 from deadsnakes repo (#19788)
- Fix R builds on CI (#19656)
- Update CD Jenkins config for include/mkldnn/oneapi/dnnl (#19725)
- Fix CI builds failing due to invalid GPG keys. (#19377)
- Disable unix-gpu-cu110 pipeline for v1.x build since we now build with cuda 11.0 in windows pipelines. (#19828)
- [BACKPORT]Enable CUDA 11.0 on nightly + CUDA 11.2 on pip (#19295)(#19764) (#19930)
- Fix nightly cd cu102 (#19940)
- Drop cu9x in cd (#19902)
- update cudnn from 7 to 8 for cu102 (#19522)
- update cudnn from 7 to 8 for cu102 (#19506)
- Attempt to fix v1.x cd by installing new cuda compt package (#19959)
- [FEATURE]Migrating all CD pipelines to Ninja build + fix cu112 CD pipeline (#19974)
- Fix nightly CD for python docker image releases (#19774)
- [CD] Fix nightly docker missing lib (#20120)
- [CD] Fix CD cu102 110 112 cuda compatibility (#20116)
- Disable codecov. (#20175)
- Static build for mxnet-cu110 (#19272)
CI fixes (#20648)
Subgraph API
- Move block.optimize_for backend_opts to kwargs (#19386)
- Backport Enable Numpy support for Gluon Block optimize_for to v1.x (#19456)
- Save/Load Gluon Blocks & HybridBlocks (#19565)
- Fixed setting attributes in reviewSubgraph (#19274)
- Fix for optimize_for multiple subgraph properties issue (#19263) (#20142)
- Reuse params from cached_op_args (#20221)
...
- adding License in libmxnet make config .sym and .ver files (#19937)
- add missing license fix from master to v1.x (#19916)
- Fix license for blockingconcurrentqueue (#19910)
- update notice year (#19893)
- Backport [LICENSE] Reorganize rat-excludes file to ease license auditing (#19743) (#19799)
- Update LICENSE (#19704)
- [LICENSE] Change intgemm to a submodule instead of fetch. (#19407)
License updates per ASF feedback (#20377)
License updates per feedback (#20428)
Remove the "copyright by contributors" line in source files (#20493)
Make sure files with 2 licenses are listed properly in LICENSE. (#20492)
Add copyright detection and removal in license checker (#20498)
[LICENSE] updated licenses for files under both ASF + BSD 3-clause (#20495)
modify erfinv implementation based on scipy (#20517)
[License] Remove mistakenly placed ASF headers (#20520)
Just remove image classification CPP example from source tarball. (#20530)
license updates (#20598)
LICENSE and CI fixes for 1.9 release (#20626)
Port #20709 to v1.9.x branch (#20722)
Website improvements
- add djl and autogluon to website (#19981) (#19995)
- add website artifacts pipeline (#19397)
- v1.x website patch (#19192)
...
- Fix take gradient (#20166)
- Pip Build: use runtime.Features instead of manual check for mkldnn headers (#19195) (#19928)
- Fix AmpCast for float16 (#19749)
- Backport extension bug fixes to v1.x (#19469) (#19503)
- Fix MKLDNN BatchNorm with even number of channels (#19150) #19299 #19425 (#19445)
- Fix to quantization dshape bug (#19501)
- Upate ps-lite to fix the zmq not found issue (#20248)
- update short desc for pip (#20237)
- set osx deploy target for v1.x wheels (#20127)
- downloading MNIST dataset from alternate URL (#20014)
- pass version param (#19982)
- Remove unmaintained BLC (#19801)
- Update setup.py for darwin builds (#19130) (#19927)
- Unskip Flaky test_gluon_data tests (#19919)
- WAR the dataloader issue with forked processes holding stale references (#19924)
- For ECR, ensure we sanitize region input from environment variable (#19882)
- Migrate to use ECR as docker cache instead of dockerhub (#19654)
- provide a faster PrefetchedDataLoader (#19748)
- Update dgl_graph.cc (#19827)
- initial commit (#19757)
- [PERFORMANCE] Layer normalization code from Marian for CPU (#19601)
- fixed macros with name (#19669)
- Fix local variable 'optimizer' referenced before assignment (#19666)
- Support destructors for custom stateful ops (#19607)
- Remove obsolete six dependency (#19620)
- Backport Faster pointwise fusion graph pass (#19269) (#19413)
- Don't use namespace for pow() function, since it is built into cuda math library, and cast the second argument so it will find an acceptable form. (#19532)
- backport #19037 (#19514)
- Allow eliminating common subexpressions when temp space is used (#19487)
- Fix h5py version on (#nto)
- Relaxing type requirements for broadcast_like (#17977) (#19447)
- initial disclaimer update (#19402) (#19415)
- backport slice assign large tensor fix (#19399)
- Fix SoftReLU fused operator numerical stability (#17849) (#19391)
- backport #19393 to v1.x (#19396)
- Remove extra --build-arg causing docker command to fail. (#19411)
- backport fixes in master branch (#19356)
- Remove build_ccache_wrappers invocation from R-package unittests (#19305)
- Refactor cmake cpp-package & add missing inference/imagenet_inference (#19228)
- Backport PRs in v1.7.x missing from v1.x to v1.8.x (#19262) (#19281)
- fixing breaking change introduced in #17123 when batch_axis=0 (#19267)
- Backport of #19078 (#19095)
- added key for samskalicky (#19224)
- delete executor before reallocating it memory (#19214)
- Nightly Large Tensor test cherrypicks (#19194)
- Updated v1.x to version 1.9 after branching v1.8.x (#19196)
- Fix flaky test #19197 by avoiding case that 0.45 mapped to 0.5 (#19201)
- Tweeking syntax to be closer to other tests (#19186)
- Add code signing key (#20276)
stop closing opened libs (#20523)
Add MKLDNN Deconvolution 1D and 3D support (#20137)