Commits

Commits on Feb 20, 2018

Add link to arXiv page (horovod#179 )
alsrgv
committedFeb 20, 2018
Add reference to arXiv publication (horovod#178 )
alsrgv
committedFeb 20, 2018

Commits on Feb 14, 2018

Bump version for new patch release (horovod#177 )
alsrgv
committedFeb 14, 2018
Add hvd.local_size() API (horovod#176 )

ppwwyyxx
authored and
alsrgv
committedFeb 14, 2018

Commits on Feb 13, 2018

Upgrade the dockerfile to use tensorflow 1.5.0, cuda 9, cudnn 7, and nccl 2.1.4. (horovod#170 )

rongou
authored and
alsrgv
committedFeb 13, 2018

Commits on Feb 6, 2018

Fix spelling error in docker documentation (horovod#164 )
alsrgv
committedFeb 6, 2018

Commits on Feb 1, 2018

Update benchmark to the official TF benchmark (horovod#158 )
alsrgv
committedFeb 1, 2018

Commits on Jan 18, 2018

Add a reference Horovod Dockerfile (horovod#150 )
alsrgv
committedJan 18, 2018

Commits on Jan 9, 2018

Add Keras ImageNet training example and LearningRateScheduleCallback (horovod#144 )
alsrgv
committedJan 9, 2018

Commits on Jan 8, 2018

Update Keras github url in .travis.yml (horovod#143 )

ziky90
authored and
alsrgv
committedJan 8, 2018

Commits on Jan 5, 2018

Add TensorFlow Estimator example, update existing examples (horovod#140 )
alsrgv
committedJan 5, 2018

Commits on Jan 3, 2018

Readme updates: logo, references (horovod#138 )
alsrgv
committedJan 3, 2018

Commits on Jan 2, 2018

Dockerize Travis as TensorFlow will require new GLIBC (horovod#136 )
alsrgv
committedJan 2, 2018

Commits on Dec 1, 2017

Run nightly test against master Keras (horovod#116 )
alsrgv
committedDec 1, 2017

Commits on Nov 22, 2017

Add -stdlib=libc++ to CFLAGS for MacOS (horovod#104 )
alsrgv
committedNov 22, 2017

Commits on Nov 16, 2017

Replace -oversubscribe with -map-by slot (horovod#95 )
alsrgv
committedNov 16, 2017

Commits on Nov 13, 2017

Modify LRWarmupCallback to scale LR from initial_lr / hvd.size() to initial_lr (horovod#88 )
alsrgv
committedNov 13, 2017

Commits on Nov 9, 2017

Bugfix: Initialize MetricAverageCallback variables (horovod#85 )
alsrgv
committedNov 9, 2017

Commits on Nov 7, 2017

Add support for boolean broadcast() and allgather() (horovod#79 )
alsrgv
committedNov 7, 2017

Commits on Nov 6, 2017

Add an error message that indicates outdated MPI to the Troubleshooting page (horovod#80 )
alsrgv
committedNov 6, 2017
Add horovod.__version__ (horovod#78 )
alsrgv
committedNov 6, 2017
Various README updates (horovod#76 )
alsrgv
committedNov 6, 2017

Commits on Nov 3, 2017

First-class Keras support (horovod#68 )
alsrgv
committedNov 3, 2017

Commits on Nov 2, 2017

Add support for new tf.sysconfig.get_compile_flags() and tf.sysconfig.get_link_flags() (horovod#66 )
alsrgv
committedNov 2, 2017

Commits on Oct 31, 2017

Add requirement for TensorFlow 1.1.0+ (horovod#64 )
alsrgv
committedOct 31, 2017

Commits on Oct 30, 2017

Update Horovod on GPU docs (horovod#61 )
alsrgv
committedOct 30, 2017

Commits on Oct 28, 2017

Various documentation updates (horovod#59 )
alsrgv
committedOct 28, 2017

Commits on Oct 19, 2017

Python 3 compatibility fix (horovod#45 )

brian8128
authored and
alsrgv
committedOct 19, 2017

Commits on Oct 17, 2017

README updates (horovod#44 )
alsrgv
committedOct 17, 2017
Copyedit concepts.md (horovod#43 )

sblotner
authored and
alsrgv
committedOct 17, 2017

Commits on Oct 12, 2017

Add grad=None check in DistributedOptimizer (horovod#38 )
alsrgv
committedOct 12, 2017
Replace one-second sleep with MPI barries in NCCL initialization (horovod#26 )
alsrgv
committedOct 12, 2017

Commits on Oct 11, 2017

Better error check for NCCL init (horovod#37 )
alsrgv
committedOct 11, 2017
Support for Cray MPI: custom command to discover MPI compilation flags (horovod#36 )
alsrgv
committedOct 11, 2017

Commits on Oct 7, 2017

Bugfix TensorFlow master build for case when CUDA libraries are present (horovod#32 )
alsrgv
committedOct 7, 2017