303 Commits (aa46caf3d95f75b3c540290f0ff8882672287d9d)

Author	SHA1	Message	Date
Yiqun Liu	112f16143b	Add an option to enable the cache of expected kernel in train phase. (#16724) ... * Add an option to enable the cache of expected kernel in train phase. test=develop * Change the default value of cache_expected_kernel to true.	6 years ago
chengduo	55b15db5af	Add unit test for fuse all_reduce ops (#16699) ... * test fuse all_reduce	6 years ago
Yiqun Liu	3fe8cb0dd7	Enable the runtime_context_cache pass in train phase (#16640) ... * Try to enable the runtime_context_cache pass in train phase. * Put the append of runtime_context_cache pass ahead of multi_dev passes. test=develop	6 years ago
Yan Xu	b4c3a6aa0b	[Imperative] implement imperative NCCLParallelContext (#16477) ... add NCCLParallelContext for parallel dygraph	6 years ago
chengduo	b75a69bad6	Add Stream for fetch op handle (#16600) ... * expose fuse broadcast ops	6 years ago
乔龙飞 Qiao Longfei	21622ca30b	Merge pull request #16172 from jacquesqiao/add-async-ssa-graph-executor-communicator ... Add async ssa graph executor communicator	6 years ago
sneaxiy	10249c0b78	Merge develop ... test=develop	6 years ago
Qiao Longfei	adf272bcec	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into add-async-ssa-graph-executor-communicator ... test=develop	6 years ago
sneaxiy	33473890f3	Merge develop ... test=develop	6 years ago
dongdaxiang	f612877797	add incubate for unified API	6 years ago
dongdaxiang	cc4def6ba5	fix some conflict for compilation	6 years ago
xjqbest	824b84d185	add DataSet and InMemoryDataFeed, support load data into memory and shuffle data	6 years ago
dongdaxiang	be757096da	add pybind for fleet	6 years ago
Qiao Longfei	d8974e6da0	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into add-async-ssa-graph-executor-communicator ... test=develop	6 years ago
chengduo	1096746cbf	Fuse Adam And SGD ops (#15933) ... * fuse optimizer	6 years ago
sneaxiy	2c836ff914	check default grad maker ... test=develop	6 years ago
Zeng Jinle	69cb9792ea	Merge pull request #16506 from sneaxiy/revert-16424-fix_allocator_bug ... Revert "Fix allocator bug"	6 years ago
chengduo	ed61d67c73	Fix the interface of Pass::Apply (#16484) ... * modify the interface of Pass::Allay test=develop * Polish code test=develop * Fix Travis CI test=develop * fix Pass::Apply interface test=develop * Fix Travis CI test=develop	6 years ago
Zeng Jinle	174d0d0b90	Revert "Fix allocator bug" ... add include headers to fix travis-ci test=develop	6 years ago
gongweibao	eb83abeac3	Add DGC(Deep Gradient Compression) interface. (#15841)	6 years ago
Zeng Jinle	644e8af4cf	Merge pull request #16424 from sneaxiy/fix_allocator_bug ... Fix allocator bug	6 years ago
Zeng Jinle	c7c6eeb44e	Merge pull request #16409 from sneaxiy/feature/advance_gc ... Enhance gc to support deleting tensor buffer in advance	6 years ago
wopeizl	c300b1ba69	Tensor index (#16223) ... * extend the slice function for python test=develop	6 years ago
Xin Pan	f8c279b11c	Merge pull request #16454 from panyx0718/imperative2 ... polish deepCF model to support real dataset	6 years ago
Qiao Longfei	30618409db	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into add-async-ssa-graph-executor-communicator	6 years ago
chengduo	4f2278f032	Add doc for CPUPlace CUDAPlace CUDAPinPlace (#16442) ... test=develop	6 years ago
sneaxiy	78fb3a62e0	fix env variable settting bug ... test=develop	6 years ago
sneaxiy	2d92b6be98	merge develop ... test=develop	6 years ago
Xin Pan	fd24ab47ab	polish ... test=develop	6 years ago
sneaxiy	a7d0ac50b8	Merge develop	6 years ago
sneaxiy	7000ec85d9	fix some op grad maker ... fix ctest eager deletion disable bug test=develop	6 years ago
sneaxiy	f8ed2c229e	try to fix ci error ... test=develop	6 years ago
sneaxiy	c20db6357b	split PR ... test=develop	6 years ago
sneaxiy	2f54d9f995	Merge develop ... test=develop	6 years ago
sneaxiy	a93a9eef8f	add op registry type ... refine gc code test=develop	6 years ago
sneaxiy	953214ad97	add more unittest ... modify allocator strategy remove changes of legacy buddy_allocator test=develop	6 years ago
chengduo	f26ba5bddd	Fuse AllReduce (#15921) ... * fuse all_reduce test=develop * add fuse_parameter_groups_size test=develop * Polish code test=develop * Fix travis-ci test=develop * Add SetGroupAccordingToLayers and SetGroupAccordingToGroupSize test=develop * Add SetGroupAccordingToMemorySize test=develop * fix multi_devices_graph test=develop * reset params_grads test=develop * Polish code test=develop	6 years ago
sneaxiy	fd23262e0c	merge develop, fix conflict ... test=develop	6 years ago
Qiyang Min	c7f1f3ed0c	Merge pull request #16214 from velconia/imperative_infer_var_type ... Implement imperative infer var type	6 years ago
sneaxiy	161b8ddcaa	Merge develop	6 years ago
Qiyang Min	8e4ad008fb	Merge pull request #16198 from velconia/imperative_train_speed ... Improve imperative mode training speed	6 years ago
minqiyang	36dce65bb3	Take DataType and VarType apart ... test=develop	6 years ago
qingqing01	8ad672a287	Support sync batch norm. (#16121) ... * Support Sync Batch Norm. * Note, do not enable it in one device. Usage: build_strategy = fluid.BuildStrategy() build_strategy.sync_batch_norm = True binary = fluid.compiler.CompiledProgram(tp).with_data_parallel( loss_name=loss_mean.name, build_strategy=build_strategy)	6 years ago
minqiyang	7355d41834	1. Add imperative gperf profiler ... 2. Add binutils 2.27 in manylinux support test=develop	6 years ago
sneaxiy	ac0e0f5181	merge develop ... test=develop	6 years ago
minqiyang	42e96a029f	Accelerate CPU part	6 years ago
sneaxiy	682f2dbf29	merge develop ... test=develop	6 years ago
Yan Xu	30568473ec	fix broadcast on mp mode (#15951) ... * fix broadcast with mp mode * polish code test=develop * fix bcast strategy test=develop * fic cpplint test=develop * fix py3 failed test=develop * fix comment test=develop * update comment test=develop	6 years ago
baojun	e3c37bd564	remove const_cast and refactor ngraph engine code (#15925) ... * remove concast_cast and refactor code test=develop * reduce flag use test=develop	6 years ago
Qiyang Min	1f4aa7a202	Imperative remove all descs (#16045) ... * Remove Desc in Forward Pass * Refactor VarBase * Add dbg info * Only check type in imperative mode * Polish code and support optimizer test=develop * Fix stop gradient problem in PyLayer test=develop	6 years ago