8186 Commits (266cdf7d016a9ec7581a4174d74f0b35d849e180)

Author	SHA1	Message	Date
gongweibao	266cdf7d01	Fix dgc bug. (#16709)	6 years ago
Qiyang Min	7e560558c2	Imperative fix bugs (cherry pick to 1.4) (#16680) ... * Fix imperative bugs test=release/1.4 * Polish code test=release/1.4	6 years ago
Jiabin Yang	29f3441615	Cherry pick/fix transformer (#16620) ... * Imperative deep-first backward process (#16605) * Fix bug of gradient interface * shrink transformer * Right transformer * Change from width-first backward to deep-first backward process test=develop * Reverse iterator op's input test=develop * Polish code * Change the iteration direction in ingrads' map slots test=develop * Polish code test=develop * test=develop, cherry-pick fix for transformer in dygraph * test=develop, fix transformer in dygraph /	6 years ago
Jiabin Yang	89d09b8319	Cherry pick 1.4/ptb fix (#16607) ... * test=develop, ptb_rnn fix op * test=release/1.4, refine code * test=release/1.4, fix ci failed error	6 years ago
gongweibao	065ffcce6f	fix dgcclipnorm bug test=develop (#16629)	6 years ago
chengduo	ab6600e18c	Fix bug of FastThreadedExecutor (#16666) ... test=release/1.4	6 years ago
chengduo	4cc614412c	[Cherry-pick]Fix the bug of all_reduce_deps_pass (#16648) ... * fix the bug of all_reduce_deps_pass test=release/1.4	6 years ago
guru4elephant	76b49f02ee	Merge pull request #16539 from guru4elephant/train_with_pipe_reader_merge_develop ... Train with pipe reader merge develop	6 years ago
Qiyang Min	d8d73ff3db	Merge pull request #15584 from velconia/imperative_lr_scheduler ... Support imperative learning rate scheduler	6 years ago
qingqing01	1ebd7434d5	Add linear learning warmup method in learning rate scheduler. (#16563) ... * Add linear learning warmup method This warmup lr can be combinated with other learning rate strategies. For example: decayed_lr = fluid.layers.linear_lr_warmup( fluid.layers.piecewise_decay(boundaries, lr_steps), warmup_steps, start_lr, end_lr)	6 years ago
Wu Yi	22b02bfa62	Batch norm cudnn accurate (#16545) ... * fix cudnn batch norm accuracy test=develop * fix cudnn batch norm accuracy test=develop * disable failed test for later fix test=develop	6 years ago
zhaoyuchen2018	0b0abdbcd4	Merge pull request #16041 from zhaoyuchen2018/docrefine ... Fix document array_read code error	6 years ago
dongdaxiang	a53c8cd5a7	fix infer_from_dataset docs and abastract class problem ... test=develop	6 years ago
xjqbest	a99c8d0c29	fix client to client communication bug ... test=develop	6 years ago
lujun	2b4d887460	Merge pull request #16551 from wopeizl/remove_useless_check ... remove the useless check	6 years ago
Kaipeng Deng	3d939d32ee	Merge pull request #16023 from heavengate/kl_div_loss ... KL div loss: add kldiv_loss op	6 years ago
Kaipeng Deng	54474637ae	Merge pull request #16057 from heavengate/softmax_axis ... Add attr 'axis' for softmax	6 years ago
Kaipeng Deng	63ac947e2f	Merge pull request #16135 from heavengate/shift ... Add temporal_shift op for TSM model	6 years ago
chengduo	bb80dae7d0	Add DecoupledWeightDecay (#16427) ... * Add DecoupledWeightDecay	6 years ago
dongdaxiang	3829eac27b	fix API spec about infer_from_dataset ... test=develop	6 years ago
whs	73c4f2b7b6	Fix distillation for soft label. (#16538) ... test=develop	6 years ago
minqiyang	64b0929417	Polish code ... test=develop	6 years ago
peizhilin	9c6eb1aa46	remove the useless check test=develop	6 years ago
wopeizl	e014950e87	add slice support for dim < 0 (#16494) ... * add slice support for dim < 0 test=develop	6 years ago
dongdaxiang	87027a2eef	fix API.spec problem and executor's docstring ... test=develop	6 years ago
minqiyang	34426e761e	Polish code ... test=develop	6 years ago
minqiyang	fb7c787d34	Fix conflicts ... test=develop	6 years ago
dongdaxiang	ade9337486	fix API.spec ... test=develop	6 years ago
minqiyang	3e57981294	Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into imperative_lr_scheduler ... test=develop	6 years ago
Zhen Wang	4c1ec41de3	Merge pull request #16531 from wanghaoshuang/quan_ck ... [slim] Fix checkpoint of quantization strategy.	6 years ago
AIFollowers	e18ab78f67	add model_stat.py (#16512) ... * Add a tool to summary model's PARAMS, FLOPs in paddle/fluid/contrib.	6 years ago
dongdaxiang	720647e17f	rebase current develop and fix conflict ... test=develop	6 years ago
dongdaxiang	93c3c7f9b3	fix dataset testcase problem ... test=develop	6 years ago
dongdaxiang	d739bab844	fix async_executor problem and remove some unnecessary testcase, fix trainer_desc import problem ... test=develop	6 years ago
dongdaxiang	60b7bf6fa6	add infer_from_dataset for inference	6 years ago
dongdaxiang	88880d9b69	fix import trainer_desc_pb2 error ... test=develop	6 years ago
dongdaxiang	0030eb2a61	fix distributed building ... test=develop	6 years ago
xjqbest	1497ce388d	fix code style of test_dataset.py ... test=develop	6 years ago
xjqbest	7cdd57a474	fix code style of test_dataset.py ... test=develop	6 years ago
xjqbest	748d54cb46	fix code style of test_dataset.py ... test=develop	6 years ago
xjqbest	1073b4d8f9	fix code style of test_dataset.py ... test=develop	6 years ago
dongdaxiang	45eb6f0765	run pre-commit check files and fix code style problem ... test=develop	6 years ago
xjqbest	e57ac5ed17	fix code style ... test=develop	6 years ago
xjqbest	97c74e60c3	fix code style ... test=develop	6 years ago
xjqbest	a38b98cb32	fix code style & runtime error ... test=develop	6 years ago
dongdaxiang	8e14d8f900	add data_generator package into setup.py	6 years ago
dongdaxiang	17790188d0	make role maker and distributed optimizer private	6 years ago
dongdaxiang	d87ba58c14	refine document of python API, make device_worker and trainer's API private ... test=develop	6 years ago
dongdaxiang	b95b80bc76	add doc string for executor and update API.spec ... test=develop	6 years ago
xjqbest	d52586a97d	add doc string ... test=develop	6 years ago