Commit Graph

2881 Commits (835ba4ec81275c1e15abda724bcceb0cbae3ae9c)

Author SHA1 Message Date
Leo Chen 48600d7f17
Add op function generator for dygraph (#21569)
6 years ago
Zhang Ting 548efcd2e4
Fix unit tests to avoid check_grad checking failures (#21554)
6 years ago
liym27 be6a639655
Add CI for checking Input/Output/Attr of modified Ops (#21522)
6 years ago
guofei 835ea12cde Control flow API: while_loop (#21276)
6 years ago
Leo Chen 4f81d1bd5f
Refine VarBase init function (#21587)
6 years ago
lijianshe02 56882ce432
change input data type and decrease max_relative_error value in test_check_grad for grop_nom_op test test=develop (#21608)
6 years ago
Huihuang Zheng 1dcf6a7212
Add Much Complex Test and Fix Bugs for Control Flow cond API (#21532)
6 years ago
hutuxian c5aec2fe68
Paddlebox Related to Framework (#21586)
6 years ago
liym27 9da7e6b4d4
add file check_op_desc.py and add interface to get default value. (#21530)
6 years ago
Feiyu Chan 2057df7ac0 add fluid.layers.gelu & doc (#21515)
6 years ago
Jacek Czaja 9ce0e29dc3 [MKL-DNN] Batch norm mkl-dnn NHWC support (#21553)
6 years ago
hong 08483a6819
Add dygraph linear warm up decay (#21186)
6 years ago
lilong12 de46b15951
Unify the rank of prelu alpha to 4, corresponding to [N, C, H, W], except for the all mode
6 years ago
Leo Chen cdd46d7e02
Split VarBase from Python Variable for Dygraph (#21359)
6 years ago
Youwei Song cdba41af4d dygraph Embedding layer use lookuptable v2 (#21209)
6 years ago
wangchaochaohu 4c9b3dafa7
fill_constant_batch_size_like OP precious problem fix (#21337)
6 years ago
Zhang Ting b1da35261b
fix unit tests that do not need to inherit OpTest (#21460)
6 years ago
WangXi 768f9242e9 Fix dgc clip & rampup step, test=develop (#21491)
6 years ago
lilong12 e75ded08a0
fix the compatiable problem between PY2 and PY3 (issue#20749) (#20942)
6 years ago
Zeng Jinle a3535812f6
add _use_system_allocator to some op tests, test=develop (#21504)
6 years ago
Jacek Czaja 18a5d30754 [MKL-DNN] Conv2d and Conv2d transpose MKL-DNN NHWC support (#21466)
6 years ago
zhongpu 6ebf0f47b8 support SelectedRows in dygraph, test=develop (#21078)
6 years ago
lilong12 0bc8bdf724
set dim[0] to -1 if dim[0] < 0 during compiling for c_allgather op (#21402)
6 years ago
ruri 94bef03539
Revert "Add masked select api (#21172)" (#21456)
6 years ago
Zeng Jinle 87ab93af01
fix adam fp64, test=develop (#21423)
6 years ago
hutuxian 7e68bc896b
refactor AUC OP and add its CUDA Kernel (#21336)
6 years ago
Zhang Ting 101240d2c1
fix PythonAPI test in Op unittest, test=develop (#21462)
6 years ago
wawltor dbbe6e9cb6
fix the device supported of the op unique and unique_with_counts. (#21395)
6 years ago
Zhang Ting 3df13ab40c
fix PythonAPI test in Op unittest, test=develop (#21455)
6 years ago
Jie Fang 5e813b53c5 nhwc optimization for batchnorm (#21090)
6 years ago
Chen Weihang 664f958a02
Fix optimizer op infershape failed in dygraph multi-cards mode (#21374)
6 years ago
Huihuang Zheng 630be31952
Fix Cond Bug for Nested Control Flow (#21340)
6 years ago
Jacek Czaja cd43c4440e [MKL-DNN] LRN and Pool2d (FWD) NHWC support (#21375)
6 years ago
zhaoyuchen2018 b16274556a
Add dscending for argsort (#21400)
6 years ago
hong ac8546701d
Add dygraph execution context (#20157)
6 years ago
hutuxian a6b089c614
add macro to ban windows (#21422)
6 years ago
Kaipeng Deng ebfb720a63
add Adam beta1/beta2 support Variable (#21234)
6 years ago
Zeng Jinle 09696d5df8
Use system allocator in OpTest (#21335)
6 years ago
ruri 007c997572
Add masked select api (#21172)
6 years ago
Kaipeng Deng 67c836fb5c
batch_norm momentum support variable (#21246)
6 years ago
hutuxian 47a82e38e3
Support data_norm gpu kernel (#21325)
6 years ago
Youwei Song d5ff79e55e Support numpy bridge (enabled by default in dygraph mode) (#20983)
6 years ago
Michał Gallus 5d7d548275 INT8 Fully-connected (#17641)
6 years ago
lilong12 41d13209d7
add the framework support for distfc (#21197)
6 years ago
hong a214a3081b
change download log format (#21290)
6 years ago
GaoWei8 234060f88f Add fc padding to improve mkl GEMM's performance when N and K are multiple of 128. (#20972)
6 years ago
ruri 6cfcbe0510
reduce interp op input size to pass CI, test=develop (#21341)
6 years ago
Jacek Czaja f4cf028a8c [MKL-DNN] Error throwing for NHWC layout for MKL-DNN ops (#21207)
6 years ago
Michał Gallus ed9ceb9f98 Refactor MKL-DNN ElementwiseMul (#21061)
6 years ago
zhaoyuchen2018 08c19c585d
Improve argsort performance. (#21267)
6 years ago
WangXi 8ac7687e36 Fix dgc accuracy by mv regularization to local (#21278)
6 years ago
Zeng Jinle b9f8ae8494
Add global value getter setter (#21285)
6 years ago
Dong Daxiang 691ced87c0
Refactor fetch handler (#21264)
6 years ago
Yi Liu f1b09ba30e
adapt test_collective_base.py for only two GPU cards available. (#21307)
6 years ago
gongweibao ed2a185248
optimize nhwc for tensor core in ConvOp and ConvGradOp (#20597)
6 years ago
Liufang Sang f0b1518438 add dequantize_abs_max op and modify lookup_table op (#20899)
6 years ago
hutuxian a6ce2306f9 support cvm_op run in gpu (#21300)
6 years ago
Chen Weihang 952508527a
Polish some PE code details (#21274)
6 years ago
xujiaqi01 319d2ba925
fix fs_client_param bug (#21212)
6 years ago
zhongpu fa4d055098 fix bug for python/paddle/fluid/tests/unittests/test_elementwise_mul_op.py, test=develop (#21289)
6 years ago
zhongpu c4ede95c74 open dygraph op test, test=develop (#19787)
6 years ago
lijianshe02 382cf5d7e3
add input type and input data type check for Print_op test=develop (#21250)
6 years ago
liym27 b0fc822747 Add control flow api: case (#21114)
6 years ago
danleifeng 0e7baabe59 extend elementwise broadcast function (#20957)
6 years ago
yaoxuefeng b5d8ba8394
fix data_norm op to avoid impractical normalization result test=develop (#21152)
6 years ago
Zeng Jinle cdb3d27985
Fix warn of gcc8 (#21205)
6 years ago
danleifeng 3fe63d6780 add store_true to use_paddlecloud argument in launch.py (#21168)
6 years ago
liym27 9247528252 Control flow API: switch_case (#21103)
6 years ago
guofei 56b5d14704 Fix the error of init variable in StaticRNN when stop_gradient=ON (#21118)
6 years ago
WangXi 3c98ec90ce Fix INF bug of softmax_cross_entropy_op (#21165)
6 years ago
Zeng Jinle 0f30d3a213
fix dygraph trace bug, test=develop (#21193)
6 years ago
xujiaqi01 23876de55b
fix cache table bug, add save_paddle_inference_model, fix hdfs util bug (#21052)
6 years ago
Kaipeng Deng 98b59cb82c
fix elementwise_mod float point kernel. test=develop (#21183)
6 years ago
Zeng Jinle 5fdfbe3413
Add friendly dygraph trace API (#21091)
6 years ago
whs cfdd1fc2cd
Fix warpctc in padding mode. (#21033)
6 years ago
Tao Luo 3976bbe2ce
add input type and dtype check template, and update some APIs check (#21161)
6 years ago
gongweibao a5fc291fe5 Use 2 cards for hallreduce unit test. (#21085)
6 years ago
zhaoyuchen2018 4a544762a2
Add Asypadding for conv fusion. (#21041)
6 years ago
WangXi de5d3ff688 Fix dgc buffer illegal & reuse velocity (#21012)
6 years ago
lilong12 53148e0696
modify the implementation of save_persistables and save_inference_model for fleet collective mode (#20802)
6 years ago
ceci3 f62a929151
fix instance norm (#21042)
6 years ago
lilong12 e249d9a3e2
fix the computation for dx (grad for x) for prelu operation. (#20949)
6 years ago
Huihuang Zheng e64d55f04e
Add basic Python Cond Layer (#21050)
6 years ago
Huihuang Zheng dcf371b685
Disable cudnn_conv in unit tests. (#21080)
6 years ago
Yiqun Liu 35f17ae28f
Add the check of lod_level between compile-time and runtime. (#20961)
6 years ago
joanna.wozna.intel 77c2083586 Add transpose2 INT8 for mkl-dnn (#19424)
6 years ago
LielinJiang 06063b7001 add op locality_aware_nms, test=develop (#20976)
6 years ago
liym27 26a6e27afe fix bug in pool/conv/conv_transpose: UpdatePaddingAndDilation, _get_padding_with_SAME and conv2dtranspose_forward_naive. (#20997)
6 years ago
Adam 3fda695bb0 Add support for asymetric padding in MKLDNN pool, conv and conv_transpose (#21062)
6 years ago
Huihuang Zheng 1957192f05
Add select_input_op and select_output_op (#21016)
6 years ago
hong 72e0969b27
fix uniform random (#21009)
6 years ago
liym27 f0e95a6049 Polish error messages of pool_2d/3d and add Raises in English document. test=develop (#21017)
6 years ago
Huihuang Zheng 4cf96cd307
Add grad_name Property for Class Variable (#20991)
6 years ago
xujiaqi01 1d1a07937a
simplify master+patch,remove ins when size != merge_size or has conflict slot (#20913)
6 years ago
Zhang Ting de9bec607e lrn supports channel_last input, test=develop (#20954)
6 years ago
Dong Daxiang a6747a6ef1
add launch_ps module so that we can launch a parameter server trainin… (#20936)
6 years ago
Leo Chen 2c3c579b9b tensor.set() supports array list and remove unused code, test=develop (#20959)
6 years ago
Leo Chen 9974e40787 Update Tensor.set() to support float16 (#19964)
6 years ago
123malin 20cdff0e02
Optimize decay (#20816)
6 years ago
Chengmo 16596f6498
Fix Paddle Cloud role maker (#20860)
6 years ago
liym27 59de8e1214 Compatible int32 and int64 for attr in concat/split/unsqueeze. test=develop (#20912)
6 years ago
Zhang Ting 8d1e9f0f7e maxout supports channel_last input (#20846)
6 years ago
hong 8c4573a3cb
GradMaker for dygraph (#19706)
6 years ago
Yiqun Liu 16e4d02675
Refine the cache of program, context and scope in executor. (#18483)
6 years ago
Wilber b489760099
fix jit_matmul bug test=develop (#20886)
6 years ago
hong ff0886a92a
save load problem fix and new feature add (#20823)
6 years ago
Youwei Song 2058bab1c0 Add Sequential api (#20789)
6 years ago
liym27 6802539a2e support Tensor for split and concat, support -1 in num_or_sections, add check num_or_sections (#20780)
6 years ago
Yiqun Liu 6fcfd32e6c
Check and correct the output's lod_level in DynamicRNN related operators (#19144)
6 years ago
Zeng Jinle da9e9dd07f
fix py_reader combination ut, test=develop (#20861)
6 years ago
liym27 84d221b667 improve unsqueeze op to support int, Tensor for argument axes (#20824)
6 years ago
silingtong123 03d7f3ddb2 Make shape tensor support int32 (#20757)
6 years ago
Huihuang Zheng 95ba4bd2ab
Add shape and type check at read_op (#20754)
6 years ago
Aurelius84 aacd16dbb4 add pyramid_hash_op (#20698)
6 years ago
whs c8e49be2f1
Fix roi_perspective_transform op (#20764)
6 years ago
xujiaqi01 48669aa8f0
fix several sparse table issuses (#20686)
6 years ago
Zhang Ting 5a8d885d72 All elements in attr(shape) of crop_tensor can be -1 and int32/64 kernel registered (#20756)
6 years ago
Tao Luo efbdad0596
make search_compute support avx default (#20779)
6 years ago
WangXi 250e72d254 Fix DGC algorithm flow to make it the same as paper (#20758)
6 years ago
Zeng Jinle 8ff6b289bd
[Dygraph to static graph]JIT/Trace (#20775)
6 years ago
石晓伟 37cd43545a
update the infer shape of matmul, test=develop (#20717)
6 years ago
gongweibao e425124041
Wait pserver to complete initialization. (#20777)
6 years ago
gongweibao 8088395a84
Set unique port to every distribute test to avoid potential port conflicts (#20759)
6 years ago
wangchaochaohu 0687bcd64f
Refine getitem of Variable (#20729)
6 years ago
Zhang Ting 483d0512ce Add choice of CUDA Place and remove fluid.layers.data for python API test of resize Ops (#20689)
6 years ago
danleifeng 79e08ecebf add assertions on whether elementwise_div divison is zero (#20618)
6 years ago
123malin 95e90aa102
test=develop, add communicator_is_sgd_optimizer flag (#20677)
6 years ago
WangXi 507afa8a8a Fix dgc nan by stripping nccl from sparseReduce. (#20630)
6 years ago
gongweibao c1710e91b2
Disable GRPC_ARG_ALLOW_REUSEPORT to avoid potencial problem. (#20690)
6 years ago
gongweibao 78431dc7bc
Add detail log on parallel dygraph unit test when failed. (#20660)
6 years ago
gongweibao f3f52fc1e2
Retry when failed to bind address. (#20642)
6 years ago
qingqing01 01eddc1a04
Support fp16 in GPU impl of fused_elemwise_activation_op. (#20636)
6 years ago
Tao Luo ea20f8be23
refine error message of input TypeError in framework.py (#19970)
6 years ago
Chengmo 940c6ff1c8
Fix communicator slow bug & fix communicator stop bug (#20366)
6 years ago
liym27 ae4ddbc821 add fill_constant input(shape) dtype check. test=develop (#20620)
6 years ago
123malin b4a3b75002
bug fix: invalid learning rate decay in pserver async mode (#20325)
6 years ago
WangXi cadc6a9704 fix dgc test and bug when not set trainers_endpoints_, test=develop (#20617)
6 years ago
石晓伟 a4753f3a79 Optimize error message of mean_op and matmul_op (#20413)
6 years ago
mapingshuo f55d1c6867
Fleet: deal with special case: strategy is None (#20359)
6 years ago
翟飞跃 36acfaeeda Add fused_embedding_seq layer into fluid.contrib (#19771)
6 years ago
zhupengyang 01209b5161 add input type and dtype check for cast_op (#20070)
6 years ago
Youwei Song 67fcb0c923 fix api fluid.cuda_pinned_places (#20424)
6 years ago
gongweibao bf6470c71e
Add detail logs on resnet unit test (#20558)
6 years ago
chengduo 36c85ef492
Add sub-scope check in RecurrentOp (#20468)
6 years ago
zhupengyang 5e65c753ea enhance input type chec for concat (#20584)
6 years ago
JesseyXujin 2ff18e537f
add expand_as op, test=develop (#20565)
6 years ago
Zeng Jinle 40effc61af
Refine py_reader exit (#20331)
6 years ago
Guo Sheng bd99df715a
Fix basic_gru and docs of gru_unit and dynamic_gru (#19393)
6 years ago
Yuan Shuai 172e91c008
Refine error message of transpose_op (#20437)
6 years ago
hong19860320 f567214b6a
Fix the error message of zeros op (#20476)
6 years ago