Commit Graph

5169 Commits (b43b46e619ad2aaccf960c7dd4db7f6d15d8a399)

Author SHA1 Message Date
littletomatodonkey 1c08a2136e
test=develop, add addmm op () 5 years ago
zhaoyuchen2018 7b5e23c034
OP(fusion_gru) error message enhancement. test=develop () 5 years ago
Chengmo 8c0bdde934
Add Tdm sampler op in Contrib () 5 years ago
GaoWei8 517929f148
Op (reorder_lod_tensor_by_rank) error message enhancement () 5 years ago
Wilber 1ac9db4354
error message enhancement for fusion_seqpool_concat_op. test=develop () 5 years ago
Wilber 286c2e0ede
error message enhancement for py_func op. () 5 years ago
zhongpu 37fcf03af7
Op (Save/Load) error message enhancement, test=develop () 5 years ago
Zhen Wang 84cd45f674
Solve the conflict of ops with the same name, test for CI. () 5 years ago
wangguanzhong c2f5a3ad34
enhance the error message of roi_align, test=develop () 5 years ago
silingtong123 cec234b1aa
test=develop, error message of tree_conv OP enhancement () 5 years ago
Kaipeng Deng b465bb0de7
fix adaptive_pool2d/pool3d error message. test=develop () 5 years ago
Zhaolong Xing f345607115
Refine transpose flatten concat error message () 5 years ago
Zeng Jinle 7f3e0eaad1
refine error msg, test=develop () 5 years ago
zhaoyuchen2018 42d67dacb6
OP(minus) error message enhancement. test=develop () 5 years ago
Huihuang Zheng a82ce2b1bb
API/OP (ConditionalBlock) error message enhancement () 5 years ago
Yiqun Liu 4489f0d304
Op(fetch) error message enhancement. () 5 years ago
Zhen Wang 2cf27260ae
OP(fake_quantize) error message enhancement () 5 years ago
Zhen Wang 1cf64e00fc
improve error messages of fake_dequantize_op. test=develop () 5 years ago
mozga-intel 3baaee9aab
Remove: NGraph engine from PDPD repository () 5 years ago
wangchaochaohu 81e8fd4a3e
API(fluid.layers.array_length) error message enhancement () 5 years ago
wangguanzhong 6bb8206d03
enhance the error message of box_clip, test=develop () 5 years ago
liym27 8987946fe2
Api/Op (select_input/select_ouput) error message enhancement. () 5 years ago
Wilber 5f22478a93
error message enhancement for repeated fc. test=develop () 5 years ago
Wilber a5bdf485d5
fill op error message enhancement. test=develop () 5 years ago
GaoWei8 2c4b57e94b
Op (concat) error message enhancement () 5 years ago
GaoWei8 66cae9157e
Op (lod_reset) error message enhancement () 5 years ago
liym27 dc225ed2fc
OP (tensor_array_read_write) error message enhancement. test=develop () 5 years ago
GaoWei8 c068512f34
Implement a new C++ operator where and API tensor.where () 5 years ago
Yiqun Liu 55d0c8fde7
Enhance the error message of feed_op. () 5 years ago
Huihuang Zheng 71b5f1d2b2
OP (recurrent) error message enhancement () 5 years ago
Adam 62aff0a7ac
Add DNNL GELU kernels () 5 years ago
silingtong123 009c049e82
Add randint op API () 5 years ago
qingqing01 6162cf2f2e
Make optimizer consistent in dygraph and static-graph and remove some LOG-INFO. () 5 years ago
wangchaochaohu 29c4fae112
Tensor value support () 5 years ago
Chengmo 426912df5a
Add Index sample OP () 5 years ago
zhangchunle 638d924d89
Op (FusionSquaredMatSub) error message enhancement. () 5 years ago
ShenLiang c706ff20a3
fix conflict, test=develop () 5 years ago
Chen Weihang 75bd350710
Implement StaticModelRunner to support dygraph fine-tune static graph pre-training model () 5 years ago
cc 9297f49e4b
[OP] Add randperm op () 5 years ago
Kaipeng Deng d223a24904
Fix inplace_abn compile error on Windows () 5 years ago
Tao Luo 0b583235f5
Revert "Solve the conflict of ops with the same name. ()" () 5 years ago
wawltor 6577f91b74
Add the sum op to API 2.0, add some parameters for new api 5 years ago
WuHaobo c4d0305239
add tril op and triu op () 5 years ago
yongqiangma eb035f24d1
add unbind op () 5 years ago
zhangchunle fd9b7bdb3d
Op (FusedEmbeddingSeqPool) error message enhancement. () 5 years ago
Chen Weihang 16315d3d9e
Delete Ref & VectorRef and add GetDataSafely () 5 years ago
Zhen Wang abe3e6906d
Solve the conflict of ops with the same name. () 5 years ago
wawltor 0b092d05f1
Add the argmax op to API 2.0, and update some parameters 5 years ago
wawltor 915341e3de
Add the zeros, ones, ones_like, zeros_like for api 2.0, test=develop () 5 years ago
Zhen Wang 56b50c97f8
Add allclose_op () 5 years ago
kinghuin 948c57d84b
move sin, sqrt, tanh, atan to paddle.tensor.math and add a new parameter "out" () 5 years ago
Chengmo a2e9af5663
Add Tdm child OP in contrib () 5 years ago
Wilber 9676ac1c5c
Add flip op. () 5 years ago
Feiyu Chan 81f1402f6c
Add functional convolutions in paddle.nn.functional () 5 years ago
Feiyu Chan bcafe3179a
add MKL computation back to gelu's non-approximate part () 5 years ago
zhongpu dbfbd7eac4
support Exhaustive search in dygraph () 5 years ago
zhaoyuchen2018 01d7ccd4b6
Fix elementwise compile error, test=develop () 5 years ago
channings a2e10930cf
update linspace, equal operators to API 2.0 () 5 years ago
zhaoyuchen2018 4fe9ca6959
improve elementwise performance. () 5 years ago
Chen Weihang 7f1ad510bd
Add op inout check macro to simplify error message writing () 5 years ago
zhongpu bfb07aafe8
Revert "Exhaustive search ()", test=develop () 5 years ago
liym27 b7b0b3595b
Add unittest for transformer prediction in dygraph_to_static () 5 years ago
zhongpu 48144e4099
Exhaustive search () 5 years ago
Adam da7c73f847
Delete is_test attribute from activation operators () 5 years ago
Kaipeng Deng 21d95be0db
Add inplace abn op () 5 years ago
wangchaochaohu 69e3f99362
refine the error message () 5 years ago
xujiaqi01 3a45767d49
add fleet pslib pull and push sparse op and push dense op () 5 years ago
GaoWei8 20eed5401a
Change fluid.layers.where‘s C++ operator name () 5 years ago
Jacek Czaja 012886df79
[DNNL] Softmax mkldnn op inplace support () 5 years ago
zhaoyuchen2018 58615a6272
Improve elementwise performance. () 5 years ago
Zhaolong Xing 430b0099c9
[Paddle-TRT]: Ernie Dynamic shape support. () 5 years ago
liym27 6af480ca33
Support int64 for op assign_value. test=develop () 5 years ago
Zeng Jinle 53e6f8e1da
rename macro, test=develop () 5 years ago
Zeng Jinle bba740710d
add cuda resource pool for BufferedReader, test=develop () 5 years ago
Zeng Jinle 7d8d50b6cc
rename no_need_buffer_vars macro, test=develop () 5 years ago
Liufang Sang a486a739e1
fix compile error in win gpu () 5 years ago
Zeng Jinle 7ca77a90ac
add Tensor::IsSharedBufferWith method, test=develop () 5 years ago
Zeng Jinle b8886bf122
rename no_need_buffer_vars_macro, test=develop () 5 years ago
wangchaochaohu b721e23b25
transpose cudnn using cudnn v7 api () 5 years ago
Adam 4f5e4540f8
Improve SGD jit code to work with large data () 5 years ago
Liufang Sang 4db031902d
add dequantize_log_op and make pyramid hash support int8 weight () 5 years ago
Zeng Jinle 9258e96094
fix read op comments, test=develop, test=document_fix () 5 years ago
Zeng Jinle acfc9b8a70
Reader sequential and inference partial feed () 5 years ago
Wilber 95b356a069
update embedding_eltwise_layernorm fuse and kernel. test=develop () 5 years ago
Zeng Jinle a31d7328b7
Add dygraph double grad implementation () 5 years ago
songyouwei 2e2da7124b
high-performance dygraph slice () 5 years ago
Sylwester Fraczek abee05a8c8
added mkldnn swish activation () 5 years ago
Zhaolong Xing 8c6fde9e69
fix align error () 5 years ago
Liufang Sang 915b892a15
Fix div zero in fake quantize op () 5 years ago
Feiyu Chan 01ab8a0619
add approximation for gelu, test=develop () 5 years ago
Adam 5842ae6785
Revert "Change ShareDataWith() to TensorCopy() in conv_mkldnn ()" () 5 years ago
GaoWei8 1dc1f9270e
Fix lod error of concat op for axis = 0 () 5 years ago
Zhang Ting 714b0076b6
Override GetKernelTypeForVar to avoid device transform, test=develop () 5 years ago
wangchaochaohu 112e3edbf6
fix the conv group problem test=develop () 5 years ago
wangchaochaohu 3757e0687c
Add Unittest for backward of fusion group () 5 years ago
chengjuntao 63f3ada7b9
fix bug which input shape () 5 years ago
wangchaochaohu f0d193a23c
Cast fusion for fusion group () 5 years ago
yaoxuefeng 29a7a52d38
Fix instag () 5 years ago
wawltor f154d5860f
Speed up the matmul op, use the gemm replace the batch gemm () 5 years ago
Adam 056edf3929
Change ShareDataWith() to TensorCopy() in conv_mkldnn () 5 years ago
Zhaolong Xing 8d6dc102fe
[Ernie GPU Optimize]: Embedding_eltwise_layernorm Fuse () 5 years ago
guofei 3d8571e884
modify assign op and add unittest of assign op () 5 years ago
Zeng Jinle d33c4343e1
Imperative tracer refactoring () 5 years ago
tangwei12 ad9c8f6d2d
fix communicator when break under pyreder mode () 5 years ago
mapingshuo 5ba9dfc16a
add lookup_table_dequant_op () 5 years ago
Zhaolong Xing dd67d44a50
[Paddle-TRT] : (Part1) Dynamic shape support () 5 years ago
tangwei12 07e13b84cd
remove vlog, test=develop () 5 years ago
Wilber f686310d81
fix concat_mkldnn op. test=develop () 5 years ago
Zhaolong Xing 1a533ed2de
[BUG]: Multihead matmul op's ouput size should be BxSx(N*H) () 5 years ago
石晓伟 ddb9b46fec
change the function in op_teller, test=develop () 5 years ago
tianshuo78520a 433cef03e5
fix typo word () 5 years ago
Kaipeng Deng ebc7ffc300
fix detection_map. test=develop () 5 years ago
zhaoyuchen2018 72dde4abde
Refine adam op to improve performance, test=develop () 5 years ago
wangguanzhong f2d1cd119a
fix lod level, test=develop () 5 years ago
FlyingQianMM 79d712346f
Correct CPU gradients of the argsort op () 5 years ago
guofei ae8b5f11a3
Change ShareDataWith() to TensorCopy() in ref_by_trainer_id () 5 years ago
chengjuntao 15c2667143
register fp16 for assign op () 5 years ago
dyning 1c0653462d
fix generate_mask_labels lod level () 5 years ago
GaoWei8 ba140222d6
fix compile&runtime lod_equality of lod_reset () 5 years ago
ShenLiang 3132681e8a
add partial_sum op in contrib () 5 years ago
ShenLiang e136661304
add partial_concat op in contrib () 5 years ago
tianshuo78520a d2ba91aad1
fix typo words () 5 years ago
Yibing Liu 6e7bfe30a6
register fp16 kernel for some ops () () 5 years ago
tangwei12 66a3150135
SYNC with communicaotor () 5 years ago
Yiqun Liu 22bbd54719
Add the support of fp16 in fusion_group () 5 years ago
Huihuang Zheng adfa5b8354
Add PADDLE_ENFORCE to Check Sequence Length of RecurrentOp () 5 years ago
lidanqing d926214535
[UT coverage] improve the mul_mkldnn_op line coverage () 5 years ago
Zhaolong Xing a06d75a280
[Paddle-TRT] Refine the error log about runtime batch and max_batch_size. () 5 years ago
Adam 608447bfd5
Update MKLDNN to v1.2 () 5 years ago
Adam ab610a34ff
transpose_mkldnn code change to meet Paddle standards () 5 years ago
Jiawei Wang 8f035fb637
Add TopK Op Grad CPU&GPU Kernel test=develop () 5 years ago
Steffy-zxf 90ee366653
update ops's unittest data type from float32 to float64 and shape over 100 () 5 years ago
Zhaolong Xing 8acd745c25
[Ernie GPU Optim]: Fuse three fc to multihtead matmul () 5 years ago
Guo Sheng 31b5464632
Add support for dynamic_decode(while) training. () 5 years ago
Wojciech Uss 4cddb43c5c
Add support for Ernie NLP model to the Slim QAT () 5 years ago
Double_V 58d99247f4
support slice double grad, test=develop () 5 years ago
hutuxian 1a7962be97
Paddlebox about box_wrapper () 5 years ago
huzhiqiang 9e29d3ebed
【OpPorting Example】DEMO OF FIX COMPILE&RUNTIME LOD_EQUALITY () 5 years ago
zhaoyuchen2018 54970444ce
Improve transpose performance with tile sm copy, test=develop () 5 years ago
Wilber a90fa54092
Compile without nccl deps. [1/2] () 5 years ago
Wilber de009152a7 Compile without nccl deps. [2/2] () 5 years ago
Yiqun Liu 4b2227e958 Fix dismatch of std::max's arguments type on windows. () 5 years ago
Wilber 870f465887 fix test_fusion_seqpool_concat lod level between compile and runtime () 5 years ago
Zhong Hui a61d09527b
Fix the integer overflow problem of sequence2batch () 5 years ago
cc 197913ebe1
Add weight quantization in post_training_quanzitaion () 5 years ago
Tao Luo 7c9ce097f1
refine reshape_op shape error message () 5 years ago
LielinJiang 2b1386b2b2
optimize performance of interpolate op () 5 years ago
Yiqun Liu 44b45b9f07
Correct the use of DeviceContext in unittest sequence_pooling_test and sequence_padding_test () 5 years ago
Wilber 7bc4b09500
add WITH_NCCL option for cmake. () 5 years ago
Tao Luo 943cb8c664
fix sigmoid cudnn bug () 5 years ago