bf461fa524
Improving error report message for sequence_expand op ( #27245 )
...
* improve err report, test=develop
5 years ago
bbad3414e8
Enhance the error messages for files in operators/math
...
Enhance the error messages for files in operators/math
5 years ago
aae41c6fca
refine error message related to paddle-TRT ( #27256 )
5 years ago
d708b21074
Update amp_check_finite_and_scale_op and add an updating_loss_scaling op for static graph amp training. ( #26240 )
...
* update amp_check_finite_and_scale_op for static_amp.
* use amp_check_finite_and_scale in static graph amp.
* update grads to zero when grads own infinite values(as for amp_checkout_finite_and_scale op).
* add update_loss_scaling op in cpp.
* add update_loss_scaling_op unit test.
* update the doc of the check_finite_and_unscale op
* Update the process of gradients updating skipping if the gradients have infinite values.
* update the way to zero grads.
* update test_update_loss_scaling_op.py
* add log info when find infinite grads.
* add the unit test for UpdateLossScaling Layer.
5 years ago
cc3f4b813a
Add int8 GRU kernel ( #27220 )
...
* Add int8 GRU kernel with UTs
* Lint fixes
* More lint fixes
5 years ago
9437ce36c4
Error description optimize for math dir
...
Error description optimize for math dir
5 years ago
5c1bafbbc6
use eval to improve performance, test=develop ( #25459 )
5 years ago
5c4eed66fd
Fix GRU mkldnn kernel fail on look_table_v2 ( #27198 )
...
* Fix the lookup_table_v2 failed on GRU mkldnn kernel issue
test=develop
* fix according to reviews, removed x_num_col_dims
test=develop
* update gru model. change according to reviews
test=develop
* change according to reviews
test=develop
5 years ago
33ff833af2
fix loaded no params layer run error ( #27241 )
5 years ago
1b84c0bf43
Lite subgraph refine predictor ( #27167 )
5 years ago
2e59769612
add empty op (c++, python, unit test) ( #26659 )
5 years ago
c5f957ae38
add double grad for tile op and expand_v2 op ( #27114 )
...
* add double grad for tile, test=develop
* add double grad for expand_v2 op, test=develop
5 years ago
58a88ba9af
add double grad for expand ( #27183 )
...
* add double grad for expand, test=develop
5 years ago
7c7fbd3218
fix error msg of fused_embedding_fc_lstm_op, test=develop ( #27231 )
5 years ago
e005861598
[oneDNN]Introducing oneDNN 1.6 ( #27137 )
...
* - introducing oneDNN 1.6
test=develop
* - Removed redundant code
test=develop
5 years ago
5bd84b22c4
revert divide ( #27202 )
5 years ago
fde5cfe881
fix the CudaPinMemory bug for the equal op ( #27176 )
...
fix the CudaPinMemory bug for the equal op and add the test case for the equal op
5 years ago
cc3306f7c8
restruct logsumexp to speed up compiling ( #27191 )
5 years ago
50e60e8779
update error info for selected_rows_functor
...
update error info for selected_rows_functor
5 years ago
5d039f4086
modified the implement of Lars optimizer ( #26733 )
...
add lars to fleet meta optimizer
5 years ago
c71d79b1d2
[cuda11 support] change the CMakeLists to support the cuda11 ( #27124 )
5 years ago
43b0445b29
Add double grad in reduce sum ( #27115 )
...
* set default value to strategy in distributed_optimizer test=develop
5 years ago
ed292695c5
optimize the error message for math dir
...
optimize the error message for math dir
5 years ago
4558d395e9
fix Norm op error ( #26771 )
...
* fix frobenius_norm error, rm p=0 2-axis support. test=develop
5 years ago
4d7d661249
Fix kl and summary bug ( #27132 )
...
* fix summary rnn
* fix kl_div bug when input shape is [1] and reduction is batchmean
5 years ago
eb01976037
[2.0 API]Add checker in grid_sample_grad op ( #27126 )
5 years ago
a28ae86e11
Enhance ops to support LoD as input for dygraph detection models. ( #25316 )
...
* enhance collect_op for dygraph, test=develop
* enhance detection ops with lod, test=develop
* support none bbox left in generate_proposals, test=develop
* unfiy MultiLevelRoisNum, test=develop
* update core.ops, test=develop
* add op register for new input & output, test=develop
5 years ago
8df5b4d608
Add correlation api to contrib ( #27015 )
...
* add correlation api to contrib
5 years ago
1b102dd552
optimize the error message for unpooling.cc
...
fix the error message for the unpooling.cc
5 years ago
58f3ef982a
fix typo for interp_v2,test=develop ( #26843 )
...
* fix typo for interp_v2,test=develop
* align with torch, test=develop
* add area mode, test=develop
* fix bug, test=develop
* format notes, test=develop
* update for converage, test=develop
* fix bilinear, test=develop
* fix bicubic, test=develop
* fix typo, test=develop
* fix coverage, test=develop
* fix helper.input_dtype, test=develop
* polish notes, test=develop
* polish notes, test=develop
* polish notes, test=develop
5 years ago
5af81f833c
fix gpu kernel for numel Op ( #27085 )
5 years ago
19ca6d9dd2
add .part to speed up compile ( #27044 )
5 years ago
4ff16eb201
Add padding cudnn interface ( #26370 )
...
* add lstm cudnn of padding data and refine cudnn codes
5 years ago
8857e3911f
add the dynamic dtype check for the argmin/argma
...
update the check for the dtype check for the argmin, argmax
5 years ago
041f4ab842
refine linspace Op for dtype setting( #27071 )
5 years ago
9aa39584fe
fix cuda generator hard-coded offset step ( #27027 )
5 years ago
f6653c71e9
[oneDNN] Fix to conv2d grad with groups ( #27006 )
...
* - Added fix to mobilenet
* - compilation fix
* - Fix to conv2d grad oneDNN with groups
test=develop
5 years ago
a72752263b
support heter-xpu-ps ( #27018 )
...
support heter-xpu-ps
5 years ago
2660ea379d
Fix cuda kernel of affine grid ( #27003 )
...
test=develop
5 years ago
ff3dc8ac73
fix the remainder ( #26995 )
5 years ago
7f3e6ca596
add cuda generator ( #26786 )
5 years ago
95e1434bb2
Add bfloat16 data type ( #25402 )
5 years ago
29b844ad5e
Fix clip op attr ( #26924 )
5 years ago
e480168fae
fix dropout bug in backward when input is 1d tensor ( #26837 )
...
* fix dropout bug in backward when input is 1d tensor, test=develop
* add test case and refine error message, test=develop
* refine error message, test=develop
5 years ago
5e874cc333
- Cosmetic fixes to align with PADDLE_ENFORCE guidelines ( #26891 )
...
test=develop
5 years ago
5205748481
fix eigen in push sparse; fix hadoop command ( #26872 )
...
* fix eigen in push sparse; fix hadoop command
test=develop
* add log in load_combine_op
test=develop
5 years ago
0a29fc85d6
fix the argmin,argmax op for the paddlepaddle 2.0
...
* fix the argmin,argmax op for the paddlepaddle 2.0， add checkPoint for the argmax/argmin
5 years ago
d0962abd20
supplement bug fix of parameter server ( #26217 )
...
* fix fluid.embedding
5 years ago
60ffc22026
Refine bernoulli and unsqueeze op ( #26842 )
...
* add check for bernoulli and register bool for unsqueeze
* follow comments
5 years ago
ebc5f99789
add embedding 2.0 ( #26649 )
...
* add embedding 2.0
* add embedding support input int32
5 years ago
40378edfa8
Add the AddCheckpoint macro to softplus op ( #26809 )
5 years ago
11fb8a1c10
Refine cudnn softmax ( #25757 )
...
* refine cudnn softmax
5 years ago
7ee70a47b8
update the doc for the some ops
...
update the doc for the some ops, ceil asin, atan
5 years ago
0f1ad9b06c
leaky_relu and hardshrink add checkpoint for behavior changed ( #26802 )
5 years ago
7f2aa2db3c
【paddle.fleet】Support Heter Parameter Server ( #25998 )
...
* Support Heter Parameter Server
5 years ago
a1b99fae07
Adadelta Optimizer ( #26590 )
...
* add doc; notest
* fix doc; notest
* update doc; notest
* refine optimizer && adam
* refine optimizer; notest
* add adam
* fix doc
* fix doc && add adamw; notest
* add error message
* bug fix
* refine rmsprop && adamax
* fix ci
* buf fix
* update comment
* unify arguments place; notest
* fix ut, test=develop
* bug fix
* fix conflicts, test=develop
* add examples code
* bug fix
* fix comments
* fix sample code
* add sample code for Optimizer
* add adamax ut, test=develop
* fix rmsprop ut, test=develop
* add ut for optimizer.py and adamw.py
* first commit of adadelta optimizer
* fix learning rate
* fix adadelta doc and add sgd momentum
* remove unused fluid
* fix codestyle
* Update test_adam_op.py
* Update test_adam_op.py
* fix SGD in 2 unittests
* fix SGD in 2 unittests
* fix ci
* fix ut
Co-authored-by: MRXLT <xlt2024@gmail.com>
Co-authored-by: mapingshuo <mps2012@yeah.net>
5 years ago
346689c6f1
Register conv_transpose Op version for compatible Op upgrades ( #26745 )
...
* fix bug
* add version check
* fix docs, test=document_fix
* fix formula, test=document_fix
5 years ago
7afb1df11e
Decouple weights and bias from fc primitive in MKLDNN cache ( #26708 )
...
* decouple weights and bias from fc primitive in cache
* removed reduntant update of pointers
5 years ago
844583c8fd
Refine paddle.manual_seed ( #26496 )
...
* refine manual seed
* fix ci problem
* fix unittests
* fix unittest
* set is_init_py=false in manual_seed
* fix unittest
* fix bernoulli_op
* fix(unittest): change random_seed to manual_seed
* 🐞 fix(unittest): fix manual_seed
* trigger ci
* fix test_sentiment
* fix test_imperative_save_load
* fix test_uniform_random_op
* fix test_uniform_random_op
* fix test_jit_save_load
* merge develop
* fix manual_seed
* fix manual_seed
* use global engine
* use shared_ptr
* fix double free
* fix bug
* fix bug
* fix bug
* fix test bug
* fix test bug
* fix test bug
* fix ci
5 years ago
29494d703d
fix remainder, floor_div ( #26732 )
...
* fix remainder, floordiv
5 years ago
5f524efe56
modify error report message, test=develop ( #26743 )
5 years ago
4561fc37e2
Add check point for gather Op ( #26696 )
5 years ago
1ec30cb160
register cumsum Op version for compatible Op upgrades ( #26734 )
...
register cumsum Op version for compatible Op upgrades
5 years ago
c282db3a93
add broadcast feature for elementwise logical op
...
add broadcast feature for elementwise logical op
5 years ago
63eef7632e
Fix clip input check ( #26683 )
...
* Fix clip input check
* Fix default min/max value
* Allow both max and min to be None
* Register op change
* Revert OP signature change
5 years ago
f311d3c1cf
Fix pow api type error with python side method, merge elementwise_pow and pow. ( #26163 )
...
As the title
5 years ago
e4cc6a28b0
Norm op support 2-axis ( #26492 )
5 years ago
89d7d86684
add intepolte_v2 ( #26520 )
...
* add intepolte_v2
* fix linear interp
* polish unittest, test=develop
* update code samples to 2.0 API, test=develop
* remove warning, test_develop
* add name in attrs, test=develop
* polish code, test=develop
* change Align to align, test=develop
* fix unittest in py3,test=develop
* fix coverage, test=develop
* fix coverage, test=develop
* fix for windows ci, test=develop
* fix coverage, test=develop
5 years ago
97cebfa4d3
add dtype for unique ( #26655 )
...
* update doc, test=document_fix
* add attr(dtype)
* refine code
5 years ago
1c68138327
[api 2.0] add collective op for cpu using gloo and paddle.distributed.* apis ( #26552 )
...
add collective op for cpu using gloo and paddle.distributed.* apis
5 years ago
559e43eee4
Small change in conv2d and quantize pass ( #26671 )
5 years ago
8986a82131
fix adaptive gpu grad bug, add doc refine ( #26660 )
5 years ago
286eca2d9e
update the code for the topk v2
...
add the top v2 for the paddlepaddle api 2.0
5 years ago
f82384113b
Fix atomicAdd in grid sample op and affine grid op ( #26647 )
...
test=develop
5 years ago
32ba8602c6
Enhance py_func error info message. ( #26557 )
5 years ago
0a895bc0df
improve unique op ( #26537 )
...
* add unique_v2 op
* remove unique_v2 op
* update doc
5 years ago
a004dfde3d
Use atomicAdd defined in paddle fromework ( #26631 )
...
test=develop
5 years ago
c80fcf901e
reduce_mean error if keepdim=True and reduce_all=True ( #26614 )
5 years ago
a065a24232
【2.0 API】Enhance affine grid operator ( #26385 )
...
* Enhance affine grid operator:
1. Add cuda kernel
2. Add align corners options
test=develop
* Move new affine_grid api to functional
test=develop
* Add CUDA kernel for affine_grid.
test=develop
* Add more unitest for grid sample API
test=develop
5 years ago
6f69fbc8ea
fix elu grad whne alpha less then zero, test=develop ( #26543 )
5 years ago
786373ba29
Use atomicAdd defined in paddle framework ( #26628 )
...
test=develop
5 years ago
1f82c0cd62
[Api2.0] add pixel shuffle ( #26071 )
5 years ago
79539cf198
【2.0 API】Add CUDA kernel and enhance options for grid_sample ( #26576 )
...
This PR enhance CPU kernel and add new CUDA kernel to make grid_sample support:
- align_corners: with bool type.
- padding mode: which can be in ['zeros', 'reflect', 'border']
- Interpolation mode: which ca be in ['bilinear', 'nearest']
The old CPU and CUDNN version only support align_corners=true, padding_mode='zeros' and interpolation_mode='bilinear'.
The behavior of the new version op in default mode is compatible with the old version.
5 years ago
8645591d66
support fp64 in huber_loss cuda kernel ( #26583 )
5 years ago
efee426742
support generator seed in related kernals test=develop ( #26495 )
5 years ago
bf4a4636f1
change to use bce_loss op, add shape check for bce_loss
...
change to use bce_loss op, add numel check for bce_loss.
5 years ago
0e81626081
add div, floor_div, remainder ( #26562 )
...
* add div, floor_div, remainder
5 years ago
24566e951c
Support empty bbox in bipartite math op ( #26488 )
5 years ago
199b0c7c1b
Add isfinite v2 op ( #26344 )
...
add the isnan, isfinite, isinf api for the paddle 2.0
5 years ago
ebf9b2125e
add paddle.gather for API2.0 ( #26455 )
5 years ago
9219b79104
gather_nd Op for API 2.0 refine ( #26540 )
5 years ago
9b14117cac
logsumexp: impl kernel, refine docs ( #26307 )
5 years ago
5c2b9258a6
Fix (de/re)quantize cache keys ( #26549 )
5 years ago
6b28456ed0
add the argmax, argmin for the api2.0
...
* add the new api and op for the argmax, argmin
5 years ago
d26ae9ad87
Update conv_transpose api ( #26427 )
...
* update conv_transpose api
5 years ago
faa9b97b78
fix cscatter, test=develop ( #26554 )
5 years ago
45711dade7
【API】rename div to divide, add floor_divide, remainder ( #26434 )
5 years ago
4e0c6d91aa
add paddle.tensor.linalg.diag API, diag_v2 OP and CUDA kernel
...
add paddle.tensor.linalg.diag API, diag_v2 OP and CUDA kernel.
5 years ago
f8863e0603
leaky_relu and LeakyReLU: alpha->negative_slope ( #26216 )
5 years ago
c609066074
Add Matmul op ( #26411 )
...
* add matmul_v2
5 years ago
aa2a9b5d89
add bernoulli op ( #26511 )
...
* add bernoulli op
* fix cuda kernel and add unit test
* refine doc
* fix uniform
5 years ago
f3909020de
Add mechanism for blocking oneDNN cache clearing ( #26502 )
...
* Add mechanism for blocking oneDNN cache clearing
* Review changes and Add thread guards
5 years ago
b6eb37f5b3
add error message for cholesky ( #26444 )
...
* add error message
5 years ago
138ecf24aa
support Baidu Kunlun AI Accelerator ( #25959 )
...
* support Baidu AI Accelerator
* test=kunlun
* minor
* test=kunlun
* support xpu op in separate file
* test=kunlun
* update XPU error message and remove duplicated code
* test=kunlun
* minor
* test=kunlun
* minor
* test=kunlun
5 years ago
4f259354d2
mod cvm test=develop ( #25146 )
...
* mod cvm test=develop
* mod code format test=develop
5 years ago
e167e87974
【API2.0】add masked_select Op for API2.0 ( #26374 )
5 years ago
6e5670b8bd
mean: not support int32, int64; add check for axis ( #26401 )
5 years ago
4ad504e7c7
hardshrink: support threshold < 0 ( #26403 )
5 years ago
e92f770c42
Add collective ops (reduce) ( #26340 )
5 years ago
bdb805505e
【API2.0】add numel API for paddle test=develop ( #26311 )
5 years ago
2073ffc04d
Enhance the data type of linspace API ( #26247 )
5 years ago
40d193ed17
Add the ReLU6, Tanhshrink, SELU, Softplus, Softshrink and Softsign for the api 2.0 ( #26376 )
5 years ago
f00f982a02
add cub impl for arg max, min ( #25941 )
...
test=develop
5 years ago
6914a12f82
rename the inputs of allclose ( #26360 )
...
* rename input
* add unittest, test=develop
* use paddle.data instead of fluid.data, test=develop
5 years ago
bcf03273f6
add pad func ( #26106 )
...
* add pad func
* add pad
* test=develop, add pad op and apis
* restore pad2d
* test=develop, fix paddl declare
* fix pad interface
* test=develop, fix pad
* test=develop, add all pad api and cos_sim
* test=develop, remove padding default value
* test=develop, rename var to tensor
* test=develop, add more tests
* test=develop, rename tovar to totensor
* test=develop, fix init
* test=develop, add more test
* test=develop, add more tests
5 years ago
eeeef957c7
Fix ps gpu ( #26218 )
...
* support ps-gpu
5 years ago
6cbeafb6c0
add zero norm, inf norm support for p_norm op ( #26364 )
...
* add zero norm, inf norm support for p_norm op
* fix the invalid argument check, fix the dtype problem in test case.
5 years ago
1fbee267d4
remove scope in cudnn lstm ( #25188 )
5 years ago
3f816bc8b4
[Quantization] Conv2d_transpose and mul support channnelwise quantization ( #25639 )
...
* Conv2d_transpose and mul support channnelwise quantization, test=develop
* Skip collecting out threshold for output tensor of which the type is not fp32 or fp64, test=develop
* Fix error in test_user_defined_quantization, test=develop
* Add depthwise_conv_bn_fuse, test=develop
* Add conv_transpose_bn_fuse_pass for post_training_quant, test=develop
5 years ago
638bbb6153
Improve expand as ( #26290 )
...
align expand_as op to expand.
5 years ago
586a6dd358
log_softmax and LogSoftmax: impl kernel and refind docs ( #26088 )
5 years ago
23261ff44b
add cpu random Generator ( #26013 )
5 years ago
69742bd9a4
Enable mkldnn layout conversion ( #25778 )
...
* enable mkldnn layout conversion
* review fix: remove tmp_place
* fix test mkldnn swish
* add UT for PrepareData CPU->MKLDNN
* add #ifdef PADDLE_WITH_MKLDNN
* Force-push commit
Co-authored-by: grygielski <adam.grygielski@gmail.com>
5 years ago
6d22f5c73e
Add PADDLE_ENFORCE in nll loss cuda kernel ( #26294 )
...
* add nll loss API, update demo code of the comment
5 years ago
241b44db14
[API 2.0] adaptive expand op to use shape instead of expand_times ( #26206 )
...
* adaptive expand op to 2.0 (align to torch.expand) , test=develop
5 years ago
fbd4d3cc97
[API 2.0] add paddle.tile op ( #26245 )
...
* add tile_op, test=develop
5 years ago
a2d3e5c03b
Fix `paddle.abs` docstring ( #25942 )
...
test=document_fix
remove activation wording
5 years ago
22165934bc
Fix `paddle.acos` docstring ( #25958 )
...
test=develop,test=document_fix
remove activation wording
5 years ago
a5b5b00e02
Fix `paddle.asin` docstring ( #25967 )
...
test=develop,test=document_fix
remove activation wording
5 years ago
c758765769
Fix `paddle.atan` docstring ( #25968 )
...
test=develop,test=document_fix
remove activation wording
tanh -> tan
5 years ago
c4e480efc5
Fix `paddle.cos` docstring ( #25969 )
...
test=develop,test=document_fix
explain input/out put range and out of boundary behavior
5 years ago
2d6cc0b125
support the tuple for attribute of axis in min, max for api2.0
...
Update the code for the min,max, test=develop
5 years ago
ffe52b4452
[OpDevOptimize] Add common infershape functions ( #26096 )
...
* add unchaged infershape function
* add broadcast infershape function
* fix bug
* rename infershape functions
* add UnaryOpUnchangedInferShapeCheckAxis
* add error message
* add test for common infer shape functions
* dont update existed ops
* dont update op_desc.h
* add more test
* add error check, refine error message
5 years ago
2d95280e1f
Feature/Enable Auto-Mixed-Precision in dynamic graph ( #24903 )
...
* add auto_cast, test=develop
* add loss scaler, test=develop
* add comments, test=develop
* refine code, test=develop
* refine code, test=develop
* do not set flags automatically, test=develop
* fix custom op bug, test=develop
* add more test, test=develop
* refine enable logic, test=develop
* enable amp test with GPU, test=develop
* add unittest
* add test for found_inf
* follow comments
* follow comments
* remove global variable, use singleton
* add some notes
* update comments
* update comments
* update comments
* add use_dynamic_loss_scaling argument
* refine found_inf
* refine found_inf
5 years ago
9c17b3c9f8
Add the max, min, maximum, minimum api for the API 2.0
...
* Add the max, min, maximum, minimum api for the API 2.0, test=develop
5 years ago
1be6bf45ae
Add assign to fusion_group and enhance inplace execution in fusion_group. ( #26121 )
5 years ago
8caee2ad51
【paddle.fleet】add the support for multi-node training for pipeline ( #25907 )
...
* add the support for multi-node training
5 years ago
bf2db646de
fix cumsum op for API 2.0, optimize performance
...
update cumsum api and fix up the cumsum op
5 years ago
1893cd6bb8
Add oneDNN relu6 op ( #26037 )
...
* Add oneDNN relu6 op
* Lint fixes
5 years ago
50f149a48e
fix cudnn workspace size problem during inference. ( #26021 )
...
test=develop
5 years ago
3c8daa9b89
Add pin memory control for BufferedReader ( #26026 )
...
* add pin memory control
* fix buffered reader init problem
* fix unittest error
* add unittest for coverage
5 years ago
ad4a0466a5
Add cuda pinned place branch in slice op GetExpectedKernelType ( #26027 )
...
* add cuda pinned place branch
* add unittest
* add skip when not gpu
5 years ago
e853ece0a2
update document template for unary elementwise layers ( #25896 )
...
1. update document template for unary elementwise layers(a.k.a. activation layer);
2. remove generate_op_noattr and use generate_activation instead; remove redundant function copies;
3. minor update for docstring to fix rst format errors.
4. fix doc for Rsqrt OP
5. add sample code for each activation separately;
6. remove the unused deprecated decorator.
5 years ago
734cf1c3e9
Change use_quantizer attribute name and data type ( #25838 )
...
* Change use_quantizer attribute name and data type
* Fix problem with setting attribute
* Add changes due to review
* Small change in function
* Restore use_quantizer attr for compatibility
5 years ago
5258d53d65
refine unsqueeze, test=develop ( #25470 )
...
* refine unsqueeze, test=develop
* update unsqueeze, test=develop
* refine unsqueeze, test=develop
* refine unsqueeze, test=develop
* update
* remove None, test=develop
* follow comments
* support bool
* update doc
* follow comments
* merge develop
5 years ago
224620071b
add new flatten op test=develop ( #25393 )
5 years ago
68c6160e63
Add oneDNN fusion_gru kernel ( #25594 )
...
* Add oneDNN fusion_gru kernel and fix fc+gru pass
test=develop
* Formatting changes
test=develop
* Lint fixes
test=develop
* Add memory::format_tag::any to GRU weights
test=develop
* Fix build with CUDA
* Fix build with CUDA v2
5 years ago
dca56f47f5
fix invalid read of pnorm gradient function
...
fix invalid read of pnorm gradient function and delete the unused code
5 years ago
358bc06c72
[CUDNN8 support] : support CUDNN8 ( #25664 )
...
* cunn8 support
test=develop
* fix ci error
test=develop
5 years ago
5970871a64
add eltwise clip cuda impl. ( #25689 )
...
test=develop
5 years ago
b717895f64
Fix registering trt plugin ( #25744 )
...
* develop dynamic shape serilization
* add test param for gelu
* fix bugs
* delete redundant comments
* debug
* fix conflict. test=develop
* fix bug. test=develop
* add trt dynamic shape serialized support
* fix ernie serialized bug
test=develop
* fix codestyle
test=develop
* fix bug
test=develop
* fix bug.test=develop
* modify cmakelist test=develop
* fix bug
test=develop
* fix error message. test=develop
* fix trt register plugin based on pr#25003
* add trt dynload
* fix deserialization bug of not finding plugin registration
* refine code style
* recover engine key in tensorrt_subgraph_pass
* for ci coverage
* add unittest for deserialization
Co-authored-by: haozech <chenhaoze94@gmail.com>
5 years ago
a697e94693
Update the code of the compare ops for the broadcast function
...
Update the code for the compare ops for the broadcast function
5 years ago
ff717d5158
Add support for tuple of concat Op test=develop ( #25800 )
5 years ago
253fd407e8
Fix/distibuted heart beat ( #25902 )
...
* disable heart beat UT
5 years ago
d11c140e28
fix dump, fix cvm check ( #25400 )
...
* fix dump, fix cvm check
test=develop
* fix
test=develop
* fix
test=develop
* fix
test=develop
5 years ago
6486fe8a94
improve GPU performance of transpose, test=develop ( #25862 )
5 years ago
2d24f56a7a
avoid data transfer, test=develop ( #25810 )
5 years ago
bca303165a
fix inverse bug ( #25641 )
...
* fix inverse bug, test=develop
* fix the untest, test=develop
* add singular checking, test=develop
* fix the utest, test=develop
* use memory::copy, test=develop
* fix bost_get, test=develop
* fix position, test=develop
5 years ago
e52dae6ef6
Using input.place() in GetExpectedKernel in slice_op ( #25595 )
...
* modify GetExpectedKernelType
* use input place
* add ENFORCE check
5 years ago
595a719795
Update the api for the compare_ops
...
Update the code for the compare_ops, update the api and doc
5 years ago
32b9577b2a
refine the split op for API 2.0 test=develop ( #25320 )
5 years ago
ce506930c3
Fix the bug that Input(Offsets) and attr(offsets) cannot be set at the same time. ( #24975 )
...
* bug fix, test=develop
5 years ago
2d9dbd31ad
Fix/mkl dnn ( #25835 )
5 years ago
caa90a6510
Integrated Trainer of Parameter Server (API add `fluid.contrib.layers.sparse_embedding` only) ( #22957 )
...
* Integrated Trainer of Parameter Server
5 years ago
42189be67b
[Quant] Remove the output for moving_average_abs_max_scale op ( #25697 )
...
* Remove the output for moving_average_abs_max_scale op, test=develop
5 years ago
23d1228c4d
remove ProgramTranslator.save_inference_model ( #25740 )
...
* remove ProgramTranslator.save_inference_model
* adapt save_quantized_model
* revert buffer check implemention
* remove useless import function
5 years ago
1b3081b1b4
Simplify BufferedReader to improve DataLoader performance ( #25648 )
...
* simplify buffered reader to improve DataLoader performance
* fix 22 failed unittests
* fix cuda pinned context condition
* fix test_reader_reset failed
* fix two failed unittests
* change unittest place
* polish error messaage
* polish cast op GetExpecctedKernelType
* remove debug info in unittest
5 years ago
e0a9115e28
fix random compile failure due to missing file ( #25661 )
5 years ago
1aaa26f102
add dnnl sigmoid (logistic) activation ( #25745 )
5 years ago
1e4ab728fb
refine the concat Op for API 2.0 test=develop ( #25307 )
5 years ago
98899b73d2
Fix FC + GRU fuse pass ( #25687 )
5 years ago
4ec1251a1e
Refine squeeze, test=develop ( #25281 )
...
* refine squeeze, test=develop
* update squeeze, test=develop
* refine compile-time infershape, test=develop
* add more unittest, test=develop
* follow comments, test=develop
* add update_api, test=develop
* follow comments, test=develop
5 years ago
e5bbffa84c
Add NOMINMAX define due to windows.h max/min macro conflict ( #25637 )
...
test=develop
5 years ago
70cee22fde
New features, add sinh and cosh op, test=develop ( #25495 )
...
* New features, add sinh and cosh op, test=develop
* remove duplicate test function and remove out paramters, test=develop
* Add out paramters temporary, remove later. test=develop
* remove out args, PR 25570, test=develop
* remove TestParameter, test=developx
* add test api for static dygraph, test=develop
* add backword unittests for sinh and cosh, test=develop
5 years ago
a1350744eb
register fp16 kernel, test=develop ( #25630 )
5 years ago
5453a912fe
add fp64 support in sequence_pool, test=develop ( #25662 )
...
add fp64 support in sequence_pool, test=develop
5 years ago
6e86fd3750
fix concat dimension ( #25606 )
...
Fix the condition of concat dimension judgment.
5 years ago
95fa383df2
optimize embedding cuda kernel lookup_table_v2,test=develop ( #25587 )
5 years ago
7206417259
supports xpu runtime, test=develop ( #25554 )
...
* update ResetHolder, test=develop
* add TensorShare for lite engine, test=develop
* tensor data changed from copying to sharing, test=develop
* supports xpu runtime, test=develop
* fix code styles, test=develop
5 years ago
30d1ff3bb4
call cublasGemmStridedBatchedEx when using fp16, test=develop ( #25553 )
5 years ago
ca1185d06b
[Dy2Stat] Fix scope in run_program_op ( #25579 )
...
* add reinforcement learning model test=develop
* align backward test=develop
* add gym in paddle_build.sh test=develop
* rm pip install in script test=develop
* refine paddle_build.sh test=develop
* fix sed error in macOS test=develop
* polish code test=develop
* fix scope problem
* refine code by reviewer comment
5 years ago
7dbc441eab
[oneDNN] cache cosmetics improvement ( #25576 )
5 years ago
e362095e45
fix softmax with cross entropy out of bound; test=develop ( #25549 )
5 years ago
d8fe517bf8
Add Support for SelectedRows for Transpose OP and Fix a Bug That SelectedRows Cannot be Supported in SimNet ( #25536 )
...
This PR fixes a bug that SelectedRows cannot be supported in SimNet. The reason of this bug is that dygraph basic_engine didn't copy var's type when the var needs to be accumulated during backward. So when a var is SelectedRows and needs to be accumulated, like SimNet which calls net for two times, the var's type will be changed to default LoDTensor thus bug happens. To fix it, we just also copy the type.
Without this PR, the accumulated SelectedRows parameters in dygraph will be changed into LoDTensor. So when we fixed the bug of supporting SelectedRows in SimNet, we found `test_imperative_lod_tensor_to_selected_rows` failed and threw the error that SelectedRows was not supported for Transpose OP. To fix it, too, this PR also added support for SelectedRows for Transpose OP.
5 years ago
848aca7ae8
[CI] [Lite-Subgraph] CI add lite subgraph check. ( #25346 )
5 years ago
71c71e684c
fix logical_* ops' doc ( #25479 )
...
* fix doc of logical_* op.
* fix doc of op pow.
* fix comment syntax error9D
* fix operator reciprocal demo.
* fix logical_* ops' doc. test=develop,test=document_fix
* bug fix. test=develop,test=document_fix
* bug fix. test=develop,test=document_fix
* bug fix. test=develop,test=document_fix
* bug fix. test=develop,test=document_fix
5 years ago
4717bdbcfb
Fix hang in seq_topk_avg_pooling op ( #25522 )
...
* fix topk_avg_pool hang test=develop
* refactor get_topk_pos test=develop
* add check of channel_num and num_k test=develop
* add TopKPosPaddingId test=develop
5 years ago
7129f544f0
Add bilateral_slice op ( #25401 )
...
* add bilateral slice op
5 years ago
ca725c82f2
improve fp16 performance of slice_grad, test=develop ( #25523 )
5 years ago
5d3766ff3d
modify flip test=develop ( #25312 )
...
According to paddle 2.0 standard
1, change flip api attr name 'dim' to 'axis'.
2, support empty axis
3, change example code to imperative mode.
5 years ago
41d2247275
[Dy2static] Refactor ProgramTranslator save_inference_model API ( #24989 )
...
* experimental refactoring, test=develop
* add TranslatedLayer & remove StaticModelRunner, test=develop
* revert tracedlayer change, test=develop
* fix test_mnist unittest error, test=develop
* add doc & examples, test=develop
* polish doc details, test=develop
* add imperative.jit module, test=develop
* change TranslatedLayer pos, test=develop
* adjust jit module import path, test=develop
* polish doc based review result
* add SaveLoadConfig.separate_params to save paraams separately
* add Layer.buffer support, test=develop
* polish doc details based review result, test=develop
* polish details baesd review comments, test=develop
* add empty str check for param, test=develop
* add unittests, test=develop
* polish details based review comment, test=develop
* remove blanks in comment, test=develop
* polish doc details, test=develop
* update imperative doc link, test=develop
* add api attr for load, test=develop
5 years ago
aaa7cbd56f
modify trace api test=develop ( #25397 )
5 years ago
f9ac5fb992
[Dy2stat] Fix Memory Optimization in run_program_op and Add SimNet as Unit Test ( #25383 )
...
Add Similarity Net as unit test. During the unit test, we found three problems:
1. The run_program_op has memory optimization error when running dy2stat net multiple times.
2. The support for SelectedRows can cause problem in dy2stat.
3. The return grammar has problem.
This PR fixes the 1. problem but modify codes for the 2. 3. problems to make PR smaller. I will fix those two problems in the next PR(s)
5 years ago
c42d662e2a
modify roll test=develop ( #25321 )
5 years ago
548cdbc544
Quantization-aware training for dygraph ( #24634 )
...
* Add the imperative quantization aware training.
* This is the python part of Imperative QAT. test=develop
5 years ago
0b54d54fd8
Fix index overflow bug of the CUDA kernel loop increment ( #25435 )
...
* fix softmax_with_cross_entropy cuda kernel overflow bug, test=develop
* replace old macro & for condition, test=develop
* polish details, test=develop
5 years ago
e528392de9
[Paddle-TRT] SkipLayernorm vectorized memory optimization ( #25117 )
...
* add explicit specialization
* add skiplayernorm vector load if available
* test=develop
5 years ago
5b573c58e2
randperm API: remove out, devive, stop_gradient; add name ( #25410 )
5 years ago
bb45af02ac
add the c++ part of Imperative QAT. test=develop ( #25446 )
5 years ago
050a9bf79d
[oneDNN] LRN cleanup ( #25416 )
5 years ago
1974aadcf0
fix concat shape error ( #25414 )
...
* fix concat shape error
test=develop
5 years ago
4b3778a3ee
Revert/barrier for sync ( #25417 )
...
* add retry for prefetch
* Revert "Fix/sync barrier (#25016 )"
This reverts commit be6a315fbd
5 years ago
52be62c5ae
fix instance norm in dy ( #24717 )
...
* fix bn & in in dy, test=develop
* update instance_norm,test=develop
* fix bugs,test=develop
* add more case in unittest,test=develop
* fix,test=develop
* fix,test=develop
5 years ago
eb3173e2b6
rand API: remove out, device, stop_gradient; add name ( #25246 )
5 years ago
6de75082cb
fix test_hsigmoid windows ci ( #25311 )
5 years ago
f593c3fb2f
fix the formula of floor OP and ceil OP ( #25292 )
5 years ago
bc7610583b
use eval() to improve CPU performance ( #25243 )
5 years ago
74468bf428
add mish op. ( #24565 )
...
* add mish op. test=develop
5 years ago
6d6efafeeb
Add `matrix_nms_op` ( #24400 )
...
* Add `matrix_nms_op`
test=develop
* Make ci happy
test=develop
* Exit early when no detection
test=develop
* Fix license year
test=develop
* Output index as well
test=develop
* Match nms2 lod behavior and add `return_index` flag
test=develop
* Make CI happy
test=develop
* Fix wording
test=develop
5 years ago
e85fcaa712
Fix fluid.embedding in Distributed Training ( #25174 )
...
* test=develop, fix_embedding
5 years ago
c00f827843
Avoid data transforming ShapeTensor from CPU to GPU in fill_constant op. ( #25267 )
5 years ago
f1a9593d69
test=develop, bug fix for index_select and roll op ( #25251 )
5 years ago
c2e072587c
test=develop fix generate_proposals's error ( #25227 )
5 years ago
4c964abdf7
support build on arm. test=develop ( #25212 )
5 years ago
1458cc0c68
Fix bug: Don't check dims if contain_unknown_dim of cross_entropy_grad_op in compile time ( #25221 )
5 years ago
68e93d8a17
Fix beam_search InferShape ( #25169 )
...
* fix beam_search infershape, test=develop
* fix beam search op unittest, test=develop
5 years ago
bd0b38e671
Refactor of conv fp32 oneDNN operator ( #25137 )
...
* Refactor of conv fp32 oneDNN operator
test=develop
* Formatting fix
test=develop
* Return Enforces
test=develop
* GetWeights improvements
test=develop
5 years ago
19c4db1b56
don't re-generate header file if content doesn't change ( #25130 )
...
* don't re-generate header file if content doesn't change. test=develop
* add copy_if_different function. test=develop
5 years ago
a7944904d3
[oneDNN]elementwise_add and elementwise_mul int8 support ( #24984 )
...
* Start implementing int8 eltwise add
test=develop
* - Fix to Michal PR
* - Fix
test=develop
* - Lint fixes
test=develop
* - Added checking if elementwise_mul can be used
test=develop
* - Added attribs to skip_attrs_set
test=develop
* - Improved broadcasting
test=develop
- fixes to compilation
- fix
- fix
- Lint fixes
test=develop
* - removed redundant condition
test=develop
Co-authored-by: Michal Gallus <michal.gallus@intel.com>
5 years ago
fa657b3dbb
fix bug of prelu when rank not equal 4, test=develop ( #25067 )
...
* fix bug of prelu when rank not equal 4, test=develop
* fix prelu inference, test=develop
* fix api, test=develop
* fix shape when mode is chennel, test=develop
* remove debug code, test=develop
* add unittest, test=develop
5 years ago
479c8834f7
[Paddle-TRT] Fixes #24731 , opt for SoftmaxKernelWithEltadd kernel, test=develop ( #24834 )
...
* blockReduce opt
* launch threads align to warpSize
* reduce unnecessary shared memory for broadcast reduced value
* vectorize SoftmaxKernelWithEltadd
* add fp16 constrain
* test=develop
5 years ago
028de857d4
fix dtype error of compare op, test=develop ( #25059 )
5 years ago
770c11a117
fix make device_context error ( #25045 )
...
* test=develop
* test=develop
* fix bug
* test=develop
* test=develop
5 years ago
be6a315fbd
Fix/sync barrier ( #25016 )
...
* fix sync barrier with barrier monitor, test=develop
5 years ago
8db66fc3f6
fix cos_sim, test=develop ( #25017 )
5 years ago
621b638550
improve performance of instance_norm, test=develop ( #25005 )
5 years ago
613303dbf6
refine the slice Op to improve the performance of xlnet for fp16 training ( #24967 )
5 years ago
d152d7231e
clear old var in scope, test=develop ( #24976 )
5 years ago
0eb1b0bc01
Add support the 5d, 6d tensor support for the reduce ops
...
Add the support the 5d,6d tensor support for the reduce ops;
Add the same time, the compile time, it was 22 minutes, it was 21 minutes after fixed.
5 years ago
24e24987f0
fixes the place info in the Print op ( #24934 )
...
fixes the CUDAPlace info in the Print op
5 years ago
6be0ee159e
Support LoDTensorArray in reverse_op ( #24797 )
...
* Support LoDTensorArray in reverse_op test=develop
* polish en doc and unittest code test=develop
* refine sample code test=develop
* add example of LoDTensorArray test=develop
* fix typo test=develop
5 years ago
a7cb97a1a5
Fix/isfinite on windows ( #24927 )
...
* refine isfinite, test=develop
* use namespace std of isfinite, test=develop, test=win_gpu
5 years ago
4c01d6d53e
Enhance checking in some operator. ( #24473 )
5 years ago
6e10022781
add queue_generator_op, dequeue_op, enqueue_op and ut ( #24481 )
...
* add queue_generator_op, dequeue_op, enqueue_op and ut, test=develop
5 years ago
1e818158f5
Feature/add amp_checkout_finite_and_scale op ( #24875 )
...
* add amp_check_finite_and_scale op, test=develop
* add cpu kernel, test=develop
* use bool, test=develop
* follow comments, test=develop
5 years ago
a6beb96dd0
FTRL with sparse update, test=develop ( #22092 )
5 years ago
d1062d5278
Replace all errors thrown by LOG(FATAL) with PADDLE_THROW ( #24759 )
...
* remove REPLACE_ENFORCE_GLOG compile option & add ci rule prohibit LOG(FATAL) using, test=develop
* remove ci test case, test=develop
* replace all LOG(FATAL) & polish message, test=develop
* fix typo, test=develop
* polish error info detail, test=develop
5 years ago
23a85f030c
Remove old mkldnn_elementwise_mul test ( #24855 )
...
test=develop
5 years ago
b67ded04f2
Support gradient accumulation of fp16 in imperative mode ( #24823 )
...
* support gradient accumulation of fp16 in imperative mode, test=develop
* enhance coverage test, test=develop
* follow comments, test=develop
5 years ago
704cad6a66
Add histc op ( #24562 )
...
* add histc operator, test=develop
* update english doc to 2.0 API, test=develop
* update API from histc to histogram, test=develop
Co-authored-by: root <root@yq01-gpu-255-129-15-00.epc.baidu.com>
5 years ago
12bffdc086
Enhance error message of checkpoint_notify_op, fake_init_op gen_nccl_id_op and listen_and_serv_op ( #24554 )
...
test=develop
5 years ago
b490e41c1d
Add isCached() mechanism for BatchNorm and LRN oneDNN operators ( #24798 )
...
* Add isCached() mechanism for BatchNorm and LRN oneDNN operators
test=develop
* Formatting fix
test=develop
5 years ago
a7e21cbed3
Move input_size check into RunTime phrase of gru_unit_op and refine error message ( #24776 )
...
* Add IsRuntime judgement in GRUUnit test=develop
* add IsRuntime judgement is GradOp test=develop
* Refine Error Message of SelecteInput/Output test=develop
* refine Error Message of RNNMemoryHelperOp test=develop
5 years ago
f66594a558
fix bug that diag API can't use on Windows( #24762 )
5 years ago
c0911fdd32
rename inplace/no_need_buffer inferer, part4, test=develop ( #24781 )
5 years ago
be82de4c79
polish two error message, test=develop ( #24778 )
5 years ago
b0e7439fbc
rename inplace/no_need_buffer inferer, part2, test=develop ( #24733 )
5 years ago
a6fbba65ff
rename inplace/no_need_buffer inferer, part3, test=develop ( #24734 )
5 years ago
355caee18b
fix conv_transpose Op fp16 error test=develop ( #24695 )
5 years ago
56a714a19b
Add isCached() machinism to oneDNN pooling primitive ( #24724 )
5 years ago
lilong12
Zhong Hui
Pei Yang
Zhen Wang
Adam
Jack Zhou
Zhang Ting
lidanqing
Chen Weihang
Wilber
furnace
Qi Li
Jacek Czaja
ShenLiang
wawltor
zhupengyang
Steffy-zxf
JZ-LIANG
wangchaochaohu
Qinghe JING
kinghuin
yongqiangma
LielinJiang
whs
wangguanzhong
xiaoting
GaoWei8
yaoxuefeng
Chengmo
joanna.wozna.intel
Yang Zhang
huangjun12
Thunderbrook
Leo Chen
tangwei12
hong19860320
Jiawei Wang
Wojciech Uss
LutaoChu
joejiong
Bai Yifan
ruri
Guanghua Yu
qingqing01
WangXi
ShenLiang
QingshuChen
Zhaolong Xing
littletomatodonkey
cc
Sylwester Fraczek
Yiqun Liu
Feiyu Chan
xujiaqi01
Zhang Ting
Aurelius84
Zhou Wei
cnn
mapingshuo
donproc
石晓伟
hong
Huihuang Zheng
Shibo Tao
zlsh80826
ceci3
WuHaobo
Kaipeng Deng
123malin
FDInSky
liym27
liu zhengxi
tianshuo78520a
Aurelius84
leesusu
Michał Gallus
Qi Li
Yi Liu