Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add flash attn for af2 #8

Open
wants to merge 428 commits into
base: develop
Choose a base branch
from
Open

Add flash attn for af2 #8

wants to merge 428 commits into from
This pull request is big! We’re only showing the most recent 250 commits.

Commits on May 5, 2023

  1. Mv cpp_extension test dir (PaddlePaddle#53330)

    @tianshuo78520a
    tianshuo78520a committed May 5, 2023
    Configuration menu
    Copy the full SHA
    e85fbac View commit details
    Browse the repository at this point in the history

  2. 【Hackathon No.61】uniform_random 算子FP16/BF16单测完善 (PaddlePaddle#52949)

    @co63oc
    co63oc committed May 5, 2023
    Configuration menu
    Copy the full SHA
    b02de1b View commit details
    Browse the repository at this point in the history

  3. [XPU] Fix the out_max of the branch in xpu_conv2d op(PaddlePaddle#53343)

    @sprouteer
    sprouteer committed May 5, 2023
    Configuration menu
    Copy the full SHA
    d27f15e View commit details
    Browse the repository at this point in the history

  4. [XPU] Fusion of gather and assign operators to fused_mt op for reduci… 

    …ng memory usage (PaddlePaddle#53262)
    @shentanyue
    shentanyue committed May 5, 2023
    Configuration menu
    Copy the full SHA
    2039115 View commit details
    Browse the repository at this point in the history

  5. remove some [-Wunused-parameter]warning (PaddlePaddle#53397) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 5, 2023
    Configuration menu
    Copy the full SHA
    58435ae View commit details
    Browse the repository at this point in the history

  6. [Dygraph] Fix bugs in dp_pp_comm_overlap for HybridParallel (PaddlePa… 

    …ddle#53384)
    @haohongxiang
    haohongxiang committed May 5, 2023
    Configuration menu
    Copy the full SHA
    0d9a23b View commit details
    Browse the repository at this point in the history

  7. Revert "【Hackathon No.52】为 Paddle dist 算子实现 float16 数据类型支持 (PaddlePad… 

    …dle#50915)" (PaddlePaddle#53527)

This reverts commit 9c40653.
    @jinyouzhi
    jinyouzhi committed May 5, 2023
    Configuration menu
    Copy the full SHA
    d463f8e View commit details
    Browse the repository at this point in the history

Commits on May 6, 2023

  1. move UniformRawKernel to legacy (PaddlePaddle#53158) 

    * move UniformRawKernel to legacy

* Update uniform_kernel.cc

* Update uniform_kernel.cu

* Update uniform_kernel.cc

* Update uniform_kernel.cu

* Update uniform_kernel.h

* Update uniform_kernel.cc

* Empty Commit to setup deployments
    @zhangyuqin1998
    zhangyuqin1998 committed May 6, 2023
    Configuration menu
    Copy the full SHA
    13e2e10 View commit details
    Browse the repository at this point in the history

  2. rem npu in test (PaddlePaddle#53469) 

    * rem npu in test

* restore some code
    @KimBioInfoStudio
    KimBioInfoStudio committed May 6, 2023
    Configuration menu
    Copy the full SHA
    a499731 View commit details
    Browse the repository at this point in the history

  3. Add trt pow converter. (PaddlePaddle#53462) 

    * Add trt pow converter.

* update to use AddConstantLayer

* add dims=0 ut
    @jiweibo
    jiweibo committed May 6, 2023
    Configuration menu
    Copy the full SHA
    5a44bf7 View commit details
    Browse the repository at this point in the history

  4. Rename randint_raw and move it to legacy (PaddlePaddle#53157) 

    * Rename randint_raw and move it to legacy

* Update fetch_v2_op.cc

* Update randint_kernel.cc

* Update randint_kernel.cu

* Empty Commit to setup deployments
    @zhangyuqin1998
    zhangyuqin1998 committed May 6, 2023
    Configuration menu
    Copy the full SHA
    3e7be9c View commit details
    Browse the repository at this point in the history

  5. [inference][trt] add reduce_all and reduce_any (PaddlePaddle#53088)

    @zhangjun
    zhangjun committed May 6, 2023
    Configuration menu
    Copy the full SHA
    12406ca View commit details
    Browse the repository at this point in the history

  6. fix brpc double link (PaddlePaddle#53512) 

    * polish

* polish

* polish

* polish

* polish

* polish

* polish

* polish

* polish

* polish

* polish
    @liuzhenhai93
    liuzhenhai93 committed May 6, 2023
    Configuration menu
    Copy the full SHA
    03fe3ce View commit details
    Browse the repository at this point in the history

  7. use int64 to calc dim for c softmax (PaddlePaddle#53541) 

    * use int64 to calc dim for c softmax

* fix complie bug
    @FeixLiu
    FeixLiu committed May 6, 2023
    Configuration menu
    Copy the full SHA
    da963ea View commit details
    Browse the repository at this point in the history

  8. Rewirte the reshape of temp_mask and temp_bias.

    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    08a8b75 View commit details
    Browse the repository at this point in the history

  9. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    6a65ee0 View commit details
    Browse the repository at this point in the history

  10. [XPU] substitute new api kernel for combinatorial adaptive avg_pool2d… 

    …_grad kernel (PaddlePaddle#53528)
    @RuohengMa
    RuohengMa committed May 6, 2023
    Configuration menu
    Copy the full SHA
    eda8df7 View commit details
    Browse the repository at this point in the history

  11. Update commit and fix reduce_dim.

    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    4682c0d View commit details
    Browse the repository at this point in the history

  12. XPU Support external stream (PaddlePaddle#53334)

    @csy0225
    csy0225 committed May 6, 2023
    Configuration menu
    Copy the full SHA
    99399f3 View commit details
    Browse the repository at this point in the history

  13. Add fused_gate_attention API. (PaddlePaddle#53432) 

    * Add fused_gate_attention API.

* Implement FusedDropout API.

* Fix doc and add unittest.

* Skip for non-gpu device.

* Add unittest.
    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    b729512 View commit details
    Browse the repository at this point in the history

  14. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    165afab View commit details
    Browse the repository at this point in the history

  15. API support use_flash_attn.

    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    dd2860e View commit details
    Browse the repository at this point in the history

  16. [IR] OpTrait & OpInterface & OpInfo (PaddlePaddle#52846) 

    * add OpTrait OpInterface ValueIterator TypeList

* refine code

* refine code

* refine code

* add opinfo

* add typeid copy constructor

* add trait interface construct method for opinfo_impl

* add trait interface construct method for opinfo_impl

* add trait interface construct method for opinfo_impl

* add trait interface construct method for opinfo_impl

* add trait interface construct method for opinfo_impl

* add create

* add member func for opinfo

* fix compile bug

* add op interface in ircontext

* fix compile bug

* fix compile bug

* refine code

* fix compile bug

* add ut

* refine ut

* refine code of opinfo_impl

* delete unused code

* add dyncast for operation

* refine comment

* refine opinfo_impl

* delete unused code

* refine code by comment

* refine code

* refine code

* refine code for registerOp

* refine opfin create

* refine code of search method of ircontext

* refine op attribute

* change opinfo_map key from type_id to string
    @zhangbo9674
    zhangbo9674 committed May 6, 2023
    Configuration menu
    Copy the full SHA
    d91d758 View commit details
    Browse the repository at this point in the history

  17. Use copy_if_different to avoid recompilation of generated cutlass (Pa… 

    …ddlePaddle#53531)

kernels.
    @umiswing
    umiswing committed May 6, 2023
    Configuration menu
    Copy the full SHA
    f5476da View commit details
    Browse the repository at this point in the history

  18. 【prim】Elementwise double grad (PaddlePaddle#53014) 

    * add mul doubel grad

* add sub_double_grad

* add add sub high test

* add mutiply test

* modify other unsqueeze

* delete api.yaml

* only for make ci run

* midify unsqueeze

* modify unsqueeze

* tmp

* modify operants gen

* review modify

* modify review

* debug

* debug

* modify ci cross boundary

* delete log
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 6, 2023
    Configuration menu
    Copy the full SHA
    a5a0e8f View commit details
    Browse the repository at this point in the history

  19. fix strided_slice ut (PaddlePaddle#53553) 

    * fix strided_slice ut

* remove check_dygraph
    @USTCKAY
    USTCKAY committed May 6, 2023
    Configuration menu
    Copy the full SHA
    1d8c82b View commit details
    Browse the repository at this point in the history

  20. fix conv1d_transpose insert quant node bug (PaddlePaddle#53320)

    @xiaoluomi
    xiaoluomi committed May 6, 2023
    Configuration menu
    Copy the full SHA
    ca174ea View commit details
    Browse the repository at this point in the history

  21. [inference][trt] add lookup_table op trt converter, use trt gather la… 

    …yer (PaddlePaddle#53554)

* add lookup_table op trt converter

* update
    @yuanlehome
    yuanlehome committed May 6, 2023
    Configuration menu
    Copy the full SHA
    08b44e6 View commit details
    Browse the repository at this point in the history

  22. Add PADDLE_THROW in take_along_axis kernel when the datatype of index… 

    … is wrong. (PaddlePaddle#53556)
    @Xreki
    Xreki committed May 6, 2023
    Configuration menu
    Copy the full SHA
    b65e932 View commit details
    Browse the repository at this point in the history

Commits on May 8, 2023

  1. Fix compiling error on CI.

    @Xreki
    Xreki committed May 8, 2023
    Configuration menu
    Copy the full SHA
    fe80730 View commit details
    Browse the repository at this point in the history

  2. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 8, 2023
    Configuration menu
    Copy the full SHA
    184cf9a View commit details
    Browse the repository at this point in the history

  3. [cutlass] Avoid rewrite generated kernels in recompilation. (PaddlePa… 

    …ddle#53575)
    @umiswing
    umiswing committed May 8, 2023
    Configuration menu
    Copy the full SHA
    a299153 View commit details
    Browse the repository at this point in the history

  4. fix tag commit id with newest

    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    65c6ed1 View commit details
    Browse the repository at this point in the history

  5. Fix typos, test=document_fix (PaddlePaddle#53540)

    @co63oc
    co63oc committed May 8, 2023
    Configuration menu
    Copy the full SHA
    acefdeb View commit details
    Browse the repository at this point in the history

  6. Fix sample code of poisson_nll_loss (PaddlePaddle#53551)

    @LyndonKong
    LyndonKong committed May 8, 2023
    Configuration menu
    Copy the full SHA
    3fd2e76 View commit details
    Browse the repository at this point in the history

  7. Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into… 

    … add_flash_attn_for_af2
    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    462e36e View commit details
    Browse the repository at this point in the history

  8. [AMP] fix static promote (PaddlePaddle#53439)

    @zhangting2020
    zhangting2020 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    2bf6128 View commit details
    Browse the repository at this point in the history

  9. Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into… 

    … add_flash_attn_for_af2
    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    3458f8c View commit details
    Browse the repository at this point in the history

  10. [Paddle-TRT] add generic plugin for lookup_table_v2(embedding) op (Pa… 

    …ddlePaddle#53539)

* add embedding generic plugin， not enabled
    @yuanlehome
    yuanlehome committed May 8, 2023
    Configuration menu
    Copy the full SHA
    fca8595 View commit details
    Browse the repository at this point in the history

  11. Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into… 

    … add_flash_attn_for_af2
    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    be44a91 View commit details
    Browse the repository at this point in the history

  12. [test]mv fluid op mkldnn to test/cpp/fluid/mkldnn (PaddlePaddle#53458)

    @gouzil
    gouzil committed May 8, 2023
    Configuration menu
    Copy the full SHA
    a01b20d View commit details
    Browse the repository at this point in the history

  13. add ut for lookup_table op trt converter (PaddlePaddle#53563)

    @yuanlehome
    yuanlehome committed May 8, 2023
    Configuration menu
    Copy the full SHA
    2f50338 View commit details
    Browse the repository at this point in the history

  14. Merge branch 'add_flash_attn_for_af2' of https://github.com/JamesLim-… 

    …sy/Paddle into add_flash_attn_for_af2
    @Xreki
    Xreki committed May 8, 2023
    Configuration menu
    Copy the full SHA
    a9ba1ba View commit details
    Browse the repository at this point in the history

  15. Merge branch 'add_flash_attn_for_af2' of https://github.com/JamesLim-… 

    …sy/Paddle into add_flash_attn_for_af2
    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    c0f497a View commit details
    Browse the repository at this point in the history

  16. [Dy2St]Following update of register_hook for static mode (PaddlePaddl… 

    …e#53572)
    @yangguohao
    yangguohao committed May 8, 2023
    Configuration menu
    Copy the full SHA
    f3f3d57 View commit details
    Browse the repository at this point in the history

  17. Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into… 

    … add_flash_attn_for_af2
    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    bfe5a8c View commit details
    Browse the repository at this point in the history

  18. Merge branch 'add_flash_attn_for_af2' of https://github.com/JamesLim-… 

    …sy/Paddle into add_flash_attn_for_af2
    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    0b7fda0 View commit details
    Browse the repository at this point in the history

  19. change flash_attn commit id

    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    ac3ff47 View commit details
    Browse the repository at this point in the history

  20. throw warning at __getitem__, not slice_utils (PaddlePaddle#53579)

    @zoooo0820
    zoooo0820 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    fe91940 View commit details
    Browse the repository at this point in the history

  21. [inference][trt]Unary operation support 0d (PaddlePaddle#53506) 

    * fix trt Unary operation do not support 0d when TRT < 8.6

* update unary ut

* add rsqrt to unary_list

* move rsqrt to act_list
    @zhangjun
    zhangjun committed May 8, 2023
    Configuration menu
    Copy the full SHA
    10f9249 View commit details
    Browse the repository at this point in the history

  22. docs: Fix code example vsplit in docstring (PaddlePaddle#53558)

    @Asthestarsfalll
    Asthestarsfalll committed May 8, 2023
    Configuration menu
    Copy the full SHA
    e988251 View commit details
    Browse the repository at this point in the history

  23. hack 1-D tensor to Scalar (PaddlePaddle#53552)

    @zoooo0820
    zoooo0820 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    186f5e0 View commit details
    Browse the repository at this point in the history

  24. [XPU] Optimize fp16 xpu models (PaddlePaddle#53523)

    @wz1qqx
    wz1qqx committed May 8, 2023
    Configuration menu
    Copy the full SHA
    0a59825 View commit details
    Browse the repository at this point in the history

  25. [BUG] fix paddle.to_tensor/Tensor.item/Tensor.numpy BF16 bug (PaddleP… 

    …addle#53567)
    @zhwesky2010
    zhwesky2010 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    2aedd9d View commit details
    Browse the repository at this point in the history

  26. Fix timeout v2 (PaddlePaddle#53514)

    @sljlp
    sljlp committed May 8, 2023
    Configuration menu
    Copy the full SHA
    26c3077 View commit details
    Browse the repository at this point in the history

  27. Fix core dumped in training when check_nan_inf=1

    @AnnaTrainingG
    AnnaTrainingG committed May 8, 2023
    Configuration menu
    Copy the full SHA
    7dcf5e5 View commit details
    Browse the repository at this point in the history

  28. rm npu (PaddlePaddle#53566)

    @Liyulingyue
    Liyulingyue committed May 8, 2023
    Configuration menu
    Copy the full SHA
    6d396ac View commit details
    Browse the repository at this point in the history

  29. Fix the calculation of y_grad in divide_backward (PaddlePaddle#53582)

    @ZzSean
    ZzSean committed May 8, 2023
    Configuration menu
    Copy the full SHA
    b6c0407 View commit details
    Browse the repository at this point in the history

  30. fix: newExe didnot gc transferred variable (PaddlePaddle#53545)

    @kangguangli
    kangguangli committed May 8, 2023
    Configuration menu
    Copy the full SHA
    f74237c View commit details
    Browse the repository at this point in the history

  31. 【BugFix】fix err of api to_tensor, which caused by numpy version upd… 

    …ate (PaddlePaddle#53534)

* fix

* update code

* pre-commit

* remove scale check (0-D tensor is usable)

* fix data dtype err

* fix numpy default dtype diff

* fix data dtype

* fix data dtype

* update

* fix coverage
    @feifei-111
    feifei-111 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    116fcad View commit details
    Browse the repository at this point in the history

  32. A copy of PaddlePaddle#51438 (PaddlePaddle#53587)

    @jiweibo
    jiweibo committed May 8, 2023
    Configuration menu
    Copy the full SHA
    70180df View commit details
    Browse the repository at this point in the history

  33. fix op unitest for fused_gate_attn

    @JamesLim-sy
    JamesLim-sy committed May 8, 2023
    Configuration menu
    Copy the full SHA
    ce937f6 View commit details
    Browse the repository at this point in the history

  34. add complex support for optest (PaddlePaddle#53356) 

    * add complex support for  optest

* add complex grad test

* append one

* move some debug info

* move some debug info

* move some debug info

* move some debug info

* add more complex test

* Fix naming ambiguity

* Revert "add more complex test"

This reverts commit dbcb051.

* change backward gradient, add TODO
    @GGBond8488
    GGBond8488 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    e522ceb View commit details
    Browse the repository at this point in the history

  35. [Zero-Dim] Support p_norm/reduce_sum_p output 0D (PaddlePaddle#53421)

    @zhwesky2010
    zhwesky2010 committed May 8, 2023
    Configuration menu
    Copy the full SHA
    e4bf1a8 View commit details
    Browse the repository at this point in the history

Commits on May 9, 2023

  1. remove some [-Wunused-parameter]warning and WITH_DISTRIBUTE flag (Pad… 

    …dlePaddle#53532)

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    727fa27 View commit details
    Browse the repository at this point in the history

  2. Fix xpu2 kp compile error (PaddlePaddle#53548)

    @zhangbopd
    zhangbopd committed May 9, 2023
    Configuration menu
    Copy the full SHA
    8d340ee View commit details
    Browse the repository at this point in the history

  3. fix error sample code in static.nn.loss.nce (PaddlePaddle#53588)

    @RedContritio
    RedContritio committed May 9, 2023
    Configuration menu
    Copy the full SHA
    af2ad8d View commit details
    Browse the repository at this point in the history

  4. [Paddle-TRT] Del 2 useless pass (PaddlePaddle#53414) 

    * delete delete_fill_constant_op_pass and unsqueeze2_eltwise_fuse_pass
    @zhoutianzi666
    zhoutianzi666 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    aec4e38 View commit details
    Browse the repository at this point in the history

  5. fix eval branch of prim vjp of batch_norm in amp mode (PaddlePaddle#5… 

    …3598)
    @cyber-pioneer
    cyber-pioneer committed May 9, 2023
    Configuration menu
    Copy the full SHA
    eb12e62 View commit details
    Browse the repository at this point in the history

  6. [Zero-Dim] add 0D test for linalg.norm/linalg.cond (PaddlePaddle#53592) 

    * add 0D test for linalg and linalg.cond

* remove p_norm test

* Update test_zero_dim_tensor.py

* Update test_zero_dim_tensor, test=allcase

* add 0D op test for cond and pnorm,test=allcase

* fix conda error
    @GGBond8488
    GGBond8488 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    6029e02 View commit details
    Browse the repository at this point in the history

  7. [cutlass] Add generated files to .gitignore (PaddlePaddle#53607)

    @umiswing
    umiswing committed May 9, 2023
    Configuration menu
    Copy the full SHA
    9682b04 View commit details
    Browse the repository at this point in the history

  8. add logaddexp api (PaddlePaddle#52284) 

    * add logaddexp

* update
    @zhiboniu
    zhiboniu committed May 9, 2023
    Configuration menu
    Copy the full SHA
    72cb09e View commit details
    Browse the repository at this point in the history

  9. Support trt cuda graph. (PaddlePaddle#53406)

    @jiweibo
    jiweibo committed May 9, 2023
    Configuration menu
    Copy the full SHA
    ea0abf9 View commit details
    Browse the repository at this point in the history

  10. Support static graph code-gen for unpool3d (PaddlePaddle#53479)

    @sanbuphy
    sanbuphy committed May 9, 2023
    Configuration menu
    Copy the full SHA
    dd90f10 View commit details
    Browse the repository at this point in the history

  11. fix ut timeout, test=document_fix (PaddlePaddle#53629)

    @XieYunshen
    XieYunshen committed May 9, 2023
    Configuration menu
    Copy the full SHA
    0f1b077 View commit details
    Browse the repository at this point in the history

  12. 【Hackathon 4th No.30】为 Paddle 新增 paddle.sparse.sum 稀疏 API (PaddlePadd… 

    …le#51406)
    @zrr1999
    zrr1999 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    14c642c View commit details
    Browse the repository at this point in the history

  13. [PHI kernels] Bind XPU kernels (PaddlePaddle#53336) 

    * bind sparse_coo_tensor, reduce_max/max_int32, range/arange_int32, equal_bool, scatter_grad_float32, nearest_interp_int64 kernels

* add more unit tests; modify compilation logic of xpu sparse kernels
    @RuohengMa
    RuohengMa committed May 9, 2023
    Configuration menu
    Copy the full SHA
    7e9c87c View commit details
    Browse the repository at this point in the history

  14. rem tools/infer_prune_patches (PaddlePaddle#53596)

    @KimBioInfoStudio
    KimBioInfoStudio committed May 9, 2023
    Configuration menu
    Copy the full SHA
    a37ef76 View commit details
    Browse the repository at this point in the history

  15. [Zero-Dim] add 0D Tensor UT case for XPU and expand kernel support 0D (… 

    …PaddlePaddle#53555)

* [Zero-Dim] add 0D Tensor UT case for XPU

* fix comment

* remove some unnecessary UT
    @zhwesky2010
    zhwesky2010 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    e588f2d View commit details
    Browse the repository at this point in the history

  16. remove some [-Wunused-parameter]warning (PaddlePaddle#53617) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    bafc346 View commit details
    Browse the repository at this point in the history

  17. [Hackthon 4th No.7] add paddle.unflatten API (PaddlePaddle#53055)

    @Ainavo
    Ainavo committed May 9, 2023
    Configuration menu
    Copy the full SHA
    9cd0a5b View commit details
    Browse the repository at this point in the history

  18. update openblas version (PaddlePaddle#52983)

    @jiweibo
    jiweibo committed May 9, 2023
    Configuration menu
    Copy the full SHA
    9244ceb View commit details
    Browse the repository at this point in the history

  19. [static op generation] coalesce_tensor (PaddlePaddle#53570) 

    * [phi][api] add autogen code coalesce_tensor

* [phi][api]fix args

* [phi][api] supplement attrs
    @gouzil
    gouzil committed May 9, 2023
    Configuration menu
    Copy the full SHA
    eaed168 View commit details
    Browse the repository at this point in the history

  20. [CINN]Adjust Bert unittest loss ground truth (PaddlePaddle#53628) 

    [CINN]Adjust Bert unittest loss ground truth, see: PaddlePaddle/CINN#1357
    @Aurelius84
    Aurelius84 committed May 9, 2023
    Configuration menu
    Copy the full SHA
    45ce0ad View commit details
    Browse the repository at this point in the history

  21. Add compare accuracy api (PaddlePaddle#53430)

    zhangkaihuo committed May 9, 2023
    Configuration menu
    Copy the full SHA
    4907485 View commit details
    Browse the repository at this point in the history

Commits on May 10, 2023

  1. Support static graph code-gen for yolo_loss (PaddlePaddle#52946)

    @sanbuphy
    sanbuphy committed May 10, 2023
    Configuration menu
    Copy the full SHA
    3be7a6c View commit details
    Browse the repository at this point in the history

  2. Fix the index calculation in cross_entroy_kernel. (PaddlePaddle#53659)

    @Xreki
    Xreki committed May 10, 2023
    Configuration menu
    Copy the full SHA
    26fe2dc View commit details
    Browse the repository at this point in the history

  3. fix two minor issues (PaddlePaddle#53658)

    @WintersMontagne10335
    WintersMontagne10335 committed May 10, 2023
    Configuration menu
    Copy the full SHA
    ee1aa69 View commit details
    Browse the repository at this point in the history

  4. Add elementwise_heaviside tests (PaddlePaddle#53549)

    @co63oc
    co63oc committed May 10, 2023
    Configuration menu
    Copy the full SHA
    7a8635d View commit details
    Browse the repository at this point in the history

  5. Add einsum to the default white_list. (PaddlePaddle#53586)

    @Xreki
    Xreki committed May 10, 2023
    Configuration menu
    Copy the full SHA
    2eea311 View commit details
    Browse the repository at this point in the history

  6. Revert "Optimize the implementation of the argsort operator. (PaddleP… 

    …addle#47738)" (PaddlePaddle#53631)

This reverts commit 9e9b705.
    @Vvsmile
    Vvsmile committed May 10, 2023
    Configuration menu
    Copy the full SHA
    aafaad9 View commit details
    Browse the repository at this point in the history

  7. [CustomDevice] fix reducer when input on cpu (PaddlePaddle#53662)

    @ronny1996
    ronny1996 committed May 10, 2023
    Configuration menu
    Copy the full SHA
    e077678 View commit details
    Browse the repository at this point in the history

  8. remove some [-Wunused-parameter] warning and WITH_DISTRIBUT flags (Pa… 

    …ddlePaddle#53650)

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 10, 2023
    Configuration menu
    Copy the full SHA
    65e57a7 View commit details
    Browse the repository at this point in the history

  9. scale, square, sum, swish trt op converter support zero dim (PaddlePa… 

    …ddle#53660)
    @yuanlehome
    yuanlehome committed May 10, 2023
    Configuration menu
    Copy the full SHA
    6a279df View commit details
    Browse the repository at this point in the history

  10. [NPU] PP for npu (PaddlePaddle#53501) 

    * revert p2p communication for xpu

* pp for npu

* update

* update

* fix xpuplace

* add ut for sync send

* Revert "fix xpuplace"

This reverts commit f89c1d7.

* add ut for pp sync send

* rm unusable ut

* update
    @sljlp
    sljlp committed May 10, 2023
    Configuration menu
    Copy the full SHA
    f023d42 View commit details
    Browse the repository at this point in the history

  11. [CINN] fix check_cinn in Optest check condition error bug (PaddlePadd… 

    …le#53676)
    @thisjiang
    thisjiang committed May 10, 2023
    Configuration menu
    Copy the full SHA
    c828934 View commit details
    Browse the repository at this point in the history

  12. Fix bug in log_softmax kernel when lastdim is larger than 100000 (Pad… 

    …dlePaddle#53654)
    @ZzSean
    ZzSean committed May 10, 2023
    Configuration menu
    Copy the full SHA
    0f319f8 View commit details
    Browse the repository at this point in the history

  13. [LAUNCH] add log overwrite flag (PaddlePaddle#53608) 

    * add log overwrite flag

* use strtobool
    @kuizhiqing
    kuizhiqing committed May 10, 2023
    Configuration menu
    Copy the full SHA
    7f39bcd View commit details
    Browse the repository at this point in the history

  14. [static op generation] lstsq (PaddlePaddle#53290)

    @Liyulingyue
    Liyulingyue committed May 10, 2023
    Configuration menu
    Copy the full SHA
    4f33f44 View commit details
    Browse the repository at this point in the history

  15. [XPU]Conv transpose fp16 && fix unittest (PaddlePaddle#53626) 

    * fix as review, add fp16 conv2d_transpose

* fix unittest of bn and reduce_mean

* fix bn unittest

* fix ci

* fix ci
    @wz1qqx
    wz1qqx committed May 10, 2023
    Configuration menu
    Copy the full SHA
    38d664b View commit details
    Browse the repository at this point in the history

  16. no value clip for parallel cross entropy (PaddlePaddle#53547)

    @FeixLiu
    FeixLiu committed May 10, 2023
    Configuration menu
    Copy the full SHA
    65a3a58 View commit details
    Browse the repository at this point in the history

  17. add index_put api (PaddlePaddle#52886) 

    * add index_put api

* fix value broadcast in backward and add test case in static

* add timeout=120s for index_put

* add op_compat for index_put

* add inplace index_put test

* add test case when index tensor in indices is int32 when indices.size less than x.dims

* add index_put api backward in cpu place

* add backward test case

* refactor code to delete some duplicated code

* replace reshape with resize for decrease extra memcpy

* add datatype flag in backward yaml

* fix bug in documentation

* Update python/paddle/tensor/manipulation.py

---------

Co-authored-by: Ligoml <39876205+Ligoml@users.noreply.github.com>
    @Courtesy-Xs @Ligoml
    Courtesy-Xs and Ligoml committed May 10, 2023
    Configuration menu
    Copy the full SHA
    f3393f4 View commit details
    Browse the repository at this point in the history

  18. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 10, 2023
    Configuration menu
    Copy the full SHA
    23c108f View commit details
    Browse the repository at this point in the history

Commits on May 11, 2023

  1. 【prim】add dygraph error code when close prim flag for op who has comp… 

    …osite implement but no grad kernel (PaddlePaddle#53610)

* add no prim no gradOp error code

* delete prim_white_list throw error

* delete invoke_forward_api  throw error

* delete invoke_forward_api  throw error

* review

* review
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    fb8ea98 View commit details
    Browse the repository at this point in the history

  2. Merge branch 'add_flash_attn_for_af2' of https://github.com/JamesLim-… 

    …sy/Paddle into add_flash_attn_for_af2
    @Xreki
    Xreki committed May 11, 2023
    Configuration menu
    Copy the full SHA
    ad3f70a View commit details
    Browse the repository at this point in the history

  3. Fix div error when dtype is int64 in static mode (PaddlePaddle#53705) 

    * Fix div error when dtype is int64 in static mode

* Fix out dtype
    @0x45f
    0x45f committed May 11, 2023
    Configuration menu
    Copy the full SHA
    00ded2e View commit details
    Browse the repository at this point in the history

  4. 昇腾和寒武纪相关代码退场 npu相关代码退场2 (PaddlePaddle#53568)

    @Liyulingyue
    Liyulingyue committed May 11, 2023
    Configuration menu
    Copy the full SHA
    0d45ac7 View commit details
    Browse the repository at this point in the history

  5. Try to crop the flash-attention lib.

    @Xreki
    Xreki committed May 11, 2023
    Configuration menu
    Copy the full SHA
    7ff9f5e View commit details
    Browse the repository at this point in the history

  6. [XPU] update log for bkcl function calls. (PaddlePaddle#53609) 

    * [XPU] update log for bkcl function calls.

* minor update

* revert unnecessary modifications.
    @houj04
    houj04 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    d67d74c View commit details
    Browse the repository at this point in the history

  7. [XPU] update dependency for xccl. (PaddlePaddle#53697) 

    * [XPU] update dependency for xccl.

* remove unnecessary codes.
    @houj04
    houj04 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    44aebd4 View commit details
    Browse the repository at this point in the history

  8. [Doc] remove execution_strategy doc (PaddlePaddle#53668) 

    * remove execution_strategy docstring

* remove doc of num_iteration_per_run; test=document_fix
    @kangguangli
    kangguangli committed May 11, 2023
    Configuration menu
    Copy the full SHA
    49de9de View commit details
    Browse the repository at this point in the history

  9. up index warning level (PaddlePaddle#53691) 

    * up warning level

* numpy still vlog-0
    @zoooo0820
    zoooo0820 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    6ec8d85 View commit details
    Browse the repository at this point in the history

  10. [XPU] add depthwise_conv2d_transpose (PaddlePaddle#53680) 

    * add_depthwise_conv2d_transpose

* Update test_depthwise_conv2d_transpose_op_xpu.py

删除print语句
    @SaltFish11
    SaltFish11 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    08b6f5d View commit details
    Browse the repository at this point in the history

  11. [Paddle-Inference] Support trt 0dims of expand_as_v2 and mish. (Paddl… 

    …ePaddle#53627)

* support_expand_mish
    @xiaoxiaohehe001
    xiaoxiaohehe001 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    aebff6d View commit details
    Browse the repository at this point in the history

  12. [test]mv fluid [controlflow,detection,dlnne,tensorrt] tests to tests (P… 

    …addlePaddle#53470)

* [test]mv fluid controlflow detection dlnne tensorrt tests to tests

* [test]clean dlnne

* [test] fix test_tensorrt_engine_op

* [test] try fix path error

* [test] RollBACK test_tensorrt_engine_op

* [test] RollBACK test_tensorrt_engine_op

* [test]add todo

* Empty-Commit; test=document_fix
    @gouzil
    gouzil committed May 11, 2023
    Configuration menu
    Copy the full SHA
    8075752 View commit details
    Browse the repository at this point in the history

  13. [KUNLUN]Revert "revert p2p communication for xpu (PaddlePaddle#53496)" ( 

    PaddlePaddle#53633)

* Revert "revert p2p communication for xpu (PaddlePaddle#53496)"

This reverts commit eda0c58.

* update
    @sljlp
    sljlp committed May 11, 2023
    Configuration menu
    Copy the full SHA
    4a97ba5 View commit details
    Browse the repository at this point in the history

  14. add unitest for reshpe 0 dims (PaddlePaddle#53685)

    @zhoutianzi666
    zhoutianzi666 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    32dae48 View commit details
    Browse the repository at this point in the history

  15. remove a part of npu (PaddlePaddle#53677)

    @jjyaoao
    jjyaoao committed May 11, 2023
    Configuration menu
    Copy the full SHA
    314d041 View commit details
    Browse the repository at this point in the history

  16. Revert elementwise (PaddlePaddle#53663) 

    * modify concat_grad add sum comp rule

* delete default mul_double_grad

* delete high grad test

* recover yaml

* modify yaml
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    b4024aa View commit details
    Browse the repository at this point in the history

  17. Update openblas version. (PaddlePaddle#53708)

    @jiweibo
    jiweibo committed May 11, 2023
    Configuration menu
    Copy the full SHA
    9555ae8 View commit details
    Browse the repository at this point in the history

  18. [XPU][PHI Kernels] add pad op for xpu (PaddlePaddle#53684) 

    * add pad op for xpu

* add pad op for xpu

* add pad op for xpu
    @lj970926
    lj970926 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    6f28eb7 View commit details
    Browse the repository at this point in the history

  19. move DataLoader code to paddle.io (PaddlePaddle#48699) 

    * move DataLoader to paddle.io. test=develop
    @heavengate
    heavengate committed May 11, 2023
    Configuration menu
    Copy the full SHA
    793f3b9 View commit details
    Browse the repository at this point in the history

  20. [CustomDevice] fix recompute (PaddlePaddle#53718)

    @ronny1996
    ronny1996 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    2f56b6d View commit details
    Browse the repository at this point in the history

  21. remove some [-Wunused-parameter] warning (PaddlePaddle#53683) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    dbb6269 View commit details
    Browse the repository at this point in the history

  22. [inference Zero-Dim]add equal, elementwise_op trt 0d (PaddlePaddle#53704 

    )
    @zhangjun
    zhangjun committed May 11, 2023
    Configuration menu
    Copy the full SHA
    04e5e7b View commit details
    Browse the repository at this point in the history

  23. [inference][trt]add trt sparse weights switch (PaddlePaddle#53562)

    @zhangjun
    zhangjun committed May 11, 2023
    Configuration menu
    Copy the full SHA
    4a69a53 View commit details
    Browse the repository at this point in the history

  24. Correct the condition of whether can use flash-attn.

    @Xreki
    Xreki committed May 11, 2023
    Configuration menu
    Copy the full SHA
    e92a9bb View commit details
    Browse the repository at this point in the history

  25. fix doc of compare_accuracy (PaddlePaddle#53661)

    zhangkaihuo committed May 11, 2023
    Configuration menu
    Copy the full SHA
    5417382 View commit details
    Browse the repository at this point in the history

  26. [inference Zero-Dim]prelu trt converter support zero dim tensor (Padd… 

    …lePaddle#53634)

* prelu op trt converter support zero dim
    @yuanlehome
    yuanlehome committed May 11, 2023
    Configuration menu
    Copy the full SHA
    82c7388 View commit details
    Browse the repository at this point in the history

  27. add cinn bf16 support (PaddlePaddle#53637) 

    添加CINN与Paddle框架的BFloat16类型映射
    @lanxianghit
    lanxianghit committed May 11, 2023
    Configuration menu
    Copy the full SHA
    3888682 View commit details
    Browse the repository at this point in the history

  28. revise 'Examples' of LBFGS to create right docs(cn), test=docs_preview ( 

    PaddlePaddle#53375)
    @lijialin03
    lijialin03 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    dc003fa View commit details
    Browse the repository at this point in the history

  29. [Inference Zero-Dim] Support trt 0dim of gelu, hard_swish, hard_sigmo… 

    …id and leaky_relu (PaddlePaddle#53714)

* support_act
* delete_silu
    @xiaoxiaohehe001
    xiaoxiaohehe001 committed May 11, 2023
    Configuration menu
    Copy the full SHA
    b150b16 View commit details
    Browse the repository at this point in the history

Commits on May 12, 2023

  1. [XPU] bkcl_broadcast support int64_t (PaddlePaddle#53720)

    @houj04
    houj04 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    13cdaab View commit details
    Browse the repository at this point in the history

  2. fix gpu mem alloc: use phi::memory_utils::Alloc (PaddlePaddle#53721)

    @Wangzheee
    Wangzheee committed May 12, 2023
    Configuration menu
    Copy the full SHA
    6cd7609 View commit details
    Browse the repository at this point in the history

  3. fix er error msg of index_put Op (PaddlePaddle#53717)

    @Courtesy-Xs
    Courtesy-Xs committed May 12, 2023
    Configuration menu
    Copy the full SHA
    92db839 View commit details
    Browse the repository at this point in the history

  4. fix jacobian and hessian's docstring (PaddlePaddle#53732) 

    * fix jacobian and hessian's docstring

* fix hessian's docstring

* fix hessian's docstring
    @HydrogenSulfate
    HydrogenSulfate committed May 12, 2023
    Configuration menu
    Copy the full SHA
    3e3297c View commit details
    Browse the repository at this point in the history

  5. 【Prim】support higher order autodiff for dy2static+composite (PaddlePa… 

    …ddle#53171)

* [Dy2St]Fix x grad names when high order gradient

* Polish error msg

* Add inputs var to backward in dy2st

* Fix error

* Get grad names for backward API

* Fix save load

* Polish code

* Add ut

* [prim] fix not support optional grad bugs in higher order autodiff

* [prim] remove duplicate fill_any_like caused by infershape_for_composite

* fix _strip_grad_suffix_ bugs in higher-order autodiff

* [prim] create output for test_static_prim.cc

---------

Co-authored-by: 0x45f <wangzhen45@baidu.com>
    @cxxly @0x45f
    cxxly and 0x45f committed May 12, 2023
    Configuration menu
    Copy the full SHA
    b73594b View commit details
    Browse the repository at this point in the history

  6. Skip fake alloc in static build for some communication OPs (PaddlePad… 

    …dle#53593)

* Skip fake alloc in static build for depend and nop op

* Skip communication op

* Skip sync op
    @From00
    From00 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    58916e3 View commit details
    Browse the repository at this point in the history

  7. Revert "[CINN]Adjust Bert unittest loss ground truth (PaddlePaddle#53628 

    )" (PaddlePaddle#53731)

This reverts commit 45ce0ad.
    @Aurelius84
    Aurelius84 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    95ae5d5 View commit details
    Browse the repository at this point in the history

  8. fix doc eror of index_put in develop (PaddlePaddle#53727) 

    * fix doc eror of index_put in develop

* fix doc error for index_put; test=document_fix; test=docs_preview
    @Courtesy-Xs
    Courtesy-Xs committed May 12, 2023
    Configuration menu
    Copy the full SHA
    4e416c9 View commit details
    Browse the repository at this point in the history

  9. move pow2_decay_with_linear_warmup kernel to phi (PaddlePaddle#53741) 

    * update

* update
    @huangjiyi
    huangjiyi committed May 12, 2023
    Configuration menu
    Copy the full SHA
    348565b View commit details
    Browse the repository at this point in the history

  10. [PHI] update xpu api version; bind reduce_any_bool xpu kernel; remove… 

    … unnecessary header (PaddlePaddle#53716)
    @RuohengMa
    RuohengMa committed May 12, 2023
    Configuration menu
    Copy the full SHA
    0603777 View commit details
    Browse the repository at this point in the history

  11. sequence_mask functionalization (PaddlePaddle#53478) 

    * sequence_mask functionalization

* fix sequence_mask test
    @GreatV
    GreatV committed May 12, 2023
    Configuration menu
    Copy the full SHA
    d2b1e3c View commit details
    Browse the repository at this point in the history

  12. [inference zero dim] softmax, stack op trt converter support zero dim (… 

    …PaddlePaddle#53729)

* softmax support

* support stack
    @yuanlehome
    yuanlehome committed May 12, 2023
    Configuration menu
    Copy the full SHA
    05d3fc8 View commit details
    Browse the repository at this point in the history

  13. [Inference] Update switch stream logical. (PaddlePaddle#53589)

    @jiweibo
    jiweibo committed May 12, 2023
    Configuration menu
    Copy the full SHA
    eb97f4f View commit details
    Browse the repository at this point in the history

  14. Remove the softmax_out argument.

    @Xreki
    Xreki committed May 12, 2023
    Configuration menu
    Copy the full SHA
    df8c302 View commit details
    Browse the repository at this point in the history

  15. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 12, 2023
    Configuration menu
    Copy the full SHA
    fc3c281 View commit details
    Browse the repository at this point in the history

  16. [CustomDevice] add inference MP support, PART0 (PaddlePaddle#53719) 

    * [CustomDevice] add inference MP support, PART0

* update
    @ronny1996
    ronny1996 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    d03bbef View commit details
    Browse the repository at this point in the history

  17. 【prim】add forward output for Silu grad signature (PaddlePaddle#53632) 

    * add rules

* modify silu_grad input

* modify kernel signature

* modify kernel signature

* code style

* review
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    3846111 View commit details
    Browse the repository at this point in the history

  18. Remove is_causal.

    @Xreki
    Xreki committed May 12, 2023
    Configuration menu
    Copy the full SHA
    d01c89c View commit details
    Browse the repository at this point in the history

  19. [XPU] remove clip of c_softmax_with_cross_entropy_op (PaddlePaddle#53734 

    )
    @houj04
    houj04 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    1019b26 View commit details
    Browse the repository at this point in the history

  20. fix dtype missmatch error (PaddlePaddle#53712)

    @zhangting2020
    zhangting2020 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    772b490 View commit details
    Browse the repository at this point in the history

  21. test(prim-cinn): split test_resnet and test_bert into three tests (Pa… 

    …ddlePaddle#53723)

* test(prim-cinn): split test_resnet and test_bert into three tests

* test(prim-cinn): fix cmake file to run prim test in CINN-CI
    @6clc
    6clc committed May 12, 2023
    Configuration menu
    Copy the full SHA
    60cf9b5 View commit details
    Browse the repository at this point in the history

  22. [CustomDevice] add inference MP support, PART1 (PaddlePaddle#53702)

    @ronny1996
    ronny1996 committed May 12, 2023
    Configuration menu
    Copy the full SHA
    c497b43 View commit details
    Browse the repository at this point in the history

  23. fix add_n kernel of large shape (PaddlePaddle#53749)

    @zhiqiu
    zhiqiu committed May 12, 2023
    Configuration menu
    Copy the full SHA
    4d39cc7 View commit details
    Browse the repository at this point in the history

  24. 【Hackathon 4 No.20】Add i0 / i0e to paddle (PaddlePaddle#52058) 

    * added base code for i0 and i0e

* added grad base code for i0 and i0e

* added i0 and i0e python code

* added ops and backward yaml config

* added i0 and i0e cpu kernel, but not test.

* added i0 and i0e code and unitest files

* added test files

* added i0/i0e gpu implementation code

* updated code style

* updated code style

* fixed unitests code

* updated i0 with eigen3

* fixed bug and added more test cases

* refactor: fixed static graph bug

* refactor: removed i0 and i0e from op_compat

* refactor: updated code style

* refactor: updated op_compat.yaml

* refactor: updated op_compat.yaml

* refactor: fixed op name mapping and optimize unittest case

* refactor: manually implement i0 / i0e

* refactor: added grad kernel for i0 / i0e,didn't finish

* Update math.py

* refactor: added equation to doc in English and added comments for computing i0 / i0e gradient

* refactor: removed eigen implementation

* refactor: finished i0 / i0e cpu and gpu op

* refactor: updated code style

* fix: find  a bug but not fix

* fix: incorrect unittest cases

* update: updated code style and remove my file

* update: updated unittest case

* fix: fixed sign error

* fix: fixed mistakes when merging

* refactor: updated code style

* refactor: remove unused code

* refactor: updated code style
    @PommesPeter
    PommesPeter committed May 12, 2023
    Configuration menu
    Copy the full SHA
    ce256f7 View commit details
    Browse the repository at this point in the history

Commits on May 13, 2023

  1. Revert elementwise add (PaddlePaddle#53745) 

    * modify concat_grad add sum comp rule

* delete default mul_double_grad

* delete high grad test

* recover yaml

* modify yaml

* recover add_double_grad prim
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 13, 2023
    Configuration menu
    Copy the full SHA
    b75d8c7 View commit details
    Browse the repository at this point in the history

Commits on May 14, 2023

  1. fix build error (PaddlePaddle#53790) 

    * fix build error

* fix build error

* fix
    @tianshuo78520a
    tianshuo78520a committed May 14, 2023
    Configuration menu
    Copy the full SHA
    3e90a46 View commit details
    Browse the repository at this point in the history

Commits on May 15, 2023

  1. move OneHotRawKernel to legacy (PaddlePaddle#53200) 

    * move OneHotRawKernel to legacy

* fix
    @zhangyuqin1998
    zhangyuqin1998 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    34122e3 View commit details
    Browse the repository at this point in the history

  2. Tranpose layout (PaddlePaddle#53351) 

    * update

* Update backward.h

* Update composite_backward_api.h

* Update tensor_utils.cc

* Update backward.cc

* update

* stype

* update

* add ctest

* code stype
    @AnnaTrainingG
    AnnaTrainingG committed May 15, 2023
    Configuration menu
    Copy the full SHA
    3dce9f0 View commit details
    Browse the repository at this point in the history

  3. [XPU][PHI] bind index_sample_grad xpu kernel (PaddlePaddle#53753)

    @RuohengMa
    RuohengMa committed May 15, 2023
    Configuration menu
    Copy the full SHA
    8105607 View commit details
    Browse the repository at this point in the history

  4. relocate python/paddle/fluid/regularizer.py (PaddlePaddle#53106) 

    * relocate regularizer.py

* fix bug

* fix bug

* fix bug

* relocate the import

* replace _regularization_coeff with coeff

* remove the L1DecayRegularizer and L2DecayRegularizer
    @longranger2
    longranger2 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    00e415d View commit details
    Browse the repository at this point in the history

  5. [CI]Fix test_bert_primm_cinn gt loss value (PaddlePaddle#53796)

    @Aurelius84
    Aurelius84 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    359f43a View commit details
    Browse the repository at this point in the history

  6. bug fixes (PaddlePaddle#53798)

    @risemeup1
    risemeup1 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    a822a08 View commit details
    Browse the repository at this point in the history

  7. Fix bug of hybrid_parallel_optimizer, amp use scaler.minimize(), (Pad… 

    …dlePaddle#53773)

however it can't deal with group of parameter_list of dict.
    @GhostScreaming
    GhostScreaming committed May 15, 2023
    Configuration menu
    Copy the full SHA
    5152971 View commit details
    Browse the repository at this point in the history

  8. [PHI]Add Filter for get_kernel_signatures.py (PaddlePaddle#53760) 

    * delete log

* filter some kernel signature
    @YuanRisheng
    YuanRisheng committed May 15, 2023
    Configuration menu
    Copy the full SHA
    b428e8f View commit details
    Browse the repository at this point in the history

  9. [UnitTest]Fix deprecate fluid.regularizer test=document_fix (PaddlePa… 

    …ddle#53805)
    @Aurelius84
    Aurelius84 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    3d4d7c1 View commit details
    Browse the repository at this point in the history

  10. add check ops for prim (PaddlePaddle#52302) 

    * add check ops for prim

* fix pow and concat composite registration

* modify log

* add note and remove useless code

* remove useless code

* modify program to check

* remove useless note
    @Charles-hit
    Charles-hit committed May 15, 2023
    Configuration menu
    Copy the full SHA
    3d6bd6a View commit details
    Browse the repository at this point in the history

  11. fix bug of test_pad_op for cinn (PaddlePaddle#53772)

    @zyfncg
    zyfncg committed May 15, 2023
    Configuration menu
    Copy the full SHA
    a9c3e32 View commit details
    Browse the repository at this point in the history

  12. remove some [-Wunused-paramter]warning (PaddlePaddle#53681) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    96188fc View commit details
    Browse the repository at this point in the history

  13. remove some [-Wunsed-parameter]warning (PaddlePaddle#53679) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    ca2ea16 View commit details
    Browse the repository at this point in the history

  14. remove some [-Wunsed-parameter] warning (PaddlePaddle#53687) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    8ed01e8 View commit details
    Browse the repository at this point in the history

  15. remove some [-Wunsed-parameter] warning (PaddlePaddle#53689) 

    * test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    3e1fffe View commit details
    Browse the repository at this point in the history

  16. [BUG] fix windows kernel dispatch of _lzcnt bug (PaddlePaddle#53728)

    @zhwesky2010
    zhwesky2010 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    972daa4 View commit details
    Browse the repository at this point in the history

  17. Reduce inference library size and compile time (PaddlePaddle#53369) 

    * Reduce inference library size and compile time

* resolve conflicts
    @chalsliu
    chalsliu committed May 15, 2023
    Configuration menu
    Copy the full SHA
    0ef5180 View commit details
    Browse the repository at this point in the history

  18. Silu double grad (PaddlePaddle#53605) 

    * add rules

* modify no kernel yaml parse

* success op generate

* success test_silu_double

* modify bug

* modify static error

* modify silu_grad input

* modify kernel signature

* modify kernel signature

* code style

* code style

* review

* delete opinfo modify
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    94c3880 View commit details
    Browse the repository at this point in the history

  19. [inference Zero-Dim][trt] Add Zero-Dim tensor support for clip, cast,… 

    … flatten_contiguous_range (PaddlePaddle#53769)

* [inference Zero-Dim][trt]clip,cast,flatten_contiguous_range trt op converter support zero dim
    @bukejiyu
    bukejiyu committed May 15, 2023
    Configuration menu
    Copy the full SHA
    cc9aeda View commit details
    Browse the repository at this point in the history

  20. [CustomDevice] add inference MP support, PART2 (PaddlePaddle#53701)

    @ronny1996
    ronny1996 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    e04f8d4 View commit details
    Browse the repository at this point in the history

  21. [CustomDevice] add inference MP support, PART3 (PaddlePaddle#53703)

    @ronny1996
    ronny1996 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    56fded1 View commit details
    Browse the repository at this point in the history

  22. move dequantize kernel to phi (PaddlePaddle#53739) 

    * update

* fix bug

* fix output type def
    @huangjiyi
    huangjiyi committed May 15, 2023
    Configuration menu
    Copy the full SHA
    efd410c View commit details
    Browse the repository at this point in the history

  23. [AMP]fix embedding model weight type mismatch error (PaddlePaddle#53770) 

    * fix embedding model weight type mismatch error

* Update fp16_utils.py

---------

Co-authored-by: Zhang Ting <zhangting_2017@163.com>
    @shaojiewang @zhangting2020
    shaojiewang and zhangting2020 committed May 15, 2023
    Configuration menu
    Copy the full SHA
    848deec View commit details
    Browse the repository at this point in the history

Commits on May 16, 2023

  1. [PHI] fix duplicate instantiation of reduce_any when compiling with X… 

    …PU_KP (PaddlePaddle#53806)
    @RuohengMa
    RuohengMa committed May 16, 2023
    Configuration menu
    Copy the full SHA
    2174e91 View commit details
    Browse the repository at this point in the history

  2. fix pinv api for divide zero (PaddlePaddle#53815)

    @andyjiang1116
    andyjiang1116 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    434343c View commit details
    Browse the repository at this point in the history

  3. [CustomDevice] fix BatchNorm (PaddlePaddle#53820)

    @ronny1996
    ronny1996 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    926b886 View commit details
    Browse the repository at this point in the history

  4. fix simple typos (PaddlePaddle#53783) 

    * correct 1th to 1st

* correct 1th to 1st

* fix typo

* fix typos
    @MahmoudAshraf97
    MahmoudAshraf97 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    847c48a View commit details
    Browse the repository at this point in the history

  5. Fix typos, test=document_fix (PaddlePaddle#53803)

    @co63oc
    co63oc committed May 16, 2023
    Configuration menu
    Copy the full SHA
    4f7dfd0 View commit details
    Browse the repository at this point in the history

  6. Fix typos in paddle/phi/core/generator.cc (PaddlePaddle#53802)

    @co63oc
    co63oc committed May 16, 2023
    Configuration menu
    Copy the full SHA
    79c84ba View commit details
    Browse the repository at this point in the history

  7. [phi] move stft to phi - Step 1 (PaddlePaddle#53517) 

    * [phi]mv StftKernel to phi

* [phi] fix KernelSignature

* [phi]fix arr error

* [phi] Disable check_dygraph

* [phi]fix include

* [phi] rewrite mutable_data, add output register

* [phi] fix  Alloc

* [phi] fix Alloc again

* [phi] fix mutable_data

* [phi] fix onesided_out Resize
    @gouzil
    gouzil committed May 16, 2023
    Configuration menu
    Copy the full SHA
    00c21ab View commit details
    Browse the repository at this point in the history

  8. [inference][trt]Remove unused code from teller.cc (PaddlePaddle#53758) 

    * remove unused code
    @zhangjun
    zhangjun committed May 16, 2023
    Configuration menu
    Copy the full SHA
    2a94b81 View commit details
    Browse the repository at this point in the history

  9. [AMP] Allow to switch whether to use promote strategy to choose kerne… 

    …l for O2 training. (PaddlePaddle#53742)

* Allow to switch whether to use promote strategy to choose kernel for O2 training.

* Fix comparing error and add unittest.
    @Xreki
    Xreki committed May 16, 2023
    Configuration menu
    Copy the full SHA
    db407bf View commit details
    Browse the repository at this point in the history

  10. [Inference] clean unused code/target for reduce inference so volume (… 

    …PART I) (PaddlePaddle#53762)

* remove prelu land ookuip_table plugin, adjust .h include location

* clean code and adjust some .h

* update
    @yuanlehome
    yuanlehome committed May 16, 2023
    Configuration menu
    Copy the full SHA
    51ecd93 View commit details
    Browse the repository at this point in the history

  11. Add fill_constant_batch_size_like tests (PaddlePaddle#53736)

    @co63oc
    co63oc committed May 16, 2023
    Configuration menu
    Copy the full SHA
    98100fd View commit details
    Browse the repository at this point in the history

  12. [dygraph]remove legacy code : _in_eager_mode_ and _in_eager_without_d… 

    …ygraph_check() (PaddlePaddle#53761)

* remove _in_eager_mode_

* remove _in_eager_mode_
    @liudongxue01
    liudongxue01 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    b133317 View commit details
    Browse the repository at this point in the history

  13. 【Hackathon4 No.61】remainder 算子FP16/BF16单测完善 (PaddlePaddle#52920)

    @enkilee
    enkilee committed May 16, 2023
    Configuration menu
    Copy the full SHA
    481511a View commit details
    Browse the repository at this point in the history

  14. Add Japanese README (PaddlePaddle#53726) 

    * Add Japanese README

* Update README_ja.md
    @eltociear
    eltociear committed May 16, 2023
    Configuration menu
    Copy the full SHA
    ad45b36 View commit details
    Browse the repository at this point in the history

  15. [static op generation] InstanceNorm (PaddlePaddle#53340) 

    * mv InstanceNorm

* modify op_version.yaml

* modify add Operator:: in get_expected_kernel_func.cc

* rm gradexpectedkernel

* add extra

* add float epsilon=1e-5
    @Liyulingyue
    Liyulingyue committed May 16, 2023
    Configuration menu
    Copy the full SHA
    7b81092 View commit details
    Browse the repository at this point in the history

  16. support auto generation V2 abs (PaddlePaddle#53341)

    @enkilee
    enkilee committed May 16, 2023
    Configuration menu
    Copy the full SHA
    b86bbe8 View commit details
    Browse the repository at this point in the history

  17. static graph autogen code support for softmax op (PaddlePaddle#53581) 

    * static graph autogen code support for softmax op

* bug fixed

* fix PR-CI-Windows error

* fix CI error

* bug fixed

* fix conflicts
    @GreatV
    GreatV committed May 16, 2023
    Configuration menu
    Copy the full SHA
    312f018 View commit details
    Browse the repository at this point in the history

  18. Move fused batchnorm to Phi (PaddlePaddle#53476) 

    * trans fused batch norm Compute function

* trans batch norm register info to phi

* trans fused batch norm grad Compute

* trans batch norm grad register info

* add sig file

* update sig file

* Update fused_bn_activation_kernel.cu

* Update fused_bn_activation_grad_kernel.cu

* fix

* Rename fused_bn_activation_kernel_grad.cu to fused_bn_activation_kernel.cu

* fix

* fix

* fix CudnnDataType error

* fix

* fix include

* update

* add #if

* add fused bn act to cmakelist.txt

* update  cmakelist

* fix #ifdef error

* add timeout set

* add env set

* fix

* fix

* Update fused_bn_activation_sig.cc
    @AndSonder
    AndSonder committed May 16, 2023
    Configuration menu
    Copy the full SHA
    5e5481d View commit details
    Browse the repository at this point in the history

  19. fix _strip_grad_suffix_ bugs when input patten is 'x@GRAD@RENAME'

    @cxxly
    cxxly committed May 16, 2023
    Configuration menu
    Copy the full SHA
    0689e2a View commit details
    Browse the repository at this point in the history

  20. 昇腾和寒武纪相关代码退场 npu相关代码退场3 (PaddlePaddle#53699) 

    * rm npu

* rm use_npu

* rm npuid

* rm use_npu

* rm npuid

* delete npupinned

* roll back sth.

* roll back sth.

* delete npupinned

* roll back sth.

* roll back sth.

* rm npu

* rollback something

* rollback npu identity

* rollback npu identity
    @Liyulingyue
    Liyulingyue committed May 16, 2023
    Configuration menu
    Copy the full SHA
    5b054d2 View commit details
    Browse the repository at this point in the history

  21. move cudnn_lstm kernel to phi (PaddlePaddle#53730) 

    * update

* fix bug

* test

* test

* update

* update mutable_data

* fix bug

* update

* fix bug

* update output type reg

* update

* update
    @huangjiyi
    huangjiyi committed May 16, 2023
    Configuration menu
    Copy the full SHA
    52889e3 View commit details
    Browse the repository at this point in the history

  22. [AMP] support OD level for static (PaddlePaddle#53768)

    @AnnaTrainingG
    AnnaTrainingG committed May 16, 2023
    Configuration menu
    Copy the full SHA
    c2c3bd4 View commit details
    Browse the repository at this point in the history

  23. [XPU] Add sigmoid_elementmul_xpu_fuse_pass (PaddlePaddle#53580)

    @sprouteer
    sprouteer committed May 16, 2023
    Configuration menu
    Copy the full SHA
    32e36b1 View commit details
    Browse the repository at this point in the history

  24. 【static】modify backward prune logic for EmptygradOpMaker (PaddlePaddl… 

    …e#53746)

* add rules

* modify no kernel yaml parse

* success op generate

* success test_silu_double

* modify bug

* modify static error

* modify silu_grad input

* modify kernel signature

* modify kernel signature

* code style

* code style

* review

* delete opinfo modify

* modify gradOpMaker

* modify gradOpMaker

* modify genarated-j2

* add approve rules

* modify aytograd_functional_static_test
    @xiaoguoguo626807
    xiaoguoguo626807 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    69161a9 View commit details
    Browse the repository at this point in the history

  25. Fix some tests for issuse 52842 (PaddlePaddle#53795) 

    * polish

* polish
    @liuzhenhai93
    liuzhenhai93 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    c33ba9d View commit details
    Browse the repository at this point in the history

  26. add timer to pp (PaddlePaddle#53831)

    @FeixLiu
    FeixLiu committed May 16, 2023
    Configuration menu
    Copy the full SHA
    0ab7f94 View commit details
    Browse the repository at this point in the history

  27. 【PaddlePaddle Hackathon 4 No.34】为 Paddle 优化 Lerp OP 在 GPU 上的性能 (Paddl… 

    …ePaddle#53154)

* modify lerp_kernel.cu

* pre-commit

* fix some CI issues

* fix some CI issues

* fix some CI issues

* fix some CI issues

* fix some CI issues

* fix some CI issues

* fix some CI issues

* fix some CI issues

* Add files via upload

fix some CI issues
    @WintersMontagne10335
    WintersMontagne10335 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    e592534 View commit details
    Browse the repository at this point in the history

  28. remove some [-Wunused-parameter] warning and fix a file to pass cppli… 

    …nt (PaddlePaddle#53814)

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop

* test,test=develop
    @Galaxy1458
    Galaxy1458 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    10a38b4 View commit details
    Browse the repository at this point in the history

  29. 【Hackathon No57】add bf16 for mode (PaddlePaddle#53195) 

    * add bf16 for mode

* remove random seed 666

* try to fix op_type error

* test for me

* try to fix op_type

* fix redundancy code

* add fp,bf for lastdim

* fix some error

* simplify code

* fix shape error

* optype error

* fix skipif bf16
    @Difers
    Difers committed May 16, 2023
    Configuration menu
    Copy the full SHA
    640cff0 View commit details
    Browse the repository at this point in the history

  30. [Zero-Dim] update 0d tensor api en doc, test=document_fix (PaddlePadd… 

    …le#53823)
    @zhwesky2010
    zhwesky2010 committed May 16, 2023
    Configuration menu
    Copy the full SHA
    50f0acc View commit details
    Browse the repository at this point in the history

  31. Add huber_loss tests (PaddlePaddle#53535)

    @co63oc
    co63oc committed May 16, 2023
    Configuration menu
    Copy the full SHA
    74b91bc View commit details
    Browse the repository at this point in the history

Commits on May 17, 2023

  1. fix bug of unpool in the module 'enet' of paddleseg (PaddlePaddle#53840)

    @heavyrain-lzy
    heavyrain-lzy committed May 17, 2023
    Configuration menu
    Copy the full SHA
    d6d3de7 View commit details
    Browse the repository at this point in the history

  2. update openblas version (PaddlePaddle#53748) 

    * update openblas version

* update
    @jiweibo
    jiweibo committed May 17, 2023
    Configuration menu
    Copy the full SHA
    8965366 View commit details
    Browse the repository at this point in the history

  3. 【Hackathon 4 No.21】Add i1 / i1e to paddle (PaddlePaddle#53210) 

    * Add i1 and i1e op

* resolve merge conflicts
    @LyndonKong
    LyndonKong committed May 17, 2023
    Configuration menu
    Copy the full SHA
    a63fb4c View commit details
    Browse the repository at this point in the history

  4. [fluid] decoupling abn op (PaddlePaddle#53826)

    @gouzil
    gouzil committed May 17, 2023
    Configuration menu
    Copy the full SHA
    38e5cd0 View commit details
    Browse the repository at this point in the history

  5. test(cinn): fix resnet50 precision (PaddlePaddle#53649)

    @6clc
    6clc committed May 17, 2023
    Configuration menu
    Copy the full SHA
    56e8aff View commit details
    Browse the repository at this point in the history

  6. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 17, 2023
    Configuration menu
    Copy the full SHA
    91a0ea5 View commit details
    Browse the repository at this point in the history

  7. [IR] Program & Parameter & PaddleDialect (PaddlePaddle#53557) 

    * add program parameter dialect_interface

* fix op create bug

* add ir parameter convert pd variable methods

* refine code

* fix bug

* refine by ut

* refine ut

* delete unused code

* refine code

* refine code by comment

* reset WITH_NEW_IR

* refine op attribute map

* refine program and op create

* refine program and op create
    @zhangbo9674
    zhangbo9674 committed May 17, 2023
    Configuration menu
    Copy the full SHA
    78967ad View commit details
    Browse the repository at this point in the history

  8. remove fluid memory_usage_calc&model_stat&op_frequence (PaddlePaddle#… 

    …53838)
    @zoooo0820
    zoooo0820 committed May 17, 2023
    Configuration menu
    Copy the full SHA
    2cb2801 View commit details
    Browse the repository at this point in the history

  9. Polish codes.

    @Xreki
    Xreki committed May 17, 2023
    Configuration menu
    Copy the full SHA
    f6be954 View commit details
    Browse the repository at this point in the history

  10. Supports offline compilation of Paddle third-party libraries (PaddleP… 

    …addle#53744)

* optimize logsumexp in small data scale

* fix

* fix

* add #pragma once

* compile protobuf offline

* add submodlu gflags

* check_submodules

* check_submodules

* add_submodule protobuf

* add_submodule_protobuf

* add_submodule

* add .gitmodules

* add_submodules

* fix_compiler error

* support offline compile

* support offline compile

* support offline_compile

* remove cub

* remove brpc

* support offline compile

* support offline compile

* canning patching on cryptopp

* modify .gitigonre of cryptopp

* test

* offline compile

* add_submodule zlib

* modify .gitmodules

* modify .gitmodules

* fix setup.py bug

* delete submodule cryptopp

* fix windows compile bug

* fix xxhash compile problem

---------

Co-authored-by: Asthestarsfalll <1186454801@qq.com>
Co-authored-by: Asthestarsfalll <72954905+Asthestarsfalll@users.noreply.github.com>
    @risemeup1 @Asthestarsfalll
    3 people committed May 17, 2023
    Configuration menu
    Copy the full SHA
    734dc44 View commit details
    Browse the repository at this point in the history

  11. [CustomDevice] suport device_guard for custom device (PaddlePaddle#53808 

    )

* suport device_guard for npu

* fix comment

* fix typo
    @YanhuiDua
    YanhuiDua committed May 17, 2023
    Configuration menu
    Copy the full SHA
    9e045ee View commit details
    Browse the repository at this point in the history

  12. [CINN] extend cinn single test timeout from 150 to 200, test=document… 

    …_fix (PaddlePaddle#53908)
    @thisjiang
    thisjiang committed May 17, 2023
    Configuration menu
    Copy the full SHA
    4f1bf19 View commit details
    Browse the repository at this point in the history

Commits on May 18, 2023

  1. [AMP]Master grad in static graph (PaddlePaddle#53362) 

    * add master gradients on static graph

* add unit test for bf16 master grad static graph

* use float16 as v100 test dtype

* only skip GPU which do not support bf16

* use linear layer to test master grad

* 1.push master grad creation before all optimizer ops; 2.remove useless unittest; 3.use a function to create master grad states
    @shaojiewang
    shaojiewang committed May 18, 2023
    Configuration menu
    Copy the full SHA
    972581d View commit details
    Browse the repository at this point in the history

  2. Fix typos in send_v2_op.cu.cc (PaddlePaddle#53904)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    65ce688 View commit details
    Browse the repository at this point in the history

  3. Fix typos, test=document_fix (PaddlePaddle#53916)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    92121d1 View commit details
    Browse the repository at this point in the history

  4. fix -Werror=format-security (PaddlePaddle#53886)

    @engineer1109
    engineer1109 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    6d7076c View commit details
    Browse the repository at this point in the history

  5. rm cmake npu (PaddlePaddle#53869) 

    * rm cmake npu

* Update generic.cmake

* Update generic.cmake
    @Liyulingyue
    Liyulingyue committed May 18, 2023
    Configuration menu
    Copy the full SHA
    79ce3fa View commit details
    Browse the repository at this point in the history

  6. rm tools npu (PaddlePaddle#53870) 

    * rm tools npu

* Update get_pr_ut.py

* Update get_pr_ut.py
    @Liyulingyue
    Liyulingyue committed May 18, 2023
    Configuration menu
    Copy the full SHA
    d294eef View commit details
    Browse the repository at this point in the history

  7. [XPU] do not call check_nccl_version_for_p2p under xpu (PaddlePaddle#… 

    …53862)

* [XPU] do not call check_nccl_version_for_p2p under xpu

* refine code.
    @houj04
    houj04 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    5d638fe View commit details
    Browse the repository at this point in the history

  8. [Fix Typo] Fix gpu_info.h, Wheter->Whether (PaddlePaddle#53564)

    @jiahy0825
    jiahy0825 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    236e742 View commit details
    Browse the repository at this point in the history

  9. support auto generate for op layer_norm (PaddlePaddle#53178) 

    * simplify layer_norm_op.cc

* support auto generate for op layer_norm

* update unittest for composite_layer_norm

* remove layer_norm_op.cc from scripts

* replace layer_norm_op with generated_op

* add get_expected_kernel for layer_norm

* update cmake kernel register function for layer_norm_mkldnn_op
    @RedContritio
    RedContritio committed May 18, 2023
    Configuration menu
    Copy the full SHA
    4f07b65 View commit details
    Browse the repository at this point in the history

  10. Fix typos in executor_statistics.cc (PaddlePaddle#53917)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    1ac28b6 View commit details
    Browse the repository at this point in the history

  11. [CustomOp Unittest] Fix XPU unittest, discard static backward (Paddle… 

    …Paddle#53899)
    @jiahy0825
    jiahy0825 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    2d0c694 View commit details
    Browse the repository at this point in the history

  12. [Dy2static-Fallback] add set_eval_frame function in pybind. (PaddlePa… 

    …ddle#52006)

* [Dy2static-Fallback] add set_eval_frame function in pybind.
1. add set_eval_frame function in pybind.

* add unittest for eval frame hooker.

* [support py38]

* fix-GeneratorExit error in eval frame hooker

* support python == 3.9

* support 3.10

* fix some comments
    @2742195759
    2742195759 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    7b1695a View commit details
    Browse the repository at this point in the history

  13. Del test_async_read_write in CPU (PaddlePaddle#53882) 

    * fix

* fix
    @tianshuo78520a
    tianshuo78520a committed May 18, 2023
    Configuration menu
    Copy the full SHA
    acb5039 View commit details
    Browse the repository at this point in the history

  14. Fix typos in elementwise dir (PaddlePaddle#53907)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    2782b29 View commit details
    Browse the repository at this point in the history

  15. move sequence_mask op InferShape func (PaddlePaddle#53782) 

    * move sequence_mask op InferShape func

* add dtype infer
    @GreatV
    GreatV committed May 18, 2023
    Configuration menu
    Copy the full SHA
    a862deb View commit details
    Browse the repository at this point in the history

  16. add fp16 and bf16 for trunc (PaddlePaddle#53876)

    @longranger2
    longranger2 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    d8407c5 View commit details
    Browse the repository at this point in the history

  17. Fix typos, test=document_fix (PaddlePaddle#53927)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    e916e80 View commit details
    Browse the repository at this point in the history

  18. Fix typos (PaddlePaddle#53912)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    117e951 View commit details
    Browse the repository at this point in the history

  19. Add segment_pool tests (PaddlePaddle#53785)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    0bed220 View commit details
    Browse the repository at this point in the history

  20. move fusion_group kernel to phi (PaddlePaddle#53781)

    @huangjiyi
    huangjiyi committed May 18, 2023
    Configuration menu
    Copy the full SHA
    26da689 View commit details
    Browse the repository at this point in the history

  21. Fused elementwises kernels and ops (PaddlePaddle#51427) 

    * Fused elementwises kernels and ops

* change fuse pass name

* adjust .pbtxt files

* adjust quantization attributes

* add missing arguments and fix others, review fixed

* simplify fused kernel registration

* fix elementwise unit tests

* reuse one fused elementwise op

* adjust proto

* Add supported datatypes

* Change 'Scale' to 'scale' in tests, change some tests to onednn

* Revert breaking changes

* Fix unit tests

* Delete obsolete test cases

* Delete commented out code

* Fix codestyle

* delete temporary condition

* fix conflicts and delete duplicate fusing

* Fix code after merge

* Move tests to new directory

* fix tests volatility

* Rename test_elementwise_add_onednn_op.py to test_elementwise_add_mkldnn_op.py

* Update CMakeLists.txt add mkldnn op test

---------

Co-authored-by: Silv3S <slawomir.siwek@intel.com>
    @HulekJakub @Silv3S
    HulekJakub and Silv3S committed May 18, 2023
    Configuration menu
    Copy the full SHA
    fb4a6ec View commit details
    Browse the repository at this point in the history

  22. remove CopyWithContext limitation (PaddlePaddle#53771)

    @engineer1109
    engineer1109 committed May 18, 2023
    Configuration menu
    Copy the full SHA
    d53d8fd View commit details
    Browse the repository at this point in the history

  23. Fix qkv_transpose_out's shape and scaling of Q * K.

    @Xreki
    Xreki committed May 18, 2023
    Configuration menu
    Copy the full SHA
    ba84941 View commit details
    Browse the repository at this point in the history

  24. Add einsum tests (PaddlePaddle#53722)

    @co63oc
    co63oc committed May 18, 2023
    Configuration menu
    Copy the full SHA
    c3c8579 View commit details
    Browse the repository at this point in the history

  25. Merge branch 'develop' into add_flash_attn_for_af2

    @Xreki
    Xreki committed May 18, 2023
    Configuration menu
    Copy the full SHA
    bee8537 View commit details
    Browse the repository at this point in the history

  26. Update commit of flash-attention.

    @Xreki
    Xreki committed May 18, 2023
    Configuration menu
    Copy the full SHA
    3747978 View commit details
    Browse the repository at this point in the history