- add inplace attribute, for memory reuse - https://github.com/PaddlePaddle/Paddle/pull/10665 - eigen ci testing - https://github.com/PaddlePaddle/Paddle/pull/10731 - fix compile testing - https://github.com/PaddlePaddle/Paddle/pull/10667 - https://github.com/PaddlePaddle/Paddle/pull/10629 - https://github.com/PaddlePaddle/Paddle/pull/10450 - https://github.com/PaddlePaddle/Paddle/pull/10657 - benchmarkSuite framework, for consistent checking - https://github.com/PaddlePaddle/Paddle/pull/10646 - https://github.com/PaddlePaddle/Paddle/pull/10624 - make uint8 support in data_type transform - https://github.com/PaddlePaddle/Paddle/pull/10715 - refine batch norm op - https://github.com/PaddlePaddle/Paddle/pull/10502 

• Build Scripts:
  • Add a new feature to run a single uni test: https://github.com/PaddlePaddle/Paddle/pull/10678
  • Generate dockerfile after each build: https://github.com/PaddlePaddle/Paddle/pull/10719
  • Cleanup old build scripts: https://github.com/PaddlePaddle/Paddle/pull/10721
• Teamcity CI
  • Using different machines to debug the “test_all_ops” failure.
  • [WIP] Try to enable ccache to speed up build
  • [WIP] Debug build “C_API” failure in nightly build
• Fix document error: https://github.com/PaddlePaddle/Paddle/pull/10726

• Fix data_feeder LoD bug
  • https://github.com/PaddlePaddle/Paddle/pull/10677
• Simplify recognize digits example code
  • https://github.com/PaddlePaddle/Paddle/pull/10722
• [WIP] Make the creation of LoDTensor more user friendly in book examples
  • https://github.com/PaddlePaddle/Paddle/issues/10735
• Fix signed/unsigned comparison warning
  • https://github.com/PaddlePaddle/Paddle/pull/10724
• Review:
  • https://github.com/PaddlePaddle/Paddle/pull/10678#pullrequestreview-120882860
  • https://github.com/PaddlePaddle/Paddle/pull/10601#pullrequestreview-119618917
  • https://github.com/PaddlePaddle/Paddle/pull/10587#pullrequestreview-119572393

• mkldnn：
  • plan the July 5th roadmap with Intel:
    • implement the inference optimization of OCR Vehicle Plate Recognition model at first. https://github.com/PaddlePaddle/Paddle/issues/10685
    • Profile ocr Vehicle Plate Recognition model, 8%~10% slower than the online version.
  • [merge] update mklml version, disable building tests and examples when install mkldnn, fix compiler error on gcc48，fix dead link of mkldnn doc: https://github.com/PaddlePaddle/Paddle/pull/10571，https://github.com/PaddlePaddle/Paddle/pull/10567， https://github.com/intel/mkl-dnn/pull/238
• [merge] implement convert tensorrt relu op, and its unit-test: https://github.com/PaddlePaddle/Paddle/pull/10495
• code review：
  • mkldnn:
    • [merge] Reusing of softmax mkldnn primitives：https://github.com/PaddlePaddle/Paddle/pull/10576
    • Patch mkldnn for build on gcc 4.8.2： https://github.com/PaddlePaddle/Paddle/pull/10616
    • [bug] remove light_imdb & tiny_imdb: https://github.com/PaddlePaddle/models/pull/907
    • [bug] Add profile in aishell example: https://github.com/PaddlePaddle/models/pull/910
    • crnn-ctc label was not found：https://github.com/PaddlePaddle/models/pull/912
  • inference:
    • https://github.com/PaddlePaddle/Paddle/pull/10549
    • https://github.com/PaddlePaddle/Paddle/pull/10623
    • https://github.com/PaddlePaddle/Paddle/pull/10660

• Test some tcp framework performance
  • https://github.com/PaddlePaddle/Paddle/pull/10120
• Help OCR groud to solve
  • Coredump after call inference api.
  • Slowdown when link with fluid library.

• Bug fixes and enhancements
  • https://github.com/PaddlePaddle/Paddle/pull/10709
  • https://github.com/PaddlePaddle/Paddle/pull/10702
  • https://github.com/PaddlePaddle/Paddle/pull/10666
  • https://github.com/PaddlePaddle/Paddle/pull/10640
  • https://github.com/PaddlePaddle/Paddle/pull/10620
• [API] Trainer.train
  • https://github.com/PaddlePaddle/Paddle/pull/10620
• Reader related jobs:
  • Support uint8 of tensor
    • https://github.com/PaddlePaddle/Paddle/pull/10683
  • [WIP] RandomCropOp

• aws tool on vgg16 dist train as example integration with CE
  • tool adjustment done
  • vgg16 dist train error https://github.com/PaddlePaddle/Paddle/issues/10720
• Trying to bring nightly back online
  • https://github.com/PaddlePaddle/Paddle/issues/10732
  • https://github.com/PaddlePaddle/Paddle/issues/10728
• reviews
  • https://github.com/PaddlePaddle/Paddle/pull/10719
  • https://github.com/PaddlePaddle/Paddle/pull/10726
  • https://github.com/PaddlePaddle/Paddle/pull/10718

• [Merged] Adapt the convertor to tensorrt backend
  • https://github.com/PaddlePaddle/paddle-onnx/pull/47
• Add Inception_v4 model config in Fluid API
  • https://github.com/PaddlePaddle/models/pull/900
• Add inception-v4 to supported models in onnx convertor
  • https://github.com/PaddlePaddle/paddle-onnx/pull/49
• [DeepASR] Add profile in aishell example
  • https://github.com/PaddlePaddle/models/pull/910
• [WIP] Add multiple groups in conv_transpose_op to support face detection model

Code Review:

• https://github.com/PaddlePaddle/Paddle/pull/10621
• https://github.com/PaddlePaddle/paddle-onnx/pull/48

• benchmark scripts polish https://github.com/PaddlePaddle/Paddle/pull/10707
• fix possible memory leek https://github.com/PaddlePaddle/Paddle/pull/10663
• NCCL2 dist train perf tests: https://docs.google.com/spreadsheets/d/1D5Xc_TfGfMV5aKh4ZJS_b4js3Mnn06H1Po0iuECZLr4/edit?pli=1#gid=1453458907

• NMT:
  • Add beamsearch decoder using while_op in Transformer (WIP).
    • https://github.com/PaddlePaddle/models/pull/915
  • Refine Transformer code.
    • https://github.com/PaddlePaddle/models/pull/894
• Review:
  • https://github.com/PaddlePaddle/Paddle/pull/10527
  • https://github.com/PaddlePaddle/Paddle/pull/10641

• Enhanced is_empty_op
  • https://github.com/PaddlePaddle/Paddle/pull/10704
• Enhancing assign_value_op
• Found a bug of conditional block
  • https://github.com/PaddlePaddle/Paddle/issues/10711
• Simple RNN beam search

• Enhance reduce op
  • https://github.com/PaddlePaddle/Paddle/pull/10708
• Add dice loss:
  • https://github.com/PaddlePaddle/Paddle/pull/10717

• PR

  • Balance parameter_opt between cards
    • https://github.com/PaddlePaddle/Paddle/pull/10546
  • Feature/fuse reduce op
    • https://github.com/PaddlePaddle/Paddle/pull/10615
  • Update SE-Resnext
    • https://github.com/dzhwinter/benchmark/pull/99
  • The current SE-Resnext 152 benchmark result on P40
    • https://github.com/PaddlePaddle/Paddle/issues/10706
  • Fix pe bug [prevent loss.grad optimized by mem_opt]
    • https://github.com/PaddlePaddle/Paddle/pull/10645
  • Refine fetch op handle
    • https://github.com/PaddlePaddle/Paddle/pull/10671

• Review

  • https://github.com/PaddlePaddle/Paddle/pull/10683
  • https://github.com/PaddlePaddle/Paddle/pull/10713
  • https://github.com/PaddlePaddle/Paddle/pull/10709
  • https://github.com/PaddlePaddle/Paddle/pull/10687
  • https://github.com/PaddlePaddle/Paddle/pull/10619
  • https://github.com/PaddlePaddle/Paddle/pull/10572

• Fix a profiler race condition
  • https://github.com/PaddlePaddle/Paddle/pull/10621
• Delete prefetch_ctx_ after use
  • https://github.com/PaddlePaddle/Paddle/pull/10659
• Our tests interfere with each other and cause random failure
  • https://github.com/PaddlePaddle/Paddle/pull/10547
• Revert "CI: rerun failed tests. (#10536)
  • https://github.com/PaddlePaddle/Paddle/pull/10543
• Add a multi-dim add layer test.
  • https://github.com/PaddlePaddle/Paddle/pull/10437
• Add comment to explain how to run inference test
  • https://github.com/PaddlePaddle/Paddle/pull/10430
• timeline for distributed training
  • https://github.com/PaddlePaddle/Paddle/pull/10380
• add some instructions for running vgg distributedly
  • https://github.com/PaddlePaddle/Paddle/pull/10128
• allow inference test to generate timeline
  • https://github.com/PaddlePaddle/Paddle/pull/10526
• CI merge with latest develop to avoid stale PR
• code review for distributed training, multi-gpu training and inference.