2018 05 17

dongzhihong

• 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

Lei Wang

• 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

kexinzhao

• 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

Bai Yifan

• Put detection op in a new directory
  • https://github.com/PaddlePaddle/Paddle/pull/10595
• Write a introduction to paddlecloud for vis
  • http://wiki.baidu.com/pages/viewpage.action?pageId=491475037
• [WIP] roi_pool bug fix
  • https://github.com/PaddlePaddle/Paddle/pull/10700

qiaolongfei

• fluid high level api

  • Update trainer api https://github.com/PaddlePaddle/Paddle/pull/10653
  • user need to set feed order for Trainer.train and Trainer.test https://github.com/PaddlePaddle/Paddle/pull/10679

• Fluid C++ trainer lib and demo(support Abacus). https://github.com/PaddlePaddle/Paddle/issues/10574

  • Add cpp trainer lib and demo https://github.com/PaddlePaddle/Paddle/pull/10681
  • add boost to inference lib https://github.com/PaddlePaddle/Paddle/pull/10660
  • code clean https://github.com/PaddlePaddle/Paddle/pull/10687
  • add pybind.h to inference lib for static link https://github.com/PaddlePaddle/Paddle/pull/10690
  • add mkldnn to paddle lib https://github.com/PaddlePaddle/Paddle/pull/10705

tonyyang-svail

CI

1. move test_network_with_dtype out of test_all_ops: https://github.com/PaddlePaddle/Paddle/pull/10675
2. speed up test label semantic roles: https://github.com/PaddlePaddle/Paddle/pull/10718

API:

1. fit a line: https://github.com/PaddlePaddle/Paddle/pull/10301

Chenxi

• 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

wanghaoshuang

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

Dang qingqing

• Fixed-point quantization
  • Develop quantize transpiler. https://github.com/PaddlePaddle/Paddle/pull/10693
  • [WIP] update doc.
• Face Detection:
  • Discuss and make plan with Face Team.
  • [WIP] develop face detection model.
• Others:
  • Discuss and make place with OCR Team
  • follow up issue from VIS:
    • https://github.com/PaddlePaddle/Paddle/issues/10638
    • https://github.com/PaddlePaddle/Paddle/issues/10627 (by @yifan)
  • Fix bug in reshape op:
    • https://github.com/PaddlePaddle/Paddle/pull/10713
  • Refine smooth L1 loss.
    • https://github.com/PaddlePaddle/Paddle/pull/10641
• Code review:
  • https://github.com/PaddlePaddle/Paddle/pull/10709
  • https://github.com/PaddlePaddle/Paddle/pull/10606
  • https://github.com/PaddlePaddle/Paddle/pull/10595
  • https://github.com/PaddlePaddle/Paddle/pull/10519

Yan Xu

• [review] overlap send op and backward ops, https://github.com/PaddlePaddle/Paddle/pull/10550
• manylinux build refine
  • https://github.com/PaddlePaddle/Paddle/pull/10697
  • https://github.com/PaddlePaddle/Paddle/pull/10694
  • https://github.com/PaddlePaddle/Paddle/pull/10585
• pserver bug fix, https://github.com/PaddlePaddle/Paddle/pull/10626

Tang Wei

• [WIP] PServer checkpoint
• https://github.com/PaddlePaddle/Paddle/issues/10376
• https://github.com/PaddlePaddle/Paddle/pull/10532

nicky

• Add support to SNPE converter script - Use Gemm to replicate MatMul, still encounter other error
• Try fluid to caffe converter with alexnet model to integrate to demo
• Simplify Fluid API with Machine Translation

sidgoyal78

• PR: Recommendation system impl + changes to trainer API: https://github.com/PaddlePaddle/Paddle/pull/10535
• PR: Adding FP16 support to load_combine_op: https://github.com/PaddlePaddle/Paddle/pull/10601
• Discussion with Nicky and Daming on new API and changes to Trainer.

Wei Xing

documents:

• updated some new fluid api documents to paddlepaddle.org

  • https://github.com/PaddlePaddle/Paddle/pull/10637

• Update all *.whl links to latest version in install using pip for 0.12.0 and develop documentation.

  • https://github.com/PaddlePaddle/Paddle/pull/10712

• deadlinks fixed

  • https://github.com/PaddlePaddle/Paddle/pull/10594

[WIP] hsigmoid operator

tensor-tang

• Review

  • [Merged] Update mklml version https://github.com/PaddlePaddle/Paddle/pull/10571

  • [Merged] Reusing of softmax mkldnn primitives: https://github.com/PaddlePaddle/Paddle/pull/10576

  • [WIP] Update activations for MKL-DNN https://github.com/PaddlePaddle/Paddle/pull/10597

• Compare Inference Perf BTW CPU and MKLDNN https://github.com/PaddlePaddle/Paddle/issues/10651

• [WIP] Enable MKLDNN inference test https://github.com/PaddlePaddle/Paddle/pull/10701

Yan Chunwei

• Inference

  • init Inference top APIs
  • feature/analysis node representation
  • OpConverter change BlockDesc to proto::BlockDesc
  • Feature/engine support load ONNX model
  • Feature/Inference data flow graph for analysis
  • Inference Working Status WIKI

• CE

  • Add code format to README.md

• PR https://github.com/PaddlePaddle/Paddle/wiki/Inference-Working https://github.com/PaddlePaddle/Paddle/pull/10495#pullrequestreview-120091219 https://github.com/PaddlePaddle/Paddle/pull/10657#pullrequestreview-120114309 https://github.com/Superjomn/paddle-ce-latest-kpis/pull/10#pullrequestreview-120479422 https://github.com/PaddlePaddle/Paddle/pull/10690#pullrequestreview-120549081

guochaorong

• Paddle development and test standard environment

  • entrance
    • Docker: http://wiki.baidu.com/pages/viewpage.action?pageId=491475809
    • vm: http://wiki.baidu.com/pages/viewpage.action?pageId=493068823

• Cpu slow problem

  • Issue：https://github.com/PaddlePaddle/Paddle/issues/10639#issuecomment-389455809  

• Set CPU monitor for CE models

  • Issue：https://github.com/PaddlePaddle/continuous_evaluation/issues/46
  • pr：https://github.com/Superjomn/paddle-ce-latest-kpis/pull/11/files  

• support multi card for CE models

  • issue: https://github.com/PaddlePaddle/continuous_evaluation/issues/45
  • pr：https://github.com/Superjomn/paddle-ce-latest-kpis/pull/14

Yan Chunwei

• Inference

  • init Inference top APIs
  • feature/analysis node representation
  • OpConverter change BlockDesc to proto::BlockDesc
  • Feature/engine support load model
  • Feature/Inference data flow graph for analysis
  • Inference Working Status WIKI

• CE

  • Update README.md

• PR https://github.com/PaddlePaddle/Paddle/wiki/Inference-Working https://github.com/PaddlePaddle/Paddle/pull/10495#pullrequestreview-120091219 https://github.com/PaddlePaddle/Paddle/pull/10657#pullrequestreview-120114309 https://github.com/Superjomn/paddle-ce-latest-kpis/pull/10#pullrequestreview-120479422 https://github.com/PaddlePaddle/Paddle/pull/10690#pullrequestreview-120549081

varunarora

• Designed a new prototype for PaddlePaddle.org homepage; currently under review
• Assisted Nicky in solving Resnet op issues to convert to DLC
• Began writing new documentation for ProgramDesc overview
• Setup an AWS node for VisuaLDL playground for Daming
• Added few more notes for installation to paddle-onnx (https://github.com/PaddlePaddle/paddle-onnx/pull/48)
• [non-Paddle investigation of SVAIL speech infra problems]

daming-lu

• Simplify Fluid API:
  • https://github.com/PaddlePaddle/Paddle/pull/10738
  • https://github.com/PaddlePaddle/Paddle/pull/10301
  • https://github.com/PaddlePaddle/Paddle/pull/10674
• VisualDL:
  • setup playground server:
    • ec2-13-250-98-219.ap-southeast-1.compute.amazonaws.com