0.1.10

Changelog

v0.1.10 (21/08/2020)

• [bug] Fix do_bias_correction for Adam #3501
  修复 Adam 在 do_bias_correction 为 True 时的 bug

v0.1.9

0.1.9

v0.1.9 (20/08/2020)

cuDNN integration enhancement and bug fixes

• [bug] Fix cudnn conv math type fallback #3460

  设置 cudnn conv math type

• [enhancement] Add cuDNN implementation for bias_add #3489

  增加 bias add 调用 cudnn 的实现

• [enhancement] Remove useless code in cudnn conv util #3480

  删除 cudnn conv util 中的冗余代码

Eager execution enhancement

• [feature] manipulate lazy interface blobs in eager #3226

  在 eager 中操作 lazy blob

CPU-only support

纯 CPU 支持的相关改动

• [feature] Dev pure cpu #3398
• [enhancement] Dev bn ln cpu #3431
• [feature] bert cpu #3476
• [enhancement] Dev Layer Norm CPU and Batch Norm CPU training #3433
• [bug] fix some bug in dev pure cpu #3410
• [enhancement] conv2d cpu support groups #3488
• [enhancement] Dev bn training cpu #3445

Preprocessing

数据预处理的改动

• [feature] image.random_crop #3442
• [enhancement] Refactor image resize #3450

Op/Kernel optimizations

• [enhancement] Optimize transpose performance #3487

  优化 transpose 性能

• [enhancement] Refactor slice op #3444

  优化 slice op

• [feature] Fused softmax kernel #3496

  融合版本的 softmax

• [enhancement] Optimize conv performance #3477

  优化 conv 性能

System

• [bug] Add B21SubTskGphBuilder #3360

  修复多机消耗更多内存的 bug

Python front end

• [enhancement] Add typing.List global function return annotation #3443

  增加 oneflow function 的返回值签名

Documentation

增加/修订文档

• [enhancement] add api docs #3483
• [enhancement] Dev add docstring2 #3305

Development and tooling

开发 / Debug 工具相关的修改

• [feature] Dev boxing log #3466

  增加 boxing 日志

• [feature] add dqn to model tests #3473

  DQN 加入集成测试

• [enhancement] py fmt respects gitignore #3492

• [feature] Add OF_CUDA_CHECK/OF_CUDNN_CHECK/OF_CUBLAS_CHECK/OF_CURAND_CHECK #3446

  优化 CUDA CHECK 的错误信息

0.1.8-benchmark1: Optimize transpose performance (#3487)

Former-commit-id: 809793c4c7af831de8116a752c9874c75a6c9bb2

0.1.8-benchmark: Add ofrecord_image_classification_reader (#3419)

* ofrecord_image_classification_reader

* refine

* use num_decode_threads_per_machine

* Specify image_feature_name and label_feature_name

* Extract LoadWorker and DecodeWorker

* Remove image_feature_name/label_feature_name default value

* underlaying_=>base

* set decode_buffer_size_per_thread to 32

Former-commit-id: 526766a72e7bbb1ebb1897c8dca90c254b1576cf

v0.1.8: Final Cut (#3354)

* bump version

* rm pip install

Co-authored-by: tsai <caishenghang@1f-dev.kbaeegfb1x0ubnoznzequyxzve.bx.internal.cloudapp.net>
Former-commit-id: f1e94dec265dfc4bfd7e7aa66dfac2525a3910e4

v0.1.7

bump version (#3348)

Co-authored-by: tsai <caishenghang@1f-dev.kbaeegfb1x0ubnoznzequyxzve.bx.internal.cloudapp.net>
Former-commit-id: 9e3dae5b910b8ab42976e4ffe6d0d9af636f55f8

v0.1.6

bump version (#3326)

Co-authored-by: tsai <caishenghang@1f-dev.kbaeegfb1x0ubnoznzequyxzve.bx.internal.cloudapp.net>
Former-commit-id: 1e535df35b533c6d0329ba5daa0a6ba89e769593

v0.1.5

bump version (#3317)

Co-authored-by: tsai <caishenghang@1f-dev.kbaeegfb1x0ubnoznzequyxzve.bx.internal.cloudapp.net>
Former-commit-id: 2b651408e19af73cf2407215eb55d789a73ef9e7

version v0.1.4

bump version (#3291)

Former-commit-id: 7925cfcb887dd6bd41e3dd3e5570bec7bddca970

version v0.1.3

bump version 0.1.3 (#3281)

Former-commit-id: bb373a6f3a27ce9ad276f9aae5149584b1663ded