Highlight

• Provide a new set of API for Tensorflow and PyTorch. It is more intuitive and user-friendly.
• Support FlinkML estimator style API in Python, you can check out the quick start examples for Tensorflow 1.15 and PyTorch.
• Support iterative training with data from Flink table.
• Dependencies management of Python API is now seamlessly integrated with the dependencies management of PyFlink, detail can be found here.

What's Changed

• Bump dl-on-flink version by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/700
• Update quick start doc by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/695
• Bump hadoop-common from 2.8.0 to 2.10.1 by @dependabot in https://github.com/flink-extended/dl-on-flink/pull/697
• Bump junit from 4.12 to 4.13.1 by @dependabot in https://github.com/flink-extended/dl-on-flink/pull/688
• Bump httpclient from 4.5.6 to 4.5.13 by @dependabot in https://github.com/flink-extended/dl-on-flink/pull/684
• Recommend using virtualenv for quick start by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/703
• Rename modules from flink-ml-* to dl-on-flink-* by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/702
• Add inference example to quick start by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/705
• Run example in github action nighlty by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/707
• Include pybind as submodule by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/708
• Add github workflow to prepare release by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/709
• Remove extra space of commands in quick start by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/710
• Increase timeout waiting for AM in testing by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/712
• Fix unstable unit test AppMasterServerTest#multiRegisterNode by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/714
• Use the archive url to download flink binary by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/715
• Add checkstyle and spotless by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/718
• Refactor RoleUtils and TFUtils API by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/716
• NodeUtils set python path properties with entry file path at first by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/721
• Install dev wheels for nightly build by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/722
• AMInputFormat wait for am node finish before return from reachedEnd by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/724
• Support Tensorflow on Flink iterative batch training by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/723
• Add Python TFUtils API by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/726
• Update TensorBoardPythonRunner by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/730
• Update DL on Flink Tensorflow Python API by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/731
• Use RowCSVCoding as TF default encoding/decoding class by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/732
• Run python unit test in GitHub workflow by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/734
• Support pyFlink env configuration by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/733
• Bump Supported Tensorflow version to 2.4.x by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/736
• Update Project Copyright Header by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/737
• Introduce the New PyTorch API by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/735
• Rename PytorchContext to PyTorchContext by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/738
• Update quick start examples by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/739
• Fix prepare release workflow by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/741
• Specify protobuf version for dl-on-flink-tensorflow wheel build by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/742
• DL on Flink support FlinkML API by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/743
• Fix examples by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/745
• Avoid tight loop in NodeOperator#onEpochWatermarkIncremented by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/747
• Support adding extra cluster config properties to estimator by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/746
• Bump fastjson from 1.2.49 to 1.2.83 by @dependabot in https://github.com/flink-extended/dl-on-flink/pull/748
• Add flink ml estimator examples by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/750
• Update NodeOperator and NodeInputFormat close timeout by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/749
• Fix unstable failover test by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/751
• NodeServer#stopMLRunner should not swallow the InterruptedException by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/752
• Add Java examples for Tensorflow and PyTorch by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/753
• Add DL on Flink API docs by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/754
• NodeServer is interrupted periodically if it is not terminated on time by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/755
• Specify grpcio version upper bound by @Sxnan in https://github.com/flink-extended/dl-on-flink/pull/756

Full Changelog: https://github.com/flink-extended/dl-on-flink/compare/0.4.0...0.5.0-rc0