Releases: alibaba/Alink
Releases · alibaba/Alink
Alink version 1.6.2
Release version 1.6.2
Alink version 1.6.1
Optimize performance and fixed bugs.
Alink version 1.6.0
Optimize performance and fixed bugs.
Alink version 1.5.8
- Add more rules about exception.
- Add more graph algorithms.
Alink version 1.5.7
- Add pipeline model support in model stream.
- Refine online learning.
- Fixed some bugs.
Alink version 1.5.6
- Add partition in ak, csv and parquet source/sink.
- Add model stream as initial model in online learning
- Add lazyVizDive and lazyVizStatistics
- Add hbase connector
- Add custom path in resource plugin
Alink version 1.5.5
- Add datahub catalog.
- Add Kernel Dense Estimate
- Fixed some bugs
Alink version 1.5.4
- Add parquet source
- Add Export2FileSinkStreamOp https://www.yuque.com/pinshu/alink_tutorial/book_java_3_2_5
- Add Model File Path on predictop and pipeline.
- Add TFRecordDataset source/sink
- Add ONNX predict operators
Alink version 1.5.3
- Add xgboost wrapper using plugin
- Add PyTorch model predictor, close #207.
- Support pyalink on vvp. https://www.yuque.com/pinshu/alink_tutorial/flink_vvp
- Support pyalink on dsw. https://www.yuque.com/pinshu/alink_tutorial/pai_dsw
- Support pyalink on pai-designer. https://www.yuque.com/pinshu/alink_tutorial/pai_designer
- Add serializer of MTable and Tensor types
- Add code of pyalink, close #208.
- Fix #199, #193.
Alink version 1.5.2
- Improve the performance of model stream.
- Add ToTensor, ToVector, ToMTable and support tensor, vector, mtable types in csv source.
- Keras sequential operators now support string/int types; Improve plugin mechanism in TF predictor.
- Add redis to plugin and add lookup redis operator.
- PyAlink: StreamOperator print function now supports specifying port.