Releases: logicalclocks/hops-util-py
Releases · logicalclocks/hops-util-py
Release hops-util-py v0.10.0.0
- [featurestore] visualization of feature statistics
- [featurestore] support petastorm datasets
- [featurestore] support Hive partitioning
- [serving] REST API wrapper
- [serving] add support for SkLearn
- [kafka] Inference logging utilities
- [devices] nvidia-ml-py logging
- [devices] add support for ROCM
- [hdfs] Add API call to add .ipynb file
- integration tests
- bugfixes
Release hops-util-py v0.9.0.0
- Bugfixes
- New spark 2.4.0 features to the feature store: support image and avro datatypes
Release hops-util-py v0.8.0.0
Includes a python client for Hopsworks feature store and bug fixes
Release hops-util-py v0.6.0.1
Includes bugfix for publishing experiments to elastic for projects with uppercase characters in the name
Release hops-util-py v0.6.0.0
-
Added support for CollectiveAllReduceStrategy, ParameterServerStrategy, MirroredStrategy.
-
Added support for kafka.
-
Extend hdfs module to wrap large parts of the Pydoop API.
-
Introduce experiment module for launching machine learning experiments.
-
Document public API.
Release hops-util-py v0.7.0.0
- Add support for Random Search using the experiment module