Watson Studio Local release versions on various platforms
Learn about the most recent releases of Watson Studio Local and the respective features implemented.
| Platform | Watson Studio Local version | Release date | Comments |
|---|---|---|---|
| POWER | 1.2.3.4 1.2.3.2 1.2.3.1 |
March 2021 March 2020 April 2019 |
|
| x86-64 | 1.2.3.5 1.2.3.3 1.2.3.1 |
March 2021 June 2020 March 2019 |
|
| HDP | 1.2.3.0 | October 2018 | |
| ICP for Data Science 2.1.0.2 | 1.2.0.3 | June 2018 | |
| ICP for Data Science 2.1.0.2 on POWER | 1.2.1.0 | September 2018 | |
| Linux on z | 1.2.0.3 | October 2018 | Supports RHEL and Ubuntu. GPU, R Shiny app, and Brunel visualization not supported. |
| IIAS | 1.2.0.3 | November 2018 |
Version 1.2.3.1 features
- x86 features
-
- Installation configuration parameters secure the Watson Studio Local platform and users can remove the sshd and netcat services from applicable pods
- Watson Studio Local platform is secured with updates to various components across all docker images
- Cloudera Distribution for Hadoop version 5.16 and 6.0.1 support
- POWER features
- Silent installation support
Version 1.2.3 features
- SAML support
- NFS support
- overlay2 support
- Track usage scripts
- Hive support in HDP Version 3.0.1
- R jobs
- Hadoop batch score and evaluate models
- New Hadoop utility methods for models
- Separate usage percentages in Admin Console
- Admin Console pod search
- Add Spark jars
Version 1.2.2 features
- IBM Data Science Experience renamed to IBM Watson Studio
- IBM Deployment Manager renamed to IBM Watson Machine Learning
- Unicode (UTF-8) support for data sets
- Access to HDFS data in HDP 3.0
- Cluster troubleshooting and repair utility
- Installation performance monitoring utility
- OpenShift support (Beta)
Version 1.2.1 features
- Jupyter terminal (
) - Any DSX Local user can install a library or package in the root conda environment and save it in a custom image
- My Images page
- Push custom images to a Hadoop cluster for remote jobs
- Model groups
- Roles for a project release: Deployment Admin, Admin, Developer, and Viewer
- Project release Workers tab
- Remote deployments (Beta)
- Git merge conflict resolution
- Separate Commit and Push
- Restricted libraries and a Libraries page
- Test a data source connection
- Preview remote data sets for all supported JDBC data sources and custom JDBC data sources
- Browse schema for most table types
- SQL Query object type
- Microsoft SQL Server support
- Hyperledger Composer support
- Refine data on a remote JDBC or HDFS data set
- Data Refinery jobs
- SSHD service panel in the Admin Console
- Conda Management panel in the Admin Console
- Search, filter, and sort projects. Bookmark web browser URLs for project searches
- Support for Cognos Dashboards Embedded add-on
- Support for Watson Explorer oneWEX add-on
- Support for Spark Canvas add-on (Beta)
Version 1.2.0 features
- Model Management and Deployment; requires license and deployment nodes
- RHEL docker and RHEL packages now prerequisites instead of being included
- Batch score or evaluate models created in RStudio
- Batch score WML models with CSV files
- Associate scripts with a model
- Support for DSX Hadoop Integration clusters. Hadoop integration page
- Decision Optimization Community Edition
- Create data source to a custom JDBC
- Apache Spark version 2.2.1 option
- Support for unmanaged resources
- All Active Environments page
- Support for BitBucket Git repo
- Commit history and tags for Git
- Select files in project push
- Import a clone of an external Git repository
- Script editor
- Test script as API option to generate a curl command
- IBM Data Platform Manager is now called the Admin Console
- Run configuration scripts from the Admin Console
- Audit record of user login attempts
- SPSS Modeler worker
- Stop jobs
- Data Refiner (Beta)
Version 1.1.3.1 features
- Jobs
- Image management
- Create data sources for IBM Big SQL, Cloudera Hive, Cloudera HDFS data, and a non-secure HDP cluster
- Runtimes now called Environments. Environments can be associated with GPUs
- Support for GPUs by NVIDIA in Azure, AWS and Softlayer. Support for GPU on Red Hat Enterprise Linux x86
- Batch score or evaluate a model from the assets page, which creates a script that can be scheduled to run as a job
- Model versioning
- Support for model types: scikit-learn with pickle format, scikit-learn with joblib format, XGBoost, Keras TensorFlow, and WML
- Publish models to the Model Management dashboard
- Publish assets
- Assets now saved in the cluster file system
- Load balancer option in wdp.conf file install
- Preview an R Shiny app
- Automatic add-on install
- Support for the SPSS Modeler Flows add-on with advanced visualizations
Version 1.1.2.2 features
- Library projects
- Project runtimes that can reserve CPU cores and RAM memory. View all runtimes page
- Create data sources for Hortonworks HDP: HDFS - HDP and Hive - HDP. Python utility functions to transfer files between the HDP cluster and the DSX Local cluster
- Deploy models that can batch score remote data sets to an output CSV file
- Support for model types: scikit-learn (for developing machine learning in a Python notebook) and XGBoost libraries
- Import and score third-party vendor models using the Custom Batch or Custom Online option
- Train models on a remote Spark using Livy REST APIs
- Publish models to the Watson ML service
- Project tree view (
) - Add more than one GitHub access token
- Support for Flow add-on
- Support for H20 Flow add-on