lindong28 commented on a change in pull request #493:
URL: https://github.com/apache/flink-web/pull/493#discussion_r775789628
##########
File path: downloads.md
##########
@@ -200,6 +200,20 @@ You can add the following dependencies to your `pom.xml`
to include Apache Flink
The `statefun-sdk` dependency is the only one you will need to start
developing applications.
The `statefun-flink-harness` dependency includes a local execution environment
that allows you to locally test your application in an IDE.
+### Apache Flink ML
+
+You can add the following dependencies to your `pom.xml` to include Apache
Flink ML in your project.
+
+```xml
+<dependency>
+ <groupId>org.apache.flink</groupId>
+ <artifactId>flink-ml-uber_2.12</artifactId>
+ <version>{{ site.FLINK_ML_VERSION_STABLE }}</version>
+</dependency>
+```
+
+The `flink-ml-uber_2.12` is a whole set of dependencies required for both
using and developing algorithms. To only use existing algorithms, users could
also only depend on `flink-ml-lib_2.12` and to only develop new algorithms,
user could also only depend on `flink-ml-core` and `flink-ml-iteration` (if the
algorithm is iterable).
Review comment:
This paragraph seems a bit difficult to read.
Could we use the following words:
`flink-ml-uber_2.12` provides all the dependencies needed to use Flink ML.
Alternatively, advanced users could use the following artifacts to import a
minimal set of Flink ML dependencies for their target use-cases:
- Use artifact `flink-ml-core` in order to develop custom ML algorithms.
- Use artifacts `flink-ml-core` and `flink-ml-iteration` in order to develop
custom ML algorithms which require iteration.
- Use artifact `flink-ml-lib_2.12` in order to use the off-the-shelf ML
algorithms from Flink ML.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
