73aa760c0f
* Increase default timeout on Spark tests Signed-off-by: Alex Black <blacka101@gmail.com> * #8840 disable deeplearning4j-nlp-korean module for scala 2.12 Signed-off-by: Alex Black <blacka101@gmail.com> * Fix for change-scala-versions.sh Signed-off-by: Alex Black <blacka101@gmail.com> * CUDA test fixes + more timeout issues Signed-off-by: Alex Black <blacka101@gmail.com> * More CUDA Signed-off-by: Alex Black <blacka101@gmail.com> * Small fix for cuDNN subsampling + same mode Signed-off-by: Alex Black <blacka101@gmail.com> * Flaky test fix Signed-off-by: Alex Black <blacka101@gmail.com> * Reduce memory requirements for ValidateCuDNN BN test Signed-off-by: Alex Black <blacka101@gmail.com> * Fix slow/ineffirient scalnet tests Signed-off-by: Alex Black <blacka101@gmail.com> * Increase timeouts to avoid failures if CI machines are slower than expected Signed-off-by: Alex Black <blacka101@gmail.com> * Ignore flaky test (issue #8849) and increase timeout for slow CI downloads Signed-off-by: Alex Black <blacka101@gmail.com> |
||
|---|---|---|
| .. | ||
| project | ||
| src | ||
| .gitignore | ||
| .scalafmt.conf | ||
| .travis.yml | ||
| LICENSE | ||
| README.md | ||
| build.sbt | ||
| buildmultiplescalaversions.sh | ||
| pom.xml | ||
| sbt-pom.xml | ||
README.md
ScalNet
ScalNet is a wrapper around Deeplearning4J emulating a Keras like API for deep learning.
ScalNet is released under an Apache 2.0 license. By contributing code to this repository, you agree to make your contribution available under an Apache 2.0 license.
ScalNet is STILL ALPHA and we are open sourcing this in an attempt to get feedback.
Come in to gitter if you are interested in learning more.
Prerequisites
- JDK 8
- Scala 2.11.+ or 2.10.x
- SBT and Maven
How to build
ScalNet depends on Deeplearning4j and ND4J
sbt
ScalNet uses sbt, but due to resolving issues, you must have Maven available to copy some nd4j-native dependencies in your classpath, in order to run the examples.
This is automatically done in build.sbt and you don't need to do anything besides having maven installed.
If you use sbt in your own project, you will probably have to proceed the same way. When ScalNet will be using releases instead of snapshots, this won't be necessary anymore.
To build, use:
$ sbt package
Alternatively, for some quick testing or usage in Jupyter for example, run:
$ sbt assembly
To obtain a JAR file with all needed dependencies.
See the official sbt documentation for more on how to use sbt.
Maven
Althought Maven is mainly used for release management, you can use the provided pom.xml to import ScalNet as a Maven project.
Target for scala 2.11
$ change-scala-versions.sh "2.11"
$ mvn package
Target for scala 2.10
$ change-scala-versions.sh "2.10"
$ mvn package
How to use
sbt
libraryDependencies ++= "org.deeplearning4j" % "scalnet_2.11" % "0.9.2-SNAPSHOT"
Maven
<dependency>
<groupId>org.deeplearning4j</groupId>
<artifactId>scalnet_2.11</artifactId>
<version>0.9.2-SNAPSHOT</version>
</dependency>
Getting started
ScalNet uses a Keras like API, wrapping deeplearning4j to make it more easier to start with.
Also, since you can call Java code from Scala, you can still use everything from deeplearning4j.
To see what ScalNet has to offer, run one of the [examples] (https://github.com/eclipse/deeplearning4j/tree/master/scalnet/src/test/scala/org/deeplearning4j/scalnet/examples) it ships with.
Please note that those examples are not state-of-the-art in any way, they're just enough to get you started.