Request and charter for SIG JVM

sigs/jvm/CHARTER.md

@@ -0,0 +1,33 @@
+# Proposed name: SIG JVM
+
+## Objective
+
+Establish an official communication channel for discussions and suggestions related to TensorFlow in Java, or any other languages running on JVMs.
+
+Develop and maintain the TensorFlow Java client and a set of additional tools to enhance the support of machine learning development in JVM languages.
+
+## Membership
+
+Everyone interested in promoting Java or other JVM languages for training, running or serving TensorFlow models is welcome to join the group, whether it is for contributing to the code or simply for proposing ideas. To participate, request an invitation to join the mailing list. Archives of the mailing list will be publicly accessible.
+
+## Resources
+
+* Mailing list: TensorFlow SIG JVM [jvm@tensorflow.org](https://groups.google.com/a/tensorflow.org/forum/#!forum/jvm)
+* Gitter chat: [tensorflow/sig-jvm](https://gitter.im/tensorflow/sig-jvm)
+* Repositories: *might change depending on the outcome of the initial discussions*
+   * `tensorflow-jvm`: Main repository for new TensorFlow JVM-related projects
+      * `tensorflow`: Actual Java client to be moved out of the main repository
+      * `tensorflow-utils`: high-level API and utilities on top of TensorFlow Java client
+      * `tensorflow-models`: pretrained models from TensorFlow adapted to Java
+
+## Contacts
+
+* Project lead: Karl Lessard [@karllessard](https://github.com/karllessard) - karl at kubx dot ca
+* Second for the leader: Christian Tzolov [@tzolov](https://github.com/tzolov) - christiantzolov at gmail
+* TensorFlow team technical contact: James Ring [@sjamesr](https://github.com/sjamesr) - sjr at google
+* For administrative questions, contact Edd Wilder-James [@ewilderj](https://github.com/ewilderj) - ewj at google
+
+## Code of Conduct
+
+As with all forums and spaces related to TensorFlow, SIG JVM is subject to
+the [TensorFlow Code of Conduct](https://github.com/tensorflow/tensorflow/blob/master/CODE_OF_CONDUCT.md).

sigs/jvm/SIG-request.md

@@ -0,0 +1,57 @@
+# Request for SIG
+
+## What is this group for?
+
+Java is still one of the (if not the) most popular programming language found in small to large enterprises around the world. For TensorFlow, it makes it a strategic choice to reach a large audience of programmers who are interested to embed machine learning algorithms in their applications, while preserving their current technology stack and expertise.
+
+TensorFlow already has a Java client out-of-the-box, which was originaly conceived to support inference on Android devices. Many contributions were made thereafter to make it a good choice even for building, training and serving models, but there is still a lot of work to be done before it reaches a level of maturity comparable to Python.
+
+The purpose of this group is to establish an official communication channel for discussions and suggestions related to TensorFlow support in Java or any other JVM-based languages, such as Kotlin or Scala.
+
+In addition to the current Java client, the group will develop and maintain new high-level artifacts, with their own repository and release life cycle, to provide a rich set of tools to the developers.
+
+## Who will be part of it?
+
+* Group leader: Karl Lessard
+* Second for the leader: Christian Tzolov
+
+Anyone interested to help are invited to join by requesting access to the mailing list, whether they are developers who wants get involve in the code or just individuals who wants to take part of the discussions.
+
+Members of the [unofficial Java group](https://groups.google.com/forum/#!forum/tensorflow-java-dev-unofficial) might be interested to join this new official one, as the former will be closed.
+
+## What initial problems will the group tackle?
+
+* Providing a new set of high-level API and features for ML development in Java or other languages running on JVMs
+    * Having their own repositories under the TensorFlow organization, these artifacts will be released independently of TensorFlow, which will speed up the deliveries of new features to the end-users.
+    * Pull requests could be merged faster by allowing some members of the community to do code reviews
+* Establish an official communication channel for discussion related to TensorFlow on JVM
+    * Right now, discussions are spread on different channels, such as the unofficial google group or on top of different GitHub issues
+* Work with Google team to extract the current Java client out of the main repository
+    * As it has been proposed in [RFC: Modular Tensorflow](https://github.com/tensorflow/community/pull/77), there is an interest to move portions of TensorFlow unrelated to the core out of the main repository.
+    * The actual Java client is a good candidate for gaining his own repository.
+
+The outcome of the initial discussions of this group will be to identify what features should be addressed first. Here is some suggestions:
+
+* Eager execution mode in the Java client
+* `tensorflow-utils` artifact: Utility library on top of the core client to simplify usage of TensorFlow in Java, such as multi-dimensional array accessors.
+* `tensorflow-models` artifact(s): Modelisation in Java of pretrained TensorFlow models.
+
+## What modes of communication do you intend to use?
+
+* A dedicated mailing list backed by Google Groups
+* A Slack channel
+* VC calls on Google Hangout could be organized on demand
+* StackOverflow
+
+## Launch plan
+
+1. Exposing the present plan and charter to the community for review
+2. SIG set up with initial group members
+3. SIG added to community pages on tensorflow.org
+4. Leader starts off mailing list discussions about initial work items
+5. Creation of repository(ies) for the inception of the first high-level libraries to be develop by this group 
+
+# Charter
+
+Here's the link to the group [charter](CHARTER.md).
+