This is the tenative roadmap for DJL through 2021. If there is something you feel is missing from this roadmap, please raise an issue so we can discuss it.
- Quarkus Integration - Integrate with Quarkus where users can get a smaller memory profile and faster startup time
- Big Data Ecosystem Integration - Better integration with the JVM Big Data libraries including Apache Spark
- Dive into Deep Learning - Continue adapting the Dive into Deep Learning book to Java and DJL
- Model Zoo Enhancement - Improve the Model Zoo design to promote the standardization and ecosystem of Translators
- New Models
- NLP - Include new NLP models that can be used out of the box on various applications
- Tabular - Create standard tabular models and improve the creation and use of tabular datasets
- RL - Add new models for reinforcement learning and add RL specific improvements to the DJL API
- GAN - Create initial GAN support