ML in Java, YES it's possible!
summary
Claim: Java can build and run real ML pipelines today. The session demonstrates JSR‑381/VisRec producing a small image classifier saved as hotdog.dnet (trained on an M1 laptop in a few minutes), shows Deeplearning4j + ND4J for building and saving models and doing ResNet50 transfer learning (more control but verbose), and DJL wrapping TensorFlow/PyTorch with a DJL model server Docker deployment. Tribuo targets traditional enterprise ML. Precision and tooling tradeoffs are measured and the demos run end to end.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 8:54AI vs ML vs deep learning
- 15:14Java for machine learning
- 17:39VisRec for image recognition
- 27:14Deep Learning for Java
- 37:26Deep Java Library
- 43:24Tribuo for enterprise machine learning
- 46:12Choosing Java ML libraries
related talks
46mML in Java, YES it's possible!
50mML in Java, YES it's possible!
43mWhy Java and the JVM Will Dominate the Future of Machine Learning, AI, and Big Data
48mML on the JVM
58mPatterns, Predictions and Programming
27mDeveloping and Deploying ML Models with Java
02h 37mDeveloping Cloud-Native Java AI applications with DJL and LangChain4J
32mDeploying & serving optimized ML pipelines using graalvm