After using Tensorflow or Keras, understanding of ND4J or DL4J is quite possible (although not easy), but with DataVec unfortunately I have lots of problems:
- completely different design than anything else widely known (eg Spark or Pandas data frames)
- poor documentation (the docs website https://deeplearning4j.konduit.ai/datavec/overview is one video + a few links to source code or examples - too little to understand it)
After using Tensorflow or Keras, understanding of ND4J or DL4J is quite possible (although not easy), but with DataVec unfortunately I have lots of problems: