Neural Network Framework

Mathematica 12: Neural Network Framework

Version 12 completes its high-level neural network framework in terms of functionality, while improving its simplicity and performance. A variety of new layers and encoders have been added, in particular, to handle sequential data such as text or audio. Importantly, a model repository is introduced, bringing a collection of pre-trained networks to be used as is, symbolically manipulated, or fine-tuned to a specific task. Finally, performance has been improved for most hardware configurations, including the possibility to train using multiple graphic cards. This framework is arguably the easiest existing tool to build neural network applications, yet with broad capabilities and without performance compromise.

Mathematica 12: Neural Network Framework
  • Access pre-trained models. »
  • Define and visualize arbitrary nets. »
  • Manipulate neural net symbolically. »
  • Access detailed training information. »
  • Measure net performances. »
  • Train a network on multiple GPUs. »
  • Train networks on text or audio data. »
  • Create attention mechanisms. »
  • Define custom recurrent layers. »
  • Generate sequences from recurrent layers efficiently. »
  • Train convolution nets on sequences. »
  • Train transformer nets. »
  • Train capsule nets. »
  • Train self-normalizing neural nets. »
  • Access reinforcement learning environments. »

Related Examples