torchafct.rst

Adaptive Function Pytorch

First of all, it is important to understand the adaptive function in MLPro-SL, which can be found on :ref:`this page <target_bf_sl_afct>`.

In this functionality, we integrate PyTorch into MLPro-SL, thus PyTorch functionalities can be reused in MLPro-SL.

At the moment, we provide only PyTorch-based multilayer perceptron in the pool, but recurrent neural networks and transformers are planned to be included in the next version shortly.

Cross Reference

• :ref:`Howto RL-MB-001: Train and Reload Model Based Agent (Gym) <Howto MB RL 001>`
• :ref:`Howto RL-MB-002: MBRL with MPC on Grid World Environment <Howto MB RL 002>`
• :ref:`API Reference: MLPro-SL Pool of Objects <target_api_pool_sl>`