tensorflow

XuanCe supports multiple deep learning frameworks for reinforcement learning research and development. Among them, TensorFlow serves as one of the supported backends for algorithm implementation.

All APIs introduced in this section are implemented based on TensorFlow, providing high flexibility, scalability, and compatibility with TensorFlow’s computational graph and eager execution modes. Users can easily extend or modify these APIs to build custom agents, policies, or learners by leveraging TensorFlow’s powerful ecosystem, including modules such as tf.keras, tf.optimizers, and tf.data.

• agents.

• communications.

• learners.

• policies.

• representations.

• runners.

• utils.