google-deepmind/trfl
DeepMind's TensorFlow library providing reinforcement learning building blocks and loss functions for training RL agents.
Not currently ranked — collecting fresh signals.
star history
TRFL is a reinforcement learning library built on top of TensorFlow, developed by Google DeepMind. It exposes reusable algorithmic building blocks for implementing RL agents, including loss functions such as Q-learning that compute TD errors and targets. The library returns both loss tensors and auxiliary information that can be differentiated to derive RL updates, allowing integration into standard TensorFlow training pipelines.
Frequently asked
- What is google-deepmind/trfl?
- DeepMind's TensorFlow library providing reinforcement learning building blocks and loss functions for training RL agents.
- Is trfl open source?
- Yes — google-deepmind/trfl is open source, released under the Apache-2.0 license.
- What language is trfl written in?
- google-deepmind/trfl is primarily written in Python.
- How popular is trfl?
- google-deepmind/trfl has 3.1k stars on GitHub.
- Where can I find trfl?
- google-deepmind/trfl is on GitHub at https://github.com/google-deepmind/trfl.