TRFL (pronounced "truffle") is a library built on top of TensorFlow that exposes several useful building blocks for implementing Reinforcement Learning agents.

TRFL can be installed from pip directly from github, with the following command: pip install git+git://github.com/deepmind/trfl.git

TRFL will work with both the CPU and GPU version of tensorflow, but to allow for that it does not list Tensorflow as a requirement, so you need to install Tensorflow and Tensorflow-probability separately if you haven't already done so.

import tensorflow as tf
import trfl

# Q-values for the previous and next timesteps, shape [batch_size, num_actions].
q_tm1 = tf.constant([[1, 1, 0], [1, 2, 0]], dtype=tf.float32)
q_t = tf.constant([[0, 1, 0], [1, 2, 0]], dtype=tf.float32)

# Action indices, pcontinue and rewards, shape [batch_size].
a_tm1 = tf.constant([0, 1], dtype=tf.int32)
pcont_t = tf.constant([0, 1], dtype=tf.float32)
r_t = tf.constant([1, 1], dtype=tf.float32)

loss, q_learning = trfl.qlearning(q_tm1, a_tm1, r_t, pcont_t, q_t)

loss is the tensor representing the loss. For Q-learning, it is half the squared difference between the predicted Q-values and the TD targets, shape [batch_size].

Extra information is in the q_learning namedtuple, including q_learning.td_error and q_learning.target.

Most of the time, you may only be interested in the loss:

loss, _ = trfl.qlearning(q_tm1, a_tm1, r_t, pcont_t, q_t)

# You can also do this, which returns the identical `loss` tensor:
loss = trfl.qlearning(q_tm1, a_tm1, r_t, pcont_t, q_t).loss

reduced_loss = tf.reduce_mean(loss)

optimizer = tf.train.AdamOptimizer(learning_rate=0.1)
train_op = optimizer.minimize(reduced_loss)

All loss functions in this module return the loss tensor and extra information using the above convention.

Different functions may have different extra fields. Check the documentation of each function for more information.

Check out the full documentation page here.