A python implementation of SARSA-lambda and Q-lambda are model-free reinforcement learning algorithms that learn by bootstrapping from the current estimate of the value function.