ruqu-qarlp

Quantum-Assisted Reinforcement Learning Policy - exploratory quantum RL implementation with variational circuits.

Part of the ruQu quantum computing suite by ruv.io.

Features

• Variational Quantum Circuits - Parameterized quantum circuits for policy representation
• Policy Gradient Methods - REINFORCE and actor-critic implementations
• Quantum State Encoding - Amplitude and angle encoding strategies
• Experience Replay - Quantum-enhanced experience buffer
• Gradient Estimation - Parameter-shift rule for quantum gradients

Installation

[dependencies]
ruqu-qarlp = "0.1"

Quick Start

use ruqu_qarlp::{QuantumPolicy, PolicyConfig, Environment};

let config = PolicyConfig {
    n_qubits: 4,
    n_layers: 2,
    learning_rate: 0.01,
    ..Default::default()
};

let mut policy = QuantumPolicy::new(config)?;
let env = Environment::cartpole();

// Training loop
for episode in 0..1000 {
    let trajectory = policy.rollout(&env)?;
    policy.update(&trajectory)?;
}

License

MIT License - see LICENSE