This repository provides a collection of Deep Reinforcement Learning algorithms, primarily targeting researchers and practitioners in the field. It offers implementations of established algorithms for experimentation and learning, with a focus on ease of use for common RL tasks.

How It Works

The project implements Deep Reinforcement Learning algorithms using TensorFlow 1.3.0 and tflearn. The core approach involves leveraging these libraries to build and train neural network policies for agents interacting with environments, such as those provided by OpenAI Gym. This combination allows for the development of agents capable of learning complex behaviors through trial and error.

Quick Start & Requirements

• Install: pip install tensorflow==1.3.0 tflearn gym==0.9.2
• Prerequisites: Python 2.7 or Python 3.6, TensorFlow 1.3.0, tflearn, OpenAI Gym 0.9.2.
• Tested environments: Pendulum-v0 in OpenAI Gym.

Highlighted Details

• Implements Deep Deterministic Policy Gradients (DDPG).
• Code has been used to solve the inverted Pendulum task in Roboschool.
• Integrates with OpenAI Gym for environment interaction.

Maintenance & Community

The project appears to have limited recent activity. Notable usage includes contributions from @keithmgould and @janscholten. Community interaction channels are not explicitly mentioned.

Licensing & Compatibility

The repository does not explicitly state a license. This may pose compatibility issues for commercial use or integration into closed-source projects.

Limitations & Caveats

The project relies on older versions of TensorFlow (1.3.0) and Python (2.7/3.6), which may present compatibility challenges with modern hardware and software stacks. The limited scope of algorithms and environments tested could also be a constraint.