https://github.com/andi611/dqn-deep-q-network-atari-breakout-tensorflow

Training a vision-based agent with the Deep Q Learning Network (DQN) in Atari's Breakout environment, implementation in Tensorflow.
https://github.com/andi611/dqn-deep-q-network-atari-breakout-tensorflow

agent atari atari-breakout breakout deep-q-learning deep-q-learning-network deep-q-network deep-qlearning-algorithm deep-qnetworks dqn mlds mlds17-machine-learning-course mlds2018spring openai-gym openai-gym-agents openai-gym-environments openai-gym-solutions reinforcement-learning tensorflow tensorflow-models

Last synced: about 2 months ago
JSON representation

Training a vision-based agent with the Deep Q Learning Network (DQN) in Atari's Breakout environment, implementation in Tensorflow.

• Host: GitHub
• URL: https://github.com/andi611/dqn-deep-q-network-atari-breakout-tensorflow
• Owner: andi611
• License: mit
• Created: 2018-11-19T07:36:24.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2018-12-12T08:18:20.000Z (over 6 years ago)
• Last Synced: 2025-03-27T04:33:13.995Z (2 months ago)
• Topics: agent, atari, atari-breakout, breakout, deep-q-learning, deep-q-learning-network, deep-q-network, deep-qlearning-algorithm, deep-qnetworks, dqn, mlds, mlds17-machine-learning-course, mlds2018spring, openai-gym, openai-gym-agents, openai-gym-environments, openai-gym-solutions, reinforcement-learning, tensorflow, tensorflow-models
• Language: Python
• Size: 50.3 MB
• Stars: 16
• Watchers: 1
• Forks: 8
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Reinforcement Learning: Deeq Q Learning Network (DQN) Agent playing Atari Breakout

* Training a vision-based agent with the Deep Q Learning Network (DQN) in Atari's Breakout environment, implementation in Tensorflow.



## Environment

* **< Python 3.7 >**

* **< [OpenAI Gym](https://github.com/openai/gym) >**

	- Install the OpenAI Gym Atari environment:

	`$ pip3 install opencv-python gym "gym[atari]"`

	- Atari environment used: `BreakoutNoFrameskip-v4`

* **< [Tensorflow r.1.12.0](https://www.tensorflow.org/) >**

## Implementation

* Deep Q Learning Network with the following improvements:

	- **Experience Replay**

	- **Fixed Target Q-Network**

	- **TD error loss function** with: *Qtarget = reward + (1-terminal) * (gamma * Qmax(s’)
)*

* DQN network Settings (in agent_dqn.py):

![](https://github.com/andi611/Reinforcement-Learning-DQN-Deep-Q-Learning-Atari-Breakout/blob/master/model/dqn_best_setting.png)

## File Description

```

.

├── ./

|   ├── agent_dqn.py ─────────── DQN model

|   ├── atari_wrapper.py ─────── Atari wrapper

|   ├── environment.py ───────── Gym wrapper

|   ├── runner.py ────────────── Main program for training and testing

|   └── Readme.md ────────────── This file

└── model/

	├── dqn_learning_curve_compare.png ──────── Figure 1  

	├── dqn_best_setting.png ────────────────── Figure 2

	├── dqn_learning_curve.png ──────────────── Figure 3

	├── checkpoint ──────────────────────────── Tensorflow model check point

	├── model_dqn-25581.data-00000-of-00001 ─── Tensorflow model data

	├── model_dqn-25581.meta ────────────────── Tensorflow model meta

	└── model_dqn-25581.index ───────────────── Tensorflow model index

```

## Usage

* Traing the DQN Agent: `$ python3 runner.py --train_dqn`

* Testing the DQN Agent: `$ python3 runner.py --test_dqn`

* Testing the DQN Agent with **gameplay rendering**: `$ python3 runner.py --test_dqn --do_render`

## Learning Curve

* Single learning curve:

![](https://github.com/andi611/Reinforcement-Learning-DQN-Deep-Q-Learning-Atari-Breakout/blob/master/model/dqn_learning_curve.png)

* With different plotting window:

![](https://github.com/andi611/Reinforcement-Learning-DQN-Deep-Q-Learning-Atari-Breakout/blob/master/model/dqn_learning_curve_compare.png)