diff --git a/README.md b/README.md index bd10f4a..53ea2c7 100644 --- a/README.md +++ b/README.md @@ -23,34 +23,39 @@ Following asynchronous algorithms are removed now in this repo but can be found # Curves > Curves for CartPole are trivial so I didn't place it here. And there isn't any fixed random seed. The curves are generated in the same manner as OpenAI baselines (one run and smoothed by recent 100 episodes) ## DQN -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/DQN-breakout.png) +> BreakoutNoFrameSkip-v4 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/dqn_pixel_atari-180407-01414.png.png) + +## Categorical DQN +> BreakoutNoFrameskip-v4 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/categorical_dqn_pixel_atari-180407-094006.png.png) + +## Quantile Regression DQN +> BreakoutNoFrameskip-v4 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/quantile_regression_dqn_pixel_atari-180407-01604.png.png) ## A2C -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/A3C-Pong.png) +> BreakoutNoFrameskip-v4 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/a2c_pixel_atari-180407-92711.png.png) + +## N-Step Q-Learning +> BreakoutNoFrameskip-v4 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/dqn_pixel_atari-180407-01414.png.png) ## DDPG -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/DDPG.png) +> RoboschoolHopper-v1 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/ddpg_continuous-180407-234141.png.png) ## PPO -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/P3O.png) +> RoboschoolHopper-v1 +![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/ppo_continuous-180408-002056.png.png) ## Action Conditional Video Prediction - ![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/ACVP.png) **Left**: One-step prediction **Right**: Ground truth -Prediction is sampled after 110K iterations and I only implemented one-step training - -## Categorical DQN - -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/CategoricalDQN.png) - -## Quantile Regression DQN -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/QRDQN.png) - -## N-Step Q-Learning -![Loading...](https://raw.githubusercontent.com/ShangtongZhang/DeepRL/master/images/A2C-NStepQ.png) +Prediction is sampled after 110K iterations, and I only implemented one-step training # Dependency > Tested in macOS 10.12 and CentO/S 6.8 @@ -83,4 +88,6 @@ Prediction is sampled after 110K iterations and I only implemented one-step trai * [Emergence of Locomotion Behaviours in Rich Environments](https://arxiv.org/abs/1707.02286) * [Action-Conditional Video Prediction using Deep Networks in Atari Games](https://arxiv.org/abs/1507.08750) * [A Distributional Perspective on Reinforcement Learning](https://arxiv.org/abs/1707.06887) -* [Distributional Reinforcement Learning with Quantile Regression](https://arxiv.org/abs/1710.10044) \ No newline at end of file +* [Distributional Reinforcement Learning with Quantile Regression](https://arxiv.org/abs/1710.10044) +* [DeepMind Control Suite](https://arxiv.org/abs/1801.00690) +* [OpenAI Baselines](https://github.com/openai/baselines), [Ilya Kostrikov](https://github.com/ikostrikov/pytorch-a2c-ppo-acktr) \ No newline at end of file diff --git a/images/ACVP.png b/images/ACVP.png new file mode 100644 index 0000000..8ce8312 Binary files /dev/null and b/images/ACVP.png differ diff --git a/images/a2c_pixel_atari-180407-92711.png b/images/a2c_pixel_atari-180407-92711.png new file mode 100644 index 0000000..91bf116 Binary files /dev/null and b/images/a2c_pixel_atari-180407-92711.png differ diff --git a/images/categorical_dqn_pixel_atari-180407-094006.png b/images/categorical_dqn_pixel_atari-180407-094006.png new file mode 100644 index 0000000..ae945ed Binary files /dev/null and b/images/categorical_dqn_pixel_atari-180407-094006.png differ diff --git a/images/ddpg_continuous-180407-234141.png b/images/ddpg_continuous-180407-234141.png new file mode 100644 index 0000000..92b0e33 Binary files /dev/null and b/images/ddpg_continuous-180407-234141.png differ diff --git a/images/dqn_pixel_atari-180407-01414.png b/images/dqn_pixel_atari-180407-01414.png new file mode 100644 index 0000000..f645b8e Binary files /dev/null and b/images/dqn_pixel_atari-180407-01414.png differ diff --git a/images/n_step_dqn_pixel_atari-180408-001104.png b/images/n_step_dqn_pixel_atari-180408-001104.png new file mode 100644 index 0000000..cbf6c74 Binary files /dev/null and b/images/n_step_dqn_pixel_atari-180408-001104.png differ diff --git a/images/ppo_continuous-180408-002056.png b/images/ppo_continuous-180408-002056.png new file mode 100644 index 0000000..4734712 Binary files /dev/null and b/images/ppo_continuous-180408-002056.png differ diff --git a/images/quantile_regression_dqn_pixel_atari-180407-01604.png b/images/quantile_regression_dqn_pixel_atari-180407-01604.png new file mode 100644 index 0000000..92f3f28 Binary files /dev/null and b/images/quantile_regression_dqn_pixel_atari-180407-01604.png differ