meta_gradient_RL

KO | EN

gitlite — search

#javascript #react #golang #machine-learning #rust #linux #docker #nodejs #java #awesome #cli #vue #security #mcp

meta_gradient_RL

(★ 21)

Implementation of NeurIPS 2018 paper "Meta-Gradient Reinforcement Learning"

meta_gradient_RL 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Meta-Gradient Reinforcement Learning
Implementation of NeurIPS 2018 paper <a href="https://proceedings.neurips.cc/paper/2018/file/2715518c875999308842e3455eda2fe3-Paper.pdf">"Meta-Gradient Reinforcement Learning"</a>. For detailed explaination and implementation techniques, please refer to an article in Medium <a href="https://hassaann.medium.com/meta-learning-meta-gradient-reinforcement-learning-an-implementation-b62c0054aafe">"Meta Learning — Meta-Gradient Reinforcement Learning — An Implementation"</a>.

We implemented A2C as the base algorithm under CartPole-v1 environment and consider γ (i.e. the discounting factor) as the only hyperparameter to be meta-learned/meta-adapted.

The maximum reward of the CartPole-v1 environment is 500. 
We constrained the number of training steps to be `500` due to computation and time limits. For each run, the training is terminated as soon as the agent obtains rewards exceeding than `475` (`475` is the reward threshold specified by the CartPole-v1 environment).

We used neural networks to implement the actor and the value network, both of which comprised three fully connected layers with a hidden size of 40 and with ReLU as activation function. Log-softmax operation was used to normalize the sum of the output of the actor network to one. The table below is the hyper-parameters used in our experiment.

The hyperparameter we use were as follows:
| hyperparameter |           |
|----------------------------|----------------------|
| optimizer                  | Adam                 |
| learning rate              | 0.01                 |
| sample size                | 30                   |
| initial gamma              | 0.99                 |
| alpha                      | 0.0001               |

meta_gradient_RL

meta_gradient_RL 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

meta_gradient_RL

meta_gradient_RL 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README