mirror of
https://github.com/wassname/rl_2d_walker.js.git
synced 2026-08-22 12:20:24 +08:00
ccf7349ffcd0aaedcdecd38fc3feb5ceb16ff37f
boxidroid
TODO: Write a project description
Installation
TODO: Describe the installation process
Usage
TODO: Write usage instructions
Contributing
- Fork it!
- Create your feature branch:
git checkout -b my-new-feature - Commit your changes:
git commit -am 'Add some feature' - Push to the branch:
git push origin my-new-feature - Submit a pull request :D
TODO
- make walker into an env, with reset, done etc
- display save loss curves... everything from info
- make diff env's, copy humanoidwalker, humanoid, learning2run etc
- display
- display/save time for each agent
- globals.step_counter
- x progress
- smile frown?
- save checkpoints
- display/save time for each agent
- obstacles later on
- load pretrained
- reward hacking
- head high, feet on ground, legs crossing, moving forward, angles near 0
Credits
- The walker code is adapted from http://rednuht.org/genetic_walkers/
- DDPG code from metacar
Description
Teaching a humanoid to walk(ish), then displaying in your browser (using tensorflow.js and reinforcement learning)
12 MiB
Languages
Jupyter Notebook
84.1%
JavaScript
14.2%
HTML
1.5%
CSS
0.2%