Science Score: 57.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
✓DOI references
Found 1 DOI reference(s) in README -
○Academic publication links
-
○Committers with academic emails
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (9.2%) to scientific vocabulary
Repository
Advantage weighted Actor Critic for Offline RL
Basic Info
- Host: GitHub
- Owner: hari-sikchi
- License: mit
- Language: Python
- Default Branch: master
- Size: 34.2 KB
Statistics
- Stars: 50
- Watchers: 2
- Forks: 8
- Open Issues: 6
- Releases: 1
Metadata Files
README.md
pytorch-AWAC
Advantage weighted Actor Critic for Offline RL implemented in pytorch.
A cleaner implementation for AWAC built on top of spinning_up SAC, and rlkit.
If you use this code in your research project please cite us as:
@software{Sikchi_pytorch-AWAC,
author = {Sikchi, Harshit and Wilcox, Albert},
doi = {10.5281/zenodo.5121023},
title = {{pytorch-AWAC}},
url = {https://github.com/hari-sikchi/AWAC}
}
Running the code
python run_agent.py --env <env_name> --seed <seed_no> --exp_name <experiment name> --algorithm 'AWAC'
Plotting
python plot.py <data_folder> --value <y_axis_coordinate>
The plotting script will plot all the subfolders inside the given folder. The value is the y-axis that is required. 'value' can be: * AverageTestEpRet
Results
Environment: Ant-v2
Dataset: https://drive.google.com/file/d/1edcuicVv2d-PqH1aZUVbO5CeRq3lqK89/view
Plot [Compare to Figure 5 of the official paper]:
References
@article{SpinningUp2018, author = {Achiam, Joshua}, title = {{Spinning Up in Deep Reinforcement Learning}}, year = {2018} }
@article{nair2020accelerating, title={Accelerating Online Reinforcement Learning with Offline Datasets}, author={Nair, Ashvin and Dalal, Murtaza and Gupta, Abhishek and Levine, Sergey}, journal={arXiv preprint arXiv:2006.09359}, year={2020} }
Owner
- Name: Harshit Sikchi
- Login: hari-sikchi
- Kind: user
- Location: University of Texas at Austin
- Website: hari-sikchi.github.io
- Twitter: HariSikchi
- Repositories: 74
- Profile: https://github.com/hari-sikchi
Reinforcement Learning, Roboticist
Citation (citation.cff)
# YAML 1.2
---
authors:
-
affiliation: "University of Texas at Austin"
family-names: Sikchi
given-names: Harshit
-
family-names: Wilcox
given-names: Albert
cff-version: "1.1.0"
doi: 10.5281/zenodo.5121023
message: "If you use this software, please cite it using these metadata."
repository-code: "https://github.com/hari-sikchi/AWAC"
title: "pytorch-AWAC"
...
GitHub Events
Total
- Watch event: 4
Last Year
- Watch event: 4
Committers
Last synced: over 2 years ago
Top Committers
| Name | Commits | |
|---|---|---|
| Harshit Sikchi | h****i | 10 |
| hari-sikchi | h****8@g****m | 4 |
| Albert Wilcox | a****i@g****m | 2 |