xwjiang2010 cdf70c2900 [Tune] Remove legacy resources implementations in Runner and Executor. (#19773) | 2 years ago | |
---|---|---|
.. | ||
agents | 246787cdd9 Revert "[RLlib] POC: `PGTrainer` class that works by sub-classing, not `trainer_template.py`. (#20055)" (#20284) | 2 years ago |
contrib | 026bf01071 [RLlib] Upgrade gym version to 0.21 and deprecate pendulum-v0. (#19535) | 3 years ago |
env | 026bf01071 [RLlib] Upgrade gym version to 0.21 and deprecate pendulum-v0. (#19535) | 3 years ago |
evaluation | 70fe25055a [RLlib] Issue: Get single step input dict incorrect. (#20217) | 2 years ago |
examples | cdf70c2900 [Tune] Remove legacy resources implementations in Runner and Executor. (#19773) | 2 years ago |
execution | a931076f59 [RLlib] Tf2 + eager-tracing same speed as framework=tf; Add more test coverage for tf2+tracing. (#19981) | 3 years ago |
models | a931076f59 [RLlib] Tf2 + eager-tracing same speed as framework=tf; Add more test coverage for tf2+tracing. (#19981) | 3 years ago |
offline | dc51af798c [RLlib] Minor fix on json encoding during worker sampling (#20134) | 2 years ago |
policy | 70fe25055a [RLlib] Issue: Get single step input dict incorrect. (#20217) | 2 years ago |
tests | 246787cdd9 Revert "[RLlib] POC: `PGTrainer` class that works by sub-classing, not `trainer_template.py`. (#20055)" (#20284) | 2 years ago |
tuned_examples | f3397b6f48 [RLlib] Minor fixes/cleanups; chop_into_sequences now handles nested data. (#19408) | 3 years ago |
utils | 246787cdd9 Revert "[RLlib] POC: `PGTrainer` class that works by sub-classing, not `trainer_template.py`. (#20055)" (#20284) | 2 years ago |
BUILD | ebd56b57db [RLlib; documentation] "RLlib in 60sec" overhaul. (#20215) | 2 years ago |
README.md | 9b8218aabd [docs] Move all /latest links to /master (#11897) | 4 years ago |
__init__.py | d001af3e59 [RLlib] Allow `rllib rollout` to run distributed via evaluation workers. (#13718) | 3 years ago |
asv.conf.json | 5d7afe8092 [rllib] Try moving RLlib to top level dir (#5324) | 5 years ago |
evaluate.py | e6aae61487 [RLlib; testing] Fix bug in stress tests not handling >1 trials per experiment (due to grid-search in IMPALA stress tests). (#18705) | 3 years ago |
rollout.py | c5d20849ae [RLlib] Rename `rllib rollout` into `rllib evaluate` (backward compatible) to match Trainer API. (#18467) | 3 years ago |
scripts.py | c5d20849ae [RLlib] Rename `rllib rollout` into `rllib evaluate` (backward compatible) to match Trainer API. (#18467) | 3 years ago |
train.py | 38c456b6f4 [RLlib; Tune] Fix rllib/train.py script after tune.Experiment c'tor change. (#20283) | 2 years ago |
RLlib is an open-source library for reinforcement learning that offers both high scalability and a unified API for a variety of applications.
For an overview of RLlib, see the documentation.
If you've found RLlib useful for your research, you can cite the paper as follows:
@inproceedings{liang2018rllib,
Author = {Eric Liang and
Richard Liaw and
Robert Nishihara and
Philipp Moritz and
Roy Fox and
Ken Goldberg and
Joseph E. Gonzalez and
Michael I. Jordan and
Ion Stoica},
Title = {{RLlib}: Abstractions for Distributed Reinforcement Learning},
Booktitle = {International Conference on Machine Learning ({ICML})},
Year = {2018}
}
You can develop RLlib locally without needing to compile Ray by using the setup-dev.py script. This sets up links between the rllib
dir in your git repo and the one bundled with the ray
package. When using this script, make sure that your git branch is in sync with the installed Ray binaries (i.e., you are up-to-date on master and have the latest wheel installed.)