Commit 732fe5a5 authored by MasterScrat's avatar MasterScrat

Updated README

parent 5398c262
Pipeline #4903 passed with stage
in 45 minutes and 49 seconds
# 🚂 Flatland Baselines # 🚂 Flatland Baselines
This repository contains reinforcement learning baselines for the [NeurIPS 2020 Flatland Challenge](https://www.aicrowd.com/challenges/neurips-2020-flatland-challenge/). This repository contains reinforcement learning baselines for the [NeurIPS 2020 Flatland Challenge](https://www.aicrowd.com/challenges/neurips-2020-flatland-challenge/) based on RLlib.
## Provided baselines **Read the [baseline documentation](https://flatland.aicrowd.com/research/baselines.html) to see how to setup and use the baselines.**
>>> >>>
**Note:** looking for something simpler to start? We also provide a simpler Dueling Double DQN method implemented using PyTorch without relying on RLlib: **https://gitlab.aicrowd.com/flatland/flatland-examples** Looking for something simpler? We also provide a DQN method implemented from scratch using PyTorch: https://gitlab.aicrowd.com/flatland/flatland-examples
>>> >>>
Read the [baseline documentation](https://flatland.aicrowd.com/research/baselines.html) to see how the methods work in details. Notes
---
### RL Methods - The basic structure of this repository is adapted from [https://github.com/spMohanty/rl-experiments/](https://github.com/spMohanty/rl-experiments/)
- The baselines are under the MIT license
- Ape-X Main links
- PPO ---
- CCPPO
- Pure Imitation Learning: MARWIL
- Mixed IL/RL: DQfD
### Custom observations * [Flatland documentation](https://flatland.aicrowd.com/)
* [NeurIPS 2020 Challenge](https://www.aicrowd.com/challenges/neurips-2020-flatland-challenge/)
- Density observations Communication
- Combined observations ---
### Other approaches * [Discord Channel](https://discord.com/invite/hCR3CZG)
* [Discussion Forum](https://discourse.aicrowd.com/c/neurips-2020-flatland-challenge)
- Action skipping * [Issue Tracker](https://gitlab.aicrowd.com/flatland/flatland/issues/)
- Action masking \ No newline at end of file
## Setup
The setup uses conda, [install it](https://www.anaconda.com/products/individual) if necessary.
```
# with GPU support:
conda env create -f environment-gpu.yml
conda activate flatland-baseline-gpu-env
# or, without GPU support:
conda env create -f environment-cpu.yml
conda activate flatland-baseline-cpu-env
```
## Usage
Training example:
`python ./train.py -f experiments/flatland_random_sparse_small/global_obs_conv_net/ppo.yaml`
Evaluation example:
`python ./rollout.py /tmp/ray/checkpoint_dir/checkpoint-0 --run PPO --no-render
--config '{"env_config": {"test": true}}' --episodes 1000 --out rollouts.pkl`
Note that `-f` overrides all other trial-specific command-line options.
## Notes
- The basic structure of this repository is adapted from [https://github.com/spMohanty/rl-experiments/](https://github.com/spMohanty/rl-experiments/)
\ No newline at end of file
#!/usr/bin/env bash #!/usr/bin/env bash
echo "===================="
echo "MARWIL"
echo "===================="
# broken! # broken!
#echo "===================="
#echo "MARWIL"
#echo "===================="
#time python ./trainImitate.py -f experiments/tests/MARWIL.yaml #time python ./trainImitate.py -f experiments/tests/MARWIL.yaml
echo "====================" echo "===================="
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment