GoMARL

KO | EN

gitlite — search

#react #ai #typescript #llm #android #macos #java #awesome #cli #kubernetes #awesome-list #vue #mcp #open-source

(★ 35)

Codes of GoMARL accompanying the paper "Automatic Grouping for Efficient Cooperative Multi-Agent Reinforcement Learning"(NeurIPS 2023). GoMARL is a domain-agnostic MARL method that learns automatic grouping for efficient cooperation by promoting intra- and inter-group coordination.

GoMARL 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Code Instruction for GoMARL

This instruction hosts the PyTorch implementation of GoMARL accompanying the paper ["**Automatic Grouping for Efficient Cooperative Multi-Agent Reinforcement Learning**" (NeurIPS, 2023)](https://proceedings.neurips.cc/paper_files/paper/2023/hash/906c860f1b7515a8ffec02dcdac74048-Abstract-Conference.html). GoMARL is a domain-agnostic MARL method that learns automatic grouping for efficient cooperation by promoting intra- and inter-group coordination.

The implementation is based on the frameworks [PyMARL](https://github.com/oxwhirl/pymarl) and [PyMARL2](https://github.com/hijkzzz/pymarl2) with environments [SMAC](https://github.com/oxwhirl/smac) and [Gfootball](https://github.com/google-research/football). All of our SMAC experiments are based on the latest PyMARL2 utilizing SC2.4.6.10. The underlying dynamics are sufficiently different, so you cannot compare runs across various versions.

## Setup

Set up the working environment:

```shell
pip3 install -r requirements.txt
```

Set up the StarCraftII game core (SC2.4.6.10):

```shell
bash install_sc2.sh
```

Set up the Gfootball environment:

```shell
bash install_gfootball.sh
```

## Training

To train `GoMARL` on the `3s5z_vs_3s6z`(SMAC) scenario:

```shell
python3 src/main.py --config=group --env-config=sc2 with env_args.map_name=3s5z_vs_3s6z
```

Change the `map_name` accordingly for other SMAC scenarios (e.g., `MMM2`).

To train `GoMARL` on the `academy_3_vs_1_with_keeper`(Gfootball) scenario:

```
python3 src/main.py --config=group --env-config=academy_3_vs_1_with_keeper
```

Change the `env-config` accordingly for other Gfootball scenarios (e.g., `academy_counterattack_easy`).

All results will be saved in the `results` folder.

The config file `src/config/algs/group.yaml` contains default hyperparameters for GoMARL.

## Evaluation

### TensorBoard

One could set `use_tensorboard` to `True` in `src/config/default.yaml`, and the training tensorboards will be saved in the `results/tb_logs` directory, containing useful info such as test battle win rate during training.

### Saving models

Same as PyMARL, set `save_model` to `True` in `src/config/default.yaml`, and the learned model during training will be saved in the `results/models/` directory. The frequency for saving models can be adjusted by setting the parameter `save_model_interval`.

### Loading models

Saved models can be loaded by adjusting the `checkpoint_path` parameter in `src/config/default.yaml`. For instance, to load the model under path `result/model/[timesteps]/agent.th`, set `checkpoint_path` to `result/model/[timesteps]`.

### Saving Starcraft II Replay

The learned model loaded from `checkpoint_path` can be evaluated by setting `evaluate` to `True` in `src/config/default.yaml`. To save the Starcraft II replays, please make sure the configuration `save_replay` is set to `True`, and use the `episode_runner`.

Check out [PyMARL](https://github.com/oxwhirl/pymarl) documentation for more information.

## See Also

See [SMAC](https://github.com/oxwhirl/smac), [PyMARL2](https://github.com/hijkzzz/pymarl2), [PyMARL](https://github.com/oxwhirl/pymarl) and [Gfootball](https://github.com/google-research/football) for additional instructions.

GoMARL

GoMARL 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

GoMARL

GoMARL 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README