Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

https://github.com/Farama-Foundation/Metaworld

Collections of robotics environments geared towards benchmarking multi-task and meta reinforcement learning
https://github.com/Farama-Foundation/Metaworld

benchmark-environments meta-rl mujoco multi-task

Last synced: 23 days ago
JSON representation

Collections of robotics environments geared towards benchmarking multi-task and meta reinforcement learning

Lists

README 

        
# Meta-World

[![License](https://img.shields.io/badge/license-MIT-blue.svg)](https://github.com/Farama-Foundation/metaworld/blob/master/LICENSE)

![Build Status](https://github.com/Farama-Foundation/Metaworld/workflows/MetaWorld%20CI/badge.svg)



# The current version of Meta-World is a work in progress. If you find any bugs/errors please open an issue.



__Meta-World is an open-source simulated benchmark for meta-reinforcement learning and multi-task learning consisting of 50 distinct robotic manipulation tasks.__ We aim to provide task distributions that are sufficiently broad to evaluate meta-RL algorithms' generalization ability to new behaviors.



For more background information, please refer to our [website](https://meta-world.github.io) and the accompanying [conference publication](https://arxiv.org/abs/1910.10897), which **provides baseline results for 8 state-of-the-art meta- and multi-task RL algorithms**.



__Table of Contents__

- [Installation](#installation)

- [Using the benchmark](#using-the-benchmark)

  * [Basics](#basics)

  * [Seeding a Benchmark Instance](#seeding-a-benchmark-instance)

  * [Running ML1, MT1](#running-ml1-or-mt1)

  * [Running ML10, ML45, MT10, MT50](#running-a-benchmark)

  * [Accessing Single Goal Environments](#accessing-single-goal-environments)

- [Citing Meta-World](#citing-meta-world)

- [Accompanying Baselines](accompanying-baselines)

- [Become a Contributor](#become-a-contributor)

- [Acknowledgements](#acknowledgements)



## Join the Community



Metaworld is now maintained by the Farama Foundation! You can interact with our community and the new developers in our [Discord server](https://discord.gg/PfR7a79FpQ)



## Maintenance Status

The current roadmap for Meta-World can be found [here](https://github.com/Farama-Foundation/Metaworld/issues/409)



## Installation

To install everything, run:



```

pip install git+https://github.com/Farama-Foundation/Metaworld.git@master#egg=metaworld

```



Alternatively, you can clone the repository and install an editable version locally:



```sh

git clone https://github.com/Farama-Foundation/Metaworld.git

cd Metaworld

pip install -e .

```



For users attempting to reproduce results found in the Meta-World paper please use this command:

```

pip install git+https://github.com/Farama-Foundation/Metaworld.git@04be337a12305e393c0caf0cbf5ec7755c7c8feb

```



## Using the benchmark

Here is a list of benchmark environments for meta-RL (ML*) and multi-task-RL (MT*):

* [__ML1__](https://meta-world.github.io/figures/ml1.gif) is a meta-RL benchmark environment which tests few-shot adaptation to goal variation within single task. You can choose to test variation within any of [50 tasks](https://meta-world.github.io/figures/ml45-1080p.gif) for this benchmark.

* [__ML10__](https://meta-world.github.io/figures/ml10.gif) is a meta-RL benchmark which tests few-shot adaptation to new tasks. It comprises 10 meta-train tasks, and 3 test tasks.

* [__ML45__](https://meta-world.github.io/figures/ml45-1080p.gif) is a meta-RL benchmark which tests few-shot adaptation to new tasks. It comprises 45 meta-train tasks and 5 test tasks.

* [__MT10__](https://meta-world.github.io/figures/mt10.gif), __MT1__, and __MT50__ are multi-task-RL benchmark environments for learning a multi-task policy that perform 10, 1, and 50 training tasks respectively. __MT1__ is similar to __ML1__ because you can choose to test variation within any of [50 tasks](https://meta-world.github.io/figures/ml45-1080p.gif) for this benchmark.  In the original Meta-World experiments, we augment MT10 and MT50 environment observations with a one-hot vector which identifies the task. We don't enforce how users utilize task one-hot vectors, however one solution would be to use a Gym wrapper such as [this one](https://github.com/rlworkgroup/garage/blob/master/src/garage/envs/multi_env_wrapper.py)



### Basics

We provide a `Benchmark` API, that allows constructing environments following the [`gymnasium.Env`](https://github.com/Farama-Foundation/Gymnasium/blob/main/gymnasium/core.py#L21) interface.



To use a `Benchmark`, first construct it (this samples the tasks allowed for one run of an algorithm on the benchmark).

Then, construct at least one instance of each environment listed in `benchmark.train_classes` and `benchmark.test_classes`.

For each of those environments, a task must be assigned to it using

`env.set_task(task)` from `benchmark.train_tasks` and `benchmark.test_tasks`,

respectively.

`Tasks` can only be assigned to environments which have a key in

`benchmark.train_classes` or `benchmark.test_classes` matching `task.env_name`.

Please see the sections [Running ML1, MT1](#running-ml1-or-mt1) and [Running ML10, ML45, MT10, MT50](#running-a-benchmark)

for more details.



You may wish to only access individual environments used in the Metaworld benchmark for your research. See the

[Accessing Single Goal Environments](#accessing-single-goal-environments) for more details.



### Seeding a Benchmark Instance

For the purposes of reproducibility, it may be important to you to seed your benchmark instance.

For example, for the ML1 benchmark environment with the 'pick-place-v2' environment, you can do so in the following way:

```python

import metaworld



SEED = 0  # some seed number here

benchmark = metaworld.ML1('pick-place-v2', seed=SEED)

```



### Running ML1 or MT1

```python

import metaworld

import random



print(metaworld.ML1.ENV_NAMES)  # Check out the available environments



ml1 = metaworld.ML1('pick-place-v2') # Construct the benchmark, sampling tasks



env = ml1.train_classes['pick-place-v2']()  # Create an environment with task `pick_place`

task = random.choice(ml1.train_tasks)

env.set_task(task)  # Set task



obs = env.reset()  # Reset environment

a = env.action_space.sample()  # Sample an action

obs, reward, done, info = env.step(a)  # Step the environment with the sampled random action

```

__MT1__ can be run the same way except that it does not contain any `test_tasks`

### Running a benchmark

Create an environment with train tasks (ML10, MT10, ML45, or MT50):

```python

import metaworld

import random



ml10 = metaworld.ML10() # Construct the benchmark, sampling tasks



training_envs = []

for name, env_cls in ml10.train_classes.items():

  env = env_cls()

  task = random.choice([task for task in ml10.train_tasks

                        if task.env_name == name])

  env.set_task(task)

  training_envs.append(env)



for env in training_envs:

  obs = env.reset()  # Reset environment

  a = env.action_space.sample()  # Sample an action

  obs, reward, done, info = env.step(a)  # Step the environment with the sampled random action

```

Create an environment with test tasks (this only works for ML10 and ML45, since MT10 and MT50 don't have a separate set of test tasks):

```python

import metaworld

import random



ml10 = metaworld.ML10() # Construct the benchmark, sampling tasks



testing_envs = []

for name, env_cls in ml10.test_classes.items():

  env = env_cls()

  task = random.choice([task for task in ml10.test_tasks

                        if task.env_name == name])

  env.set_task(task)

  testing_envs.append(env)



for env in testing_envs:

  obs = env.reset()  # Reset environment

  a = env.action_space.sample()  # Sample an action

  obs, reward, done, info = env.step(a)  # Step the environment with the sampled random action

```



## Accessing Single Goal Environments

You may wish to only access individual environments used in the Meta-World benchmark for your research.

We provide constructors for creating environments where the goal has been hidden (by zeroing out the goal in

the observation) and environments where the goal is observable. They are called GoalHidden and GoalObservable

environments respectively.



You can access them in the following way:

```python

from metaworld.envs import (ALL_V2_ENVIRONMENTS_GOAL_OBSERVABLE,

                            ALL_V2_ENVIRONMENTS_GOAL_HIDDEN)

                            # these are ordered dicts where the key : value

                            # is env_name : env_constructor



import numpy as np



door_open_goal_observable_cls = ALL_V2_ENVIRONMENTS_GOAL_OBSERVABLE["door-open-v2-goal-observable"]

door_open_goal_hidden_cls = ALL_V2_ENVIRONMENTS_GOAL_HIDDEN["door-open-v2-goal-hidden"]



env = door_open_goal_hidden_cls()

env.reset()  # Reset environment

a = env.action_space.sample()  # Sample an action

obs, reward, done, info = env.step(a)  # Step the environment with the sampled random action

assert (obs[-3:] == np.zeros(3)).all() # goal will be zeroed out because env is HiddenGoal



# You can choose to initialize the random seed of the environment.

# The state of your rng will remain unaffected after the environment is constructed.

env1 = door_open_goal_observable_cls(seed=5)

env2 = door_open_goal_observable_cls(seed=5)



env1.reset()  # Reset environment

env2.reset()

a1 = env1.action_space.sample()  # Sample an action

a2 = env2.action_space.sample()

next_obs1, _, _, _ = env1.step(a1)  # Step the environment with the sampled random action



next_obs2, _, _, _ = env2.step(a2)

assert (next_obs1[-3:] == next_obs2[-3:]).all() # 2 envs initialized with the same seed will have the same goal

assert not (next_obs2[-3:] == np.zeros(3)).all()   # The env's are goal observable, meaning the goal is not zero'd out



env3 = door_open_goal_observable_cls(seed=10)  # Construct an environment with a different seed

env1.reset()  # Reset environment

env3.reset()

a1 = env1.action_space.sample()  # Sample an action

a3 = env3.action_space.sample()

next_obs1, _, _, _ = env1.step(a1)  # Step the environment with the sampled random action

next_obs3, _, _, _ = env3.step(a3)



assert not (next_obs1[-3:] == next_obs3[-3:]).all() # 2 envs initialized with different seeds will have different goals

assert not (next_obs1[-3:] == np.zeros(3)).all()   # The env's are goal observable, meaning the goal is not zero'd out



```



## Citing Meta-World

If you use Meta-World for academic research, please kindly cite our CoRL 2019 paper the using following BibTeX entry.



```

@inproceedings{yu2019meta,

  title={Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning},

  author={Tianhe Yu and Deirdre Quillen and Zhanpeng He and Ryan Julian and Karol Hausman and Chelsea Finn and Sergey Levine},

  booktitle={Conference on Robot Learning (CoRL)},

  year={2019}

  eprint={1910.10897},

  archivePrefix={arXiv},

  primaryClass={cs.LG}

  url={https://arxiv.org/abs/1910.10897}

}

```



## Accompanying Baselines

If you're looking for implementations of the baselines algorithms used in the Meta-World conference publication, please look at our sister directory, [Garage](https://github.com/rlworkgroup/garage).



Note that these aren't the exact same baselines that were used in the original conference publication, however they are true to the original baselines.



## Become a Contributor

We welcome all contributions to Meta-World. Please refer to the [contributor's guide](https://github.com/Farama-Foundation/Metaworld/blob/master/CONTRIBUTING.md) for how to prepare your contributions.



## Acknowledgements

Meta-World is a work by [Tianhe Yu (Stanford University)](https://cs.stanford.edu/~tianheyu/), [Deirdre Quillen (UC Berkeley)](https://scholar.google.com/citations?user=eDQsOFMAAAAJ&hl=en), [Zhanpeng He (Columbia University)](https://zhanpenghe.github.io), [Ryan Julian (University of Southern California)](https://ryanjulian.me), [Karol Hausman (Google AI)](https://karolhausman.github.io),  [Chelsea Finn (Stanford University)](https://ai.stanford.edu/~cbfinn/) and [Sergey Levine (UC Berkeley)](https://people.eecs.berkeley.edu/~svlevine/).



The code for Meta-World was originally based on [multiworld](https://github.com/vitchyr/multiworld), which is developed by [Vitchyr H. Pong](https://people.eecs.berkeley.edu/~vitchyr/), [Murtaza Dalal](https://github.com/mdalal2020), [Ashvin Nair](http://ashvin.me/), [Shikhar Bahl](https://shikharbahl.github.io), [Steven Lin](https://github.com/stevenlin1111), [Soroush Nasiriany](http://snasiriany.me/), [Kristian Hartikainen](https://hartikainen.github.io/) and [Coline Devin](https://github.com/cdevin). The Meta-World authors are grateful for their efforts on providing such a great framework as a foundation of our work. We also would like to thank Russell Mendonca for his work on reward functions for some of the environments.