Tianshou (天授) is a reinforcement learning platform based on pure PyTorch. Unlike existing reinforcement learning libraries, which are mainly based on TensorFlow, have many nested classes, unfriendly API, or slow-speed, Tianshou provides a fast-speed modularized framework and pythonic API for building the deep reinforcement learning agent with the least number of lines of code. The supported interface algorithms currently include:

• Deep Q-Network (DQN)
• Double DQN
• Dueling DQN
• Categorical DQN (C51)
• Rainbow DQN (Rainbow)
• Quantile Regression DQN (QRDQN)
• Implicit Quantile Network (IQN)
• Fully-parameterized Quantile Function (FQF)
• Policy Gradient (PG)
• Natural Policy Gradient (NPG)
• Advantage Actor-Critic (A2C)
• Trust Region Policy Optimization (TRPO)
• Proximal Policy Optimization (PPO)
• Deep Deterministic Policy Gradient (DDPG)
• Twin Delayed DDPG (TD3)
• Soft Actor-Critic (SAC)
• Discrete Soft Actor-Critic (SAC-Discrete)
• Vanilla Imitation Learning
• Batch-Constrained deep Q-Learning (BCQ)
• Conservative Q-Learning (CQL)
• Discrete Batch-Constrained deep Q-Learning (BCQ-Discrete)
• Discrete Conservative Q-Learning (CQL-Discrete)
• Discrete Critic Regularized Regression (CRR-Discrete)
• Generative Adversarial Imitation Learning (GAIL)
• Prioritized Experience Replay (PER)
• Generalized Advantage Estimator (GAE)
• Posterior Sampling Reinforcement Learning (PSRL)
• Intrinsic Curiosity Module (ICM)

Here is Tianshou's other features:

• Elegant framework, using only ~4000 lines of code
• State-of-the-art MuJoCo benchmark for REINFORCE/A2C/TRPO/PPO/DDPG/TD3/SAC algorithms
• Support vectorized environment (synchronous or asynchronous) for all algorithms Usage
• Support super-fast vectorized environment EnvPool for all algorithms Usage
• Support recurrent state representation in actor network and critic network (RNN-style training for POMDP) Usage
• Support any type of environment state/action (e.g. a dict, a self-defined class, ...) Usage
• Support customized training process Usage
• Support n-step returns estimation and prioritized experience replay for all Q-learning based algorithms; GAE, nstep and PER are very fast thanks to numba jit function and vectorized numpy operation
• Support multi-agent RL Usage
• Support both TensorBoard and W&B log tools
• Support multi-GPU training Usage
• Comprehensive documentation, PEP8 code-style checking, type checking and unit tests

In Chinese, Tianshou means divinely ordained and is derived to the gift of being born with. Tianshou is a reinforcement learning platform, and the RL algorithm does not learn from humans. So taking "Tianshou" means that there is no teacher to study with, but rather to learn by themselves through constant interaction with the environment.

“天授”意指上天所授，引申为与生具有的天赋。天授是强化学习平台，而强化学习算法并不是向人类学习的，所以取“天授”意思是没有老师来教，而是自己通过跟环境不断交互来进行学习。

Tianshou is currently hosted on PyPI and conda-forge. It requires Python >= 3.6.

You can simply install Tianshou from PyPI with the following command:

$ pip install tianshou

If you use Anaconda or Miniconda, you can install Tianshou from conda-forge through the following command:

$ conda install -c conda-forge tianshou

You can also install with the newest version through GitHub:

$ pip install git+https://github.com/thu-ml/tianshou.git@master --upgrade

After installation, open your python console and type

import tianshou
print(tianshou.__version__)

If no error occurs, you have successfully installed Tianshou.

The tutorials and API documentation are hosted on tianshou.readthedocs.io.

The example scripts are under test/ folder and examples/ folder.

中文文档位于 https://tianshou.readthedocs.io/zh/master/。

RL Platform	GitHub Stars	# of Alg. (1)	Custom Env	Batch Training	RNN Support	Nested Observation	Backend
Baselines		9	✔️ (gym)	➖ (2)	✔️	❌	TF1
Stable-Baselines		11	✔️ (gym)	➖ (2)	✔️	❌	TF1
Stable-Baselines3		7 (3)	✔️ (gym)	➖ (2)	❌	✔️	PyTorch
Ray/RLlib		16	✔️	✔️	✔️	✔️	TF/PyTorch
SpinningUp		6	✔️ (gym)	➖ (2)	❌	❌	PyTorch
Dopamine		7	❌	❌	❌	❌	TF/JAX
ACME		14	✔️ (dm_env)	✔️	✔️	✔️	TF/JAX
keras-rl		7	✔️ (gym)	❌	❌	❌	Keras
rlpyt		11	❌	✔️	✔️	✔️	PyTorch
ChainerRL		18	✔️ (gym)	✔️	✔️	❌	Chainer
Sample Factory		1 (4)	✔️ (gym)	✔️	✔️	✔️	PyTorch
Tianshou		20	✔️ (gym)	✔️	✔️	✔️	PyTorch

^{(1): access date: 2021-08-08}

^{(2): not all algorithms support this feature}

^{(3): TQC and QR-DQN in sb3-contrib instead of main repo}

^{(4): super fast APPO!}

RL Platform	Documentation	Code Coverage	Type Hints	Last Update
Baselines	❌	❌	❌
Stable-Baselines			❌
Stable-Baselines3			✔️
Ray/RLlib		➖(1)	✔️
SpinningUp		❌	❌
Dopamine		❌	❌
ACME		➖(1)	✔️
keras-rl		➖(1)	❌
rlpyt			❌
ChainerRL			❌
Sample Factory	➖		❌
Tianshou			✔️

^{(1): it has continuous integration but the coverage rate is not available}

Tianshou has its unit tests. Different from other platforms, the unit tests include the full agent training procedure for all of the implemented algorithms. It would be failed once if it could not train an agent to perform well enough on limited epochs on toy scenarios. The unit tests secure the reproducibility of our platform. Check out the GitHub Actions page for more detail.

The Atari/Mujoco benchmark results are under examples/atari/ and examples/mujoco/ folders. Our Mujoco result can beat most of existing benchmark.

We decouple all of the algorithms roughly into the following parts:

• __init__: initialize the policy;
• forward: to compute actions over given observations;
• process_fn: to preprocess data from replay buffer (since we have reformulated all algorithms to replay-buffer based algorithms);
• learn: to learn from a given batch data;
• post_process_fn: to update the replay buffer from the learning process (e.g., prioritized replay buffer needs to update the weight);
• update: the main interface for training, i.e., process_fn -> learn -> post_process_fn.

Within this API, we can interact with different policies conveniently.

This is an example of Deep Q Network. You can also run the full script at test/discrete/test_dqn.py.

First, import some relevant packages:

import gym, torch, numpy as np, torch.nn as nn
from torch.utils.tensorboard import SummaryWriter
import tianshou as ts

Define some hyper-parameters:

task = 'CartPole-v0'
lr, epoch, batch_size = 1e-3, 10, 64
train_num, test_num = 10, 100
gamma, n_step, target_freq = 0.9, 3, 320
buffer_size = 20000
eps_train, eps_test = 0.1, 0.05
step_per_epoch, step_per_collect = 10000, 10
logger = ts.utils.TensorboardLogger(SummaryWriter('log/dqn'))  # TensorBoard is supported!
# For other loggers: https://tianshou.readthedocs.io/en/master/tutorials/logger.html

Make environments:

# you can also try with SubprocVectorEnv
train_envs = ts.env.DummyVectorEnv([lambda: gym.make(task) for _ in range(train_num)])
test_envs = ts.env.DummyVectorEnv([lambda: gym.make(task) for _ in range(test_num)])

Define the network:

from tianshou.utils.net.common import Net
# you can define other net by following the API:
# https://tianshou.readthedocs.io/en/master/tutorials/dqn.html#build-the-network
env = gym.make(task)
state_shape = env.observation_space.shape or env.observation_space.n
action_shape = env.action_space.shape or env.action_space.n
net = Net(state_shape=state_shape, action_shape=action_shape, hidden_sizes=[128, 128, 128])
optim = torch.optim.Adam(net.parameters(), lr=lr)

Setup policy and collectors:

policy = ts.policy.DQNPolicy(net, optim, gamma, n_step, target_update_freq=target_freq)
train_collector = ts.data.Collector(policy, train_envs, ts.data.VectorReplayBuffer(buffer_size, train_num), exploration_noise=True)
test_collector = ts.data.Collector(policy, test_envs, exploration_noise=True)  # because DQN uses epsilon-greedy method

Let's train it:

result = ts.trainer.offpolicy_trainer(
    policy, train_collector, test_collector, epoch, step_per_epoch, step_per_collect,
    test_num, batch_size, update_per_step=1 / step_per_collect,
    train_fn=lambda epoch, env_step: policy.set_eps(eps_train),
    test_fn=lambda epoch, env_step: policy.set_eps(eps_test),
    stop_fn=lambda mean_rewards: mean_rewards >= env.spec.reward_threshold,
    logger=logger)
print(f'Finished training! Use {result["duration"]}')

Save / load the trained policy (it's exactly the same as PyTorch nn.module):

torch.save(policy.state_dict(), 'dqn.pth')
policy.load_state_dict(torch.load('dqn.pth'))

Watch the performance with 35 FPS:

policy.eval()
policy.set_eps(eps_test)
collector = ts.data.Collector(policy, env, exploration_noise=True)
collector.collect(n_episode=1, render=1 / 35)

Look at the result saved in tensorboard: (with bash script in your terminal)

$ tensorboard --logdir log/dqn

You can check out the documentation for advanced usage.

It's worth a try: here is a test on a laptop (i7-8750H + GTX1060). It only uses 3 seconds for training an agent based on vanilla policy gradient on the CartPole-v0 task: (seed may be different across different platform and device)

$ python3 test/discrete/test_pg.py --seed 0 --render 0.03

Tianshou is still under development. More algorithms and features are going to be added and we always welcome contributions to help make Tianshou better. If you would like to contribute, please check out this link.

If you find Tianshou useful, please cite it in your publications.

@article{weng2021tianshou,
  title={Tianshou: A Highly Modularized Deep Reinforcement Learning Library},
  author={Weng, Jiayi and Chen, Huayu and Yan, Dong and You, Kaichao and Duburcq, Alexis and Zhang, Minghao and Su, Hang and Zhu, Jun},
  journal={arXiv preprint arXiv:2107.14171},
  year={2021}
}

Tianshou was previously a reinforcement learning platform based on TensorFlow. You can check out the branch priv for more detail. Many thanks to Haosheng Zou's pioneering work for Tianshou before version 0.1.1.

We would like to thank TSAIL and Institute for Artificial Intelligence, Tsinghua University for providing such an excellent AI research platform.