Bump min PyTorch version

.github/workflows/ci.yml

@@ -28,7 +28,7 @@ jobs:
       run: |
         python -m pip install --upgrade pip
         # cpu version of pytorch
-        pip install torch==1.8.1+cpu -f https://download.pytorch.org/whl/torch_stable.html
+        pip install torch==1.11+cpu -f https://download.pytorch.org/whl/torch_stable.html
         pip install .[extra,tests,docs]
         # Use headless version
         pip install opencv-python-headless

README.md

@@ -51,7 +51,7 @@ Documentation is available online: [https://stable-baselines3.readthedocs.io/](h
 
 ## Integrations
 
-Stable-Baselines3 has some integration with other libraries/services like Weights & Biases for experiment tracking or Hugging Face for storing/sharing trained models. You can find out more in the [dedicated section](https://stable-baselines3.readthedocs.io/en/master/guide/integrations.html) of the documentation. 
+Stable-Baselines3 has some integration with other libraries/services like Weights & Biases for experiment tracking or Hugging Face for storing/sharing trained models. You can find out more in the [dedicated section](https://stable-baselines3.readthedocs.io/en/master/guide/integrations.html) of the documentation.
 
 
 ## RL Baselines3 Zoo: A Training Framework for Stable Baselines3 Reinforcement Learning Agents
@@ -84,7 +84,7 @@ Documentation is available online: [https://sb3-contrib.readthedocs.io/](https:/
 
 ## Installation
 
-**Note:** Stable-Baselines3 supports PyTorch >= 1.8.1.
+**Note:** Stable-Baselines3 supports PyTorch >= 1.11
 
 ### Prerequisites
 Stable Baselines3 requires Python 3.7+.

docs/guide/custom_env.rst

@@ -61,7 +61,7 @@ Then you can define and train a RL agent with:
   model = A2C('CnnPolicy', env).learn(total_timesteps=1000)
 
 
-To check that your environment follows the gym interface, please use:
+To check that your environment follows the Gym interface that SB3 supports, please use:
 
 .. code-block:: python
 
@@ -71,11 +71,11 @@ To check that your environment follows the gym interface, please use:
 	# It will check your custom environment and output additional warnings if needed
 	check_env(env)
 
-
+Gym also have its own `env checker <https://www.gymlibrary.ml/content/api/#checking-api-conformity>`_ but it checks a superset of what SB3 supports (SB3 does not support all Gym features).
 
 We have created a `colab notebook <https://colab.research.google.com/github/araffin/rl-tutorial-jnrr19/blob/master/5_custom_gym_env.ipynb>`_ for a concrete example on creating a custom environment along with an example of using it with Stable-Baselines3 interface.
 
-Alternatively, you may look at OpenAI Gym `built-in environments <https://gym.openai.com/docs/#available-environments>`_. However, the readers are cautioned as per OpenAI Gym `official wiki <https://github.com/openai/gym/wiki/FAQ>`_, its advised not to customize their built-in environments. It is better to copy and create new ones if you need to modify them.
+Alternatively, you may look at OpenAI Gym `built-in environments <https://www.gymlibrary.ml/>`_. However, the readers are cautioned as per OpenAI Gym `official wiki <https://github.com/openai/gym/wiki/FAQ>`_, its advised not to customize their built-in environments. It is better to copy and create new ones if you need to modify them.
 
 Optionally, you can also register the environment with gym, that will allow you to create the RL agent in one line (and use ``gym.make()`` to instantiate the env):
 

docs/guide/install.rst

@@ -6,7 +6,7 @@ Installation
 Prerequisites
 -------------
 
-Stable-Baselines3 requires python 3.7+ and PyTorch >= 1.8.1.
+Stable-Baselines3 requires python 3.7+ and PyTorch >= 1.11
 
 Windows 10
 ~~~~~~~~~~

docs/misc/changelog.rst

@@ -4,13 +4,14 @@ Changelog
 ==========
 
 
-Release 1.5.1a2(WIP)
+Release 1.5.1a3 (WIP)
 ---------------------------
 
 Breaking Changes:
 ^^^^^^^^^^^^^^^^^
 - Changed the way policy "aliases" are handled ("MlpPolicy", "CnnPolicy", ...), removing the former
   ``register_policy`` helper, ``policy_base`` parameter and using ``policy_aliases`` static attributes instead (@Gregwar)
+- SB3 now requires PyTorch >= 1.11
 
 New Features:
 ^^^^^^^^^^^^^
@@ -31,6 +32,7 @@ Others:
 
 Documentation:
 ^^^^^^^^^^^^^^
+- Added link to gym doc and gym env checker
 
 
 Release 1.5.0 (2022-03-25)

setup.py

@@ -75,7 +75,7 @@
     install_requires=[
         "gym==0.21",  # Fixed version due to breaking changes in 0.22
         "numpy",
-        "torch>=1.8.1",
+        "torch>=1.11",
         # For saving models
         "cloudpickle",
         # For reading logs

stable_baselines3/version.txt

@@ -1 +1 @@
-1.5.1a2
+1.5.1a3