Update to Python 3.8, and use AutoROM in chapter 18

WngWai · Oct 17, 2021 · 3289b05 · 3289b05
1 parent 47208a5
commit 3289b05
Show file tree

Hide file tree

Showing 3 changed files with 112 additions and 88 deletions.
diff --git a/18_reinforcement_learning.ipynb b/18_reinforcement_learning.ipynb
@@ -52,8 +52,7 @@
     "\n",
     "if IS_COLAB or IS_KAGGLE:\n",
     "    !apt update && apt install -y libpq-dev libsdl2-dev swig xorg-dev xvfb\n",
-    "    %pip install -q -U tf-agents pyvirtualdisplay gym[box2d]\n",
-    "    %pip install -q -U atari_py==0.2.5\n",
+    "    %pip install -q -U tf-agents pyvirtualdisplay gym[box2d,atari,accept-rom-license]\n",
     "\n",
     "# Scikit-Learn ≥0.20 is required\n",
     "import sklearn\n",
@@ -2181,6 +2180,35 @@
     "Let's use TF-Agents to create an agent that will learn to play Breakout. We will use the Deep Q-Learning algorithm, so you can easily compare the components with the previous implementation, but TF-Agents implements many other (and more sophisticated) algorithms!"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "First, we need to download and install the Atari ROMs. This can be done very easily using the `AutoROM` tool, if you accept the license:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "AutoROM will download the Atari 2600 ROMs.\n",
+      "They will be installed to:\n",
+      "\t/Users/ageron/miniconda3/envs/tf2/lib/python3.8/site-packages/AutoROM/roms\n",
+      "\n",
+      "Existing ROMs will be overwritten.\n",
+      "Done!                                                                           \n"
+     ]
+    }
+   ],
+   "source": [
+    "!{sys.prefix}/bin/AutoROM --quiet --accept-license"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -4387,7 +4415,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.10"
+   "version": "3.8.12"
   }
  },
  "nbformat": 4,

diff --git a/environment.yml b/environment.yml
@@ -3,52 +3,55 @@ channels:
   - conda-forge
   - defaults
 dependencies:
-  - atari_py=0.2.6 # used only in chapter 18
-  - box2d-py=2.3 # used only in chapter 18
-  - ftfy=6.0 # used only in chapter 16 by the transformers library
-  - graphviz # used only in chapter 6 for dot files
-  - gym=0.19 # used only in chapter 18
-  - ipython=7.28 # a powerful Python shell
-  - ipywidgets=7.6 # optionally used only in chapter 12 for tqdm in Jupyter
-  - joblib=0.14 # used only in chapter 2 to save/load Scikit-Learn models
-  - jupyter=1.0 # to edit and run Jupyter notebooks
-  - matplotlib=3.4 # beautiful plots. See tutorial tools_matplotlib.ipynb
-  - nbdime=3.1 # optional tool to diff Jupyter notebooks
-  - nltk=3.6 # optionally used in chapter 3, exercise 4
-  - numexpr=2.7 # used only in the Pandas tutorial for numerical expressions
-  - numpy=1.19 # Powerful n-dimensional arrays and numerical computing tools
-  - opencv=4.5 # used only in chapter 18 by TF Agents for image preprocessing
-  - pandas=1.3 # data analysis and manipulation tool
-  - pillow=8.3 # image manipulation library, (used by matplotlib.image.imread)
-  - pip # Python's package-management system
-  - py-xgboost=1.4 # used only in chapter 7 for optimized Gradient Boosting
-  - pyglet=1.5 # used only in chapter 18 to render environments
-  - pyopengl=3.1 # used only in chapter 18 to render environments
-  - python=3.7 # Python! Not using latest version as some libs lack support
-  - python-graphviz # used only in chapter 6 for dot files
-# - pyvirtualdisplay=2.2 # used only in chapter 18 if on headless server
-  - requests=2.26 # used only in chapter 19 for REST API queries
-  - scikit-learn=1.0 # machine learning library
-  - scipy=1.7 # scientific/technical computing library
-  - tqdm=4.62 # a progress bar library
-  - wheel # built-package format for pip
-  - widgetsnbextension=3.5 # interactive HTML widgets for Jupyter notebooks
+  - atari_py=0.2  # used only in chapter 18
+  - box2d-py=2.3  # used only in chapter 18
+  - ftfy=6.0  # used only in chapter 16 by the transformers library
+  - graphviz  # used only in chapter 6 for dot files
+  - gym=0.19  # used only in chapter 18
+  - ipython=7.28  # a powerful Python shell
+  - ipywidgets=7.6  # optionally used only in chapter 12 for tqdm in Jupyter
+  - joblib=0.14  # used only in chapter 2 to save/load Scikit-Learn models
+  - jupyter=1.0  # to edit and run Jupyter notebooks
+  - matplotlib=3.4  # beautiful plots. See tutorial tools_matplotlib.ipynb
+  - nbdime=3.1  # optional tool to diff Jupyter notebooks
+  - nltk=3.6  # optionally used in chapter 3, exercise 4
+  - numexpr=2.7  # used only in the Pandas tutorial for numerical expressions
+  - numpy=1.19  # Powerful n-dimensional arrays and numerical computing tools
+  - opencv=4.5  # used only in chapter 18 by TF Agents for image preprocessing
+  - pandas=1.3  # data analysis and manipulation tool
+  - pillow=8.3  # image manipulation library, (used by matplotlib.image.imread)
+  - pip  # Python's package-management system
+  - py-xgboost=1.4  # used only in chapter 7 for optimized Gradient Boosting
+  - pyglet=1.5  # used only in chapter 18 to render environments
+  - pyopengl=3.1  # used only in chapter 18 to render environments
+  - python=3.8  # Python! Not using latest version as some libs lack support
+  - python-graphviz  # used only in chapter 6 for dot files
+  #- pyvirtualdisplay=2.2  # used only in chapter 18 if on headless server
+  - requests=2.26  # used only in chapter 19 for REST API queries
+  - scikit-learn=1.0  # machine learning library
+  - scipy=1.7  # scientific/technical computing library
+  - tqdm=4.62  # a progress bar library
+  - wheel  # built-package format for pip
+  - widgetsnbextension=3.5  # interactive HTML widgets for Jupyter notebooks
   - pip:
-    - tensorboard-plugin-profile==2.5.0 # profiling plugin for TensorBoard
-    - tensorboard==2.6.0 # TensorFlow's visualization toolkit
-    - tensorflow-addons==0.14.0 # used only in chapter 16 for a seq2seq impl.
-    - tensorflow-datasets==4.4.0 # datasets repository, ready to use
-    - tensorflow-hub==0.12.0 # trained ML models repository, ready to use
-    - tensorflow-probability==0.14.1 # Optional. Probability/Stats lib.
-    - tensorflow-serving-api==2.6.0 # or tensorflow-serving-api-gpu if gpu
-    - tensorflow==2.6.0 # Deep Learning library
-    - tf-agents==0.10.0 # Reinforcement Learning lib based on TensorFlow
-    - tfx==1.3.0 # platform to deploy production ML pipelines
-    - transformers==4.11.3 # Natural Language Processing lib for TF or PyTorch
-    - urlextract==1.4.0 # optionally used in chapter 3, exercise 4
+    - tensorboard-plugin-profile~=2.5.0  # profiling plugin for TensorBoard
+    - tensorboard~=2.7.0  # TensorFlow's visualization toolkit
+    - tensorflow-addons~=0.14.0  # used only in chapter 16 for a seq2seq impl.
+    - tensorflow-datasets~=4.4.0  # datasets repository, ready to use
+    - tensorflow-hub~=0.12.0  # trained ML models repository, ready to use
+    - tensorflow-probability~=0.14.1  # Optional. Probability/Stats lib.
+    - tensorflow-serving-api~=2.6.0  # or tensorflow-serving-api-gpu if gpu
+    - tensorflow~=2.6.0  # Deep Learning library
+    - tf-agents~=0.10.0  # Reinforcement Learning lib based on TensorFlow
+    - tfx~=1.3.0  # platform to deploy production ML pipelines
+    - transformers~=4.11.3  # Natural Language Processing lib for TF or PyTorch
+    - urlextract~=1.4.0  # optionally used in chapter 3, exercise 4
+    - ale-py~=0.7.2  # used only in chapter 18
+    - AutoROM~=0.4.2  # used only in chapter 18
+
+# Specific lib versions to avoid conflicts
   - attrs=20.3
   - click=7.1
   - packaging=20.9
   - six=1.15
   - typing-extensions=3.7
-
diff --git a/requirements.txt b/requirements.txt
@@ -4,20 +4,20 @@
 
 
 ##### Core scientific packages
-jupyter==1.0.0
-matplotlib==3.4.3
-numpy==1.19.5
-pandas==1.3.3
-scipy==1.7.1
+jupyter~=1.0.0
+matplotlib~=3.4.3
+numpy~=1.19.5
+pandas~=1.3.3
+scipy~=1.7.1
 
 ##### Machine Learning packages
-scikit-learn==1.0
+scikit-learn~=1.0
 
 # Optional: the XGBoost library is only used in chapter 7
-xgboost==1.4.2
+xgboost~=1.4.2
 
 # Optional: the transformers library is only using in chapter 16
-transformers==4.11.3
+transformers~=4.11.3
 
 ##### TensorFlow-related packages
 
@@ -27,70 +27,63 @@ transformers==4.11.3
 # you must install CUDA, cuDNN and more: see tensorflow.org for the detailed
 # installation instructions.
 
-tensorflow==2.6.0
+tensorflow~=2.6.0
 # Optional: the TF Serving API library is just needed for chapter 19.
-tensorflow-serving-api==2.6.0 # or tensorflow-serving-api-gpu if gpu
+tensorflow-serving-api~=2.6.0 # or tensorflow-serving-api-gpu if gpu
 
-tensorboard==2.6.0
-tensorboard-plugin-profile==2.5.0
-tensorflow-datasets==4.4.0
-tensorflow-hub==0.12.0
-tensorflow-probability==0.14.1
+tensorboard~=2.7.0
+tensorboard-plugin-profile~=2.5.0
+tensorflow-datasets~=4.4.0
+tensorflow-hub~=0.12.0
+tensorflow-probability~=0.14.1
 
 # Optional: only used in chapter 13.
-# NOT AVAILABLE ON WINDOWS
-tfx==1.3.0
+tfx~=1.3.0
 
 # Optional: only used in chapter 16.
-# NOT AVAILABLE ON WINDOWS
-tensorflow-addons==0.14.0
+tensorflow-addons~=0.14.0
 
 ##### Reinforcement Learning library (chapter 18)
 
 # There are a few dependencies you need to install first, check out:
 # https://github.com/openai/gym#installing-everything
-gym[Box2D]==0.21.0
-atari-py==0.2.5
+gym[Box2D,atari,accept-rom-license]~=0.21.0
 
-# On Windows, install atari_py using:
-# pip install --no-index -f https://github.com/Kojoley/atari-py/releases atari_py
-
-tf-agents==0.10.0
+tf-agents~=0.10.0
+ale-py~=0.7.2
 
 ##### Image manipulation
-Pillow==8.3.2
-graphviz==0.17
-opencv-python==4.5.3.56
-pyglet==1.5.21
+Pillow~=8.4.0
+graphviz~=0.17
+opencv-python~=4.5.3.56
+pyglet~=1.5.21
 
-#pyvirtualdisplay # needed in chapter 16, if on a headless server
+#pyvirtualdisplay # needed in chapter 18, if on a headless server
                   # (i.e., without screen, e.g., Colab or VM)
 
 
 ##### Additional utilities
 
 # Efficient jobs (caching, parallelism, persistence)
-joblib==0.14.1
+joblib~=0.14.1
 
 # Easy http requests
-requests==2.26.0
+requests~=2.26.0
 
 # Nice utility to diff Jupyter Notebooks.
-nbdime==3.1.0
+nbdime~=3.1.0
 
 # May be useful with Pandas for complex "where" clauses (e.g., Pandas
 # tutorial).
-numexpr==2.7.3
+numexpr~=2.7.3
 
-# Optional: these libraries can be useful in the classification chapter,
-# exercise 4.
-nltk==3.6.3
-urlextract==1.4.0
+# Optional: these libraries can be useful in the chapter 3, exercise 4.
+nltk~=3.6.5
+urlextract~=1.4.0
 
 # Optional: these libraries are only used in chapter 16
-ftfy==6.0.3
+ftfy~=6.0.3
 
 # Optional: tqdm displays nice progress bars, ipywidgets for tqdm's notebook support
-tqdm==4.62.3
-ipywidgets==7.6.5
-
+tqdm~=4.62.3
+ipywidgets~=7.6.5