Quantum reinforcement learning tutorial (#541)

* QRL tutorial * QRL tutorial v2 * QRL tutorial v2.1 * QRL tutorial v2.1 * QRL tutorial v3 * QRL tutorial v3.1 * QRL tutorial v4.0 * QRL tutorial v4.1 * QRL tutorial v4.2 * QRL tutorial v4.3 * QRL tutorial v4.4 Co-authored-by: MichaelBroughton <MichaelBroughton@users.noreply.github.com>

Author: sjerbi

docs/_book.yaml

@@ -50,6 +50,8 @@ upper_tabs:
  path: /quantum/tutorials/qcnn
  - title: "Research tools"
  path: /quantum/tutorials/research_tools
+ - title: "Quantum reinforcement learning"
+ path: /quantum/tutorials/quantum_reinforcement_learning
 
  - name: API
  skip_translation: true

docs/tutorials/images/gym_CartPole.gif

@@ -18,6 +18,8 @@
 pip install jupyter nbformat==4.4.0 nbconvert==5.5.0 jupyter-client==6.1.12 ipython==7.22.0
 # Workaround for ipykernel - see https://github.com/ipython/ipykernel/issues/422
 pip install ipykernel==5.1.1
+# OpenAI Gym pip package needed for the quantum reinforcement learning tutorial
+pip install gym==0.18.0
 # Leave the quantum directory, otherwise errors may occur
 cd ..
 examples_output=$(python3 quantum/scripts/test_tutorials.py)

docs/tutorials/images/pqc_re-uploading.png

@@ -40,6 +40,8 @@ def test_notebook(self, path):
  src = re.sub(r'\!(?!=)', r'#!', src)
  # For mnist.ipynb to reduce runtime in test.
  src = re.sub('NUM_EXAMPLES ?= ?.*', 'NUM_EXAMPLES = 10', src)
+ # For quantum_reinforcement_learning.ipynb to reduce runtime in test.
+ src = re.sub('n_episodes ?= ?.*', 'n_episodes = 50', src)
  cell['source'] = src
 
  _ = nbconvert.preprocessors.execute.executenb(nb,