Modify our navigational structure to handle pytorch/standalone bifurc…

…ation
qognitive · Sep 4, 2024 · 5ef89ea · 5ef89ea
1 parent 26c9ef0
commit 5ef89ea
Show file tree

Hide file tree

Showing 20 changed files with 107 additions and 33 deletions.
diff --git a/docs/conf.py b/docs/conf.py
@@ -34,3 +34,18 @@
 
 html_theme = "sphinx_rtd_theme"
 html_static_path = ["_static"]
+
+html_theme_options = {
+    # 'analytics_id': 'G-XXXXXXXXXX',  #  Provided by Google in your dashboard
+    # 'analytics_anonymize_ip': False,
+    # 'logo_only': False,
+    'display_version': True,
+    'prev_next_buttons_location': 'bottom',
+    'style_external_links': False,
+    # Toc options
+    'collapse_navigation': True,
+    'sticky_navigation': True,
+    'navigation_depth': 4,
+    'includehidden': True,
+    'titles_only': False
+}
diff --git a/docs/examples/pytorch/index.rst b/docs/examples/pytorch/index.rst
@@ -0,0 +1,6 @@
+PyTorch QCML Model Examples
+===========================
+
+.. toctree::
+
+   /examples/pytorch/mnist
diff --git a/docs/examples/pytorch/mnist.rst b/docs/examples/pytorch/mnist.rst
@@ -0,0 +1,4 @@
+QCML PyTorch MNIST
+==================
+
+pass
diff --git a/docs/examples/coil20.rst → docs/examples/stand_alone/coil20.rst b/docs/examples/coil20.rst → docs/examples/stand_alone/coil20.rst
diff --git a/docs/examples/stand_alone/index.rst b/docs/examples/stand_alone/index.rst
@@ -0,0 +1,8 @@
+Stand Alone QCML Model Examples
+===============================
+
+.. toctree::
+
+   /examples/stand_alone/coil20
+   /examples/stand_alone/wisconsin
+   /examples/stand_alone/timeseries
diff --git a/docs/examples/timeseries.rst → docs/examples/stand_alone/timeseries.rst b/docs/examples/timeseries.rst → docs/examples/stand_alone/timeseries.rst
diff --git a/docs/examples/wisconsin.rst → docs/examples/stand_alone/wisconsin.rst b/docs/examples/wisconsin.rst → docs/examples/stand_alone/wisconsin.rst
diff --git a/docs/index.rst b/docs/index.rst
@@ -2,43 +2,36 @@ Qognitive Python Client
 =======================
 
 .. toctree::
-   :maxdepth: 2
    :hidden:
    :caption: Getting started
 
    /tutorial/install
    /tutorial/datasets
    /tutorial/models
-   /tutorial/training
-   /tutorial/inference
 
 .. toctree::
-   :maxdepth: 2
    :hidden:
    :caption: Examples
 
-   /examples/wisconsin
-   /examples/coil20
-   /examples/timeseries
+   /examples/stand_alone/index
+   /examples/pytorch/index
 
 .. toctree::
-   :maxdepth: 2
    :hidden:
    :caption: Models
 
-   /models/pauli
-   /models/ensemble
+   /models/stand_alone/pauli
+   /models/stand_alone/ensemble
+   /models/pytorch/qcml_layer
 
 .. toctree::
-   :maxdepth: 2
    :hidden:
-   :caption: Parameters
+   :caption: QCML Stand Alone Parameters
 
    /parameters/optimization
    /parameters/state
 
 .. toctree::
-   :maxdepth: 2
    :hidden:
    :caption: API Reference
 

diff --git a/docs/models/pytorch/qcml_layer.rst b/docs/models/pytorch/qcml_layer.rst
@@ -0,0 +1,4 @@
+QCML Layer
+==========
+
+The QCML Pytorch layer
diff --git a/docs/models/ensemble.rst → docs/models/stand_alone/ensemble.rst b/docs/models/ensemble.rst → docs/models/stand_alone/ensemble.rst
diff --git a/docs/models/pauli.rst → docs/models/stand_alone/pauli.rst b/docs/models/pauli.rst → docs/models/stand_alone/pauli.rst
diff --git a/docs/tutorial/models.rst b/docs/tutorial/models.rst
@@ -1,24 +1,15 @@
 Model Selection
 ===============
 
-Here we will choose one of our QCML models and parameterize it for training.  There are several QCML models that you can choose from and we are continuing to develop new models as well as improve existing ones.  You can see which models are currently available on the sidebar.
+Here we will choose one of our QCML models and parameterize it for training. There are several QCML models that you can choose from and we are continuing to develop new models as well as improve existing ones. You can see which models are currently available on the sidebar.
+about:blank#blocked
+There are two categories of models. The first is a stand alone QCML model, which has at its core a single cost function that involves the quantum state. The second is a QCML PyTorch layer which has been built so that you can integrate it with classical ML deep learning layers, and so develop a hybrid quantum-classical model whose architecture fits your needs. The cost function is defined by you which is a classical cost function where QCML will be a part of the network.
 
-We're going to use the Pauli model for this example.
+Setting up each is a little different, so we'll cover them separately.
 
-Setting the Model Parameters
-----------------------------
+.. toctree::
+   :caption: QCML Models
 
-We configure our model by a call on our ``qcml`` object, which returns the object with the model configuration. This is the same procedure for both sync and async clients.
+   /tutorial/stand_alone/index
+   /tutorial/pytorch/index
 
-.. code-block:: python
-
-    qcml = qcml.pauli(
-        operators=["X", "Y", "Z"],
-        qbits=2,
-        pauli_weight=2
-    )
-
-You can read more about the pauli model, the parameterization, and the available options on the :doc:`/models/pauli` page.
-
-.. note::
-    The operators list does have to match the dataset that you are going to be running on. This might seem redundant but it happens in lots of other machine learning models where you have to match the dimensionality of your input to the dimensions of the model - or at least do some preprocessing to get the data into the proper scale.
diff --git a/docs/tutorial/pytorch/index.rst b/docs/tutorial/pytorch/index.rst
@@ -0,0 +1,9 @@
+PyTorch QCML
+============
+
+
+.. toctree::
+
+   /tutorial/pytorch/models
+   /tutorial/pytorch/train
+   /tutorial/pytorch/inference
diff --git a/docs/tutorial/pytorch/inference.rst b/docs/tutorial/pytorch/inference.rst
@@ -0,0 +1,4 @@
+Inference with PyTorch
+======================
+
+Inference on QCML PyTorch models will be coming mid Sept 2024!
diff --git a/docs/tutorial/pytorch/models.rst b/docs/tutorial/pytorch/models.rst
@@ -0,0 +1,4 @@
+Building a PyTorch Train Function
+=================================
+
+PyTorch
diff --git a/docs/tutorial/pytorch/train.rst b/docs/tutorial/pytorch/train.rst
@@ -0,0 +1,4 @@
+PyTorch Training
+================
+
+Call the train
diff --git a/docs/tutorial/stand_alone/index.rst b/docs/tutorial/stand_alone/index.rst
@@ -0,0 +1,8 @@
+Stand Alone QCML
+================
+
+.. toctree::
+
+   /tutorial/stand_alone/models
+   /tutorial/stand_alone/train
+   /tutorial/stand_alone/inference
diff --git a/docs/tutorial/inference.rst → docs/tutorial/stand_alone/inference.rst b/docs/tutorial/inference.rst → docs/tutorial/stand_alone/inference.rst
@@ -74,6 +74,6 @@ That's it!  You have all the pieces to get started with the QCML library. You ca
 
 Try some more complicated examples and see how the model performs, or dive right into your own data.
 
-* :doc:`/examples/wisconsin`
-* :doc:`/examples/coil20`
-* :doc:`/examples/timeseries`
+* :doc:`/examples/stand_alone/wisconsin`
+* :doc:`/examples/stand_alone/coil20`
+* :doc:`/examples/stand_alone/timeseries`
diff --git a/docs/tutorial/stand_alone/models.rst b/docs/tutorial/stand_alone/models.rst
@@ -0,0 +1,24 @@
+Stand Alone Models
+==================
+
+Here we will choose one of our QCML models and parameterize it for training.  There are several QCML models that you can choose from and we are continuing to develop new models as well as improve existing ones.  You can see which models are currently available on the sidebar.
+
+We're going to use the Pauli model for this example.
+
+Setting the Model Parameters
+----------------------------
+
+We configure our model by a call on our ``qcml`` object, which returns the object with the model configuration. This is the same procedure for both sync and async clients.
+
+.. code-block:: python
+
+    qcml = qcml.pauli(
+        operators=["X", "Y", "Z"],
+        qbits=2,
+        pauli_weight=2
+    )
+
+You can read more about the pauli model, the parameterization, and the available options on the :doc:`/models/stand_alone/pauli` page.
+
+.. note::
+    The operators list does have to match the dataset that you are going to be running on. This might seem redundant but it happens in lots of other machine learning models where you have to match the dimensionality of your input to the dimensions of the model - or at least do some preprocessing to get the data into the proper scale.
diff --git a/docs/tutorial/training.rst → docs/tutorial/stand_alone/train.rst b/docs/tutorial/training.rst → docs/tutorial/stand_alone/train.rst