📖 Update documentation

docs/source/_static/css/custom.css

@@ -87,3 +87,7 @@
 .tutorials-header .header-logo {
     background-image: url("../images/torch_uncertainty.png");
 }
+
+.figure-caption {
+    text-align: center;
+}

docs/source/index.rst

@@ -43,7 +43,6 @@ the following paper:
 
    installation
    quickstart
-   introduction_uncertainty
    auto_tutorials/index
    cli_guide
    api

docs/source/introduction_uncertainty.rst

@@ -1,3 +1,5 @@
+:orphan:
+
 Introduction to Classification Uncertainty
 ==========================================
 

docs/source/quickstart.rst

@@ -15,6 +15,12 @@ Yet, we take account that their will be as many different uses of TorchUncertain
 This page provides ideas on how to benefit from TorchUncertainty at all levels: from ready-to-train lightning-based models to using only specific
 PyTorch layers. 
 
+.. figure:: _static/images/structure_torch_uncertainty.jpg
+  :alt: TorchUncertainty structure
+  :figclass: figure-caption
+
+  **TorchUncertainty structure**
+
 Training with TorchUncertainty's Uncertainty-aware Routines
 -----------------------------------------------------------
 

torch_uncertainty/metrics/regression/log10.py

@@ -17,7 +17,8 @@ def __init__(self, **kwargs) -> None:
             - :attr:`target`: :math:`(N)`
 
         Args:
-            kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
+            kwargs: Additional keyword arguments, see `Advanced metric settings
+                <https://torchmetrics.readthedocs.io/en/stable/pages/overview.html#metric-kwargs>`_.
         """
         super().__init__(**kwargs)
         self.add_state(

torch_uncertainty/metrics/regression/mse_log.py

@@ -4,7 +4,7 @@
 
 class MeanSquaredLogError(MeanSquaredError):
     def __init__(self, squared: bool = True, **kwargs) -> None:
-        r"""`Compute MeanSquaredLogError`_ (MSELog).
+        r"""Compute MeanSquaredLogError (MSELog).
 
         .. math:: \text{MSELog} = \frac{1}{N}\sum_i^N  (\log \hat{y_i} - \log y_i)^2
 
@@ -22,7 +22,8 @@ def __init__(self, squared: bool = True, **kwargs) -> None:
 
         Args:
             squared: If True returns MSELog value, if False returns EMSELog value.
-            kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
+            kwargs: Additional keyword arguments, see `Advanced metric settings
+                <https://torchmetrics.readthedocs.io/en/stable/pages/overview.html#metric-kwargs>`_.
 
         Reference:
             As in e.g. From big to small: Multi-scale local planar guidance for monocular depth estimation

torch_uncertainty/metrics/regression/relative_error.py

@@ -5,7 +5,7 @@
 
 class MeanGTRelativeAbsoluteError(MeanAbsoluteError):
     def __init__(self, **kwargs) -> None:
-        r"""`Compute Mean Absolute Error relative to the Ground Truth`_ (MAErel or ARE).
+        r"""Compute Mean Absolute Error relative to the Ground Truth (MAErel or ARE).
 
         .. math:: \text{MAErel} = \frac{1}{N}\sum_i^N \frac{| y_i - \hat{y_i} |}{y_i}
 
@@ -22,7 +22,8 @@ def __init__(self, **kwargs) -> None:
             relative mean absolute error over the state
 
         Args:
-            kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
+            kwargs: Additional keyword arguments, see `Advanced metric settings
+                <https://torchmetrics.readthedocs.io/en/stable/pages/overview.html#metric-kwargs>`_.
 
         Reference:
             As in e.g. From big to small: Multi-scale local planar guidance for monocular depth estimation
@@ -38,7 +39,7 @@ class MeanGTRelativeSquaredError(MeanSquaredError):
     def __init__(
         self, squared: bool = True, num_outputs: int = 1, **kwargs
     ) -> None:
-        r"""Compute `mean squared error relative to the Ground Truth`_ (MSErel or SRE).
+        r"""Compute mean squared error relative to the Ground Truth (MSErel or SRE).
 
         .. math:: \text{MSErel} = \frac{1}{N}\sum_i^N \frac{(y_i - \hat{y_i})^2}{y_i}
 
@@ -56,7 +57,8 @@ def __init__(
         Args:
             squared: If True returns MSErel value, if False returns RMSErel value.
             num_outputs: Number of outputs in multioutput setting
-            kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
+            kwargs: Additional keyword arguments, see `Advanced metric settings
+                <https://torchmetrics.readthedocs.io/en/stable/pages/overview.html#metric-kwargs>`_.
 
         Reference:
             As in e.g. From big to small: Multi-scale local planar guidance for monocular depth estimation

torch_uncertainty/metrics/regression/silog.py

@@ -20,7 +20,8 @@ def __init__(self, lmbda: float = 1, **kwargs: Any) -> None:
 
         Args:
             lmbda: The regularization parameter on the variance of error (default 1).
-            kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
+            kwargs: Additional keyword arguments, see `Advanced metric settings
+                <https://torchmetrics.readthedocs.io/en/stable/pages/overview.html#metric-kwargs>`_.
 
         Reference:
             Depth Map Prediction from a Single Image using a Multi-Scale Deep Network.

torch_uncertainty/metrics/regression/threshold_accuracy.py

@@ -12,7 +12,8 @@ def __init__(self, power: int, lmbda: float = 1.25, **kwargs) -> None:
             power: The power to raise the threshold to. Often in [1, 2, 3].
             lmbda: The threshold to compare the max of ratio of predictions
                 to targets and its inverse to. Defaults to 1.25.
-            kwargs: Additional arguments to pass to the metric class.
+            kwargs: Additional keyword arguments, see `Advanced metric settings
+                <https://torchmetrics.readthedocs.io/en/stable/pages/overview.html#metric-kwargs>`_.
         """
         super().__init__(**kwargs)
         if power < 0: