Updated docs and fixed spelling

pyro/infer/autoguide/guides.py

@@ -182,7 +182,7 @@ def median(self, *args, **kwargs):
 
 class AutoGuideList(AutoGuide, nn.ModuleList):
     """
-    Container class to combine multiple automatic guides.
+    Container class to combine multiple automatic or custom guides.
 
     Example usage::
 
@@ -203,7 +203,7 @@ def _check_prototype(self, part_trace):
 
     def append(self, part):
         """
-        Add an automatic guide for part of the model. The guide should
+        Add an automatic or custom guide for part of the model. The guide should
         have been created by blocking the model to restrict to a subset of
         sample sites. No two parts should operate on any one sample site.
 
@@ -286,15 +286,15 @@ def my_local_guide(*args, **kwargs):
             ...
 
         guide = AutoGuideList(model)
-        guide.add(AutoDelta(poutine.block(model, expose=['my_global_param']))
-        guide.add(my_local_guide)  # automatically wrapped in an AutoCallable
+        guide.append(AutoDelta(poutine.block(model, expose=['my_global_param']))
+        guide.append(my_local_guide)  # automatically wrapped in an AutoCallable
 
     To specify a median callable, you can instead::
 
         def my_local_median(*args, **kwargs)
             ...
 
-        guide.add(AutoCallable(model, my_local_guide, my_local_median))
+        guide.append(AutoCallable(model, my_local_guide, my_local_median))
 
     For more complex guides that need e.g. access to plates, users should
     instead subclass ``AutoGuide``.

tutorial/source/svi_part_iv.ipynb

@@ -1,6 +1,7 @@
 {
  "cells": [
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -24,6 +25,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -43,6 +45,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -65,6 +68,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -84,6 +88,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -112,6 +117,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -135,6 +141,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -146,7 +153,7 @@
     "\n",
     "Sometimes the problem is actually in your model even though you think it's in the guide. \n",
     "Conversely, sometimes the problem is in your guide even though you think it's in the model or somewhere else. \n",
-    "For these reasons it can be helpful to reduce the number of moving parts while you try to identify the underyling issue.\n",
+    "For these reasons it can be helpful to reduce the number of moving parts while you try to identify the underlying issue.\n",
     "One convenient way to do this is to replace your custom guide with a [pyro.infer.AutoGuide](http://docs.pyro.ai/en/stable/infer.autoguide.html#module-pyro.infer.autoguide). \n",
     "\n",
     "For example, if all the latent variables in your model are continuous, you can try a [pyro.infer.AutoNormal](http://docs.pyro.ai/en/stable/infer.autoguide.html#autonormal) guide.\n",
@@ -164,6 +171,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -196,6 +204,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -216,16 +225,18 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
     "### 9. Use `TraceMeanField_ELBO` if applicable\n",
     "\n",
     "The basic `ELBO` implementation in Pyro, [Trace_ELBO](http://docs.pyro.ai/en/stable/inference_algos.html?highlight=tracemeanfield#pyro.infer.trace_elbo.Trace_ELBO), uses stochastic samples to estimate the KL divergence term. \n",
-    "When analytic KL diverences are available, you may be able to lower ELBO variance by using analytic KL divergences instead. This functionality is provided by [TraceMeanField_ELBO](http://docs.pyro.ai/en/stable/inference_algos.html?highlight=tracemeanfield#pyro.infer.trace_elbo.Trace_ELBO)."
+    "When analytic KL divergences are available, you may be able to lower ELBO variance by using analytic KL divergences instead. This functionality is provided by [TraceMeanField_ELBO](https://docs.pyro.ai/en/stable/inference_algos.html?highlight=tracemeanfield#pyro.infer.trace_mean_field_elbo.TraceMeanField_ELBO)."
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -243,6 +254,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -279,6 +291,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -297,6 +310,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -306,6 +320,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -316,6 +331,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [
@@ -330,6 +346,7 @@
    ]
   },
   {
+   "attachments": {},
    "cell_type": "markdown",
    "metadata": {},
    "source": [