Update AMSGrad example (#642)

docs/_static/notebooks/amsgrad.ipynb

@@ -14,7 +14,7 @@
     "paper, the authors determine a flaw in the convergence proof of the ubiquitous ADAM optimizer. They also give an example\n",
     "of a simple function for which ADAM does not converge to the correct solution. We've seen how torchbearer can be used\n",
     "for simple function optimisation before and we can do something similar to reproduce the results\n",
-    "from the paper.\n",
+    "from the paper. We should note that this isn't a suggestion that you should necesarily use AMSGrad (the proposed solution from the paper), most work still uses either SGD with momentum or vanilla Adam and some other approaches have been proposed such as [AdamW](https://arxiv.org/abs/1711.05101) which may provide benefits. Our intention here is simply to show how this interesting failure case can be demonstrated with torchbearer.\n",
     "\n",
     "**Note**: The easiest way to use this tutorial is as a colab notebook, which allows you to dive in with no setup. We recommend you enable a free GPU with\n",
     "\n",
@@ -2077,8 +2077,15 @@
     "\n",
     "So, whatever your thoughts on the AMSGrad optimizer in practice, it's probably the sign of a good paper that you can\n",
     "re-implement the example and get very similar results without having to try too hard and (thanks to torchbearer) only\n",
-    "writing a small amount of code. The paper includes some more complex, 'real-world' examples, can you re-implement those\n"
+    "writing a small amount of code. The paper includes some more complex, 'real-world' examples, can you re-implement those?\n"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {
@@ -2102,7 +2109,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.8"
+   "version": "3.7.3"
   }
  },
  "nbformat": 4,