Update text to match Paperspace setup

- Change requirements in intro cell

Author: nmb-paperspace

get-started/PyG-SchNetGNN.ipynb

@@ -21,8 +21,8 @@
     "\n",
     "Requirements:\n",
     "\n",
-    "* A Poplar SDK environment enabled (see the Getting Started guide for your IPU system\n",
-    "* Python packages installed with `pip install -r requirements.txt`\n"
+    "* Paperspace account with access to the PyTorch IPU runtime\n",
+    "* Python packages in `requirements-pyg.txt`, installed below\n"
    ]
   },
   {
@@ -344,7 +344,7 @@
     "* [torch_geometric.data.DataLoader](https://pytorch-geometric.readthedocs.io/en/latest/modules/data.html#torch_geometric.data.DataLoader)\n",
     "\n",
     "\n",
-    "The PyG dataloader supports a form of mini-batching which is [decribed here](https://pytorch-geometric.readthedocs.io/en/latest/notes/batching.html).  Effectively each mini-batch is a concatenation of multiple graphs (molecules in QM9).  Another way to understand this is that each mini-batch is one large graph comprised of multiple disconnected sub-graphs.  The PyG dataloader will generate a `batch` vector that assigns each feature in the mini-batch into a distinct subgraph.  This is useful for message passing networks (such as SchNet) and pooling layers to produce a distinct regression prediction for each molecule. Refer to the following tutorials for additional background:\n",
+    "The PyG dataloader supports a form of mini-batching which is [described here](https://pytorch-geometric.readthedocs.io/en/latest/notes/batching.html).  Effectively each mini-batch is a concatenation of multiple graphs (molecules in QM9).  Another way to understand this is that each mini-batch is one large graph comprised of multiple disconnected sub-graphs.  The PyG dataloader will generate a `batch` vector that assigns each feature in the mini-batch into a distinct subgraph.  This is useful for message passing networks (such as SchNet) and pooling layers to produce a distinct regression prediction for each molecule. Refer to the following tutorials for additional background:\n",
     "\n",
     "* [Creating message passing networks](https://pytorch-geometric.readthedocs.io/en/latest/notes/create_gnn.html)\n",
     "* [Global Pooling Layers](https://pytorch-geometric.readthedocs.io/en/latest/modules/nn.html?highlight=pooling#global-pooling-layers)\n",
@@ -1335,7 +1335,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3.6.9 ('pygnb')",
+   "display_name": "Python 3",
    "language": "python",
    "name": "python3"
   },
@@ -1349,15 +1349,14 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.13"
+   "version": "3.8.10"
   },
-  "orig_nbformat": 4,
   "vscode": {
    "interpreter": {
     "hash": "5918304a9e1be506ae80f1c157e6e427d374131e966ae13a9bf2405fbc3138af"
    }
   }
  },
  "nbformat": 4,
- "nbformat_minor": 2
+ "nbformat_minor": 4
 }