fix typo in 10_nlp.ipynb: 'batch will tend of be of similar lengths' -> 'batch will tend to be of similar lengths' (#406)

Author: dastrobu

10_nlp.ipynb

@@ -1884,7 +1884,7 @@
    "source": [
     "Remember, PyTorch `DataLoader`s need to collate all the items in a batch into a single tensor, and a single tensor has a fixed shape (i.e., it has some particular length on every axis, and all items must be consistent). This should sound familiar: we had the same issue with images. In that case, we used cropping, padding, and/or squishing to make all the inputs the same size. Cropping might not be a good idea for documents, because it seems likely we'd remove some key information (having said that, the same issue is true for images, and we use cropping there; data augmentation hasn't been well explored for NLP yet, so perhaps there are actually opportunities to use cropping in NLP too!). You can't really \"squish\" a document. So that leaves padding!\n",
     "\n",
-    "We will expand the shortest texts to make them all the same size. To do this, we use a special padding token that will be ignored by our model. Additionally, to avoid memory issues and improve performance, we will batch together texts that are roughly the same lengths (with some shuffling for the training set). We do this by (approximately, for the training set) sorting the documents by length prior to each epoch. The result of this is that the documents collated into a single batch will tend of be of similar lengths. We won't pad every batch to the same size, but will instead use the size of the largest document in each batch as the target size. (It is possible to do something similar with images, which is especially useful for irregularly sized rectangular images, but at the time of writing no library provides good support for this yet, and there aren't any papers covering it. It's something we're planning to add to fastai soon, however, so keep an eye on the book's website; we'll add information about this as soon as we have it working well.)\n",
+    "We will expand the shortest texts to make them all the same size. To do this, we use a special padding token that will be ignored by our model. Additionally, to avoid memory issues and improve performance, we will batch together texts that are roughly the same lengths (with some shuffling for the training set). We do this by (approximately, for the training set) sorting the documents by length prior to each epoch. The result of this is that the documents collated into a single batch will tend to be of similar lengths. We won't pad every batch to the same size, but will instead use the size of the largest document in each batch as the target size. (It is possible to do something similar with images, which is especially useful for irregularly sized rectangular images, but at the time of writing no library provides good support for this yet, and there aren't any papers covering it. It's something we're planning to add to fastai soon, however, so keep an eye on the book's website; we'll add information about this as soon as we have it working well.)\n",
     "\n",
     "The sorting and padding are automatically done by the data block API for us when using a `TextBlock`, with `is_lm=False`. (We don't have this same issue for language model data, since we concatenate all the documents together first, and then split them into equally sized sections.)\n",
     "\n",
@@ -2283,4 +2283,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 2
-}
+}