Fix BART CNN/DM fine-tuning instructions (#1650)

Summary:
The first step in the CNN/DM fine-tuning instructions for BART is misleading (see facebookresearch/fairseq#1391). This PR fixes the README and adds links to facebookresearch/fairseq#1391 as well as to a repository with CNN/DM processing code adjusted for BART.
Pull Request resolved: facebookresearch/fairseq#1650

Differential Revision: D19606689

fbshipit-source-id: 4f1771f47d3650035a911ab393ab6df2193c1bf9

examples/bart/README.cnn.md

@@ -1,6 +1,8 @@
 # Fine-tuning BART on CNN-Dailymail summarization task
 
-### 1) Follow instructions [here](https://github.com/abisee/cnn-dailymail) to download and process into data-files with non-tokenized cased samples.
+### 1) Download the CNN and Daily Mail data and preprocess it into data files with non-tokenized cased samples.
+
+Follow the instructions [here](https://github.com/abisee/cnn-dailymail) to download the original CNN and Daily Mail datasets. To preprocess the data, refer to the pointers in [this issue](https://github.com/pytorch/fairseq/issues/1391) or check out the code [here](https://github.com/artmatsak/cnn-dailymail).
 
 ### 2) BPE preprocess:
 ```bash