Update dataset card with number of instances

This PR updates the dataset card with:
- instance example
- number of instances
- single split information

This also adds a note to make clear that each instance contains the full text of the document.

Related to #1.

README.md

@@ -89,16 +89,28 @@ OneStopEnglish is a corpus of texts written at three reading levels, and demonst
 
 ### Data Instances
 
-[More Information Needed]
+An instance example:
+```
+{
+  "text": "When you see the word Amazon, what’s the first thing you think...",
+  "label": 0
+}
+```
+
+Note that each instance contains the full text of the document.
 
 ### Data Fields
 
-- text: document text 
-- label: reading level of the document- ele/int/adv (Elementary/Intermediate/Advance)
+- `text`: Full document text. 
+- `label`: Reading level of the document- ele/int/adv (Elementary/Intermediate/Advance).
 
 ### Data Splits
 
-[More Information Needed]
+The OneStopEnglish dataset has a single _train_ split.
+
+| Split | Number of instances |
+|-------|--------------------:|
+| train | 567                 |
 
 ## Dataset Creation