Datasets:
machineteacher
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -54,9 +54,8 @@ The dataset is in JSON format.
|
|
54 |
|
55 |
|
56 |
## Considerations for Using the Data
|
57 |
-
Please note that this dataset contains
|
58 |
This is because this public release includes only the instances that were acquired and curated from publicly available datasets.
|
59 |
-
Specifically, it is missing roughly 13k instances in training and 1.5k instances in validation data from Simplification and Formality Transfer tasks due to licensing restrictions.
|
60 |
|
61 |
Following are the details of the subsets (including the ones we are unable to publicly release):
|
62 |
|
|
|
54 |
|
55 |
|
56 |
## Considerations for Using the Data
|
57 |
+
Please note that this dataset contains 102k instances (as opposed to the 190k instances we used in the paper).
|
58 |
This is because this public release includes only the instances that were acquired and curated from publicly available datasets.
|
|
|
59 |
|
60 |
Following are the details of the subsets (including the ones we are unable to publicly release):
|
61 |
|