Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -77,7 +77,8 @@ load_dataset("clips/mfaq", "en")
|
|
77 |
| Croatian | hr | 5,215 | 819 |
|
78 |
|
79 |
## Data Fields
|
80 |
-
|
|
|
81 |
- **id**
|
82 |
- **language**
|
83 |
- **num_pairs**: the number of FAQs on the page
|
@@ -87,8 +88,14 @@ The dataset is organized by page. Each page contains a list of questions and ans
|
|
87 |
- **answer**
|
88 |
- **language**
|
89 |
|
90 |
-
|
91 |
-
You can access the flat version of any language by appending `_flat` to the configuration (e.g. `en_flat`). The data will be returned pair-by-pair instead of page-by-page.
|
|
|
|
|
|
|
|
|
|
|
|
|
92 |
|
93 |
## Source Data
|
94 |
|
|
|
77 |
| Croatian | hr | 5,215 | 819 |
|
78 |
|
79 |
## Data Fields
|
80 |
+
#### Nested (per page - default)
|
81 |
+
The data is organized by page. Each page contains a list of questions and answers.
|
82 |
- **id**
|
83 |
- **language**
|
84 |
- **num_pairs**: the number of FAQs on the page
|
|
|
88 |
- **answer**
|
89 |
- **language**
|
90 |
|
91 |
+
#### Flattened
|
92 |
+
The data is organized by pair (i.e. pages are flattened). You can access the flat version of any language by appending `_flat` to the configuration (e.g. `en_flat`). The data will be returned pair-by-pair instead of page-by-page.
|
93 |
+
- **id**
|
94 |
+
- **language**
|
95 |
+
- **num_pairs**: the number of FAQs on the page
|
96 |
+
- **domain**: source web domain of the FAQs
|
97 |
+
- **question**
|
98 |
+
- **answer**
|
99 |
|
100 |
## Source Data
|
101 |
|