RicardoRei
commited on
Commit
·
f3ab5f3
1
Parent(s):
0a1fb01
Update README.md
Browse files
README.md
CHANGED
@@ -36,7 +36,7 @@ tags:
|
|
36 |
|
37 |
# Dataset Summary
|
38 |
|
39 |
-
This dataset contains all
|
40 |
|
41 |
The data is organised into 8 columns:
|
42 |
- lp: language pair
|
@@ -49,14 +49,13 @@ The data is organised into 8 columns:
|
|
49 |
- domain: domain of the input text (e.g. news)
|
50 |
- year: collection year
|
51 |
|
52 |
-
You can also find the original data
|
53 |
-
|
54 |
|
55 |
## Python usage:
|
56 |
|
57 |
```python
|
58 |
from datasets import load_dataset
|
59 |
-
dataset = load_dataset("RicardoRei/wmt-
|
60 |
```
|
61 |
|
62 |
There is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :
|
@@ -69,5 +68,7 @@ data = dataset.filter(lambda example: example["year"] == 2022)
|
|
69 |
data = dataset.filter(lambda example: example["lp"] == "en-de")
|
70 |
|
71 |
# split by domain
|
72 |
-
data = dataset.filter(lambda example: example["domain"] == "
|
73 |
-
```
|
|
|
|
|
|
36 |
|
37 |
# Dataset Summary
|
38 |
|
39 |
+
This dataset contains all DA human annotations from previous WMT News Translation shared tasks.
|
40 |
|
41 |
The data is organised into 8 columns:
|
42 |
- lp: language pair
|
|
|
49 |
- domain: domain of the input text (e.g. news)
|
50 |
- year: collection year
|
51 |
|
52 |
+
You can also find the original data for each year in the results section https://www.statmt.org/wmt{YEAR}/results.html e.g: for 2020 data: [https://www.statmt.org/wmt20/results.html](https://www.statmt.org/wmt20/results.html)
|
|
|
53 |
|
54 |
## Python usage:
|
55 |
|
56 |
```python
|
57 |
from datasets import load_dataset
|
58 |
+
dataset = load_dataset("RicardoRei/wmt-da-human-evaluation", split="train")
|
59 |
```
|
60 |
|
61 |
There is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :
|
|
|
68 |
data = dataset.filter(lambda example: example["lp"] == "en-de")
|
69 |
|
70 |
# split by domain
|
71 |
+
data = dataset.filter(lambda example: example["domain"] == "news")
|
72 |
+
```
|
73 |
+
|
74 |
+
Note that most data is from News domain.
|