Update DiscEvalMT.py
Browse files- DiscEvalMT.py +14 -0
DiscEvalMT.py
CHANGED
|
@@ -24,6 +24,10 @@ _CITATION = '''\
|
|
| 24 |
_DESCRIPTION = '''\
|
| 25 |
English-French hand-crafted contrastive test set to test anaphora and lexical choice
|
| 26 |
'''
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
|
| 28 |
_URLS = {
|
| 29 |
'test-lexical_choice': 'lexical_choice/eval.json',
|
|
@@ -66,6 +70,16 @@ class DiscEvalMT(datasets.GeneratorBasedBuilder):
|
|
| 66 |
),
|
| 67 |
]
|
| 68 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
def _split_generators(self, dl_manager):
|
| 70 |
downloaded_files = dl_manager.download_and_extract(_URLS)
|
| 71 |
|
|
|
|
| 24 |
_DESCRIPTION = '''\
|
| 25 |
English-French hand-crafted contrastive test set to test anaphora and lexical choice
|
| 26 |
'''
|
| 27 |
+
|
| 28 |
+
_URL='https://github.com/rbawden/discourse-mt-test-sets/tree/master'
|
| 29 |
+
|
| 30 |
+
_LICENSE = 'CC-BY-SA-4.0'
|
| 31 |
|
| 32 |
_URLS = {
|
| 33 |
'test-lexical_choice': 'lexical_choice/eval.json',
|
|
|
|
| 70 |
),
|
| 71 |
]
|
| 72 |
|
| 73 |
+
def _info(self):
|
| 74 |
+
return datasets.DatasetInfo(
|
| 75 |
+
description=_DESCRIPTION,
|
| 76 |
+
features=datasets.Features({"id": datasets.Value("int64"), "text": datasets.Value("string")}),
|
| 77 |
+
supervised_keys=None,
|
| 78 |
+
homepage=_URL,
|
| 79 |
+
citation=_CITATION,
|
| 80 |
+
license=_LICENSE,
|
| 81 |
+
)
|
| 82 |
+
|
| 83 |
def _split_generators(self, dl_manager):
|
| 84 |
downloaded_files = dl_manager.download_and_extract(_URLS)
|
| 85 |
|