improved yield
Browse files- ted_talks.py +5 -3
ted_talks.py
CHANGED
@@ -53,7 +53,10 @@ class TedTalks(datasets.GeneratorBasedBuilder):
|
|
53 |
return datasets.DatasetInfo(
|
54 |
description=_DESCRIPTION,
|
55 |
features=datasets.Features(
|
56 |
-
{
|
|
|
|
|
|
|
57 |
),
|
58 |
homepage="https://github.com/neulab/word-embeddings-for-nmt",
|
59 |
citation=_CITATION,
|
@@ -147,6 +150,5 @@ class TedTalks(datasets.GeneratorBasedBuilder):
|
|
147 |
|
148 |
if len(source_target_pairs) > 0:
|
149 |
source_sentences, target_sentences = zip(*source_target_pairs)
|
150 |
-
|
151 |
for idx, (l1, l2) in enumerate(zip(source_sentences, target_sentences)):
|
152 |
-
yield idx, {
|
|
|
53 |
return datasets.DatasetInfo(
|
54 |
description=_DESCRIPTION,
|
55 |
features=datasets.Features(
|
56 |
+
{
|
57 |
+
self.config.source: datasets.features.Value("string"),
|
58 |
+
self.config.target: datasets.features.Value("string"),
|
59 |
+
}
|
60 |
),
|
61 |
homepage="https://github.com/neulab/word-embeddings-for-nmt",
|
62 |
citation=_CITATION,
|
|
|
150 |
|
151 |
if len(source_target_pairs) > 0:
|
152 |
source_sentences, target_sentences = zip(*source_target_pairs)
|
|
|
153 |
for idx, (l1, l2) in enumerate(zip(source_sentences, target_sentences)):
|
154 |
+
yield idx, {self.config.source: l1, self.config.target: l2}
|