|
|
import datasets |
|
|
import os |
|
|
import json |
|
|
|
|
|
|
|
|
_CITATION = "" |
|
|
_DESCRIPTION = """ |
|
|
The 2014 Workshop on Statistical Machine Translation: |
|
|
https://aclanthology.org/W14-3302.pdf |
|
|
|
|
|
The scenario consists of 5 subsets, each of which is a parallel corpus between English and another language. The |
|
|
non-English languages include Czech, German, French, Hindi, and Russian. |
|
|
|
|
|
For each language pair, the validation and test set each includes around 3,000 examples, while the training set is |
|
|
usually much larger. We therefore randomly downsample the training set to speedup data processing. |
|
|
|
|
|
Task prompt structure: |
|
|
|
|
|
Translate {source_language} to {target_language}: |
|
|
{Hypothesis} = {Reference} |
|
|
|
|
|
Example from WMT14 Fr-En: |
|
|
|
|
|
Hypothesis: Assemblée générale |
|
|
Reference: General Assembly |
|
|
""" |
|
|
|
|
|
class WMT14(datasets.GeneratorBasedBuilder): |
|
|
VERSION = datasets.Version("1.0.0") |
|
|
|
|
|
BUILDER_CONFIGS = [ |
|
|
datasets.BuilderConfig(name=name, version=datasets.Version("1.0.0"), description="") |
|
|
for name in ["cs-en", "de-en", "fr-en", "hi-en", "ru-en"] |
|
|
] |
|
|
|
|
|
def _info(self): |
|
|
source_language, target_language = self.config.name.split('-') |
|
|
features = datasets.Features( |
|
|
{ |
|
|
source_language: datasets.Value("string"), |
|
|
target_language: datasets.Value("string"), |
|
|
|
|
|
} |
|
|
) |
|
|
return datasets.DatasetInfo( |
|
|
description=_DESCRIPTION, |
|
|
features=features, |
|
|
homepage="", |
|
|
license="", |
|
|
citation=_CITATION, |
|
|
) |
|
|
|
|
|
def _split_generators(self, dl_manager): |
|
|
train_json = dl_manager.download(os.path.join(self.config.name, "train.jsonl")) |
|
|
test_json = dl_manager.download(os.path.join(self.config.name, "test.jsonl")) |
|
|
val_json = dl_manager.download(os.path.join(self.config.name, "validation.jsonl")) |
|
|
|
|
|
return [ |
|
|
datasets.SplitGenerator( |
|
|
name=datasets.Split.TRAIN, |
|
|
gen_kwargs={"path": train_json}, |
|
|
), |
|
|
datasets.SplitGenerator( |
|
|
name=datasets.Split.TEST, |
|
|
gen_kwargs={"path": test_json}, |
|
|
), |
|
|
datasets.SplitGenerator( |
|
|
name=datasets.Split.VALIDATION, |
|
|
gen_kwargs={"path": val_json}, |
|
|
) |
|
|
] |
|
|
|
|
|
|
|
|
def _generate_examples(self, path): |
|
|
with open(path, encoding="utf-8") as f: |
|
|
for key, row in enumerate(f): |
|
|
yield key, json.loads(row) |
|
|
|
