iva_mt_wslot-exp / iva_mt_wslot-exp.py
cartesinus's picture
release for of en2pl with massive filtering and en2es without massive filtering
0cc561c
# coding=utf-8
"""IVA_MT_WSLOT"""
import datasets
import json
_DESCRIPTION = """\
"""
_URL = "https://github.com/cartesinus/iva_mt/raw/main/release/0.2/iva_mt_wslot-dataset-0.2.1.tar.gz"
_LANGUAGE_PAIRS = ["en-pl", "en-de", "en-es", "en-sv"]
class IVA_MTConfig(datasets.BuilderConfig):
"""BuilderConfig for IVA_MT"""
def __init__(self, language_pair, **kwargs):
super().__init__(**kwargs)
"""
Args:
language_pair: language pair, you want to load
**kwargs: keyword arguments forwarded to super.
"""
self.language_pair = language_pair
class IVA_MT(datasets.GeneratorBasedBuilder):
"""OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side."""
VERSION = datasets.Version("0.2.1")
BUILDER_CONFIG_CLASS = IVA_MTConfig
BUILDER_CONFIGS = [
IVA_MTConfig(name=pair, description=_DESCRIPTION, language_pair=pair)
for pair in _LANGUAGE_PAIRS
]
def _info(self):
src_tag, tgt_tag = self.config.language_pair.split("-")
return datasets.DatasetInfo(
features=datasets.Features(
{
"id": datasets.Value("int64"),
"locale": datasets.Value("string"),
"origin": datasets.Value("string"),
"partition": datasets.Value("string"),
"translation_utt": datasets.features.Translation(languages=(src_tag, tgt_tag)),
"translation_xml": datasets.features.Translation(languages=(src_tag, tgt_tag)),
"src_bio": datasets.Value("string"),
"tgt_bio": datasets.Value("string")
}
),
supervised_keys=(src_tag, tgt_tag),
)
def _split_generators(self, dl_manager):
lang_pair = self.config.language_pair
src_tag, tgt_tag = lang_pair.split("-")
archive = dl_manager.download(_URL)
data_dir = "/".join(["iva_mt_wslot-dataset", "0.2.1", lang_pair])
output = []
test = datasets.SplitGenerator(
name=datasets.Split.TEST,
# These kwargs will be passed to _generate_examples
gen_kwargs={
"filepath": f"{data_dir}/iva_mt_wslot-{lang_pair}-test.jsonl",
"files": dl_manager.iter_archive(archive),
"split": "test",
},
)
output.append(test)
train = datasets.SplitGenerator(
name=datasets.Split.TRAIN,
gen_kwargs={
"filepath": f"{data_dir}/iva_mt_wslot-{lang_pair}-train.jsonl",
"files": dl_manager.iter_archive(archive),
"split": "train",
},
)
output.append(train)
valid = datasets.SplitGenerator(
name=datasets.Split.VALIDATION,
# These kwargs will be passed to _generate_examples
gen_kwargs={
"filepath": f"{data_dir}/iva_mt_wslot-{lang_pair}-valid.jsonl",
"files": dl_manager.iter_archive(archive),
"split": "valid",
},
)
output.append(valid)
return output
def _generate_examples(self, filepath, files, split):
"""Yields examples."""
src_tag, tgt_tag = self.config.language_pair.split("-")
key_ = 0
lang = _LANGUAGE_PAIRS.copy()
for path, f in files:
l = path.split("/")[-1].split("-")[1].replace('2', '-')
if l != self.config.language_pair:
continue
# Read the file
lines = f.read().decode(encoding="utf-8").split("\n")
for line in lines:
if not line:
continue
data = json.loads(line)
if data["partition"] != split:
continue
yield key_, {
"id": data["id"],
"locale": data["locale"],
"origin": data["origin"],
"partition": data["partition"],
"translation_utt": {src_tag: str(data['src_utt']), tgt_tag: str(data['tgt_utt'])},
"translation_xml": {src_tag: str(data['src_xml']), tgt_tag: str(data['tgt_xml'])},
"src_bio": str(data['src_bio']),
"tgt_bio": str(data['tgt_bio'])
}
key_ += 1