test / test.py
j-krzywdziak's picture
Update test.py
cd48917
raw
history blame
4.25 kB
# coding=utf-8
# Lint as: python3
"""test set"""
import csv
import os
import json
import datasets
from datasets.utils.py_utils import size_str
from tqdm import tqdm
_CITATION = """\
@inproceedings{panayotov2015librispeech,
title={Librispeech: an ASR corpus based on public domain audio books},
author={Panayotov, Vassil and Chen, Guoguo and Povey, Daniel and Khudanpur, Sanjeev},
booktitle={Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on},
pages={5206--5210},
year={2015},
organization={IEEE}
}
"""
_DESCRIPTION = """\
Lorem ipsum
"""
_BASE_URL = "https://huggingface.co/datasets/j-krzywdziak/test/tree/main"
_DATA_URL = "https://huggingface.co/datasets/j-krzywdziak/test/blob/main/dev.tar.gz"
_PROMPTS_URLS = "https://huggingface.co/datasets/j-krzywdziak/test/raw/main/dev.tsv"
logger = datasets.logging.get_logger(__name__)
class TestConfig(datasets.BuilderConfig):
"""Lorem impsum."""
def __init__(self, name, version, **kwargs):
# self.language = kwargs.pop("language", None)
# self.release_date = kwargs.pop("release_date", None)
# self.num_clips = kwargs.pop("num_clips", None)
# self.num_speakers = kwargs.pop("num_speakers", None)
# self.validated_hr = kwargs.pop("validated_hr", None)
# self.total_hr = kwargs.pop("total_hr", None)
# self.size_bytes = kwargs.pop("size_bytes", None)
# self.size_human = size_str(self.size_bytes)
description = (
f"Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor "
f"incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud "
f"exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure "
f"dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. "
f"Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt "
f"mollit anim id est laborum."
)
super(TestConfig, self).__init__(
name=name,
version=datasets.Version(version),
description=description,
**kwargs,
)
class TestASR(datasets.GeneratorBasedBuilder):
"""Lorem ipsum."""
DEFAULT_CONFIG_NAME = "all"
BUILDER_CONFIGS = [
TestConfig(
name="Test Dataset",
version="0.0.0",
)
]
def _info(self):
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=datasets.Features(
{
"audio_id": datasets.Value("string"),
"audio": datasets.Audio(sampling_rate=16_000),
"ngram": datasets.Value("string")
}
),
supervised_keys=None,
homepage=_BASE_URL,
citation=_CITATION
)
def _split_generators(self, dl_manager):
archive_path = dl_manager.download(_DATA_URL)
# (Optional) In non-streaming mode, we can extract the archive locally to have actual local audio files:
local_extracted_archive = dl_manager.extract(archive_path) if not dl_manager.is_streaming else {}
meta_path = dl_manager.download(_PROMPTS_URLS)
return [datasets.SplitGenerator(
name=datasets.Split.TEST,
gen_kwargs={
"meta_path": meta_path,
"audio_files": dl_manager.iter_archive(local_extracted_archive)
}
)]
def _generate_examples(self, meta_path, audio_files):
"""Lorem ipsum."""
metadata = {}
with open(meta_path, encoding="utf-8") as f:
for row in f:
audio_id = row.splt("\t")[0]
ngram = row.split("\t")[1]
metadata[audio_id] = {"audio_id": audio_id,
"ngram": ngram}
inside_clips_dir = True
id_ = 0
for path, f in audio_files:
_, audio_name = os.path.split(path)
if audio_name in metadata:
audio = {"bytes": f.read()}
yield id_, {**metadata[audio_id], "audio": audio}
id_ +=1