def test_load_real_metric(self, metric_name):
with tempfile.TemporaryDirectory() as temp_data_dir:
download_config = DownloadConfig()
download_config.download_mode = GenerateMode.FORCE_REDOWNLOAD
load_metric(metric_name,
data_dir=temp_data_dir,
download_config=download_config)
def test_load_real_dataset(self, dataset_name):
with tempfile.TemporaryDirectory() as temp_data_dir:
download_config = DownloadConfig()
download_config.download_mode = GenerateMode.FORCE_REDOWNLOAD
dataset = load_dataset(dataset_name, data_dir=temp_data_dir, download_config=download_config)
for split in dataset.keys():
self.assertTrue(len(dataset[split]) > 0)
def test_load_real_dataset_local(self, dataset_name):
with tempfile.TemporaryDirectory() as temp_data_dir:
download_config = DownloadConfig()
download_config.download_mode = GenerateMode.FORCE_REDOWNLOAD
download_and_prepare_kwargs = {"download_config": download_config}
dataset = load_dataset(
"./datasets/" + dataset_name,
data_dir=temp_data_dir,
download_and_prepare_kwargs=download_and_prepare_kwargs,
)
for split in dataset.keys():
self.assertTrue(len(dataset[split]) > 0)
def test_load_real_dataset(self, dataset_name):
if "/" not in dataset_name:
logging.info("Skip {} because it is a canonical dataset")
return
with tempfile.TemporaryDirectory() as temp_data_dir:
download_config = DownloadConfig()
download_config.download_mode = GenerateMode.FORCE_REDOWNLOAD
download_and_prepare_kwargs = {"download_config": download_config}
dataset = load_dataset(
dataset_name,
data_dir=temp_data_dir,
download_and_prepare_kwargs=download_and_prepare_kwargs)
for split in dataset.keys():
self.assertTrue(len(dataset[split]) > 0)
