diff --git a/modelscope/hub/api.py b/modelscope/hub/api.py index 7468e5e3..f2ff822d 100644 --- a/modelscope/hub/api.py +++ b/modelscope/hub/api.py @@ -39,8 +39,8 @@ from modelscope.utils.constant import (DEFAULT_DATASET_REVISION, DEFAULT_MODEL_REVISION, DEFAULT_REPOSITORY_REVISION, MASTER_MODEL_BRANCH, DatasetFormations, - DatasetMetaFormats, DownloadMode, - ModelFile) + DatasetMetaFormats, DownloadChannel, + DownloadMode, ModelFile) from modelscope.utils.logger import get_logger from .utils.utils import (get_endpoint, get_release_datetime, model_id_to_group_owner_name) @@ -646,6 +646,25 @@ class HubApi: def check_local_cookies(self, use_cookies) -> CookieJar: return self._check_cookie(use_cookies=use_cookies) + def dataset_download_uv(self, dataset_name: str, namespace: str): + if not dataset_name or not namespace: + raise ValueError('dataset_name or namespace cannot be empty!') + + # get channel and user_name + channel = DownloadChannel.LOCAL.value + user_name = '' + if MODELSCOPE_ENVIRONMENT in os.environ: + channel = os.environ[MODELSCOPE_ENVIRONMENT] + if MODELSCOPE_USERNAME in os.environ: + user_name = os.environ[MODELSCOPE_USERNAME] + + url = f'{self.endpoint}/api/v1/datasets/{namespace}/{dataset_name}/download/uv/{channel}?user={user_name}' + cookies = ModelScopeConfig.get_cookies() + r = requests.post(url, cookies=cookies, headers=self.headers) + resp = r.json() + raise_on_error(resp) + return resp['Message'] + class ModelScopeConfig: path_credential = expanduser(DEFAULT_CREDENTIALS_PATH) diff --git a/modelscope/msdatasets/ms_dataset.py b/modelscope/msdatasets/ms_dataset.py index 0c537df7..5c8ea59f 100644 --- a/modelscope/msdatasets/ms_dataset.py +++ b/modelscope/msdatasets/ms_dataset.py @@ -274,6 +274,8 @@ class MsDataset: try: api.on_dataset_download( dataset_name=download_dataset, namespace=namespace) + api.dataset_download_uv( + dataset_name=download_dataset, namespace=namespace) except Exception as e: logger.error(e) diff --git a/modelscope/utils/constant.py b/modelscope/utils/constant.py index 2729b75a..f0a97dbd 100644 --- a/modelscope/utils/constant.py +++ b/modelscope/utils/constant.py @@ -238,6 +238,14 @@ class DownloadMode(enum.Enum): FORCE_REDOWNLOAD = 'force_redownload' +class DownloadChannel(enum.Enum): + """ Channels of datasets downloading for uv/pv counting. + """ + LOCAL = 'local' + DSW = 'dsw' + EAIS = 'eais' + + class UploadMode(enum.Enum): """ How to upload object to remote. """ diff --git a/requirements/framework.txt b/requirements/framework.txt index 17fbd8a3..a86c0cc5 100644 --- a/requirements/framework.txt +++ b/requirements/framework.txt @@ -1,7 +1,7 @@ addict attrs -# version beyond 2.6.0 introduces compatbility issue and is being resolved -datasets<=2.6.0 +# version beyond 2.5.2 introduces compatbility issue and is being resolved +datasets<=2.5.2 easydict einops filelock>=3.3.0 diff --git a/tests/msdatasets/test_dataset_upload.py b/tests/msdatasets/test_dataset_upload.py index 3d35d480..d91f24d7 100644 --- a/tests/msdatasets/test_dataset_upload.py +++ b/tests/msdatasets/test_dataset_upload.py @@ -8,7 +8,8 @@ import zipfile from modelscope.msdatasets import MsDataset from modelscope.msdatasets.utils.dataset_utils import list_dataset_objects from modelscope.utils import logger as logging -from modelscope.utils.constant import DEFAULT_DATASET_REVISION, ModelFile +from modelscope.utils.constant import (DEFAULT_DATASET_REVISION, DownloadMode, + ModelFile) from modelscope.utils.test_utils import test_level logger = logging.get_logger(__name__) @@ -104,7 +105,10 @@ class DatasetUploadTest(unittest.TestCase): @unittest.skipUnless(test_level() >= 1, 'skip test in current test level') def test_ds_download_dir(self): - test_ds = MsDataset.load(self.dataset_name, self.namespace) + test_ds = MsDataset.load( + self.dataset_name, + namespace=self.namespace, + download_mode=DownloadMode.FORCE_REDOWNLOAD) assert test_ds.config_kwargs['split_config'].values() @unittest.skipUnless(test_level() >= 1, 'skip test in current test level')