Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Enhance] Support download dataset from openxlab #223

Merged
merged 2 commits into from
Aug 28, 2023
Merged

[Enhance] Support download dataset from openxlab #223

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
603084b
Update openxlab
HAOCHENYE Aug 9, 2023
6a83523
update
HAOCHENYE Aug 9, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 23 additions & 10 deletions mim/commands/download.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -210,38 +210,51 @@ def _download_dataset(package: str, dataset: str, dest_root: str) -> None:
f'already updated it and still get this error, please report an '
f'issue to {package}')
with open(dataset_index_path) as f:
datasets_meta = yaml.load(f, Loader=yaml.SafeLoader)
dataset_metas = yaml.load(f, Loader=yaml.SafeLoader)

if dataset not in datasets_meta:
if dataset not in dataset_metas:
raise KeyError(f'Cannot find {dataset} in {dataset_index_path}. '
'here are the available datasets: '
'{}'.format('\n'.join(datasets_meta.keys())))
dataset_meta = datasets_meta[dataset]

'{}'.format('\n'.join(dataset_metas.keys())))
dataset_meta = dataset_metas[dataset]
# OpenMMLab repo will define the `dataset-index.yml` like this:
# openxlab: true
# voc2007:
# dataset: PASCAL_VOC2007
# download_root: data
# data_root: data
# script: tools/dataset_converters/scripts/preprocess_voc2007.sh

# In this case, the top level key "voc2007" means the "Dataset Name" passed
# In this case:
# `openxlab` means download the dataset with `openxlab` cli, otherwise
# use the `odl` cli. Although `odl` cli will not be maintained in the
# future, we still keep it here for compatibility.

# The top level key "voc2007" means the "Dataset Name" passed
# to `mim download --dataset {Dataset Name}`

# The nested field "dataset" means the argument passed to `odl get`
# If the value of "dataset" is the same as the "Dataset Name", downstream
# repos can skip defining "dataset" and "Dataset Name" will be passed
# to `odl get`
src_name = dataset_meta.get('dataset', dataset)

use_openxlab = dataset_metas.get('openxlab', False)
src_name = dataset_meta.get('dataset', dataset)
# `odl get` will download raw dataset to `download_root`, and the script
# will process the raws data and put the prepared data to the `data_root`
download_root = dataset_meta['download_root']
os.makedirs(download_root, exist_ok=True)

color_echo(f'Start downloading {dataset} to {download_root}...', 'blue')
subprocess.check_call(['odl', 'get', src_name, '-d', download_root],
stdin=sys.stdin,
stdout=sys.stdout)
if use_openxlab:
subprocess.check_call(
['openxlab', 'dataset', 'get', src_name, '-d', download_root],
stdin=sys.stdin,
stdout=sys.stdout)
else:
subprocess.check_call(['odl', 'get', src_name, '-d', download_root],
stdin=sys.stdin,
stdout=sys.stdout)

if not osp.exists(download_root):
return
Expand Down
Loading