def test_open_missing(erepo):
# Remove cache to make foo missing
shutil.rmtree(erepo.dvc.cache.local.cache_dir)
repo_url = "file://" + erepo.dvc.root_dir
with pytest.raises(OutputFileMissingError):
api.read(erepo.FOO, repo=repo_url)
def test_read_with_subrepos(tmp_dir, scm, local_cloud, local_repo):
tmp_dir.scm_gen("foo.txt", "foo.txt", commit="add foo.txt")
subrepo = tmp_dir / "dir" / "subrepo"
make_subrepo(subrepo, scm, config=local_cloud.config)
with subrepo.chdir():
subrepo.scm_gen({"lorem": "lorem"}, commit="add lorem")
subrepo.dvc_gen({"dir": {"file.txt": "file.txt"}}, commit="add dir")
subrepo.dvc_gen("dvc-file", "dvc-file", commit="add dir")
subrepo.dvc.push()
repo_path = None if local_repo else f"file:///{tmp_dir}"
subrepo_path = os.path.join("dir", "subrepo")
assert api.read("foo.txt", repo=repo_path) == "foo.txt"
assert (
api.read(os.path.join(subrepo_path, "lorem"), repo=repo_path)
== "lorem"
)
assert (
api.read(os.path.join(subrepo_path, "dvc-file"), repo=repo_path)
== "dvc-file"
)
assert (
api.read(os.path.join(subrepo_path, "dir", "file.txt"), repo=repo_path)
== "file.txt"
)
def test_missing(remote_url, tmp_dir, dvc):
tmp_dir.dvc_gen("foo", "foo")
run_dvc("remote", "add", "-d", "upstream", remote_url)
# Remove cache to make foo missing
shutil.rmtree(dvc.cache.local.cache_dir)
with pytest.raises(FileMissingError):
api.read("foo")
def test_missing(repo_dir, dvc_repo, remote_url):
run_dvc("add", repo_dir.FOO)
run_dvc("remote", "add", "-d", "upstream", remote_url)
# Remove cache to make foo missing
shutil.rmtree(dvc_repo.cache.local.cache_dir)
with pytest.raises(FileMissingError):
api.read(repo_dir.FOO)
def test_missing(tmp_dir, dvc, remote):
tmp_dir.dvc_gen("foo", "foo")
# Remove cache to make foo missing
remove(dvc.cache.local.cache_dir)
api.read("foo")
remove("foo")
with pytest.raises(FileMissingError):
api.read("foo")
def test_missing(remote_url, tmp_dir, dvc):
tmp_dir.dvc_gen("foo", "foo")
run_dvc("remote", "add", "-d", TEST_REMOTE, remote_url)
ensure_dir(dvc, remote_url)
# Remove cache to make foo missing
remove(dvc.cache.local.cache_dir)
api.read("foo")
remove("foo")
with pytest.raises(FileMissingError):
api.read("foo")
def test_open_not_cached(dvc):
metric_file = "metric.txt"
metric_content = "0.6"
metric_code = "open('{}', 'w').write('{}')".format(metric_file,
metric_content)
dvc.run(
metrics_no_cache=[metric_file],
cmd=('python -c "{}"'.format(metric_code)),
)
with api.open(metric_file) as fd:
assert fd.read() == metric_content
os.remove(metric_file)
with pytest.raises(FileMissingError):
api.read(metric_file)
def test_api_missing_local_cache_exists_on_remote(tmp_dir, scm, dvc,
as_external, remote, files,
to_read):
tmp_dir.dvc_gen(files, commit="DVC track files")
dvc.push()
# Remove cache to make foo missing
remove(dvc.odb.local.cache_dir)
remove(first(files))
repo_url = f"file://{tmp_dir}" if as_external else None
file_content = get_in(files, to_read.split(os.sep))
assert api.read(to_read, repo=repo_url) == file_content
def test_open_external(repo_dir, dvc_repo, erepo, remote_url):
erepo.dvc.scm.checkout("branch")
_set_remote_url_and_commit(erepo.dvc, remote_url)
erepo.dvc.scm.checkout("master")
_set_remote_url_and_commit(erepo.dvc, remote_url)
erepo.dvc.push(all_branches=True)
# Remove cache to force download
shutil.rmtree(erepo.dvc.cache.local.cache_dir)
# Using file url to force clone to tmp repo
repo_url = "file://" + erepo.dvc.root_dir
with api.open("version", repo=repo_url) as fd:
assert fd.read() == "master"
assert api.read("version", repo=repo_url, rev="branch") == "branch"
def resnext101_32x4d(num_classes=1000, pretrained='imagenet'):
model = ResNeXt101_32x4d(num_classes=num_classes)
if pretrained is not None:
settings = pretrained_settings['resnext101_32x4d'][pretrained]
assert num_classes == settings['num_classes'], \
"num_classes should be {}, but is {}".format(settings['num_classes'], num_classes)
weights = dvcapi.read('model_weights/resnext101_32x4d-29e315fa.pth',
remote='gsremote',
mode="rb",
encoding=None)
model_weights = torch.load(io.BytesIO(weights))
model.load_state_dict(model_weights)
model.input_space = settings['input_space']
model.input_size = settings['input_size']
model.input_range = settings['input_range']
model.mean = settings['mean']
model.std = settings['std']
return model
def test_open_external(tmp_dir, erepo_dir, cloud):
erepo_dir.add_remote(config=cloud.config)
with erepo_dir.chdir():
erepo_dir.dvc_gen("version", "master", commit="add version")
with erepo_dir.branch("branch", new="True"):
# NOTE: need file to be other size for Mac
erepo_dir.dvc_gen("version", "branchver", commit="add version")
erepo_dir.dvc.push(all_branches=True)
# Remove cache to force download
remove(erepo_dir.dvc.cache.local.cache_dir)
# Using file url to force clone to tmp repo
repo_url = f"file://{erepo_dir}"
with api.open("version", repo=repo_url) as fd:
assert fd.read() == "master"
assert api.read("version", repo=repo_url, rev="branch") == "branchver"
def test_open_external(remote_url, erepo_dir):
_set_remote_url_and_commit(erepo_dir.dvc, remote_url)
with erepo_dir.chdir():
erepo_dir.dvc_gen("version", "master", commit="add version")
with erepo_dir.branch("branch", new="True"):
# NOTE: need file to be other size for Mac
erepo_dir.dvc_gen("version", "branchver", commit="add version")
erepo_dir.dvc.push(all_branches=True)
# Remove cache to force download
shutil.rmtree(erepo_dir.dvc.cache.local.cache_dir)
# Using file url to force clone to tmp repo
repo_url = "file://{}".format(erepo_dir)
with api.open("version", repo=repo_url) as fd:
assert fd.read() == "master"
assert api.read("version", repo=repo_url, rev="branch") == "branchver"
def init_detector(config, device='cuda:0'):
"""Initialize a detector from config file. DVC weights loading.
Args:
config (str or :obj:`mmcv.Config`): Config file path or the config
object.
Returns:
nn.Module: The constructed detector.
"""
if isinstance(config, str):
config = Config.fromfile(config)
elif not isinstance(config, Config):
raise TypeError('config must be a filename or Config object, '
'but got {}'.format(type(config)))
config.model.pretrained = None
model = build_detector(config.model, test_cfg=config.test_cfg)
weights = dvcapi.read('model_weights/faster_rcnn_r50_c4_2x-6e4fdf4f.pth',
remote='gsremote',
mode="rb",
encoding=None)
checkpoint = torch.load(io.BytesIO(weights))
state_dict = checkpoint['state_dict']
model.load_state_dict(state_dict, strict=False)
if 'CLASSES' in checkpoint['meta']:
model.CLASSES = checkpoint['meta']['CLASSES']
else:
model.CLASSES = get_classes('coco')
model.cfg = config # save the config in the model for convenience
model.to(device)
model.eval()
return model
def init(model_path, metadata):
ctx["model"] = pickle.loads(
api.read(metadata["model_path"], metadata["dvc_repo"], mode="rb"))
ctx["pipeline"] = pickle.loads(
api.read(metadata["pipeline_path"], metadata["dvc_repo"], mode="rb"))
