def _lightly_cli(cfg, is_cli_call=True):
cfg['loader']['shuffle'] = True
cfg['loader']['drop_last'] = True
checkpoint = _train_cli(cfg, is_cli_call)
cfg['loader']['shuffle'] = False
cfg['loader']['drop_last'] = False
cfg['checkpoint'] = checkpoint
embeddings = _embed_cli(cfg, is_cli_call)
cfg['embeddings'] = embeddings
if cfg['token'] and cfg['dataset_id']:
_upload_cli(cfg)
def embed_images(checkpoint: str, config_path: str = None, **kwargs):
"""Embed images with a self-supervised model.
Calls the same function as lightly-embed. All arguments passed to
lightly-embed can also be passed to this function (see below for an
example).
Args:
checkpoint:
Path to the checkpoint file for the embedding model.
config_path:
Path to config.yaml. If None, the default configs will be used.
**kwargs:
Overwrite default configs py passing keyword arguments.
Returns:
Embeddings, labels, and filenames of the images.
Examples:
>>> import lightly
>>> my_checkpoint_path = 'path/to/checkpoint.ckpt'
>>>
>>> # embed images with default configs
>>> embeddings, _, _ = lightly.embed_images(
>>> my_checkpoint_path, input_dir='path/to/data')
>>>
>>> # embed images with separate config file
>>> my_config_path = 'my/config/file.yaml'
>>> embeddings, _, _ = lightly.embed_images(
>>> my_checkpoint_path, input_dir='path/to/data', config_path=my_config_path)
>>>
>>> # embed images with default settings and overwrites: at inference,
>>> # we can use larger input_sizes because it requires less memory.
>>> my_collate = {input_size: 256}
>>> embeddings, _, _ = lightly.embed_images(
>>> my_checkpoint_path, input_dir='path/to/data', collate=my_collate)
>>> # the command above is equivalent to:
>>> # lightly-embed input_dir='path/to/data' collate.input_size=256
"""
config_path = _get_config_path(config_path)
config_args = _load_config_file(config_path)
config_args = _add_kwargs(config_args, kwargs)
config_args['checkpoint'] = checkpoint
return _embed_cli(config_args, is_cli_call=False)
def _lightly_cli(cfg, is_cli_call=True):
cfg['loader']['shuffle'] = True
cfg['loader']['drop_last'] = True
if cfg['trainer']['max_epochs'] > 0:
checkpoint = _train_cli(cfg, is_cli_call)
else:
checkpoint = ''
cfg['loader']['shuffle'] = False
cfg['loader']['drop_last'] = False
cfg['checkpoint'] = checkpoint
embeddings = _embed_cli(cfg, is_cli_call)
cfg['embeddings'] = embeddings
if cfg['token'] and (cfg['dataset_id'] or cfg['new_dataset_name']):
_upload_cli(cfg)
def _lightly_cli(cfg, is_cli_call=True):
cfg['loader']['shuffle'] = True
cfg['loader']['drop_last'] = True
if cfg['trainer']['max_epochs'] > 0:
print('#' * 10 + ' Starting to train an embedding model.')
checkpoint = _train_cli(cfg, is_cli_call)
else:
checkpoint = ''
cfg['loader']['shuffle'] = False
cfg['loader']['drop_last'] = False
cfg['checkpoint'] = checkpoint
print('#' * 10 + ' Starting to embed your dataset.')
embeddings = _embed_cli(cfg, is_cli_call)
cfg['embeddings'] = embeddings
if cfg['token'] and (cfg['dataset_id'] or cfg['new_dataset_name']):
print('#' * 10 +
' Starting to upload your dataset to the Lightly platform.')
_upload_cli(cfg)
print('#' * 10 + ' Finished')