def __init__(self,
pipeline: "nlpaug_flow.Pipeline",
num_transformed: int = 1,
identifiers: List[Identifier] = None,
*args,
**kwargs):
assert isinstance(pipeline, nlpaug_flow.Pipeline), (
"`pipeline` must be an nlpaug Pipeline object. Please use \n"
"from nlpaug.flow import Sequential\n"
"rg.NlpAugTransformation(pipeline=Sequential(flow=[...])).")
super(NlpAugTransformation,
self).__init__(num_transformed=num_transformed,
identifiers=Identifier.range(
n=num_transformed,
_name=self.__class__.__name__,
pipeline=[
Identifier(
_name=augmenter.name,
src=augmenter.aug_src if hasattr(
augmenter, "aug_src") else None,
action=augmenter.action,
method=augmenter.method,
) for augmenter in pipeline
],
) if not identifiers else identifiers,
*args,
**kwargs)
# Set the pipeline
self.pipeline = pipeline
def __init__(self,
num_transformed=1,
alpha_sr=0.1,
alpha_ri=0.1,
alpha_rs=0.1,
p_rd=0.1):
super(EasyDataAugmentation,
self).__init__(identifiers=Identifier.range(
n=num_transformed,
_name=self.__class__.__name__,
alpha_sr=alpha_sr,
alpha_ri=alpha_ri,
alpha_rs=alpha_rs,
p_rd=p_rd,
))
# Set the parameters
self.alpha_sr = alpha_sr
self.alpha_ri = alpha_ri
self.alpha_rs = alpha_rs
self.p_rd = p_rd
# Download wordnet
self._download_wordnet()
def __init__(self, metric: Sequence[str], threshold: float):
super(RougeMatrixSentenceTransformation, self).__init__(
num_transformed=1,
identifiers=Identifier.range(n=1, _name=self.__class__.__name__),
)
self.metric = metric
self.threshold = threshold
def __init__(
self,
n_src2tgt: int = 1,
n_tgt2src: int = 1,
langs: str = "en2de",
torchhub_dir: str = None,
device: str = "cuda",
src2tgt_topk: int = 1000,
src2tgt_temp: float = 1.0,
tgt2src_topk: int = 1000,
tgt2src_temp: float = 1.0,
):
if not _fastbpe_available:
raise ImportError(
"fastBPE not available for import. Please install fastBPE with pip "
"install fastBPE."
)
super(FairseqBacktranslation, self).__init__(
identifiers=Identifier.range(
n=n_src2tgt * n_tgt2src,
_name=self.__class__.__name__,
langs=langs,
src2tgt_topk=src2tgt_topk,
src2tgt_temp=src2tgt_temp,
tgt2src_topk=tgt2src_topk,
tgt2src_temp=tgt2src_temp,
)
)
# Set the parameters
self.n_src2tgt = n_src2tgt
self.n_tgt2src = n_tgt2src
self.src2tgt_topk = src2tgt_topk
self.src2tgt_temp = src2tgt_temp
self.tgt2src_topk = tgt2src_topk
self.tgt2src_temp = tgt2src_temp
# Setup the backtranslation models
self.src2tgt, self.tgt2src = self.load_models(
langs=langs,
torchhub_dir=torchhub_dir,
# self.logdir if not torchhub_dir else torchhub_dir,
device=device,
)
def __init__(self,
apply_fn: Callable = None,
identifiers: List[Identifier] = None,
num_outputs: int = None,
*args,
**kwargs):
if not identifiers:
assert (
num_outputs
), "Must pass in num_outputs if no identifiers are specified."
# Set the identifiers for the outputs of the Operation
self._identifiers = (Identifier.range(
n=num_outputs, _name=self.__class__.__name__, **kwargs)
if not identifiers else identifiers)
# Assign the apply_fn
if apply_fn:
self.apply = apply_fn
def __init__(
self,
n_src2tgt: int = 1,
n_tgt2src: int = 1,
langs: str = "en2de",
torchhub_dir: str = None,
device: str = "cuda",
src2tgt_topk: int = 1000,
src2tgt_temp: float = 1.0,
tgt2src_topk: int = 1000,
tgt2src_temp: float = 1.0,
):
super(FairseqBacktranslation,
self).__init__(identifiers=Identifier.range(
n=n_src2tgt * n_tgt2src,
_name=self.__class__.__name__,
langs=langs,
src2tgt_topk=src2tgt_topk,
src2tgt_temp=src2tgt_temp,
tgt2src_topk=tgt2src_topk,
tgt2src_temp=tgt2src_temp,
))
# Set the parameters
self.n_src2tgt = n_src2tgt
self.n_tgt2src = n_tgt2src
self.src2tgt_topk = src2tgt_topk
self.src2tgt_temp = src2tgt_temp
self.tgt2src_topk = tgt2src_topk
self.tgt2src_temp = tgt2src_temp
# Setup the backtranslation models
self.src2tgt, self.tgt2src = self.load_models(
langs=langs,
torchhub_dir=torchhub_dir,
# self.logdir if not torchhub_dir else torchhub_dir,
device=device,
)
def test_range(self):
# Use the range function to create multiple identifiers
identifiers = Identifier.range(3, _name="MyIdentifier", param="a", param_2="b")
for i, identifier in enumerate(identifiers):
self.assertEqual(identifier, f"MyIdentifier-{i + 1}(param=a, param_2=b)")
