fairseq/tasks/legacy_masked_lm.py
Killed 1 out of 27 mutantsSurvived
Survived mutation testing. These mutants show holes in your test suite.Mutant 2265
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -24,7 +24,7 @@
from fairseq import utils
-logger = logging.getLogger(__name__)
+logger = None
@register_task('legacy_masked_lm')
Mutant 2266
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -27,7 +27,7 @@
logger = logging.getLogger(__name__)
-@register_task('legacy_masked_lm')
+@register_task('XXlegacy_masked_lmXX')
class LegacyMaskedLMTask(FairseqTask):
"""
Task for training Masked LM (BERT) model.
Mutant 2267
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -26,8 +26,6 @@
logger = logging.getLogger(__name__)
-
-@register_task('legacy_masked_lm')
class LegacyMaskedLMTask(FairseqTask):
"""
Task for training Masked LM (BERT) model.
Mutant 2268
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -35,7 +35,6 @@
dictionary (Dictionary): the dictionary for the input of the task
"""
- @staticmethod
def add_args(parser):
"""Add task-specific arguments to the parser."""
parser.add_argument('data', help='colon separated path to data directories list, \
Mutant 2269
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -38,7 +38,7 @@
@staticmethod
def add_args(parser):
"""Add task-specific arguments to the parser."""
- parser.add_argument('data', help='colon separated path to data directories list, \
+ parser.add_argument('XXdataXX', help='colon separated path to data directories list, \
will be iterated upon during epochs in round-robin manner')
parser.add_argument('--tokens-per-sample', default=512, type=int,
help='max number of total tokens over all segments'
Mutant 2270
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -38,8 +38,8 @@
@staticmethod
def add_args(parser):
"""Add task-specific arguments to the parser."""
- parser.add_argument('data', help='colon separated path to data directories list, \
- will be iterated upon during epochs in round-robin manner')
+ parser.add_argument('data', help='XXcolon separated path to data directories list, \
+ will be iterated upon during epochs in round-robin mannerXX')
parser.add_argument('--tokens-per-sample', default=512, type=int,
help='max number of total tokens over all segments'
' per sample for BERT dataset')
Mutant 2271
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -40,7 +40,7 @@
"""Add task-specific arguments to the parser."""
parser.add_argument('data', help='colon separated path to data directories list, \
will be iterated upon during epochs in round-robin manner')
- parser.add_argument('--tokens-per-sample', default=512, type=int,
+ parser.add_argument('XX--tokens-per-sampleXX', default=512, type=int,
help='max number of total tokens over all segments'
' per sample for BERT dataset')
parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
Mutant 2272
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -40,7 +40,7 @@
"""Add task-specific arguments to the parser."""
parser.add_argument('data', help='colon separated path to data directories list, \
will be iterated upon during epochs in round-robin manner')
- parser.add_argument('--tokens-per-sample', default=512, type=int,
+ parser.add_argument('--tokens-per-sample', default=513, type=int,
help='max number of total tokens over all segments'
' per sample for BERT dataset')
parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
Mutant 2273
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -41,7 +41,7 @@
parser.add_argument('data', help='colon separated path to data directories list, \
will be iterated upon during epochs in round-robin manner')
parser.add_argument('--tokens-per-sample', default=512, type=int,
- help='max number of total tokens over all segments'
+ help='XXmax number of total tokens over all segmentsXX'
' per sample for BERT dataset')
parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
parser.add_argument('--shuffle-dataset', action='store_true', default=False)
Mutant 2274
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -43,7 +43,7 @@
parser.add_argument('--tokens-per-sample', default=512, type=int,
help='max number of total tokens over all segments'
' per sample for BERT dataset')
- parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
+ parser.add_argument('XX--break-modeXX', default="doc", type=str, help='mode for breaking sentence')
parser.add_argument('--shuffle-dataset', action='store_true', default=False)
def __init__(self, args, dictionary):
Mutant 2275
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -43,7 +43,7 @@
parser.add_argument('--tokens-per-sample', default=512, type=int,
help='max number of total tokens over all segments'
' per sample for BERT dataset')
- parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
+ parser.add_argument('--break-mode', default="XXdocXX", type=str, help='mode for breaking sentence')
parser.add_argument('--shuffle-dataset', action='store_true', default=False)
def __init__(self, args, dictionary):
Mutant 2276
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -43,7 +43,7 @@
parser.add_argument('--tokens-per-sample', default=512, type=int,
help='max number of total tokens over all segments'
' per sample for BERT dataset')
- parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
+ parser.add_argument('--break-mode', default="doc", type=str, help='XXmode for breaking sentenceXX')
parser.add_argument('--shuffle-dataset', action='store_true', default=False)
def __init__(self, args, dictionary):
Mutant 2277
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -44,7 +44,7 @@
help='max number of total tokens over all segments'
' per sample for BERT dataset')
parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
- parser.add_argument('--shuffle-dataset', action='store_true', default=False)
+ parser.add_argument('XX--shuffle-datasetXX', action='store_true', default=False)
def __init__(self, args, dictionary):
super().__init__(args)
Mutant 2279
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -44,7 +44,7 @@
help='max number of total tokens over all segments'
' per sample for BERT dataset')
parser.add_argument('--break-mode', default="doc", type=str, help='mode for breaking sentence')
- parser.add_argument('--shuffle-dataset', action='store_true', default=False)
+ parser.add_argument('--shuffle-dataset', action='store_true', default=True)
def __init__(self, args, dictionary):
super().__init__(args)
Mutant 2280
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -51,7 +51,6 @@
self.dictionary = dictionary
self.seed = args.seed
- @classmethod
def load_dictionary(cls, filename):
return BertDictionary.load(filename)
Mutant 2281
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -55,7 +55,6 @@
def load_dictionary(cls, filename):
return BertDictionary.load(filename)
- @classmethod
def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
d = BertDictionary()
for filename in filenames:
Mutant 2282
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -56,7 +56,7 @@
return BertDictionary.load(filename)
@classmethod
- def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
+ def build_dictionary(cls, filenames, workers=2, threshold=-1, nwords=-1, padding_factor=8):
d = BertDictionary()
for filename in filenames:
Dictionary.add_file_to_dictionary(filename, d, tokenizer.tokenize_line, workers)
Mutant 2283
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -56,7 +56,7 @@
return BertDictionary.load(filename)
@classmethod
- def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
+ def build_dictionary(cls, filenames, workers=1, threshold=+1, nwords=-1, padding_factor=8):
d = BertDictionary()
for filename in filenames:
Dictionary.add_file_to_dictionary(filename, d, tokenizer.tokenize_line, workers)
Mutant 2284
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -56,7 +56,7 @@
return BertDictionary.load(filename)
@classmethod
- def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
+ def build_dictionary(cls, filenames, workers=1, threshold=-2, nwords=-1, padding_factor=8):
d = BertDictionary()
for filename in filenames:
Dictionary.add_file_to_dictionary(filename, d, tokenizer.tokenize_line, workers)
Mutant 2285
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -56,7 +56,7 @@
return BertDictionary.load(filename)
@classmethod
- def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
+ def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=+1, padding_factor=8):
d = BertDictionary()
for filename in filenames:
Dictionary.add_file_to_dictionary(filename, d, tokenizer.tokenize_line, workers)
Mutant 2286
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -56,7 +56,7 @@
return BertDictionary.load(filename)
@classmethod
- def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
+ def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-2, padding_factor=8):
d = BertDictionary()
for filename in filenames:
Dictionary.add_file_to_dictionary(filename, d, tokenizer.tokenize_line, workers)
Mutant 2287
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -56,7 +56,7 @@
return BertDictionary.load(filename)
@classmethod
- def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=8):
+ def build_dictionary(cls, filenames, workers=1, threshold=-1, nwords=-1, padding_factor=9):
d = BertDictionary()
for filename in filenames:
Dictionary.add_file_to_dictionary(filename, d, tokenizer.tokenize_line, workers)
Mutant 2288
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -63,7 +63,6 @@
d.finalize(threshold=threshold, nwords=nwords, padding_factor=padding_factor)
return d
- @property
def target_dictionary(self):
return self.dictionary
Mutant 2289
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -67,7 +67,6 @@
def target_dictionary(self):
return self.dictionary
- @classmethod
def setup_task(cls, args, **kwargs):
"""Setup the task.
"""
Mutant 2290
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -78,7 +78,7 @@
return cls(args, dictionary)
- def load_dataset(self, split, epoch=1, combine=False):
+ def load_dataset(self, split, epoch=2, combine=False):
"""Load a given dataset split.
Args:
Mutant 2291
--- fairseq/tasks/legacy_masked_lm.py
+++ fairseq/tasks/legacy_masked_lm.py
@@ -78,7 +78,7 @@
return cls(args, dictionary)
- def load_dataset(self, split, epoch=1, combine=False):
+ def load_dataset(self, split, epoch=1, combine=True):
"""Load a given dataset split.
Args: