SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Temporarily revert the fill-mask improvements.

Browse Source
This commit is contained in:
Lysandre
2021-06-23 17:46:24 +02:00
parent 4bdff2cdbe
commit 941b4442ba
2 changed files with 30 additions and 81 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
tests/test_pipelines_fill_mask.py
View File

@@ -78,8 +78,7 @@ class FillMaskPipelineTests(MonoInputPipelineCommonMixin, unittest.TestCase):
@require_torch
def test_torch_fill_mask_with_targets(self):
valid_inputs = ["My name is <mask>"]
# ' Sam' will yield a warning but work
valid_targets = [[" Teven", "ĠPatrick", "ĠClara"], ["ĠSam"], [" Sam"]]
valid_targets = [[" Teven", " Patrick", " Clara"], [" Sam"]]
invalid_targets = [[], [""], ""]
for model_name in self.small_models:
unmasker = pipeline(task="fill-mask", model=model_name, tokenizer=model_name, framework="pt")
@@ -90,34 +89,10 @@ class FillMaskPipelineTests(MonoInputPipelineCommonMixin, unittest.TestCase):
for targets in invalid_targets:
self.assertRaises(ValueError, unmasker, valid_inputs, targets=targets)
@require_torch
def test_torch_fill_mask_with_targets_and_topk(self):
model_name = self.small_models[0]
unmasker = pipeline(task="fill-mask", model=model_name, tokenizer=model_name, framework="pt")
targets = [" Teven", "ĠPatrick", "ĠClara"]
top_k = 2
outputs = unmasker("My name is <mask>", targets=targets, top_k=top_k)
self.assertEqual(len(outputs), 2)
@require_torch
def test_torch_fill_mask_with_duplicate_targets_and_topk(self):
model_name = self.small_models[0]
unmasker = pipeline(task="fill-mask", model=model_name, tokenizer=model_name, framework="pt")
# String duplicates + id duplicates
targets = [" Teven", "ĠPatrick", "ĠClara", "ĠClara", " Clara"]
top_k = 10
outputs = unmasker("My name is <mask>", targets=targets, top_k=top_k)
# The target list contains duplicates, so we can't output more
# than them
self.assertEqual(len(outputs), 3)
@require_tf
def test_tf_fill_mask_with_targets(self):
valid_inputs = ["My name is <mask>"]
# ' Sam' will yield a warning but work
valid_targets = [[" Teven", "ĠPatrick", "ĠClara"], ["ĠSam"], [" Sam"]]
valid_targets = [[" Teven", " Patrick", " Clara"], [" Sam"]]
invalid_targets = [[], [""], ""]
for model_name in self.small_models:
unmasker = pipeline(task="fill-mask", model=model_name, tokenizer=model_name, framework="tf")
@@ -136,7 +111,7 @@ class FillMaskPipelineTests(MonoInputPipelineCommonMixin, unittest.TestCase):
"My name is <mask>",
"The largest city in France is <mask>",
]
valid_targets = ["ĠPatrick", "ĠClara"]
valid_targets = [" Patrick", " Clara"]
for model_name in self.large_models:
unmasker = pipeline(
task="fill-mask",
@@ -209,7 +184,7 @@ class FillMaskPipelineTests(MonoInputPipelineCommonMixin, unittest.TestCase):
"My name is <mask>",
"The largest city in France is <mask>",
]
valid_targets = ["ĠPatrick", "ĠClara"]
valid_targets = [" Patrick", " Clara"]
for model_name in self.large_models:
unmasker = pipeline(task="fill-mask", model=model_name, tokenizer=model_name, framework="tf", top_k=2)