Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove handle_impossible_answer from the default PIPELINE_KWARGS in the question answering evaluator #272

Merged
merged 3 commits into from
Aug 24, 2022
Merged

Remove handle_impossible_answer from the default PIPELINE_KWARGS in the question answering evaluator #272

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5b73a53
fix handle_impossible_answer arg in question answering evaluator
fxmarty Aug 23, 2022
3f4d295
add default pipeline kwargs for all subclasses of evalautor
fxmarty Aug 23, 2022
f72a969
trigger actions
fxmarty Aug 23, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions src/evaluate/evaluator/image_classification.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,6 +50,8 @@ class ImageClassificationEvaluator(Evaluator):
Methods in this class assume a data format compatible with the [`ImageClassificationPipeline`].
"""

PIPELINE_KWARGS = {}

def __init__(self, task="image-classification", default_metric_name=None):
super().__init__(task, default_metric_name=default_metric_name)

Expand Down
10 changes: 6 additions & 4 deletions src/evaluate/evaluator/question_answering.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,7 +84,7 @@ class QuestionAnsweringEvaluator(Evaluator):
[`QuestionAnsweringPipeline`](https://huggingface.co/docs/transformers/en/main_classes/pipelines#transformers.QuestionAnsweringPipeline).
"""

PIPELINE_KWARGS = {"handle_impossible_answer": False}
PIPELINE_KWARGS = {}

def __init__(self, task="question-answering", default_metric_name=None):
super().__init__(task, default_metric_name=default_metric_name)
Expand Down Expand Up @@ -196,7 +196,7 @@ def compute(

if squad_v2_format is None:
squad_v2_format = self.is_squad_v2_format(data=data, label_column=label_column)
logger.warn(
logger.warning(
f"`squad_v2_format` parameter not provided to QuestionAnsweringEvaluator.compute(). Automatically inferred `squad_v2_format` as {squad_v2_format}."
)

Expand All @@ -205,16 +205,18 @@ def compute(
metric = self.prepare_metric(metric)

if squad_v2_format and metric.name == "squad":
logger.warn(
logger.warning(
"The dataset has SQuAD v2 format but you are using the SQuAD metric. Consider passing the 'squad_v2' metric."
)
if not squad_v2_format and metric.name == "squad_v2":
logger.warn(
logger.warning(
"The dataset has SQuAD v1 format but you are using the SQuAD v2 metric. Consider passing the 'squad' metric."
)

if squad_v2_format:
self.PIPELINE_KWARGS["handle_impossible_answer"] = True
else:
self.PIPELINE_KWARGS["handle_impossible_answer"] = False

# Compute predictions
predictions, perf_results = self.call_pipeline(pipe, **pipe_inputs)
Expand Down
4 changes: 2 additions & 2 deletions tests/test_evaluator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -455,7 +455,7 @@ def test_model_init(self):
metric="squad",
)
self.assertEqual(results["exact_match"], 0)
self.assertEqual(results["f1"], 0)
self.assertEqual(results["f1"], 100 / 3)

model = AutoModelForQuestionAnswering.from_pretrained(self.default_model)
tokenizer = AutoTokenizer.from_pretrained(self.default_model)
Expand All @@ -466,7 +466,7 @@ def test_model_init(self):
tokenizer=tokenizer,
)
self.assertEqual(results["exact_match"], 0)
self.assertEqual(results["f1"], 0)
self.assertEqual(results["f1"], 100 / 3)

def test_class_init(self):
# squad_v1-like dataset
Expand Down