Skip to content

Registry refactorisation #937

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

Merged
merged 19 commits into from
Sep 1, 2025
Merged

Registry refactorisation #937

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
0ecd7b8
init, wip
clefourrier Aug 25, 2025
cfff0e9
unrelated but these tasks were buggy
clefourrier Aug 26, 2025
197acb7
better suite management: we don't load all optional deps all the time
clefourrier Aug 26, 2025
09e3f97
upgrade
clefourrier Aug 26, 2025
0adf37e
singleton + transformer sampling fix in config
clefourrier Aug 26, 2025
bc882cd
incredible how much code was just pulled from ifeval
clefourrier Aug 26, 2025
833e67d
fix test 1
clefourrier Aug 26, 2025
fbebba7
fix test 2
clefourrier Aug 26, 2025
7037815
fix tests part 1 - also removes fewshot truncation in the task name b…
clefourrier Aug 28, 2025
65c1341
fix registry mockup
clefourrier Aug 28, 2025
fa9a4b8
fixed last tests
clefourrier Aug 28, 2025
4f69c6c
Merge branch 'main' into clem-add-ifbench
clefourrier Aug 28, 2025
27bdf89
remove ifbench
clefourrier Sep 1, 2025
a5e590c
allow old task name but logs a warning
clefourrier Sep 1, 2025
8f989fa
Update tests/utils.py
clefourrier Sep 1, 2025
4b00856
Update src/lighteval/tasks/registry.py
clefourrier Sep 1, 2025
724ebf6
Update src/lighteval/tasks/registry.py
clefourrier Sep 1, 2025
79f90fb
Update src/lighteval/tasks/registry.py
clefourrier Sep 1, 2025
64b23f4
style
clefourrier Sep 1, 2025
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
82 changes: 41 additions & 41 deletions community_tasks/filipino_evals.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,7 @@
from langcodes import Language as LangCodeLanguage
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

in community tasks, changes are unrelated, it's just a bug fix

from langcodes import standardize_tag

from lighteval.metrics.dynamic_metrics import loglikelihood_acc_metric
from lighteval.metrics.dynamic_metrics import LogLikelihoodAccMetric
from lighteval.metrics.metrics import Metrics
from lighteval.metrics.normalizations import (
LogProbCharNorm,
Expand Down Expand Up @@ -87,8 +87,8 @@
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
],
),
)
Expand Down Expand Up @@ -117,8 +117,8 @@
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
],
),
)
Expand Down Expand Up @@ -154,9 +154,9 @@
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
version=0,
Expand Down Expand Up @@ -191,9 +191,9 @@
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
hf_avail_splits=["test"],
Expand Down Expand Up @@ -275,9 +275,9 @@ def filipino_dengue_pfn(line, task_name: str) -> Doc:
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
hf_avail_splits=["train", "test"],
Expand Down Expand Up @@ -327,9 +327,9 @@ def filipino_dengue_pfn(line, task_name: str) -> Doc:
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
)
Expand Down Expand Up @@ -360,9 +360,9 @@ def filipino_dengue_pfn(line, task_name: str) -> Doc:
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
hf_avail_splits=["test"],
Expand Down Expand Up @@ -396,9 +396,9 @@ def filipino_dengue_pfn(line, task_name: str) -> Doc:
hf_subset="default",
evaluation_splits=["tl"],
metrics=[
loglikelihood_acc_metric(normalization=None),
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=None),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
],
)
for formulation in [HybridFormulation(), MCFFormulation()]
Expand Down Expand Up @@ -427,9 +427,9 @@ def filipino_dengue_pfn(line, task_name: str) -> Doc:
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=None),
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=None),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
],
),
trust_dataset=True,
Expand Down Expand Up @@ -509,9 +509,9 @@ def create_sib200_task(language: Language, formulation):
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
hf_avail_splits=["test", "validation"],
Expand Down Expand Up @@ -565,9 +565,9 @@ def prepare_stingray_semantic_appropriateness(line: dict[str, str]) -> dict[str,
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
hf_avail_splits=["test"],
Expand Down Expand Up @@ -595,9 +595,9 @@ def prepare_stingray_semantic_appropriateness(line: dict[str, str]) -> dict[str,
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
hf_avail_splits=["test"],
Expand Down Expand Up @@ -718,9 +718,9 @@ def prepare_stingray_semantic_appropriateness(line: dict[str, str]) -> dict[str,
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
version=0,
Expand Down Expand Up @@ -762,9 +762,9 @@ def create_universalner_task(language: Language, formulation):
metrics=get_metrics_for_formulation(
formulation,
[
loglikelihood_acc_metric(normalization=LogProbTokenNorm()),
loglikelihood_acc_metric(normalization=LogProbCharNorm()),
loglikelihood_acc_metric(normalization=LogProbPMINorm()),
LogLikelihoodAccMetric(normalization=LogProbTokenNorm()),
LogLikelihoodAccMetric(normalization=LogProbCharNorm()),
LogLikelihoodAccMetric(normalization=LogProbPMINorm()),
],
),
version=0,
Expand Down
Loading
Loading