Add Filipino Benchmark Community Tasks

[ljvmiranda921]

Hi! Here's our working PR for porting our changes from our own fork filbench/lighteval into HuggingFace's official community tasks :)

[NathanHB]

started reviewing, i'm launching to evals to make sure all is good so i will come back to you once it's ready. Otherwise very great job, the PR looks great !

[ljvmiranda921]

Thanks so much for the review! 👍 I have already updated and fixed the errors :)

[Copilot]

Pull Request Overview

This PR ports Filipino benchmark tasks into the HuggingFace community tasks, adding language-specific literals, a task manifest, and the full set of Filipino task configurations.

• Added custom translation literals for Cebuano and Tagalog
• Introduced an example manifest listing all Filipino community tasks
• Added a new filipino_evals.py module defining all Filipino task configs

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
src/lighteval/tasks/templates/utils/translation_literals.py	Added TranslationLiterals entries for Cebuano and Tagalog
examples/tasks/all_filipino_tasks.txt	New file listing Filipino community tasks
community_tasks/filipino_evals.py	Added Filipino task definitions across various categories

Comments suppressed due to low confidence (2)

examples/tasks/all_filipino_tasks.txt:1

• The file appears to include literal line numbers and '+' prefixes in each line, which will break downstream parsers. Remove the leading numbers and '+' characters so each line only contains the task identifier (e.g. community|readability_ceb_mcf|0|0).

community|readability_ceb_mcf|0|0

community_tasks/filipino_evals.py:1

• [nitpick] This module is very large (~800 lines) and houses many task definitions. Consider splitting tasks into category-specific submodules or separate files to improve readability and maintainability.

# MIT License