FilBench-Eval is an Open LLM Evaluation Suite for Philippine Languages. The eval runner is integrated with HuggingFace's lighteval.
AI & ML interests
Multilinguality, computational linguistics, Filipino NLP
Recent Activity
View all activity
Models and dependency parsers for Tagalog using the UD_NewsCrawl dataset
-
The UD-NewsCrawl Treebank: Reflections and Challenges from a Large-scale Tagalog Syntactic Annotation Project
Paper • 2505.20428 • Published -
filbench/UD_Tagalog-NewsCrawl
Viewer • Updated • 15.6k • 143 • 1 -
filbench/tl_spacy_baseline
Token Classification • Updated • 5 -
filbench/tl_xlm_roberta_transition
Token Classification • Updated • 9
FilBench-Eval is an Open LLM Evaluation Suite for Philippine Languages. The eval runner is integrated with HuggingFace's lighteval.
Models and dependency parsers for Tagalog using the UD_NewsCrawl dataset
-
The UD-NewsCrawl Treebank: Reflections and Challenges from a Large-scale Tagalog Syntactic Annotation Project
Paper • 2505.20428 • Published -
filbench/UD_Tagalog-NewsCrawl
Viewer • Updated • 15.6k • 143 • 1 -
filbench/tl_spacy_baseline
Token Classification • Updated • 5 -
filbench/tl_xlm_roberta_transition
Token Classification • Updated • 9
models 6
filbench/tl_mdeberta_v3_transition
Token Classification • Updated • 13
filbench/tl_hash_transition
Token Classification • Updated • 11
filbench/tl_fasttext_transition
Token Classification • Updated • 30
filbench/tl_roberta_tgl_transition
Token Classification • Updated • 17
filbench/tl_xlm_roberta_transition
Token Classification • Updated • 9
filbench/tl_spacy_baseline
Token Classification • Updated • 5