📄 Paper: ModelLens: Finding the Best Model for Your Task from Myriads of Models
· 🤗 Collection: luisrui/modellens
· 💻 Code: github.com/luisrui/ModelLens
An expanded corpus of 1,807,133 (task, dataset, model, metric, value) records covering
47,242 HuggingFace models, 10,479 datasets, 9,152 metrics, and 2,581 tasks.
v2 extends v1 by folding in
three additional leaderboard sources:
HELM (Stanford CRFM) — 294,315 rowsLiveBench — 6,136 rows
OpenCompass — 581… See the full description on the dataset page:
https://huggingface.co/datasets/luisrui/ModelLens-corpus-v2.
