Benchmark Almanac

EvaHan organizers

We track 4 benchmarks that EvaHan organizers made or helped make.

0 in use 4 with public code 0 with a leaderboard here

Benchmarks

Name Kind Released Best score Status
EvaHan 2023 Models translate classical Chinese passages into modern Chinese or English. Organizers report BLEU scores separately for each language and resource track. Translation Nov 2022 Not in cards
EvaHan 2024 Models insert sentence boundaries and punctuation into classical Chinese passages. Organizers score the two decisions separately against human annotations. Reading in many languages Dec 2023 Not in cards
EvaHan 2025 Models identify named entities in classical Chinese histories and medical texts. Organizers rank submissions by F1 on each test set. Reading in many languages Sep 2024 Not in cards
EvaHan 2026 Models transcribe printed and handwritten Chinese books and locate page elements such as illustrations and annotations. Documents, charts and text in images Nov 2025 Not in cards