zhanghaoyu a0bac8e604 feat(text_postprocessors): add parse_bracketed_answer function to extract answers enclosed in brackets ... This function is designed to extract option answers enclosed in brackets from text, and replaces the use of first_option_postprocess in the MaritimeBench dataset. This change aims to support the answer format requirements of the MaritimeBench dataset and ensure that model outputs conform to the <option> format.		2025-04-10 16:04:42 +08:00
..
cli	[Feature] Make dump-eval-details default behavior (#1999)	2025-04-08 14:42:26 +08:00
configs	feat(text_postprocessors): add parse_bracketed_answer function to extract answers enclosed in brackets	2025-04-10 16:04:42 +08:00
datasets	chore: fix line endings and formatting; add maritime_bench dataset	2025-04-10 13:40:28 +08:00
evaluator	[Feature] Add CascadeEvaluator (#1992)	2025-04-08 11:58:14 +08:00
lagent	Update CIBench (#1089)	2024-04-26 18:46:02 +08:00
metrics	[Feat] Support multi-modal evaluation on MME benchmark. (#197)	2023-08-21 15:53:20 +08:00
models	[Update] Requirements update (#1993)	2025-04-02 12:03:45 +08:00
openicl	chore: fix line endings and formatting; add maritime_bench dataset	2025-04-10 13:40:28 +08:00
partitioners	[Feature] Evaluation Results Persistence (#1894)	2025-03-05 18:33:34 +08:00
runners	[Update] Increase memory size for CPU job of VOLC Runner (#1962)	2025-03-24 11:21:14 +08:00
summarizers	[CI] fix baseline score (#2000)	2025-04-03 19:32:36 +08:00
tasks	[Refactor] Refactorize openicl eval task (#1990)	2025-04-09 15:52:23 +08:00
utils	feat(text_postprocessors): add parse_bracketed_answer function to extract answers enclosed in brackets	2025-04-10 16:04:42 +08:00
__init__.py	[BUMP] Bump version to 0.4.2 (#1997)	2025-04-02 17:47:15 +08:00
registry.py	[Feature] Add Judgerbench and reorg subeval (#1593)	2024-10-15 16:36:05 +08:00