reonokiy/vsp-oven

--- license: apache-2.0 task_categories: - visual-question-answering - information-retrieval tags: - vision-self-play - multimodal - knowledge-intensive --- # OVEN (Vision Self-Play format) Open-domain Visual Entity Recognition (OVEN), converted to unified Parquet schema. > Hu et al., Open-domain Visual Entity Recognition: Towards Recognizing Millions of Wikipedia Entities. ICCV 2023. ## Schema | Field | Type | |-------|------| | data_id | large_string | | image_id | large_string | | question | large_string | ## Splits | Split | Rows | Size | |-------|------|------| | oven_entity_test | 709,478 | 1.5 MB | | oven_entity_train | 4,926,314 | 8.9 MB | | oven_entity_val | 126,199 | 0.5 MB | | oven_query_test | 19,781 | 0.3 MB | | oven_query_train | 32,255 | 0.5 MB | | oven_query_val | 3,291 | 0.1 MB | | **Total** | **5,817,318** | | ## Usage ```python from datasets import load_dataset ds = load_dataset("reonokiy/vsp-oven") # 或加载特定 split train = load_dataset("reonokiy/vsp-oven", split="train") ```