The huggingface dataset crag-mm-2025/crag-mm-single-turn-public:v0.1.1 (link) has duplicated domain {2: "plants and gardening", 4: "plants and gardening "}. The label text only differs in one space char. It is a potential data quality issue.
The huggingface dataset crag-mm-2025/crag-mm-single-turn-public:v0.1.1 (link) has duplicated domain {2: "plants and gardening", 4: "plants and gardening "}. The label text only differs in one space char. It is a potential data quality issue.