This dataset contains multiple configurations for different tasks.
Use the dropdown menu above to select a specific configuration to view.
main: Contains the primary train and test splits. (about 1600 samples for train and 400 samples for test)
rule_candidate: Data for rule candidate generation. (about 1600 samples)
cold_start: Data for cold-start finetuning. (about 500 samples)
inference: Data for inference tasks. (about 400 samples)
grpo_input: Input data for… See the full description on the dataset page:
https://huggingface.co/datasets/yujiangw/E-commerce.
