Dataset Name: monoboard/thai-land-tax-full-triplets
Language: Thai (th)
Tasks: Legal Retrieval • RAG • Contrastive Learning • Triplet Loss • Embedding Training
This dataset provides a legally verified retrieval corpus for training Thai-language retrieval models under the Land and Buildings Tax Act (B.E. 2562) and related regulations.
Each example includes:
A legal question (query)
One or more oracle passages (pos)… See the full description on the dataset page:
https://huggingface.co/datasets/monoboard/thai-land-tax-full-triplets.
