Part of the SakThai model family — fine-tuning and evaluation corpus for instruction-following and tool-use chat.
Files: data/train.jsonl, data/test.jsonl
Format: JSONL
License: apache-2.0
Last updated: 2026-08-01
from datasets import load_dataset
ds = load_dataset("Nanthasit/sakthai-combined-v6", split="train")
test_ds = load_dataset("Nanthasit/sakthai-combined-v6", split="test")
Data Fields… See the full description on the dataset page: https://huggingface.co/datasets/Nanthasit/sakthai-combined-v6.