Skip to content

Dataset · Orphanage

Baidu Tieba post corpus for SFT

Orphanage/Baidu_Tieba_SunXiaochuan

Text data from the Chinese Baidu Tieba community for style fine-tuning of language models.

What's inside

Contains about 100,000 Chinese text records randomly scraped from the Baidu Tieba forum (Xiaochuan community). Without images and video, the data is presented in ChatGLM4 format as train.jsonl, dev.jsonl, and raw source text files.

What you can do with it

  • SFT of language models for style tuning to colloquial Chinese
  • Experiments with dialog format model fine-tuning
Tasks
Text generation, Question answering
Languages
Chinese
Size
10K–100K rows
Downloads this month
676
Likes
13

How to load it

from datasets import load_dataset

ds = load_dataset("Orphanage/Baidu_Tieba_SunXiaochuan")