Models

Below please find some of the open-source language models trained and fine-tuned for qhchina projects. They are meant as experimental tools that can be used to analyze and play with Chinese texts.

BERT-JFJB-violence-0.1 – bert-base-chinese fine-tuned on the People’s Liberation Army Daily (Jiefangjun Bao or JFJB) for detecting violent and non-violent content in Chinese texts.

BERT-EmotionIntensity-0.1 – bert-base-chinese fine-tuned on the Douban dataset for measuring emotional intensity in Chinese texts,