We use perplexity to measure the efficiency of our language model. We collect the experiment corpus from daily news on Chinese TV System (CTS) website.
確定! 回上一頁