https://colab.research.google.com/drive/1MfnvRjF-nCgaPh-Oa9z98HRrz5pmOzKu?usp=sharing
https://colab.research.google.com/drive/1MfnvRjF-nCgaPh-Oa9z98HRrz5pmOzKu?usp=sharing
https://drive.google.com/drive/folders/1JEAssw1tw0bfyHUepdX1E69v1v9hypPT?usp=drive_link
https://drive.google.com/drive/folders/1JEAssw1tw0bfyHUepdX1E69v1v9hypPT?usp=drive_link
LLM uses words but using words to train model requires a massive amount of data, and with insufficient data will lead to weird and inconsistent outcome. But in NLP we uses term to train model for consistence and readable output.
https://github.com/fxsjy/jieba
https://github.com/fxsjy/jieba
use: 精确模式