Hello. I'm Tanaka from the Information Processing Laboratory of Yamada Tanaka Shoten. I study every day what a model learns and what it can be used for in unsupervised learning.
I created a language model with RNN and LSTM and tried to divide it. The learning data is 5 epochs of the Japanese version of wikipedia.
You can try the word-separation function by accessing the site below. I tried adding the "maybe a typo" function as a bonus.
Recommended Posts