Evaluation of text generation using perplexity.

Looking at a lot of methods, I want to use the degree of confusion to compare the results of model generation and human results. I plan to use the LSTM model to do the training set with human text...

01 January 1970 585 1 View