wizare

有没有办法实现 类似transformer包中的RobertaTokenizer

2

Roberta采用的是**byte level的BPE(BBPE)**编码。因此相比于BERT只需要一个vocab.txt，RoBRETa需要merges.txt和vocab.json。

Confusion of your paper

1

Hi, seanie. Your paper is interesting, and I want to try your method for other text generation tasks. However, I am confused about some points in your paper: 1. How...

input data format

15

I am confused about the input data format, i.e., encodings.npy and offsets.npy is each element in encodings.npy a one-hot vector? Can you provide a detailed demo of them?

Some problem about the implement of Cascaded Model

2

I want to use your Cascaded Model. In your usage description, there are 5 steps in total. The first two steps I successfully run， but fail in third steps. In...