File size: 1,977 Bytes
4709f98 496cca9 4709f98 f027f2f 4709f98 51fa942 8c0e82f 51fa942 f027f2f 8c0e82f 51fa942 f027f2f 1f6288a f027f2f 1f6288a 4709f98 51fa942 4709f98 51fa942 4709f98 efc7e07 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 |
---
language: zh
tags:
- wobert
---
## 介绍
### tf版本
https://github.com/ZhuiyiTechnology/WoBERT
### pytorch版本
https://github.com/JunnYu/WoBERT_pytorch
## 安装(主要为了安装WoBertTokenizer)
注意：transformers版本需要>=4.7.0
WoBertTokenizer的实现与RoFormerTokenizer是一样的，因此使用RoFormerTokenizer就可以了
## 使用
```python
import torch
from transformers import BertForMaskedLM as WoBertForMaskedLM
from transformers import RoFormerTokenizer as WoBertTokenizer
pretrained_model_or_path_list = [
"junnyu/wobert_chinese_plus_base", "junnyu/wobert_chinese_base"
]
for path in pretrained_model_or_path_list:
text = "今天[MASK]很好，我[MASK]去公园玩。"
tokenizer = WoBertTokenizer.from_pretrained(path)
model = WoBertForMaskedLM.from_pretrained(path)
inputs = tokenizer(text, return_tensors="pt")
with torch.no_grad():
outputs = model(**inputs).logits[0]
outputs_sentence = ""
for i, id in enumerate(tokenizer.encode(text)):
if id == tokenizer.mask_token_id:
tokens = tokenizer.convert_ids_to_tokens(outputs[i].topk(k=5)[1])
outputs_sentence += "[" + "||".join(tokens) + "]"
else:
outputs_sentence += "".join(
tokenizer.convert_ids_to_tokens([id],
skip_special_tokens=True))
print(outputs_sentence)
# RoFormer 今天[天气||天||心情||阳光||空气]很好，我[想||要||打算||准备||喜欢]去公园玩。
# PLUS WoBERT 今天[天气||阳光||天||心情||空气]很好，我[想||要||打算||准备||就]去公园玩。
# WoBERT 今天[天气||阳光||天||心情||空气]很好，我[想||要||就||准备||也]去公园玩。
```
## 引用
Bibtex：
```tex
@techreport{zhuiyiwobert,
title={WoBERT: Word-based Chinese BERT model - ZhuiyiAI},
author={Jianlin Su},
year={2020},
url="https://github.com/ZhuiyiTechnology/WoBERT",
}
``` |