This commit is contained in:
tlk-dsg 2021-10-09 14:42:23 +08:00
parent 27148a7453
commit b67c87ffb5
3 changed files with 22 additions and 12 deletions

View File

@ -2,10 +2,18 @@
<a href="https://github.com/zjunlp/deepke"> <img src="pics/logo.png" width="400"/></a>
<p>
<p align="center">
<a href="http://121.36.172.141">
<a href="https://deepke.openkg.cn">
<img alt="Documentation" src="https://img.shields.io/badge/DeepKE-website-green">
</a>
<b>简体中文 | <a href="https://github.com/zjunlp/DeepKE/blob/test_new_deepke/README_ENGLISH.md">English</a> </b>
<a href="https://pypi.org/project/deepke/#files">
<img alt="PyPI" src="https://img.shields.io/pypi/v/deepke">
</a>
<a href="https://github.com/zjunlp/DeepKE/blob/master/LICENSE">
<img alt="GitHub" src="https://img.shields.io/github/license/zjunlp/deepke">
</a>
</p>
<p align="center">
<b>简体中文 | <a href="https://github.com/zjunlp/DeepKE/blob/test_new_deepke/README_ENGLISH.md">English</a></b>
</p>
<h1 align="center">
@ -45,16 +53,18 @@ DeepKE包括了三个模块可以进行关系抽取、实体命名识别以
> python == 3.8
- torch >= 1.5
- torch == 1.5
- hydra-core == 1.0.6
- tensorboard >= 2.0
- matplotlib >= 3.1
- transformers >= 2.0
- jieba >= 0.39
- scikit-learn >= 0.22
- pytorch-transformers >= 1.2.0
- seqeval >= 0.0.5
- tqdm >= 4.31.1
- tensorboard == 2.4.1
- matplotlib == 3.4.1
- transformers == 3.4.0
- jieba == 0.42.1
- scikit-learn == 0.24.1
- pytorch-transformers == 1.2.0
- seqeval == 1.2.2
- tqdm == 4.60.0
- nltk == 3.6.3
1. **命名实体识别NER**
@ -71,7 +81,7 @@ DeepKE包括了三个模块可以进行关系抽取、实体命名识别以
常规模块为预训练模型可进入其目录修改数据集以及conf文件夹下的目录```python run.py```即可训练,```python predict.py```即可预测。
**[FEW-SHOT](https://github.com/zjunlp/DeepKE/tree/test_new_deepke/example/ner/few-shot)**
少样本模块使用了LightNER模型可进入其目录模型加载和保存位置以及配置可以在shell脚本中修改```bash run_conll2003.sh```训练conll2003,```bash run_fewshot.sh "mit-movie" False```不加载模型直接进行few-shot训练,```bash run_fewshot.sh "mit-movie" True```加载模型进行few-shot训练,```bash run_predict.sh```进行预测。
少样本模块使用了LightNER模型可进入其目录模型加载和保存位置以及配置可以在shell脚本中修改```python run.py```训练conll2003,```python run.py +train=few_shot```直接进行few-shot训练,若要加载模型修改few_shot.yaml中的load_path,```python predict.py```即可预测。
2. **关系抽取RE**

Binary file not shown.

Before

Width:  |  Height:  |  Size: 120 KiB

After

Width:  |  Height:  |  Size: 85 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 740 KiB

After

Width:  |  Height:  |  Size: 214 KiB