https://github.com/luozhouyang/transformers-keras

Transformer-based models implemented in tensorflow 2.x(using keras).
https://github.com/luozhouyang/transformers-keras

albert bert masked-language-models ner question-answering sequence-classification sequence-labeling simcse tensorflow-keras text-classification token-classification

Last synced: 5 months ago
JSON representation

Transformer-based models implemented in tensorflow 2.x(using keras).

• Host: GitHub
• URL: https://github.com/luozhouyang/transformers-keras
• Owner: luozhouyang
• License: apache-2.0
• Created: 2019-05-06T11:39:29.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2021-11-23T02:39:01.000Z (almost 4 years ago)
• Last Synced: 2025-04-03T19:40:15.816Z (6 months ago)
• Topics: albert, bert, masked-language-models, ner, question-answering, sequence-classification, sequence-labeling, simcse, tensorflow-keras, text-classification, token-classification
• Language: Python
• Homepage:
• Size: 696 KB
• Stars: 75
• Watchers: 4
• Forks: 13
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# transformers-keras

![Python package](https://github.com/luozhouyang/transformers-keras/workflows/Python%20package/badge.svg)

[![PyPI version](https://badge.fury.io/py/transformers-keras.svg)](https://badge.fury.io/py/transformers-keras)

[![Python](https://img.shields.io/pypi/pyversions/transformers-keras.svg?style=plastic)](https://badge.fury.io/py/transformers-keras)

基于`tf.keras`的Transformers系列模型实现。

所有的`Model`都是keras模型，可以直接用于训练模型、评估模型或者导出模型用于部署。

在线文档：[transformers-keras文档](https://transformers-keras.readthedocs.io/zh_CN/latest/index.html)

本库功能预览：

* 加载各种预训练模型的权重

* 掩码语言模型(Masked Language Model)解决方案

* 句向量(Sentence Embedding)的解决方案

* 抽取式问答任务(Question Answering)的解决方案

* 序列分类任务(Sequence Classification)的解决方案

* 序列标注任务(Token Classification)的解决方案

* 知识蒸馏(Knowledge Distillation)的框架

以上任务的数据处理、模型构建都完成了，可以一键开始训练。这里有两个例子：

* [使用BERT微调文本分类任务](https://transformers-keras.readthedocs.io/zh_CN/latest/start.html#id4)

* [使用BERT微调问答任务](https://transformers-keras.readthedocs.io/zh_CN/latest/start.html#id5)

更多使用方法和介绍，请查看在线文档：[transformers-keras文档](https://transformers-keras.readthedocs.io/zh_CN/latest/index.html)

## 联系方式

> 由于二维码过期限制，加微信讨论群请先加我微信邀请加入。

* 微信: luozhouyang0528

* 邮箱: zhouyang.luo@gmail.com