Kashgari

Overview | Performance | Quick start | Documentation | 中文文档 | Contributing

🎉🎉🎉 We are proud to announce that we entirely rewrote Kashgari with tf.keras, now Kashgari comes with easier to understand API and is faster! 🎉🎉🎉

Overview

Kashgari is a simple and powerful NLP Transfer learning framework, build a state-of-art model in 5 minutes for named entity recognition (NER), part-of-speech tagging (PoS), and text classification tasks.

• Human-friendly. Kashgari's code is straightforward, well documented and tested, which makes it very easy to understand and modify.
• Powerful and simple. Kashgari allows you to apply state-of-the-art natural language processing (NLP) models to your text, such as named entity recognition (NER), part-of-speech tagging (PoS) and classification.
• Built-in transfer learning. Kashgari built-in pre-trained BERT and Word2vec embedding models, which makes it very simple to transfer learning to train your model.
• Fully scalable. Kashgari provides a simple, fast, and scalable environment for fast experimentation, train your models and experiment with new approaches using different embeddings and model structure.
• Production Ready. Kashgari could export model with SavedModel format for tensorflow serving, you could directly deploy it on the cloud.

Our Goal

• Academic users Easier experimentation to prove their hypothesis without coding from scratch.
• NLP beginners Learn how to build an NLP project with production level code quality.
• NLP developers Build a production level classification/labeling model within minutes.

Performance

Task	Language	Dataset	Score	Detail
Named Entity Recognition	Chinese	People's Daily Ner Corpus	94.46 (F1)	Text Labeling Performance Report

Tutorials

Here is a set of quick tutorials to get you started with the library:

• Tutorial 1: Text Classification
• Tutorial 2: Text Labeling
• Tutorial 3: Language Embedding

There are also articles and posts that illustrate how to use Kashgari:

• 15 分钟搭建中文文本分类模型
• 基于 BERT 的中文命名实体识别（NER)
• BERT/ERNIE 文本分类和部署
• 五分钟搭建一个基于BERT的NER模型
• Multi-Class Text Classification with Kashgari in 15 minutes

Quick start

Requirements and Installation

🎉🎉🎉 We renamed again for consistency and clarity. From now on, it is all kashgari. 🎉🎉🎉

The project is based on Python 3.6+, because it is 2019 and type hinting is cool.

Backend	pypi version	desc
TensorFlow 2.x	pip install 'kashgari>=2.0.0'	coming soon
TensorFlow 1.14+	pip install 'kashgari>=1.0.0,<2.0.0'	current version
Keras	pip install 'kashgari<1.0.0'	legacy version

Find more info about the name changing.

Example Usage

Let's run an NER labeling model with Bi_LSTM Model.

from kashgari.corpus import ChineseDailyNerCorpus
from kashgari.tasks.labeling import BiLSTM_Model

train_x, train_y = ChineseDailyNerCorpus.load_data('train')
test_x, test_y = ChineseDailyNerCorpus.load_data('test')
valid_x, valid_y = ChineseDailyNerCorpus.load_data('valid')

model = BiLSTM_Model()
model.fit(train_x, train_y, valid_x, valid_y, epochs=50)

"""
_________________________________________________________________
Layer (type)                 Output Shape              Param #
=================================================================
input (InputLayer)           (None, 97)                0
_________________________________________________________________
layer_embedding (Embedding)  (None, 97, 100)           320600
_________________________________________________________________
layer_blstm (Bidirectional)  (None, 97, 256)           235520
_________________________________________________________________
layer_dropout (Dropout)      (None, 97, 256)           0
_________________________________________________________________
layer_time_distributed (Time (None, 97, 8)             2056
_________________________________________________________________
activation_7 (Activation)    (None, 97, 8)             0
=================================================================
Total params: 558,176
Trainable params: 558,176
Non-trainable params: 0
_________________________________________________________________
Train on 20864 samples, validate on 2318 samples
Epoch 1/50
20864/20864 [==============================] - 9s 417us/sample - loss: 0.2508 - acc: 0.9333 - val_loss: 0.1240 - val_acc: 0.9607

"""

Run with GPT-2 Embedding

from kashgari.embeddings import GPT2Embedding
from kashgari.corpus import ChineseDailyNerCorpus
from kashgari.tasks.labeling import BiGRU_Model

train_x, train_y = ChineseDailyNerCorpus.load_data('train')
valid_x, valid_y = ChineseDailyNerCorpus.load_data('valid')

gpt2_embedding = GPT2Embedding('<path-to-gpt-model-folder>', sequence_length=30)
model = BiGRU_Model(gpt2_embedding)
model.fit(train_x, train_y, valid_x, valid_y, epochs=50)

Run with Bert Embedding

from kashgari.embeddings import BERTEmbedding
from kashgari.tasks.labeling import BiGRU_Model
from kashgari.corpus import ChineseDailyNerCorpus

bert_embedding = BERTEmbedding('<bert-model-folder>', sequence_length=30)
model = BiGRU_Model(bert_embedding)

train_x, train_y = ChineseDailyNerCorpus.load_data()
model.fit(train_x, train_y)

Sponsors

Support this project by becoming a sponsor. Your issues and feature request will be prioritized.[Become a sponsor]

Contributors ✨

[](#contributors-)

Thanks goes to these wonderful people (emoji key):

Eliyar Eziz 📖 ⚠️ 💻

Alex Wang 💻

Yusup 💻

Thanks for your interest in contributing! There are many ways to get involved; start with the contributor guidelines and then check these open issues for specific tasks.

Feel free to join the WeChat group if you want to more involved in Kashgari's development.

Reference

This library is inspired by and references following frameworks and papers.

• flair - A very simple framework for state-of-the-art Natural Language Processing (NLP)
• anago - Bidirectional LSTM-CRF and ELMo for Named-Entity Recognition, Part-of-Speech Tagging
• Chinese-Word-Vectors

This project follows the all-contributors specification. Contributions of any kind welcome!