RisanLi/Text_Classification

TextCNN with Tensorflow and tf.Keras

Introduction

• This code implements kim's Convolutional Neural Networks for Sentence Classification simply
• Use tf.keras API to model in TextCNN.py

Requirements

• Python 3
• tensorflow 1.12
• Numpy
• Pandas
• gensim

1. Description

type	module	function	notes
-	data_helper.py	utils and data processing
-	TextCNN.py	model
-	train.py	start of whole project
-	get_vocab.py	process Google w2v file
d	dataset	datasets

2. Training

python train.py

3. Trick

• half the learning rate every 30 epochs