项目作者: LuxxxLucy

项目描述 :
(in 2017) implement paper for character-level and word-level text generation
高级语言: Python
项目地址: git://github.com/LuxxxLucy/text_generation_SeqRNN.git
创建时间: 2017-08-10T05:55:25Z
项目社区:https://github.com/LuxxxLucy/text_generation_SeqRNN

开源协议:MIT License

下载


char-RNN

still under developing

This is a Python3/ Keras / Tensorflow implementation of paper Generating Sequences With Recurrent Neural Networks

Can use it to do character level text generation and word-level

For character level text generation, I used the linux source code.

For word-level text generation, I used works from shakespeare

Setup

To run this code you need the following:

  • a machine with multiple GPUs
  • Python3
  • Numpy, Keras

Training the model

Use the main_entry.py script to train the model. To train the default model on CIFAR-10 simply use:

  1. python3 main_entry.py

usage

indicate the training dataset by

  • ‘—data_set linux_data’: linux source code (character level)
  • ‘—data_set shakespeare’: shakespeare’s plays (word-level)