https://github.com/dalinvip/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch

In PyTorch Learing Neural Networks Likes CNN、BiLSTM
https://github.com/dalinvip/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch

cnn-bilstm cnn-model gru highway-cnn lstm-model model-bilstm pytorch torchtext

Last synced: 6 days ago
JSON representation

In PyTorch Learing Neural Networks Likes CNN、BiLSTM

• Host: GitHub
• URL: https://github.com/dalinvip/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch
• Owner: dalinvip
• License: apache-2.0
• Created: 2017-07-28T11:45:58.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2023-03-20T02:58:24.000Z (almost 2 years ago)
• Last Synced: 2025-01-19T09:03:18.903Z (13 days ago)
• Topics: cnn-bilstm, cnn-model, gru, highway-cnn, lstm-model, model-bilstm, pytorch, torchtext
• Language: Python
• Homepage:
• Size: 24.6 MB
• Stars: 1,297
• Watchers: 23
• Forks: 344
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
## Introduction ##

- A classification task implement in pytorch, contains some neural networks in [models](https://github.com/bamtercelboo/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch/tree/master/models).

* Recenely,  I've released the code. 

	* **old-version-17** release [here](https://github.com/bamtercelboo/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch/releases/tag/pytorch0.3.1-old_version_17)  

	* **pytorch version == 0.3.1** release on [here](https://github.com/bamtercelboo/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch/releases/tag/pytorch0.3.1)  

- This is a version of my own architecture  ---  [pytorch-text-classification](https://github.com/bamtercelboo/pytorch_text_classification)  

- **BERT For Text Classification**  --- [PyTorch_Bert_Text_Classification](https://github.com/bamtercelboo/PyTorch_Bert_Text_Classification)  

## Requirement ##

	pyorch : 1.0.1

	python : 3.6

	torchtext: 0.2.1

	cuda : 8.0 (support cuda speed up, can chose, default True)

## Usage ##

 

modify the config file, see the Config directory([here](https://github.com/bamtercelboo/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch/tree/master/Config)) for detail.  

	1、python main.py

	2、python main.py --config_file ./Config/config.cfg 

	3、sh run.sh

## Model ##

Contains some neural networks implement in pytorch, see the [models](https://github.com/bamtercelboo/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch/tree/master/models) for detail.

## Data ##

SST-1 and SST-2.

## Result ##

I haven't adjusted the hyper-parameters seriously, you can also see train log in [here](https://github.com/bamtercelboo/cnn-lstm-bilstm-deepcnn-clstm-in-pytorch/tree/master/result).  

The following test set accuracy are based on the best dev set accuracy.    

| Data/Model | % SST-1 | % SST-2 |  

| ------------ | ------------ | ------------ |  

| CNN | 46.1086 | 84.2943 |  

| Bi-LSTM | 47.9186 | 86.3262 |  

| Bi-GRU | 47.6923 | 86.7655 |  

## Reference ##

- [基于pytorch的CNN-LSTM神经网络模型调参小结](http://www.cnblogs.com/bamtercelboo/p/7469005.html "基于pytorch的CNN-LSTM神经网络模型调参小结")

- [Convolutional Neural Networks for Sentence Classification](https://arxiv.org/pdf/1408.5882.pdf)

-  [Context-Sensitive Lexicon Features for Neural Sentiment Analysis](https://arxiv.org/pdf/1408.5882.pdf)