10000 GitHub - wbyiml/ms_DPCNN
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

wbyiml/ms_DPCNN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 

Repository files navigation

ms_DPCNN

wbyiml/ms_DPCNN

利用MindSpore实现DPCNN文本分类网络

数据集存放在resource目录内
使用resource/makeDataset.py将数据集转换为train.txt, test.txt
resourece
——glove.6B
————glove.6B.{50-300}d.txt
——rt-polaritydata
————processed (训练时产生的预处理文件)
————rt-polarity.pos 数据集正样本
————rt-polarity.neg 数据集负样本
————train.txt makeDataset.py生成
————test.txt makeDataset.py生成

glove embedding文件:https://pan.baidu.com/s/14BvOavIY0IqMfbJlU0_rZw
提取码:nyxg

预训练模型:https://pan.baidu.com/s/1eerBF6UWSYZnOInhGvKf7g
提取码:9at4

训练: CUDA_VISIBLE_DEVICES=0 python train.py --device_target GPU --data_path ./resource/rt-polaritydata --glove_path ./resource/glove.6B
使用预训练模型训练: CUDA_VISIBLE_DEVICES=0 python train.py --device_target GPU --data_path ./resource/rt-polaritydata --glove_path ./resource/glove.6B --pretrained pretrained/aclimdb.ckpt
评估: CUDA_VISIBLE_DEVICES=0 python eval.py --device_target GPU --data_path ./resource/rt-polaritydata --glove_path ./resource/glove.6B --ckpt_path outputs/dpcnn-20_149.ckpt

使用预训练模型进行调优,训练结果精度:0.7998
结果模型:https://pan.baidu.com/s/1v7OZ0Y2IGN4jU7Cr2ufBJg
提取码:4erp

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

0