Skip to content

use Py_torch achieve GRU, scale_dot_attention, transform,IF you are a getting start people, you may want see this.

Notifications You must be signed in to change notification settings

GW-S/tranform_pytorch

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 

Repository files navigation

I hava finish the simple transform model,this achievement is very base.because Iuse ground floor api, not achieve GRU only by nn.GRU.

In this file ,I achieve GRU,scaleDotAttention,additiveAttention,and transform,and a attention visualize.

but transform may hava some wrong.So cant be use indirectly,But when you can't find GRU achievement, you can see this git.

我实现了一个GRU,不是直接调库的实现,是用底层的代码实现的。

除了GRU,还有神经网络Attention,scaleDotAttenton,还有最后的transform.

还有一个英文教程能帮你们理解这个东西,如果想学习的话,可以联系我微信:sgw951225,但实话说,要付费的,五块钱吧,都中国人,你懂的。如果需要写模型的话可以联系我微信。或者需要承接做深度学习,机器学习大作业的,都可以联系我,承接数据分析及自然语言处理业务。

tip:This is some useful resource about this task: Attention is all you need: http://blog.stupidme.me/transformer-attention-is-all-you-need/ https://yq.aliyun.com/articles/342508?utm_content=m_39938 https://zhuanlan.zhihu.com/p/47282410 构建模型的几种方法: xbiquge.cc/book/42381/

使用技术: torch.cat,bmm,matmul,cuda,tril,mask_fill_ anaconda导出库等对我新鲜的技术

About

use Py_torch achieve GRU, scale_dot_attention, transform,IF you are a getting start people, you may want see this.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published