Skip to content

TensorFlow implementation of A Fast and Accurate Dependency Parser using Neural Networks

Notifications You must be signed in to change notification settings

bryanlimy/dnn-dependency-parser

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Neural Dependency Parser

A Neural Dependency Parser using TensorFlow based on A Fast and Accurate Dependency Parser using Neural Networks.

How to run

python3 model.py --optimizer=adam --lr=0.001 --activation=relu --hidden=4 --hidden_size=500 -l2_beta=10e-8

Hyper-parameters Options

  • learning rate --lr
  • number of hidden layers --hidden
  • hidden layer size --hidden_size
  • number of epochs --epochs
  • l2 regularization beta -l2_beta
  • activation --activation can be relu or cube
  • optimizer --optimizer, can be adam or adagrad

Result

Original Result Adam ReLU 4 layers
LAS 90.7 90.1
UAS 92.0 91.3

About

TensorFlow implementation of A Fast and Accurate Dependency Parser using Neural Networks

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages