Skip to content

Dynamic Key-Value Memory Networks for Knowledge Tracing

Notifications You must be signed in to change notification settings

jennyzhang0215/DKVMN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

24 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DKVMN

Dynamic Key-Value Memory Networks for Knowledge Tracing

Built With

  • MXNet - The framework used
  • Both Python2 and Python3 are supported

Prerequisites

Model Architecture

DKVMN Architecture DKVMN Code

Data format

The first line is the number of exercises a student attempted. The second line is the exercise tag sequence. The third line is the response sequence.

   15
   1,1,1,1,7,7,9,10,10,10,10,11,11,45,54
   0,1,1,1,1,1,0,0,1,1,1,1,1,0,0

Hyperparameters

--gpus: the gpus will be used, e.g "0,1,2,3"

--max_iter: the number of iterations

--test: enable testing

--train_test: enable testing after training

--show: print progress

--init_std: weight initialization std

--init_lr: initial learning rate

--final_lr: learning rate will not decrease after hitting this threshold

--momentum: momentum rate

--maxgradnorm: maximum gradient norm

--final_fc_dim: hidden state dim for final fc layer

--n_question: the number of unique questions in the dataset

--seqlen: the allowed maximum length of a sequence

--data_dir: data directory

--data_name: data set name

--load: model file to load

--save: path to save model

Training

python main.py --gpus 0

Testing

python main.py --gpus 0 --test True

Reference Paper

Jiani Zhang, Xingjian Shi, Irwin King, Dit-Yan Yeung. Dynamic Key-Value Memory Networks for Knowledge Tracing. In Proceedings of the 26th International Conference on World Wide Web, 2017: 765-774.

About

Dynamic Key-Value Memory Networks for Knowledge Tracing

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages