Skip to content

Latest commit

 

History

History
65 lines (46 loc) · 1.55 KB

README.md

File metadata and controls

65 lines (46 loc) · 1.55 KB

KoreanLM-demo

Prerequisites

Required Torch libraries:

torch, nn, cutorch, cunn, lua-utf8

Trained model:

Download lm-rnn2048x2-v2-wd0-dw1.zip from 1) [googledrive] 2) Titan:/data5/demo-sylee/lm-rnn2048x2-v2-wd0-dw1.zip

unzip to: exp/lm-rnn2048x2-v2-wd0-dw1/lm_char_epoch43.00.t7

How to execute:

$th demo_main_sylgen.lua

It basically requires CUDA

Demo Examples:

using CUDA on GPU 0...
--- Loading vocab : exp/vocab.t7
Word vocab size: 2189
-- Building network : exp/lm-rnn2048x2-v2-wd0-dw1/lm_char_epoch43.00.t7
======================================
Please enter a front part of sentence :
더
LSTM : resetStates
LSTM : resetStates


더 많은 것을 보고 있다<eos>

======================================
Please enter a front part of sentence :
위
LSTM : resetStates
LSTM : resetStates


위원회는 이 사업을 추진하고 있다<eos>

======================================
Please enter a front part of sentence :
이날
LSTM : resetStates
LSTM : resetStates


이날 오전 10시 30분께 이 사건을 수사하고 있다<eos>

======================================
Please enter a front part of sentence :
비가
LSTM : resetStates
LSTM : resetStates


비가 내리는 것은 이번이 처음이다<eos>
======================================
Please enter a front part of sentence :
유럽의
LSTM : resetStates
LSTM : resetStates


유럽의 경우 이 지역에서 이 지역에서 가장 많은 수출이 가능하다<eos>