forked from huashiyiqike/LSTM-MATLAB
-
Notifications
You must be signed in to change notification settings - Fork 0
/
aStart.m
executable file
·51 lines (39 loc) · 1.03 KB
/
aStart.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
clear
addpath(genpath('./dependence'))
addpath([pwd '/data/'])
format long
result.resume = 0;
result.savemat = 0;
result.savepic = 0;
result.savePath = './results/';
result.dataPath = './data/';
result.picPath = './pic/';
result.linespec = {'r.-','b*-','go-','rs-','b^-','gd-','r--'};
speed.usegpu = 'cpu_double';
speed.usecluster =0;
speed.numcluster=4;
speed.machine = 'lqc';
speed.gradientchecking= 0;
options.maxIter = 1e5 ;
options.display = 'off';
options.LS_init = 2;
options.maxFunEvals =50;
options.Corr = 10;
% options.optTol = 1e-66;
% options.TolX =1e-66;
% options.TolFun=1e-66;
problem.name='adding';
problem.datapath='/home/lq/timit/';
problem.numsamples=20000 ; % not enough
problem.batchsize= 200 ; % not enough
problem.T= 8;
problem.Ttest= problem.T ;
problem.numtest=3000;
problem.gate_size = 2;
problem.numMmcell=2;
problem.continualPredict=0;
problem.bias1 = 1; % other bias
problem.bias2 = 1; % output bias
sgd.alpha = 0.1;
sgd.momentum = 0.9;
Main(result,speed,options,problem,sgd)