为什么训练次数只有100？

[hedonghui]

我设置的训练次数是4000，为什么刚训练到100就停了呢？下面是我的程序，请各位高手帮我看看，有什么问题。还有如果训练误差满足后，直接输入检测样本来检验么？还是要先保存或者什么的。谢谢了！
p=[0.3298 0.3351 0.4120 0.3076 0.5120 0.3187 0.4102 0.4521 0.5523 0.3234 0.3910 0.5523;
0.4988 0.5266 0.4107 0.4505 0.5020 0.3679 0.3107 0.3586 0.4488 0.4266 0.4833 0.4461;
0.7702 0.5406 0.3012 0.6820 0.5294 0.6690 0.4352 0.4494 0.4602 0.5124 0.5468 0.5678;
0.4854 0.5926 0.4744 0.6456 0.573 0.4952 0.5020 0.4910 0.5758 0.4994 0.4860 0.5972;
0.4001 0.4976 0.4423 0.5102 0.4868 0.4762 0.4425 0.3986 0.5523 0.5003 0.4021 0.5107;
0.4604 0.5384 0.4312 0.5686 0.5194 0.4994 0.4552 0.3804 0.5312 0.4800 0.4688 0.5594;
0.5706 0.4922 0.4860 0.5130 0.5430 0.5608 0.4310 0.5430 0.4342 0.5726 0.4788 0.4892;
0.8383 0.5501 0.5428 0.5399 0.5217 0.4401 0.6011 0.4461 0.4988 0.4609 0.5527 0.5630];
>> p=p';
>> t=[0.4427 0.4343 0.6130 0.5358 0.4143 0.5049 0.5482 0.5266];
>> n=9;
>> net=newff(minmax(p),[9 1],{'tansig','purelin'},'traingdx');
>> net.trainParam.epchs=4000;
>> net.trainParam.goal=0.00001;
>> net=train(net,p,t);
TRAINGDX, Epoch 0/100, MSE 3.79778/1e-005, Gradient 8.34795/1e-006
TRAINGDX, Epoch 25/100, MSE 0.284816/1e-005, Gradient 1.33362/1e-006
TRAINGDX, Epoch 50/100, MSE 0.0464717/1e-005, Gradient 0.309116/1e-006
TRAINGDX, Epoch 75/100, MSE 0.0132153/1e-005, Gradient 0.061598/1e-006
TRAINGDX, Epoch 100/100, MSE 0.00165359/1e-005, Gradient 0.0168489/1e-006
TRAINGDX, Maximum epoch reached, performance goal was not met.

[anbcjys]

net.trainParam.epochs=4000;
你少了个o