Created by: LoganZhou

想使用LSTM网络来进行训练，参考trainer_config.lstm.py文件对trainer_config.py进行修改，但训练报错

I0225 17:12:06.709349 17327 Trainer.cpp:175] trainer mode: Normal
I0225 17:12:06.770145 17327 PyDataProvider2.cpp:243] loading dataprovider dataprovider::process
I0225 17:12:06.784571 17327 PyDataProvider2.cpp:243] loading dataprovider dataprovider::process
I0225 17:12:06.786155 17327 GradientMachine.cpp:135] Initing parameters..
I0225 17:12:06.796674 17327 GradientMachine.cpp:142] Init parameters done.
F0225 17:12:58.758431 17327 LstmLayer.cpp:155] Check failed: input.sequenceStartPositions 
*** Check failure stack trace: ***
    @           0x93ca56  google::LogMessage::Fail()
    @           0x93c9a2  google::LogMessage::SendToLog()
    @           0x93c326  google::LogMessage::Flush()
    @           0x93f3c5  google::LogMessageFatal::~LogMessageFatal()
    @           0x640b87  paddle::LstmLayer::forward()
    @           0x6c9e60  paddle::NeuralNetwork::forward()
    @           0x6bc453  paddle::GradientMachine::forwardBackward()
    @           0x75786d  paddle::TrainerInternal::forwardBackwardBatch()
    @           0x757dec  paddle::TrainerInternal::trainOneBatch()
    @           0x752cf0  paddle::Trainer::trainOneDataBatch()
    @           0x7554ef  paddle::Trainer::trainOnePass()
    @           0x756900  paddle::Trainer::train()
    @           0x5c6913  main
    @     0x2b87548e9d1d  __libc_start_main
    @           0x5dcb41  (unknown)
/usr/local/Paddle-GPU/bin/paddle: line 109: 17327 Aborted                 (core dumped) ${DEBUGGER} $MYDIR/../opt/paddle/bin/paddle_trainer ${@:2}

LSTM构建部分如下：

output_label = []
link_encode = data_layer(name='link_encode', size=TERM_NUM)
for i in xrange(FORECASTING_NUM):
    link_param = ParamAttr(
        name='_link_vec.w', initial_max=1.0, initial_min=-1.0)
    link_vec = fc_layer(input=link_encode, size=emb_size, param_attr=link_param)
    lstm = simple_lstm(
        input=link_vec, size=emb_size, lstm_cell_attr=ExtraAttr(drop_rate=0.25))
    score = fc_layer(input=lstm, size=4, act=SoftmaxActivation())
    if is_predict:
        maxid = maxid_layer(score)
        output_label.append(maxid)
    else:
        label = data_layer(name='label_%dmin' % ((i + 1) * 5), size=4)
        cls = classification_cost(
            input=score, name="cost_%dmin" % ((i + 1) * 5), label=label)
        output_label.append(cls)
outputs(output_label)

之前也参考文档使用mixed_layer来进行搭建LSTM，但也是报错 想知道这样构建错在哪里，望解答，谢谢！