Created by: jamestang0219

Hello, i want to know whether this is my fault:

in your documents, you suppose us to define the name of layer when yield just like this: yield { "pixel": pixels_float, 'label': int(label) }

and i follow your recommendation to yield like this: yield {"fea_list":fea_list,"label":label}

then define the data_layer: data = data_layer(name="fea_list", size=40)

but when i start to train the model, there are some errors: I0926 14:43:32.013584 28460 Util.cpp:113] Calling runInitFunctions I0926 14:43:32.014166 28460 Util.cpp:126] Call runInitFunctions done. [INFO 2016-09-26 14:43:32,073 networks.py:960] The input order is [fea_list, label] [INFO 2016-09-26 14:43:32,073 networks.py:963] The output order is [__cost_1__] I0926 14:43:32.075378 28460 Trainer.cpp:169] trainer mode: Normal I0926 14:43:32.075925 28460 PyDataProvider2.cpp:219] loading dataprovider dataprovider_verify::process I0926 14:43:32.076998 28460 PyDataProvider2.cpp:219] loading dataprovider dataprovider_verify::process I0926 14:43:32.077082 28460 GradientMachine.cpp:134] Initing parameters.. I0926 14:43:32.077148 28460 GradientMachine.cpp:141] Init parameters done. F0926 14:43:32.077504 28460 PythonUtil.h:197] Check failed: PySequence_Check(seq_)

i try to solve this problem, and i found if i delete the define of layer in yield code, it can successfully train the model: yield fea_list,label

is this my fault or the architecture's fault?

if this is my fault, how can i solve this problem if i want fo define the name of layer in yield.