akash13singh / mxnet-for-cdl Goto Github PK

When I run the code with a rating matrix (100000, 10000), the code crashed, and the reported error is attached. When I use a rating matrix (10000, 10000), it works normal. This error seems caused by memeory out. How to fix such error?

Traceback (most recent call last):
File "cdl.py", line 89, in
num_iter, 'sgd', l_rate=0.1, decay=0.0)#,
File "/data/CF/mxnet-for-cdl-master-kb/autoencoder.py", line 175, in finetune
last_batch_handle='pad')
File "/usr/lib/python2.7/site-packages/mxnet-0.10.0-py2.7.egg/mxnet/io.py", line 579, in init
self.data = _init_data(data, allow_empty=False, default_name=data_name)
File "/usr/lib/python2.7/site-packages/mxnet-0.10.0-py2.7.egg/mxnet/io.py", line 487, in _init_data
"should be NDArray or numpy.ndarray")
TypeError: Invalid type '<type 'numpy.ndarray'>' for data, should be NDArray or numpy.ndarray
[22:57:14] /data/CF/mxnet/dmlc-core/include/dmlc/./logging.h:304: [22:57:14] /data/CF/mxnet/mshadow/mshadow/./stream_gpu-inl.h:129: Check failed: err == CUDNN_STATUS_SUCCESS (4 vs. 0) CUDNN_STATUS_INTERNAL_ERROR

Thanks again,

What could be possible future improvements on your code? Is there anyway to optimize it? Or you think this is an optimized code?

Why you used sigmoid instead of relu?

Hello,

I run the code for the first time and i got 0.55 recall when P=1. I downloaded the code again and run it, it now generates 0.05 recall! Can you tell me why?

how the file 'images' come out, mnist data or citeulike?