Error Message in mnist2's dense_to_one_hot function

[sehHeiden]

Hi, I have data in the structure of

data-->test-->0 -->1 -->train-->0 -->1

With a current total of 39834 files.

and I converted it with: python convert_to_mnist_format.py ./ data/train train and convert_to_mnist_format.py ./ data/test test.

And get new data withhin a folder called convert_MNIST.

Now I use your MNIST2 module with: mnist = mn.read_data_sets('convert_MNIST/', one_hot=True, num_classes=2)

The problem is, there is always an index error thrown by the dense_to_one_hot function in line 74.

The errormessage is: IndexError: index 120002 is out of bounds for size 120000

The value of num_classes in this function is 2 and the value of num_labels is 60000 (why?)

I don't know where this error is coming from. Do you how to solve it?