Tensorflow argmax（）for TFLearn？

Question

I used the mnist data set of tflearn to predict handwritten digits.

Everything works, but I have the labels as one_hot . Is there a function in tflearn , which does the same as argmax() in Tensorflow?

Answer 1

You can do it simply by doing this:

pred = model.predict(test_data)

print([ np.where(r==1)[0][0] for r in np.round(pred) ])

Best.