Kaggle submission formatting

Following is the way my predictions are, but how do I change it in the kaggle format or the format from where I can calculate recall, precision etc.

Original labels:
from sklearn.metrics import accuracy_score
y_test[1:10]
array([[ 0., 0., 1., 0., 0., 0., 0., 0., 0., 0.],
[ 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.],
[ 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.],
[ 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.],
[ 0., 1., 0., 0., 0., 0., 0., 0., 0., 0.],
[ 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.],
[ 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.],
[ 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.],
[ 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.]])

But my predictions are in the following form.

predictions[1:10]
array([[ 2.8738e-08, 1.0000e+00, 3.5237e-08, 3.5941e-12, 1.5171e-08, 1.2188e-08,
1.6115e-08, 7.8309e-09, 2.7810e-08, 1.7066e-08],
[ 2.0584e-04, 2.4784e-05, 2.6280e-04, 5.5377e-04, 2.9287e-05, 9.9776e-01,
4.1169e-04, 8.5111e-06, 5.0882e-05, 6.8836e-04],
[ 5.1701e-13, 1.4983e-12, 2.5944e-14, 2.0304e-10, 6.9109e-13, 1.0000e+00,
5.0489e-09, 1.1411e-14, 4.2517e-11, 2.8654e-07],
[ 5.0656e-04, 4.6778e-04, 3.5112e-03, 3.1257e-04, 8.5019e-05, 1.4760e-03,
1.1833e-06, 9.4619e-01, 1.1579e-04, 4.7338e-02],
[ 2.9502e-11, 1.7564e-11, 4.1796e-09, 8.4230e-11, 3.5063e-11, 3.4655e-10,
1.3052e-10, 9.3465e-13, 1.0000e+00, 1.1622e-08],
[ 2.6711e-06, 9.9993e-01, 4.7585e-05, 8.9812e-09, 6.1664e-06, 9.8106e-07,
6.0579e-07, 6.3763e-06, 5.3902e-06, 2.5550e-06],
[ 1.1443e-10, 1.0254e-11, 6.6635e-09, 1.0619e-09, 1.4991e-08, 2.5944e-09,
2.9823e-13, 1.8983e-09, 1.3178e-07, 1.0000e+00],
[ 1.1051e-14, 5.0164e-09, 8.5650e-11, 1.0000e+00, 3.6033e-14, 6.4902e-08,
1.0050e-13, 7.8782e-10, 1.2581e-09, 1.5517e-08],
[ 1.8374e-09, 2.0662e-09, 1.0000e+00, 1.2901e-09, 2.6832e-09, 3.6185e-13,
2.1450e-11, 3.3748e-09, 1.1634e-06, 2.8909e-10]], dtype=float32)
So how do I change these to the format as above?

And also for kaggle submission how do I map the picture and predictions?

have a look at the end of this notebook: