This PR aims to introduce a stronger test for batch size of 2 by testing classification instead of regression. In the regression task batch size 2 yields the same results as batch size 2. Also, it solves the pylint errors possibly-used-before-assignment.
This PR aims to introduce a stronger test for batch size of 2 by testing classification instead of regression. In the regression task batch size 2 yields the same results as batch size 2. Also, it solves the pylint errors
possibly-used-before-assignment
.