I am solving huge mixed-integer linear programming (MILP) optimization problem (more than 100k variables).

How to properly run benchmark comparing various MILP solvers on various common and famous MIP problems, at least in terms of speed and memory?

Using a single framework is a good start I think, such as using pyomo or cvxpy.

Edit
Report