I want one-shot differential evolution functions exposed as diffev(cost, x0, npop=4, args=(), bounds=None…

I want one-shot differential evolution functions exposed as diffev(cost, x0, npop=4, args=(), bounds=None…: a task in MiMo-V2.6-RL-harbor-code: MiMo-V2.6-RL Code (Harbor) (Harbor dataset). For a simple bounded objective such as lambda x: (x 0 - 1.5) 2 , calling diffev(objective, (-5.0, 5.0) …

The task

For a simple bounded objective such as `lambda x: (x[0] - 1.5)**2`, calling `diffev(objective, [(-5.0, 5.0)], npop=20, bounds=[(-5.0, 5.0)], gtol=40, disp=False)` should return a best parameter vector whose first value is close to `1.5`. For a two-argument objective such as `lambda x, target: (x[0] - target)**2`,…

Part of FineEnvs/MiMo-V2.6-RL-harbor-code.