Are you required to use 2018a instead of 2020a? If not, I would recommend 2020a.
The VF13 pluging allows for SQP optimization: this is needed because in a low-thrust optimization scenario, your problem is not of rank 1, so there isn't an obvious solutio.
There are other SQP solvers, such as SNOPT, IPOPT, or OpEn. Of those, IPOPT is an interior point optimizer which, I was told, isn't great for very large sparse Jacobians (1000 or more rows). SNOPT is free for students but costs a lot for corporations: it also used to be a pain in the neck to compile into GMAT so you'll probably need to reach out to Steve Hughes to compile it correctly.
More generally, astrodynamics engineers do not use GMAT to generate thrust profiles for low thrust optimization. In my experience, one approach consists in generating the low thrust optimization thrusting angles in a custom low-fidelity tool and import that into GMAT via Python.
By mid-April 2021, Nyx should have multiple shooting implemented (using OpEn) which will allow for solving low thrust optimization problems directly in high fidelity. The toolkit is validated against GMAT. (Disclaimer: I am the author of Nyx.)