Charmm gui accuracy and forcefield version in it?

GROMACS version: latest
GROMACS modification: Yes/No
Here post your question
I have simulated a peptide with alternating L and D amino acid residue in lipid bilayer using files that i got from charmm-gui input generator.
**Forcefield that i opt at last step of input generator is - CHARMM36 (is it the recent version charmm36 july 2020)
**is the force-field taken into consideration in charmm-gui is the latest version of CHARMM-36

My aim of simulating these protein with alternating L and D amino acid residue is to scrutinize the conformational space of all residue in ramachandran plot after 100ns.

**how can i achieve the appropriate result that i am seeking for ?

I am attaching the input that i got (LD14_updated folder in drive link ) and some files after 100ns md run with ramachandran plot of ALA7 and DALA12.
you can see the some conformation of these residue are in disallowed region in ramachandran plot

step7_1 is nomenclature for files after 100ns mdrun in gromacs.

Hoping for a help