Creating my own ML_AB file for MLFF

[melissa_sanseverino]

Hello!
I have been trying to use MLFF to find a potential energy surface for a system using previous ab initio MD calculations. I have taken the data from my previous calculations and compiled them into a new ML_AB file. I directly followed the instructions on VASP wiki on how to format this file and everything but I keep getting errors when I try to run these calculations. Please let me know if there is something wrong with my ML_AB file or what the issue might be in general.
Thank you in advance!

[alexey.tal]

Hello,

I see that there is a number of problems in your ML_AB file.
- The maximum number of atoms per atom type section is missing in the header of your ML_AB
- The number of atom types is not correct in line 63, should be 5
- The number of atoms section is missing in the first configuration
- In Atom types and atom numbers the total number of 48 atoms is specified. In positions and forces only 41 atom is provided.

Please carefully check that your ML_AB file is consistent with the structure described on the wiki.

[melissa_sanseverino]

Thank you so much for your help!
I have corrected all of the things you pointed out and rechecked my ML_AB file. I even asked chatgpt to find any errors in my ML_AB file that wouldn't line up with the one provided in VASP wiki and it came up empty. Unfortunately, the calculations still would not run and I am wondering what I am doing wrong.
Thank you in advance!

[melissa_sanseverino]

So sorry I posted the wrong zip file in the first reply. Attached is the correct zip file.

[alexey.tal]

You missed The number of atoms in the configurations #120-134.

[melissa_sanseverino]

Thank you so much!