Local Correlation from modified localized HF orbitals

Sorry Professor for the misunderstanding, I do not want to perform conventional Canonical CCSD.
Let me be clear on what I have achieved so far. I have a set of localized orbitals, from which I have created the MOCOEF.LOC file, thanks to MOKIT.
Now, I want to perform Local CCSD with the approach you mentioned. Could you please correct me if the keywords in MINP file is right or not?
with LNO approximation: MINP FILE BELOW,
mem=8GB
basis=cc-pVDZ
dfbasis_scf=None
dfbasis_cor=None
calc=LNO-CCSD
uncontract=on
gauss=cart
verbosity=3
cctol=12
core=0
symm=0
charge=0
mult=1
orblocguess=read

geom=xyz
4

H         -1.67479400        -0.37892300         0.00104900
H         -1.67391300         0.38159400         0.00060400
H          1.29423900        -0.01265700        -0.00810500
H          2.05446800         0.00998700         0.00645300

Upon running the code, I get this output


 
 
 Checking keyword combinations...
 
 talg=lapl is only compatible with .not.(dfbasis_cor.eq.none)
 
 Fatal error in minp.
 Program will stop.
 
 ************************ 2026-05-01 18:04:42 *************************
                   Error at the termination of mrcc.
 **********************************************************************

I am surprised to see this, as I have not used CCSD(T) but CCSD.

Local correlation / LNO is only implemented with DF, so you cannot turn off DF via dfbasis_cor=None. This is the case for all LNO methods, LNO-CCSD and so on.

The available DF basis sets are not optimized for uncontract=on. Depending on your application, you should not use uncontract=on or optimize a dedicated DF basis for the uncontracted basis set.

Thank you Professor. For the current testing purpose, I am running uncontracted basis functions. Apart from that, I have run the job using your recommendation. However, I found that "Fatal error in drpa" which I can not understand,  I am sharing the output file for your experienced observation.

You did not share the output.
Probably again you run into some unsupported combination of special features. I would recommend not to deviate too much from my recommendations.

Very sorry Professor. Here is the output file. You can see that I have only made minimal changes to the input file. 

gauss=cart is not supported for LNO-CC methods.