This contribution reports the inference and training crosscheck results of JVET-AM0175, EE1-3.2 [1]. JVET-AM0175 presents inference results of the new int16 SADL RA model with NNVC12.0 and NNVC 13.0. The DRFG model described in JVET-AM0175 utilizes the dataset Vimeo-90K for training prepared using the NNVC12.0 encoder. During the software check by the software coordinator at the JVET meeting, it was reported that the software still had some float elements in the processing of the integer model. This has been corrected by the proponents and the inference software has been modified. The cross-check results have been added to this version of the document.