thanks for the response with potential workaround. Unfortunately I do not have Xcode 11 around anymore as it was updated by install of Xcode 12. I guess I could re-download and install it alongside Xcode 12. My own workaround was to simply access the external model file and compile and load it rather than including it in project. This seems to work fine. I did file a bug for this and received the following analysis in my feedback report: "Investigation complete - Works as currently designed" Extremely non-helpful. (Also, I looked at project settings but do not see one for auto-generation of Core ML models)

update on this issue: if I access the model by compiling and then loading it in the application that works; however getting prediction from compiled model is cumbersome because there is no wrapper class. I guess I'll file a bug and see what happens. For me it is easy to replicate the problem: (1) Generate an image classifier with create ML tool in Xcode 12 (2) Create a simple Xcode project (such as a command line tool Swift project) The project will compile and run (3) Add the ML Model class from output of model generated in step (1) to the project (4) Project now fails build process with the error given above

thanks for the response; I had done the clean and updated recommended settings. In fact I created a new project and a new ML model and had the same problem. It seems that just including the model output from create ML without any references to it at all in the project code will cause the problem.

problem still occurs with the 12.0.1 update. I have tried the various suggested workarounds but none of them work for me; although as noted by others the command line git push, pull, clone, etc work ok which indicates that the problem might be related to how the default credential helper (osxkeychain) works? Any update on the status of your filed bug?

problem occurs with released version of Xcode 12 also.