Where is the compiled llama directory

[NEMESYS43]

I have successfully compiled Llama for my setup previously. I am using ROCm 5.6 with a Radeon 7900xtx. When using the Cuda build guide in the docs I can't seem to get it to successfully compile. Can't I just replace the llama binary that this library uses with the one I already have compiled? has anyone found where it is stored in this case? I have been digging around and have not been able to find any clues.

[giladgd]

@NEMESYS43 This project compiles llama.cpp with custom code to bridge it properly to nodejs, so you can't provide your own binary for this.
You can follow the instructions in this comment to compile it for AMD ROC, but note that llama.cpp's support for OpenCL is deprecated and can be removed at any time.