I noticed that master no longer builds for me (I get a SIGABRT when trying to load the jllama.cpp library). I tracked it down to the lack of LLAMA_NATIVE which you removed here. I'm not exactly sure why this is giving my Mac Studio fits (I am running a M2 Ultra with 128G of RAM on Ventura), but it is.
As per here I'm defaulting models.home to models in both the integration test run as well as via mvn exec:java execution.
Also, as per here I'm migrating the integration test to the 2-big quantization so we can use a smaller model.
This PR does a couple of things:
LLAMA_NATIVEwhich you removed here. I'm not exactly sure why this is giving my Mac Studio fits (I am running a M2 Ultra with 128G of RAM on Ventura), but it is.models.hometomodelsin both the integration test run as well as viamvn exec:javaexecution.