Revert LLAMA_NATIVE to OFF in flake.nix

[iSma]

I've noticed that since PR #4605, performance (CPU-only) took a massive dive when using the Nix flake (I went from ~4 tokens/s to <0.5). It seems that the slowdown is caused by LLAMA_NATIVE=ON. Reverting to OFF (as it was before the PR) restores the expected performance.

This regression was observed on both an i7-1165G7 and a Ryzen 3800X running NixOS.

FWIW, the llama-cpp package in nixpkgs has LLAMA_NATIVE=OFF.

I'm not sure what the implications of turning off LLAMA_NATIVE are, maybe @philiptaron and @SomeoneSerge want to chime in.

[SomeoneSerge]

option(LLAMA_NATIVE "llama: enable -march=native flag" ON)

Oh yes, we surely would prefer that OFF. Ideally, we never resort to -march=native (which generates random outputs depending on the builder's scheduler, load, and hardware), but instead model concrete targets or concrete architecture levels as part of the derivation

[philiptaron]

Dang, that's my fault in doing the transcription. Call it an "off by on" error 😅 . Thanks for the PR; LGTM.