Resolve build errors: update/remove kv cache APIs, disable building tools, latest release & upstream org ref

[brittlewis12]

Hi Marcus, thanks for continuing to maintain this excellent project!

I encountered a few small issues catching up to the latest llama.cpp release:

• there’s a newly reorganized tools/ dir in llama.cpp for certain popular examples/ -- disabled these in the build.rs to align with avoiding building tests, examples, etc.
  • llama : move end-user examples to tools directory ggml-org/llama.cpp#13249
  • my editor auto-formats (cargo fmt), so I also included the small change to build.rs
• llama_kv_cache_* APIs have been renamed llama_kv_self as you noted in compile error for update-llama-cpp-2025-05-21 branch #739 (and the kv cache view APIs were removed entirely)
  • llama : remove llama_kv_cache_view API + remove deprecated ggml-org/llama.cpp#13653

I also noticed the repo has a stale organization reference for the git submodule -- not a big deal since github redirects it, but figured worth keeping up-to-date

let me know if you have any further feedback about these changes -- eg I could imagine wanting to rename the public-facing kv_cache_* methods to kv_self_* to mirror upstream? and whatever else comes to mind.

have a lovely weekend!

[MarcusDunn]

renaming the functions to match upstream would be appreciated, but I see no reason for that to block this PR.

Always appreciative of the PRs! Thanks!