* Add ONNX RT-GenAI backend (v0.8.2) support to LLM
* Build ONNX Runtime (v1.22.0) as part of ONNX Runtime-GenAI Integration
* Introduce general KleidiAI flag (-DUSE_KLEIDIAI) and configure framework-specific KleidiAI flags
* Update README with ONNX Runtime-GenAI backend usage and build instructions
* Disable -DLLAMA_CURL by default

Change-Id: I6812a3fafb78f55043a7653c7dce267a20670cf3
Signed-off-by: Yunus Kalkan <yunus.kalkan@arm.com>

* Add missing docstring

* Rename the following readme's to resolve:
  Doxygen warning: found more than one mainpage
  1) src/cpp/config/README.md to intro.md
  2) src/cpp/frameworks/README.md to intro.md

Signed-off-by: Conor Kennedy <conor.kennedy@arm.com>
Change-Id: I7d1ebed3eef575a547ab7a5bc0dac0454180122b

* remove use of template throughout the implementation that needs us
 to provide the backend used as an argument.
* Using an opaque pointer in the top level header that can then
 be implemented by different backends.
* Formatting changes to adhere to the coding guidelines for C/C++

Change-Id: Ia47889535609f7e6c5ac6214399f403b3d63f5b4
Signed-off-by: Nina Drozd <nina.drozd@arm.com>

Liam Barry authored May 21, 2025 and Éanna Ó Catháin committed May 29, 2025

Updated function names and commit SHA

Signed-off-by: Liam Barry <liam.barry@arm.com>
Change-Id: Ib66d8ae1b437e1119e9bc0415a03177f1dbae4ca

MLECO-6048: Async mode tests for llama.cpp

See merge request !1

Liam Barry authored Apr 29, 2025 and Nina Drozd committed May 02, 2025

Support code added to allow testing of subscriber and token emission in async mode

Signed-off-by: Liam <liam.barry@arm.com>
Change-Id: I23f0e3919a525538e8f31965eeba6ea44e99ea80