Skip to content

MLECO-6388: Upstream ONNX Runtime framework integration into RTVA

Yunus Kalkan requested to merge feature/onnxrt-integration into main
  • Add llmFramework build option to select between LLM backends (e.g., llama.cpp, onnxruntime-genai)
  • Integrate llmFramework selection into native build via Gradle arguments
  • Replace shell script with Python script for pushing models and configuration files
  • Add ONNX Runtime specific configuration files (e.g., onnxConfigUser.json)
  • Update README with LLM framework usage instructions and build options
  • Update the referenced LLM repository commit to enable ONNX Runtime compatibility

Change-Id: Ie998213a79828fc647b8e9b371e6ab0bcb2b0f01 Signed-off-by: Yunus Kalkan yunus.kalkan@arm.com

Merge request reports