Configuration types (DirectML feature levels, GPU adapter selection) shared by the win-directml-java runtime modules.

BERT-class encoder pipelines (MiniLM, E5, cross-encoder reranker) on DirectML with CPU fallback, embeddings, and reranking with per-layer coalesced GPU submissions.

Neutral, swing-free Java 21 local generation runtime for Windows DirectML/WARP. Public API in com.aresstack.windirectml.inference.api (GenerationRuntime, GenerationModelHandle, catalog-driven family adapters for Qwen, SmolLM2, Gemma 3, Phi-3 and ...

Public Java 21 facade for direct in-process use of the win-directml-java local ML runtime (embeddings, batch embeddings, reranking).