llm-semantic-router/Vela-1.0-Encoder-307M-Embedding

Expand encoder retrieval coverage with matching exports and measured scores

Update encoder with additional Chinese retrieval training and matching exports

Add optional qualified AMD SDPA runtime helper

Update retrieval-distilled encoder and matching exports

Refine retrieval and multi-resolution embeddings

Improve multilingual classification and long-document retrieval

Restore product model card and update evaluation scores

Publish R3 interim encoder: classification and long retrieval gains with disclosed tradeoffs

Improve multilingual retrieval and semantic matching

Improve multilingual retrieval and semantic matching

Add ROCm-accelerated ONNX representations

Add portable FP32 ONNX models

Add original mmBERT evaluation comparison for Vela Embedding

Simplify Vela quick starts and model cards

Release Vela on the unified Encoder base

Unify Vela model cards with Semantic Router branding

Release Vela 1.0 Embedding

initial commit