| EmbeddingGemma 2 source weights and original model documentation by Google DeepMind, distributed under Apache-2.0. | |
| Host tokenization and preprocessing use Hugging Face Transformers under Apache-2.0. | |
| Core AI conversion and adapter by maxffarrell, 2026, Apache-2.0. | |
| See UPSTREAM_README.md and SOURCE.json for attribution and pinned provenance. | |