Memories embeds retrieval/inference components from these projects:

ONNX Runtime - Microsoft Corporation (MIT)
https://github.com/microsoft/onnxruntime
See ONNXRUNTIME-LICENSE.txt and ONNXRUNTIME-THIRD-PARTY-NOTICES.txt.

ort Rust bindings - pyke.io (MIT OR Apache-2.0)
https://github.com/pykeio/ort

FastEmbed - Anush008 and contributors (Apache-2.0)
https://github.com/Anush008/fastembed-rs

Tokenizers - Hugging Face and contributors (Apache-2.0)
https://github.com/huggingface/tokenizers

Jieba-rs - messense and contributors (MIT)
https://github.com/messense/jieba-rs

Exact Rust dependency versions and checksums are recorded in Cargo.lock.
The Apache-2.0 license is included as LICENSE in this distribution.

EmbeddingGemma model weights are downloaded separately by the application or
supplied by the user. They are subject to the Gemma Terms of Use and Prohibited
Use Policy, not the Memories application license. The model installer includes
the upstream model card and writes a GEMMA-NOTICE.txt alongside verified files.
https://ai.google.dev/gemma/terms
https://ai.google.dev/gemma/prohibited_use_policy
