Inference Components

llama.cpp

llama-cpp-python provides the base capability of running inference on CPU and builds on the work of llama.cpp. Both packages are available under an MIT license.

Hugging Face Transformers

Hugging Face Transformers provides the platform that we use for offering voice-to-text and for use of the Vision Language Model for document conversion. It is licensed via the Apache 2.0 License

Hugging Face Transformers runs using pytorch which is licensed using a BSD-style License

Models

The embedding model is nomic-embed-text-v1.5 and is licensed using the Apache-2.0 license.