Paddler API Client

Features

• Add paddler_client - Rust library to connect to Paddler
• Embedding, and inference performance improvements
• Update llama.cpp bindings to the latest version

Fixes

• #140 (Can't load HF-Models with unknown model architecture)
• #138 (Error when calling generate_embedding_batch with 2 or more objects)