• ✨ Reduces model size by up to 50x, enabling deployment on resource-constrained devices.
• ✨ Inference speed up to 500x faster on CPU compared to original Sentence Transformers.
• ✨ Minimal performance loss relative to original transformer models.
• ✨ Runs locally without requiring external API calls.