Multilingual E5 Large text embedding added

This releases adds in the Multilingual E5 Large model as the new default text embedder in the embed_text deployment.

Text Embedding Added

Added text embedding deployment. Initial model supported is the SigLIP text model to accompany the image embedding model. Additional text embedding models will be added in future releases.

Image Embedding

Added initial AI capability in the model repository. This is the embed_image deployment which is a Triton Inference Server Business Logic Script (BLS) for image embedding models. Currently, the SigLIP model is supported.

This shows the basic pattern that we want for each model capability added to the repository. Check out the docs/embed_image.md and docs/siglip_vision.md

• Short description of what the deployment does
• Examples of how to send data to the endpoint
• Performance load testing to demonstrate that the deployment has quality throughput
• Validation of the model to demonstrate the model performs as expected.

Ideally, the validation will utilize data cited in the model's original documentation. The goal is to show that the model is at least performing as described by the authors of the model. Of course, users of the deployment will need to do their own verification that the model performs well on their particular data.