WIP - Switch the MMQnA LVM for Gaudi to use the llava microservice instead of tgi-gaudi

[dmsuehir]

Description

After adding audio data prep/ingestion, we found that the LVM used for MMQnA on Gaudi wasn't working when the user query only contained text (no images). After investigating this I found that I wasn't getting reasonable results with text-only prompts to tgi-gaudi (even outside of using other OPEA services). Xeon uses the llava microservice from GenAIComps for the LVM instead, which uses the transformers image-to-text pipeline in the backend. This llava microservice also supports Gaudi. In this PR I am switching the LVM microservice for MMQnA for Gaudi to use llava instead of tgi-gaudi, in order to be able to get better results with text only queries. Another thing to note is that it doesn't seem like I'm able to use LLaVa-NeXT
models with microservice, so the model has to be changed as well.

Issues

RFC: https://github.com/opea-project/docs/blob/main/community/rfcs/24-10-02-GenAIExamples-001-Image_and_Audio_Support_in_MultimodalQnA.md

Type of change

List the type of change like below. Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)

Dependencies

List the newly introduced 3rd party dependency if exists.

Tests

Describe the tests that you ran to verify your changes.