Considerations for Using LLMs: External vs. In-House Solutions

Brandon J. We currently don't host our own LLMs. We might eventually. I feel like teams fall into 2 camps:

1.
Ok with external providers (OpenAI, Azure OpenAI, Vertex Palm) and they might use those providers to generate embeddings.
2.
You don't want to send your data external, the Huggingface option kind of helps with that use case.

Brandon J. We currently don't host our own LLMs. We might eventually. I feel like teams fall into 2 camps:

1.
Ok with external providers (OpenAI, Azure OpenAI, Vertex Palm) and they might use those providers to generate embeddings.
2.
You don't want to send your data external, the Huggingface option kind of helps with that use case.