feat: Adding support for Ollama local api embeddings

[KrishnaM251]

Please describe the purpose of this pull request. See title.

How to test poetry run pytest -s tests/test_endpoints.py

Have you tested this PR? Yes, but it does not work yet

Related issues or PRs Addresses #1369

Is your PR over 500 lines of code? No.

Additional context Current issue: Unable to connect to http://localhost:11434 when running indicated test. This issue occured both when using LLamaIndex's interface for creating an Ollama EmbeddingModel object, and when creating a vanilla EmbeddingModel object without LLamaIndex. Do I need to have something running at http://localhost:11434 first?

IMPORTANT The test fails when run on a file I created called ollama.json. It is containted (locally) within configs/embedding_model_configs/, however this change did not get commited (and therefore pushed) for some reason.

Here is the contents of ollama.json

{
    "embedding_endpoint_type": "ollama",
    "embedding_endpoint": "http://localhost:11434",
    "embedding_model": "mixedbread-ai/mxbai-embed-large-v1",
    "embedding_dim": 512,
    "embedding_chunk_size": 200
}

[cpacker]

@sarahwooders looks like this is the same issue popping up again with tests failing on contrib PRs

[sarahwooders]

moved to and merged with #1433