Why `LangchainLLMWrapper` and `LangchainEmbeddingsWrapper` are Required for Ollama Models in RAGAS Metrics?

[tapasig]

While using Ollama model for RAGAs evaluation metrices the two wrapper functions are required: one for llm model and another for embedding model.
Because of the interface compatibility: the RAGAs metrices expect LLMs to implement the BaseRagasLLM, which Ollama models don't natively support. LLMs to implement the BaseRagasLLM interface, which Ollama models don't natively support. May be because of Ollama's unique response schema response.content vs OpenAI's choices.message.content - kindly verify

from ragas.llms import LangchainLLMWrapper
from ragas.embeddings import LangchainEmbeddingsWrapper

wrapped_llm = LangchainLLMWrapper(ollama_llm)
wrapped_embeddings = LangchainEmbeddingsWrapper(ollama_embeddings)

ANSWER_CORRECTNESS = ragas.metrics.AnswerCorrectness(name = "ANSWER_CORRECTNESS",
                                                     weights = [0.90, 0.10],
                                                     llm = wrapped_llm,
                                                     embeddings = wrapped_embeddings
                                                     )

[karthik18495]

May I ask if this has been integrated in the latest pull request #12 ?