Proxy routing support for embeddings (similar to OpenAILike for model routing)

[kirankc]

Allow mechanism to route embedding calls to a proxy like LiteLLM. Currently Phidata supports routing model calls using OpenAILike feature, and hoping that we can do the same thing for embeddings as well.

raise OpenAIError(
openai.OpenAIError: Missing credentials. Please pass one of api_key, azure_ad_token, azure_ad_token_provider, or the AZURE_OPENAI_API_KEY or AZURE_OPENAI_AD_TOKEN environment variables.