feat: Add Azure embedders support

[vblagoje]

Why:

Add new embedding components that leverage Azure's cognitive services for text and document embedding. These new components are intended to provide users with more embedding options and utilize Azure's models.

What:

The main changes introduced in the azure_embedders branch include:

1. Addition of Azure Embedders: The AzureOpenAITextEmbedder and AzureOpenAIDocumentEmbedder classes have been added. These are similar to their OpenAI counterparts but use Azure API for embedding texts and documents.
2. Modifications in __init__.py: The __init__.py file within the embedders directory has been modified to include the newly added Azure embedder classes.
3. Unit Tests: New unit tests have been added for both the AzureOpenAITextEmbedder and AzureOpenAIDocumentEmbedder to ensure they work as expected and handle various inputs correctly.
4. Release Notes: A release note has been added to document the inclusion of new Azure embedders.

How can it be used:

The new Azure embedders can be used in the Haystack framework for tasks that involve text and document embedding. Users can utilize these embedders to convert texts or documents into vectors using Azure's cognitive services. This can be particularly useful for tasks involving semantic search, text clustering, or any application that requires understanding the semantic content of texts.

How did you test it:

1. Unit Tests: New tests were added to verify the functionality of the Azure embedders. These tests check if the embedders correctly initialize with default and custom parameters and if they return expected embeddings for given texts and documents.
2. Integration Testing: The embedders were tested in an integrated environment to ensure they interact correctly with the Azure API and the rest of the haystack components. This includes testing with actual text inputs and verifying the embeddings' quality and format.

Notes for the reviewer:

• Review the implementation of the Azure embedders to ensure they align with the project's standards and efficiently utilize Azure's cognitive services.
• Consider the unit tests' coverage and whether any additional scenarios need to be tested.
• As the embedders rely on external services (Azure), consider the latency, cost, and availability implications when using these components in production environments.

[vblagoje]

Reviewers @anakin87 @dfokina - still waiting for our DevOps for the actual Azure model deployments. Will take another day or two. Please don't start the review process yet.

[vblagoje]

@anakin87 thanks to @steppi91 , we have tested the live azure embedding models in the last c744be1 commit. This should be ready to go.

[vblagoje]

All yours for tomorrow @anakin87 🚀

[anakin87]

Good PR!
I found some little opportunities for improvement.

[anakin87]

Please make the tests pass with the suggested changes,
then this PR is good to go!!!