Python: Azure Model-as-a-Service Python connector (#6742)

### Motivation and Context

<!-- Thank you for your contribution to the semantic-kernel repo!
Please help reviewers and future users, providing the following
information:
  1. Why is this change required?
  2. What problem does it solve?
  3. What scenario does it contribute to?
  4. If it fixes an open issue, please link to the issue here.
-->
Related to: #6693

Azure Model-as-a-Service allows users to deploy certain models from the
Azure AI Studio model catalog as an API. This option also provides
pay-as-you-go access to the models hosted. Below are some of the models
that are supported:
- Microsoft Phi-3 family
- Meta Llama family (Llama 2 chat & Llama 3 instruct)
- Mistral-Small & Mistral-Large
- and more

We'd like to provide an AI connector for users of SK to use Azure
Model-as-a-Service.

### Description

<!-- Describe your changes, the overall approach, the underlying design.
These notes will help understanding how your code works. Thanks! -->
A new AI connector named `azure_ai_inference` is added to support Azure
Model-as-a-Service. This connector takes a new dependency on the Python
`azure.ai.inference` SDK.

### Contribution Checklist

<!-- Before submitting this PR, please make sure: -->

- [X] The code builds clean without any errors or warnings
- [X] The PR follows the [SK Contribution
Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md)
and the [pre-submission formatting
script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts)
raises no violations
- [X] All unit tests pass, and I have added new tests where possible
- [X] I didn't break anyone 😄

---------

Co-authored-by: Evan Mattson <[email protected]>

Author: TaoChenOSU

python/poetry.lock

@@ -48,6 +48,7 @@ weaviate-client = { version = ">=3.18,<5.0", optional = true}
 pinecone-client = { version = ">=3.0.0", optional = true}
 psycopg = { version="^3.1.9", extras=["binary","pool"], optional = true}
 redis = { version = "^4.6.0", optional = true}
+azure-ai-inference = {version = "^1.0.0b1", allow-prereleases = true, optional = true}
 azure-search-documents = {version = "11.6.0b4", allow-prereleases = true, optional = true}
 azure-core = { version = "^1.28.0", optional = true}
 azure-identity = { version = "^1.13.0", optional = true}
@@ -73,6 +74,7 @@ optional = true
 
 [tool.poetry.group.unit-tests.dependencies]
 google-generativeai = { version = ">=0.1,<0.4" }
+azure-ai-inference = {version = "^1.0.0b1", allow-prereleases = true}
 azure-search-documents = {version = "11.6.0b4", allow-prereleases = true}
 azure-core = "^1.28.0"
 azure-cosmos = "^4.7.0"
@@ -97,6 +99,7 @@ weaviate-client = ">=3.18,<5.0"
 pinecone-client = ">=3.0.0"
 psycopg = { version="^3.1.9", extras=["binary","pool"]}
 redis = "^4.6.0"
+azure-ai-inference = {version = "^1.0.0b1", allow-prereleases = true}
 azure-search-documents = {version = "11.6.0b4", allow-prereleases = true}
 azure-core = "^1.28.0"
 azure-identity = "^1.13.0"
@@ -116,10 +119,10 @@ weaviate = ["weaviate-client"]
 pinecone = ["pinecone-client"]
 postgres = ["psycopg"]
 redis = ["redis"]
-azure = ["azure-search-documents", "azure-core", "azure-identity", "azure-cosmos", "msgraph-sdk"]
+azure = ["azure-ai-inference", "azure-search-documents", "azure-core", "azure-identity", "azure-cosmos", "msgraph-sdk"]
 usearch = ["usearch", "pyarrow"]
 notebooks = ["ipykernel"]
-all = ["google-generativeai", "grpcio-status", "transformers", "sentence-transformers", "torch", "qdrant-client", "chromadb", "pymilvus", "milvus", "weaviate-client", "pinecone-client", "psycopg", "redis", "azure-search-documents", "azure-core", "azure-identity", "azure-cosmos", "usearch", "pyarrow", "ipykernel"]
+all = ["google-generativeai", "grpcio-status", "transformers", "sentence-transformers", "torch", "qdrant-client", "chromadb", "pymilvus", "milvus", "weaviate-client", "pinecone-client", "psycopg", "redis", "azure-ai-inference", "azure-search-documents", "azure-core", "azure-identity", "azure-cosmos", "usearch", "pyarrow", "ipykernel"]
 
 [tool.ruff]
 line-length = 120

python/pyproject.toml

@@ -0,0 +1,21 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+from semantic_kernel.connectors.ai.azure_ai_inference.azure_ai_inference_prompt_execution_settings import (
+    AzureAIInferenceChatPromptExecutionSettings,
+    AzureAIInferenceEmbeddingPromptExecutionSettings,
+)
+from semantic_kernel.connectors.ai.azure_ai_inference.azure_ai_inference_settings import AzureAIInferenceSettings
+from semantic_kernel.connectors.ai.azure_ai_inference.services.azure_ai_inference_chat_completion import (
+    AzureAIInferenceChatCompletion,
+)
+from semantic_kernel.connectors.ai.azure_ai_inference.services.azure_ai_inference_text_embedding import (
+    AzureAIInferenceTextEmbedding,
+)
+
+__all__ = [
+    "AzureAIInferenceChatCompletion",
+    "AzureAIInferenceChatPromptExecutionSettings",
+    "AzureAIInferenceEmbeddingPromptExecutionSettings",
+    "AzureAIInferenceSettings",
+    "AzureAIInferenceTextEmbedding",
+]

python/semantic_kernel/connectors/ai/azure_ai_inference/__init__.py

@@ -0,0 +1,45 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+from typing import Literal
+
+from pydantic import Field
+
+from semantic_kernel.connectors.ai.prompt_execution_settings import PromptExecutionSettings
+from semantic_kernel.utils.experimental_decorator import experimental_class
+
+
+@experimental_class
+class AzureAIInferencePromptExecutionSettings(PromptExecutionSettings):
+    """Azure AI Inference Prompt Execution Settings.
+
+    Note:
+        `extra_parameters` is a dictionary to pass additional model-specific parameters to the model.
+    """
+
+    frequency_penalty: float | None = Field(None, ge=-2, le=2)
+    max_tokens: int | None = Field(None, gt=0)
+    presence_penalty: float | None = Field(None, ge=-2, le=2)
+    seed: int | None = None
+    stop: str | None = None
+    temperature: float | None = Field(None, ge=0.0, le=1.0)
+    top_p: float | None = Field(None, ge=0.0, le=1.0)
+    extra_parameters: dict[str, str] | None = None
+
+
+@experimental_class
+class AzureAIInferenceChatPromptExecutionSettings(AzureAIInferencePromptExecutionSettings):
+    """Azure AI Inference Chat Prompt Execution Settings."""
+
+
+@experimental_class
+class AzureAIInferenceEmbeddingPromptExecutionSettings(PromptExecutionSettings):
+    """Azure AI Inference Embedding Prompt Execution Settings.
+
+    Note:
+        `extra_parameters` is a dictionary to pass additional model-specific parameters to the model.
+    """
+
+    dimensions: int | None = Field(None, gt=0)
+    encoding_format: Literal["base64", "binary", "float", "int8", "ubinary", "uint8"] | None = None
+    input_type: Literal["text", "query", "document"] | None = None
+    extra_parameters: dict[str, str] | None = None

python/semantic_kernel/connectors/ai/azure_ai_inference/azure_ai_inference_prompt_execution_settings.py

@@ -0,0 +1,37 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+from typing import ClassVar
+
+from pydantic import SecretStr
+
+from semantic_kernel.kernel_pydantic import HttpsUrl, KernelBaseSettings
+from semantic_kernel.utils.experimental_decorator import experimental_class
+
+
+@experimental_class
+class AzureAIInferenceSettings(KernelBaseSettings):
+    """Azure AI Inference settings.
+
+    The settings are first loaded from environment variables with
+    the prefix 'AZURE_AI_INFERENCE_'.
+    If the environment variables are not found, the settings can
+    be loaded from a .env file with the encoding 'utf-8'.
+    If the settings are not found in the .env file, the settings
+    are ignored; however, validation will fail alerting that the
+    settings are missing.
+
+    Required settings for prefix 'AZURE_AI_INFERENCE_' are:
+    - endpoint: HttpsUrl - The endpoint of the Azure AI Inference service deployment.
+                This value can be found in the Keys & Endpoint section when examining
+                your resource from the Azure portal.
+                (Env var AZURE_AI_INFERENCE_ENDPOINT)
+    - api_key: SecretStr - The API key for the Azure AI Inference service deployment.
+                This value can be found in the Keys & Endpoint section when examining
+                your resource from the Azure portal. You can use either KEY1 or KEY2.
+                (Env var AZURE_AI_INFERENCE_API_KEY)
+    """
+
+    env_prefix: ClassVar[str] = "AZURE_AI_INFERENCE_"
+
+    endpoint: HttpsUrl
+    api_key: SecretStr

python/semantic_kernel/connectors/ai/azure_ai_inference/azure_ai_inference_settings.py

@@ -0,0 +1 @@
+# Copyright (c) Microsoft. All rights reserved.

python/semantic_kernel/connectors/ai/azure_ai_inference/services/__init__.py

@@ -0,0 +1,22 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+import asyncio
+import contextlib
+from abc import ABC
+
+from azure.ai.inference.aio import ChatCompletionsClient, EmbeddingsClient
+
+from semantic_kernel.kernel_pydantic import KernelBaseModel
+from semantic_kernel.utils.experimental_decorator import experimental_class
+
+
+@experimental_class
+class AzureAIInferenceBase(KernelBaseModel, ABC):
+    """Azure AI Inference Chat Completion Service."""
+
+    client: ChatCompletionsClient | EmbeddingsClient
+
+    def __del__(self) -> None:
+        """Close the client when the object is deleted."""
+        with contextlib.suppress(Exception):
+            asyncio.get_running_loop().create_task(self.client.close())