Python : #6499 Initial Commit for Mistral Connector

.github/workflows/python-integration-tests.yml

@@ -96,6 +96,8 @@ jobs:
           AZURE_KEY_VAULT_CLIENT_ID: ${{secrets.AZURE_KEY_VAULT_CLIENT_ID}}
           AZURE_KEY_VAULT_CLIENT_SECRET: ${{secrets.AZURE_KEY_VAULT_CLIENT_SECRET}}
           ACA_POOL_MANAGEMENT_ENDPOINT: ${{secrets.ACA_POOL_MANAGEMENT_ENDPOINT}}
+          MISTRALAI_API_KEY: ${{secrets.MISTRALAI_API_KEY}}
+          MISTRALAI_CHAT_MODEL_ID: ${{ vars.MISTRALAI_CHAT_MODEL_ID }}
         run: |
           if ${{ matrix.os == 'ubuntu-latest' }}; then
             docker run -d --name redis-stack-server -p 6379:6379 redis/redis-stack-server:latest
@@ -163,6 +165,8 @@ jobs:
           AZURE_KEY_VAULT_CLIENT_ID: ${{secrets.AZURE_KEY_VAULT_CLIENT_ID}}
           AZURE_KEY_VAULT_CLIENT_SECRET: ${{secrets.AZURE_KEY_VAULT_CLIENT_SECRET}}
           ACA_POOL_MANAGEMENT_ENDPOINT: ${{secrets.ACA_POOL_MANAGEMENT_ENDPOINT}}
+          MISTRALAI_API_KEY: ${{secrets.MISTRALAI_API_KEY}}
+          MISTRALAI_CHAT_MODEL_ID: ${{ vars.MISTRALAI_CHAT_MODEL_ID }}
         run: |
           if ${{ matrix.os == 'ubuntu-latest' }}; then
             docker run -d --name redis-stack-server -p 6379:6379 redis/redis-stack-server:latest

python/pyproject.toml

@@ -66,8 +66,9 @@ redis = { version = "^4.6.0", optional = true}
 usearch = { version = "^2.9", optional = true}
 pyarrow = { version = ">=12.0.1,<17.0.0", optional = true}
 weaviate-client = { version = ">=3.18,<5.0", optional = true}
+# mistralai
+mistralai = { version = "^0.4.1"}
 
-# Groups are for development only (installed through Poetry)
 [tool.poetry.group.dev.dependencies]
 pre-commit = ">=3.7.1"
 ruff = ">=0.4.5"
@@ -128,16 +129,19 @@ usearch = "^2.9"
 pyarrow = ">=12.0.1,<17.0.0"
 # weaviate
 weaviate-client = ">=3.18,<5.0"
+# mistralai
+mistralai = "^0.4.1"
 
 # Extras are exposed to pip, this allows a user to easily add the right dependencies to their environment
 [tool.poetry.extras]
-all = ["google-generativeai", "transformers", "sentence-transformers", "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", "motor"]
+all = ["google-generativeai", "transformers", "sentence-transformers", "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", "motor", "mistralai"]
 
 azure = ["azure-ai-inference", "azure-search-documents", "azure-core", "azure-identity", "azure-cosmos", "msgraph-sdk"]
 chromadb = ["chromadb"]
 google = ["google-generativeai"]
 hugging_face = ["transformers", "sentence-transformers"]
 milvus = ["pymilvus", "milvus"]
+mistralai = ["mistralai"]
 mongo = ["motor"]
 notebooks = ["ipykernel"]
 pinecone = ["pinecone-client"]

python/samples/concepts/auto_function_calling/chat_mistral_function_calling.py

@@ -0,0 +1,174 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+import asyncio
+import os
+from functools import reduce
+from typing import TYPE_CHECKING
+
+from semantic_kernel import Kernel
+from semantic_kernel.connectors.ai.function_choice_behavior import FunctionChoiceBehavior
+from semantic_kernel.connectors.ai.mistral_ai.services.mistral_ai_chat_completion import (
+    MistralAIChatCompletion,
+    MistralAIChatPromptExecutionSettings,
+)
+from semantic_kernel.contents import ChatHistory
+from semantic_kernel.contents.chat_message_content import ChatMessageContent
+from semantic_kernel.contents.function_call_content import FunctionCallContent
+from semantic_kernel.contents.streaming_chat_message_content import StreamingChatMessageContent
+from semantic_kernel.core_plugins import MathPlugin, TimePlugin
+from semantic_kernel.functions import KernelArguments
+
+if TYPE_CHECKING:
+    from semantic_kernel.functions import KernelFunction
+
+
+system_message = """
+You are a chat bot. Your name is Mosscap and
+you have one goal: figure out what people need.
+Your full name, should you need to know it, is
+Splendid Speckled Mosscap. You communicate
+effectively, but you tend to answer with long
+flowery prose. You are also a math wizard,
+especially for adding and subtracting.
+You also excel at joke telling, where your tone is often sarcastic.
+Once you have the answer I am looking for,
+you will return a full answer to me as soon as possible.
+"""
+
+kernel = Kernel()
+
+kernel.add_service(MistralAIChatCompletion(service_id="chat", ai_model_id="mistral-large-latest"))
+
+plugins_directory = os.path.join(__file__, "../../../../../prompt_template_samples/")
+# adding plugins to the kernel
+# the joke plugin in the FunPlugins is a semantic plugin and has the function calling disabled.
+# kernel.import_plugin_from_prompt_directory("chat", plugins_directory, "FunPlugin")
+# the math plugin is a core plugin and has the function calling enabled.
+kernel.add_plugin(MathPlugin(), plugin_name="math")
+kernel.add_plugin(TimePlugin(), plugin_name="time")
+
+chat_function = kernel.add_function(
+    prompt="{{$chat_history}}{{$user_input}}",
+    plugin_name="ChatBot",
+    function_name="Chat",
+)
+# enabling or disabling function calling is done by setting the function_call parameter for the completion.
+# when the function_call parameter is set to "auto" the model will decide which function to use, if any.
+# if you only want to use a specific function, set the name of that function in this parameter,
+# the format for that is 'PluginName-FunctionName', (i.e. 'math-Add').
+# if the model or api version does not support this you will get an error.
+
+# Note: the number of responses for auto invoking tool calls is limited to 1.
+# If configured to be greater than one, this value will be overridden to 1.
+execution_settings = MistralAIChatPromptExecutionSettings(
+    service_id="chat",
+    max_tokens=2000,
+    temperature=0.7,
+    top_p=0.8,
+    function_choice_behavior=FunctionChoiceBehavior.Auto()
+)
+
+history = ChatHistory()
+
+history.add_system_message(system_message)
+history.add_user_message("Hi there, who are you?")
+history.add_assistant_message("I am Mosscap, a chat bot. I'm trying to figure out what people need.")
+
+arguments = KernelArguments(settings=execution_settings)
+
+
+def print_tool_calls(message: ChatMessageContent) -> None:
+    # A helper method to pretty print the tool calls from the message.
+    # This is only triggered if auto invoke tool calls is disabled.
+    items = message.items
+    formatted_tool_calls = []
+    for i, item in enumerate(items, start=1):
+        if isinstance(item, FunctionCallContent):
+            tool_call_id = item.id
+            function_name = item.name
+            function_arguments = item.arguments
+            formatted_str = (
+                f"tool_call {i} id: {tool_call_id}\n"
+                f"tool_call {i} function name: {function_name}\n"
+                f"tool_call {i} arguments: {function_arguments}"
+            )
+            formatted_tool_calls.append(formatted_str)
+    print("Tool calls:\n" + "\n\n".join(formatted_tool_calls))
+
+
+async def handle_streaming(
+    kernel: Kernel,
+    chat_function: "KernelFunction",
+    arguments: KernelArguments,
+) -> None:
+    response = kernel.invoke_stream(
+        chat_function,
+        return_function_results=False,
+        arguments=arguments,
+    )
+
+    print("Mosscap:> ", end="")
+    streamed_chunks: list[StreamingChatMessageContent] = []
+    async for message in response:
+        if not execution_settings.function_choice_behavior.auto_invoke_kernel_functions and isinstance(
+            message[0], StreamingChatMessageContent
+        ):
+            streamed_chunks.append(message[0])
+        else:
+            print(str(message[0]), end="")
+
+    if streamed_chunks:
+        streaming_chat_message = reduce(lambda first, second: first + second, streamed_chunks)
+        print("Auto tool calls is disabled, printing returned tool calls...")
+        print_tool_calls(streaming_chat_message)
+
+    print("\n")
+
+
+async def chat() -> bool:
+    try:
+        user_input = input("User:> ")
+    except KeyboardInterrupt:
+        print("\n\nExiting chat...")
+        return False
+    except EOFError:
+        print("\n\nExiting chat...")
+        return False
+
+    if user_input == "exit":
+        print("\n\nExiting chat...")
+        return False
+    arguments["user_input"] = user_input
+    arguments["chat_history"] = history
+
+    stream = True
+    if stream:
+        await handle_streaming(kernel, chat_function, arguments=arguments)
+    else:
+        result = await kernel.invoke(chat_function, arguments=arguments)
+
+        # If tools are used, and auto invoke tool calls is False, the response will be of type
+        # ChatMessageContent with information about the tool calls, which need to be sent
+        # back to the model to get the final response.
+        function_calls = [item for item in result.value[-1].items if isinstance(item, FunctionCallContent)]
+        if not execution_settings.function_choice_behavior.auto_invoke_kernel_functions and len(function_calls) > 0:
+            print_tool_calls(result.value[0])
+            return True
+
+        print(f"Mosscap:> {result}")
+    return True
+
+
+async def main() -> None:
+    chatting = True
+    print(
+        "Welcome to the chat bot!\
+        \n  Type 'exit' to exit.\
+        \n  Try a math question to see the function calling in action (i.e. what is 3+3?)."
+    )
+    while chatting:
+        chatting = await chat()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

python/samples/concepts/chat_completion/chat_mistral_api.py

@@ -0,0 +1,71 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+import asyncio
+
+from semantic_kernel import Kernel
+from semantic_kernel.connectors.ai.mistral_ai import MistralAIChatCompletion
+from semantic_kernel.contents import ChatHistory
+from semantic_kernel.functions import KernelArguments
+
+system_message = """
+You are a chat bot. Your name is Mosscap and
+you have one goal: figure out what people need.
+Your full name, should you need to know it, is
+Splendid Speckled Mosscap. You communicate
+effectively, but you tend to answer with long
+flowery prose.
+"""
+
+kernel = Kernel()
+
+service_id = "mistral-ai-chat"
+kernel.add_service(MistralAIChatCompletion(service_id=service_id, ai_model_id="mistral-small-latest"))
+
+settings = kernel.get_prompt_execution_settings_from_service_id(service_id)
+settings.max_tokens = 2000
+settings.temperature = 0.7
+settings.top_p = 0.8
+
+chat_function = kernel.add_function(
+    plugin_name="ChatBot",
+    function_name="Chat",
+    prompt="{{$chat_history}}{{$user_input}}",
+    template_format="semantic-kernel",
+    prompt_execution_settings=settings,
+)
+
+chat_history = ChatHistory(system_message=system_message)
+chat_history.add_user_message("Hi there, who are you?")
+chat_history.add_assistant_message("I am Mosscap, a chat bot. I'm trying to figure out what people need")
+chat_history.add_user_message("I want to find a hotel in Seattle with free wifi and a pool.")
+
+
+async def chat() -> bool:
+    try:
+        user_input = input("User:> ")
+    except KeyboardInterrupt:
+        print("\n\nExiting chat...")
+        return False
+    except EOFError:
+        print("\n\nExiting chat...")
+        return False
+
+    if user_input == "exit":
+        print("\n\nExiting chat...")
+        return False
+
+    answer = await kernel.invoke(chat_function, KernelArguments(user_input=user_input, chat_history=chat_history))
+    chat_history.add_user_message(user_input)
+    chat_history.add_assistant_message(str(answer))
+    print(f"Mosscap:> {answer}")
+    return True
+
+
+async def main() -> None:
+    chatting = True
+    while chatting:
+        chatting = await chat()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())