Make cache store typed, and improve docs

python/packages/autogen-agentchat/tests/test_group_chat.py

@@ -34,10 +34,10 @@
 from autogen_agentchat.teams._group_chat._swarm_group_chat import SwarmGroupChatManager
 from autogen_agentchat.ui import Console
 from autogen_core import AgentId, CancellationToken
-from autogen_core.models import ReplayChatCompletionClient
 from autogen_core.tools import FunctionTool
 from autogen_ext.code_executors.local import LocalCommandLineCodeExecutor
 from autogen_ext.models.openai import OpenAIChatCompletionClient
+from autogen_ext.models.replay import ReplayChatCompletionClient
 from openai.resources.chat.completions import AsyncCompletions
 from openai.types.chat.chat_completion import ChatCompletion, Choice
 from openai.types.chat.chat_completion_chunk import ChatCompletionChunk

python/packages/autogen-agentchat/tests/test_magentic_one_group_chat.py

@@ -18,7 +18,7 @@
 )
 from autogen_agentchat.teams._group_chat._magentic_one._magentic_one_orchestrator import MagenticOneOrchestrator
 from autogen_core import AgentId, CancellationToken
-from autogen_core.models import ReplayChatCompletionClient
+from autogen_ext.models.replay import ReplayChatCompletionClient
 from utils import FileLogHandler
 
 logger = logging.getLogger(EVENT_LOGGER_NAME)

python/packages/autogen-core/docs/src/reference/index.md

@@ -48,12 +48,16 @@ python/autogen_ext.agents.video_surfer
 python/autogen_ext.agents.video_surfer.tools
 python/autogen_ext.auth.azure
 python/autogen_ext.teams.magentic_one
+python/autogen_ext.models.cache
 python/autogen_ext.models.openai
+python/autogen_ext.models.replay
 python/autogen_ext.tools.langchain
 python/autogen_ext.tools.graphrag
 python/autogen_ext.tools.code_execution
 python/autogen_ext.code_executors.local
 python/autogen_ext.code_executors.docker
 python/autogen_ext.code_executors.azure
+python/autogen_ext.cache_store.diskcache
+python/autogen_ext.cache_store.redis
 python/autogen_ext.runtimes.grpc
 ```

python/packages/autogen-core/docs/src/reference/python/autogen_ext.cache_store.diskcache.rst

@@ -0,0 +1,8 @@
+autogen\_ext.cache_store.diskcache
+==================================
+
+
+.. automodule:: autogen_ext.cache_store.diskcache
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/packages/autogen-core/docs/src/reference/python/autogen_ext.cache_store.redis.rst

@@ -0,0 +1,8 @@
+autogen\_ext.cache_store.redis
+==============================
+
+
+.. automodule:: autogen_ext.cache_store.redis
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/packages/autogen-core/docs/src/reference/python/autogen_ext.models.cache.rst

@@ -0,0 +1,8 @@
+autogen\_ext.models.cache
+=========================
+
+
+.. automodule:: autogen_ext.models.cache
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/packages/autogen-core/docs/src/reference/python/autogen_ext.models.replay.rst

@@ -0,0 +1,8 @@
+autogen\_ext.models.replay
+==========================
+
+
+.. automodule:: autogen_ext.models.replay
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/packages/autogen-core/docs/src/user-guide/agentchat-user-guide/tutorial/models.ipynb

@@ -8,7 +8,9 @@
                 "\n",
                 "In many cases, agents need access to LLM model services such as OpenAI, Azure OpenAI, or local models. Since there are many different providers with different APIs, `autogen-core` implements a protocol for [model clients](../../core-user-guide/framework/model-clients.ipynb) and `autogen-ext` implements a set of model clients for popular model services. AgentChat can use these model clients to interact with model services. \n",
                 "\n",
-                "**NOTE:** See {py:class}`~autogen_core.models.ChatCompletionCache` for a caching wrapper to use with the following clients."
+                "```{note}\n",
+                "See {py:class}`~autogen_ext.models.cache.ChatCompletionCache` for a caching wrapper to use with the following clients.\n",
+                "```"
             ]
         },
         {

python/packages/autogen-core/docs/src/user-guide/core-user-guide/framework/model-clients.ipynb

@@ -327,9 +327,20 @@
    "source": [
     "## Caching Wrapper\n",
     "\n",
-    "`autogen_core` implements a {py:class}`~autogen_core.models.ChatCompletionCache` that can wrap any {py:class}`~autogen_core.models.ChatCompletionClient`. Using this wrapper avoids incurring token usage when querying the underlying client with the same prompt multiple times. \n",
+    "`autogen_ext` implements {py:class}`~autogen_ext.models.cache.ChatCompletionCache` that can wrap any {py:class}`~autogen_core.models.ChatCompletionClient`. Using this wrapper avoids incurring token usage when querying the underlying client with the same prompt multiple times.\n",
     "\n",
-    "{py:class}`~autogen_core.models.ChatCompletionCache` uses a {py:class}`~autogen_core.CacheStore` protocol to allow duck-typing any storage object that has a pair of `get` & `set` methods (such as `redis.Redis` or `diskcache.Cache`). Here's an example of using `diskcache` for local caching:"
+    "{py:class}`~autogen_core.models.ChatCompletionCache` uses a {py:class}`~autogen_core.CacheStore` protocol. We have implemented some useful variants of {py:class}`~autogen_core.CacheStore` including {py:class}`~autogen_ext.cache_store.diskcache.DiskCacheStore` and {py:class}`~autogen_ext.cache_store.redis.RedisStore`.\n",
+    "\n",
+    "Here's an example of using `diskcache` for local caching:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# pip install -U \"autogen-ext[openai, diskcache]\""
    ]
   },
   {
@@ -346,18 +357,37 @@
     }
    ],
    "source": [
-    "from typing import Any, Dict, Optional\n",
+    "import asyncio\n",
+    "import tempfile\n",
     "\n",
-    "from autogen_core.models import ChatCompletionCache\n",
+    "from autogen_core.models import UserMessage\n",
+    "from autogen_ext.cache_store.diskcache import DiskCacheStore\n",
+    "from autogen_ext.models.cache import CHAT_CACHE_VALUE_TYPE, ChatCompletionCache\n",
+    "from autogen_ext.models.openai import OpenAIChatCompletionClient\n",
     "from diskcache import Cache\n",
     "\n",
-    "diskcache_client = Cache(\"/tmp/diskcache\")\n",
     "\n",
-    "cached_client = ChatCompletionCache(model_client, diskcache_client)\n",
-    "response = await cached_client.create(messages=messages)\n",
+    "async def main() -> None:\n",
+    "    with tempfile.TemporaryDirectory() as tmpdirname:\n",
+    "        # Initialize the original client\n",
+    "        openai_model_client = OpenAIChatCompletionClient(model=\"gpt-4o\")\n",
+    "\n",
+    "        # Then initialize the CacheStore, in this case with diskcache.Cache.\n",
+    "        # You can also use redis like:\n",
+    "        # from autogen_ext.cache_store.redis import RedisStore\n",
+    "        # import redis\n",
+    "        # redis_instance = redis.Redis()\n",
+    "        # cache_store = RedisCacheStore[CHAT_CACHE_VALUE_TYPE](redis_instance)\n",
+    "        cache_store = DiskCacheStore[CHAT_CACHE_VALUE_TYPE](Cache(tmpdirname))\n",
+    "        cache_client = ChatCompletionCache(openai_model_client, cache_store)\n",
+    "\n",
+    "        response = await cache_client.create([UserMessage(content=\"Hello, how are you?\", source=\"user\")])\n",
+    "        print(response)  # Should print response from OpenAI\n",
+    "        response = await cache_client.create([UserMessage(content=\"Hello, how are you?\", source=\"user\")])\n",
+    "        print(response)  # Should print cached response\n",
+    "\n",
     "\n",
-    "cached_response = await cached_client.create(messages=messages)\n",
-    "print(cached_response.cached)"
+    "asyncio.run(main())"
    ]
   },
   {

python/packages/autogen-core/src/autogen_core/__init__.py

@@ -10,7 +10,7 @@
 from ._agent_runtime import AgentRuntime
 from ._agent_type import AgentType
 from ._base_agent import BaseAgent
-from ._cache_store import CacheStore
+from ._cache_store import CacheStore, InMemoryStore
 from ._cancellation_token import CancellationToken
 from ._closure_agent import ClosureAgent, ClosureContext
 from ._component_config import (
@@ -87,6 +87,7 @@
     "AgentRuntime",
     "BaseAgent",
     "CacheStore",
+    "InMemoryStore",
     "CancellationToken",
     "AgentInstantiationContext",
     "TopicId",

python/packages/autogen-core/src/autogen_core/_cache_store.py

@@ -1,15 +1,16 @@
-from typing import Any, Optional, Protocol
+from typing import Any, Dict, Generic, Optional, Protocol, TypeVar, cast
 
+T = TypeVar("T")
 
-class CacheStore(Protocol):
+
+class CacheStore(Protocol, Generic[T]):
     """
     This protocol defines the basic interface for store/cache operations.
 
-    Allows duck-typing with any object that implements the get and set methods,
-    such as redis or diskcache interfaces.
+    Sub-classes should handle the lifecycle of underlying storage.
     """
 
-    def get(self, key: str, default: Optional[Any] = None) -> Optional[Any]:
+    def get(self, key: str, default: Optional[T] = None) -> Optional[T]:
         """
         Retrieve an item from the store.
 
@@ -23,7 +24,7 @@ def get(self, key: str, default: Optional[Any] = None) -> Optional[Any]:
         """
         ...
 
-    def set(self, key: str, value: Any) -> Optional[Any]:
+    def set(self, key: str, value: T) -> None:
         """
         Set an item in the store.
 
@@ -32,3 +33,14 @@ def set(self, key: str, value: Any) -> Optional[Any]:
             value: The value to be stored in the store.
         """
         ...
+
+
+class InMemoryStore(CacheStore[T]):
+    def __init__(self) -> None:
+        self.store: Dict[str, Any] = {}
+
+    def get(self, key: str, default: Optional[T] = None) -> Optional[T]:
+        return cast(Optional[T], self.store.get(key, default))
+
+    def set(self, key: str, value: T) -> None:
+        self.store[key] = value

python/packages/autogen-core/src/autogen_core/models/__init__.py

@@ -1,6 +1,4 @@
-from ._cache import ChatCompletionCache
 from ._model_client import ChatCompletionClient, ModelCapabilities, ModelFamily, ModelInfo  # type: ignore
-from ._replay_chat_completion_client import ReplayChatCompletionClient
 from ._types import (
     AssistantMessage,
     ChatCompletionTokenLogprob,
@@ -17,7 +15,6 @@
 
 __all__ = [
     "ModelCapabilities",
-    "ChatCompletionCache",
     "ChatCompletionClient",
     "SystemMessage",
     "UserMessage",
@@ -32,5 +29,4 @@
     "ChatCompletionTokenLogprob",
     "ModelFamily",
     "ModelInfo",
-    "ReplayChatCompletionClient",
 ]

python/packages/autogen-core/tests/test_cache_store.py

@@ -1,6 +1,6 @@
 from unittest.mock import Mock
 
-from autogen_core import CacheStore
+from autogen_core import CacheStore, InMemoryStore
 
 
 def test_set_and_get_object_key_value() -> None:
@@ -30,3 +30,19 @@ def test_set_overwrite_existing_key() -> None:
     mock_store.get.return_value = new_value
     mock_store.set.assert_called_with(key, new_value)
     assert mock_store.get(key) == new_value
+
+
+def test_inmemory_store() -> None:
+    store = InMemoryStore[int]()
+    test_key = "test_key"
+    test_value = 42
+    store.set(test_key, test_value)
+    assert store.get(test_key) == test_value
+
+    new_value = 2
+    store.set(test_key, new_value)
+    assert store.get(test_key) == new_value
+
+    key = "non_existent_key"
+    default_value = 99
+    assert store.get(key, default_value) == default_value

python/packages/autogen-ext/pyproject.toml

@@ -46,6 +46,12 @@ video-surfer = [
     "ffmpeg-python",
     "openai-whisper",
 ]
+diskcache = [
+    "diskcache>=5.6.3"
+]
+redis = [
+    "redis>=5.2.1"
+]
 
 grpc = [
     "grpcio~=1.62.0", # TODO: update this once we have a stable version.

python/packages/autogen-ext/src/autogen_ext/cache_store/diskcache.py

@@ -0,0 +1,26 @@
+from typing import Any, Optional, TypeVar, cast
+
+import diskcache
+from autogen_core import CacheStore
+
+T = TypeVar("T")
+
+
+class DiskCacheStore(CacheStore[T]):
+    """
+    A typed CacheStore implementation that uses diskcache as the underlying storage.
+    See :class:`~autogen_ext.models.cache.ChatCompletionCache` for an example of usage.
+
+    Args:
+        cache_instance: An instance of diskcache.Cache.
+                        The user is responsible for managing the DiskCache instance's lifetime.
+    """
+
+    def __init__(self, cache_instance: diskcache.Cache):  # type: ignore[no-any-unimported]
+        self.cache = cache_instance
+
+    def get(self, key: str, default: Optional[T] = None) -> Optional[T]:
+        return cast(Optional[T], self.cache.get(key, default))  # type: ignore[reportUnknownMemberType]
+
+    def set(self, key: str, value: T) -> None:
+        self.cache.set(key, cast(Any, value))  # type: ignore[reportUnknownMemberType]

python/packages/autogen-ext/src/autogen_ext/cache_store/redis.py

@@ -0,0 +1,29 @@
+from typing import Any, Optional, TypeVar, cast
+
+import redis
+from autogen_core import CacheStore
+
+T = TypeVar("T")
+
+
+class RedisStore(CacheStore[T]):
+    """
+    A typed CacheStore implementation that uses redis as the underlying storage.
+    See :class:`~autogen_ext.models.cache.ChatCompletionCache` for an example of usage.
+
+    Args:
+        cache_instance: An instance of `redis.Redis`.
+                        The user is responsible for managing the Redis instance's lifetime.
+    """
+
+    def __init__(self, redis_instance: redis.Redis):
+        self.cache = redis_instance
+
+    def get(self, key: str, default: Optional[T] = None) -> Optional[T]:
+        value = cast(Optional[T], self.cache.get(key))
+        if value is None:
+            return default
+        return value
+
+    def set(self, key: str, value: T) -> None:
+        self.cache.set(key, cast(Any, value))

python/packages/autogen-ext/src/autogen_ext/models/cache/__init__.py

@@ -0,0 +1,6 @@
+from ._chat_completion_cache import CHAT_CACHE_VALUE_TYPE, ChatCompletionCache
+
+__all__ = [
+    "CHAT_CACHE_VALUE_TYPE",
+    "ChatCompletionCache",
+]

python/packages/autogen-core/src/autogen_core/models/_cache.py → python/packages/autogen-ext/src/autogen_ext/models/cache/_chat_completion_cache.py

@@ -3,74 +3,80 @@
 import warnings
 from typing import Any, AsyncGenerator, List, Mapping, Optional, Sequence, Union, cast
 
-from .._cache_store import CacheStore
-from .._cancellation_token import CancellationToken
-from ..tools import Tool, ToolSchema
-from ._model_client import (
+from autogen_core import CacheStore, CancellationToken
+from autogen_core.models import (
     ChatCompletionClient,
-    ModelCapabilities,  # type: ignore
-    ModelInfo,
-)
-from ._types import (
     CreateResult,
     LLMMessage,
+    ModelCapabilities,  # type: ignore
+    ModelInfo,
     RequestUsage,
 )
+from autogen_core.tools import Tool, ToolSchema
+
+CHAT_CACHE_VALUE_TYPE = Union[CreateResult, List[Union[str, CreateResult]]]
 
 
 class ChatCompletionCache(ChatCompletionClient):
     """
-    A wrapper around a ChatCompletionClient that caches creation results from an underlying client.
+    A wrapper around a :class:`~autogen_ext.models.cache.ChatCompletionClient` that caches
+    creation results from an underlying client.
     Cache hits do not contribute to token usage of the original client.
 
     Typical Usage:
 
-        Lets use caching with `openai` as an example:
+    Lets use caching on disk with `openai` client as an example.
+    First install `autogen-ext` with the required packages:
 
-        .. code-block:: bash
+    .. code-block:: bash
 
-            pip install "autogen-ext[openai]==0.4.0.dev13"
+        pip install -U "autogen-ext[openai, diskcache]"
 
-        And use it as:
+    And use it as:
 
-        .. code-block:: python
+    .. code-block:: python
 
-            # Initialize the original client
-            from autogen_ext.models.openai import OpenAIChatCompletionClient
+        import asyncio
+        import tempfile
 
-            openai_client = OpenAIChatCompletionClient(
-                model="gpt-4o-2024-08-06",
-                # api_key="sk-...", # Optional if you have an OPENAI_API_KEY environment variable set.
-            )
+        from autogen_core.models import UserMessage
+        from autogen_ext.models.openai import OpenAIChatCompletionClient
+        from autogen_ext.models.cache import ChatCompletionCache, CHAT_CACHE_VALUE_TYPE
+        from autogen_ext.cache_store.diskcache import DiskCacheStore
+        from diskcache import Cache
 
-            # Then initialize the CacheStore. Either a Redis store:
-            import redis
 
-            redis_client = redis.Redis(host="localhost", port=6379, db=0)
+        async def main():
+            with tempfile.TemporaryDirectory() as tmpdirname:
+                # Initialize the original client
+                openai_model_client = OpenAIChatCompletionClient(model="gpt-4o")
 
-            # or diskcache:
-            from diskcache import Cache
+                # Then initialize the CacheStore, in this case with diskcache.Cache.
+                # You can also use redis like:
+                # from autogen_ext.cache_store.redis import RedisStore
+                # import redis
+                # redis_instance = redis.Redis()
+                # cache_store = RedisCacheStore[CHAT_CACHE_VALUE_TYPE](redis_instance)
+                cache_store = DiskCacheStore[CHAT_CACHE_VALUE_TYPE](Cache(tmpdirname))
+                cache_client = ChatCompletionCache(openai_model_client, cache_store)
 
-            diskcache_client = Cache("/tmp/diskcache")
+                response = await cache_client.create([UserMessage(content="Hello, how are you?", source="user")])
+                print(response)  # Should print response from OpenAI
+                response = await cache_client.create([UserMessage(content="Hello, how are you?", source="user")])
+                print(response)  # Should print cached response
 
-            # Then initialize the ChatCompletionCache with the store:
-            from autogen_core.models import ChatCompletionCache
 
-            # Cached client
-            cached_client = ChatCompletionCache(openai_client, diskcache_client)
+        asyncio.run(main())
 
-        You can now use the `cached_client` as you would the original client, but with caching enabled.
-    """
+    You can now use the `cached_client` as you would the original client, but with caching enabled.
 
-    def __init__(self, client: ChatCompletionClient, store: CacheStore):
-        """
-        Initialize a new ChatCompletionCache.
+    Args:
+        client (ChatCompletionClient): The original ChatCompletionClient to wrap.
+        store (CacheStore): A store object that implements get and set methods.
+            The user is responsible for managing the store's lifecycle & clearing it (if needed).
+    """
 
-        Args:
-            client (ChatCompletionClient): The original ChatCompletionClient to wrap.
-            store (CacheStore): A store object that implements get and set methods.
-                The user is responsible for managing the store's lifecycle & clearing it (if needed).
-        """
+    def __init__(self, client: ChatCompletionClient, store: CacheStore[CHAT_CACHE_VALUE_TYPE]):
         self.client = client
         self.store = store
 

python/packages/autogen-ext/src/autogen_ext/models/replay/__init__.py

@@ -0,0 +1,5 @@
+from ._replay_chat_completion_client import ReplayChatCompletionClient
+
+__all__ = [
+    "ReplayChatCompletionClient",
+]

python/packages/autogen-core/src/autogen_core/models/_replay_chat_completion_client.py → python/packages/autogen-ext/src/autogen_ext/models/replay/_replay_chat_completion_client.py

@@ -4,20 +4,17 @@
 import warnings
 from typing import Any, AsyncGenerator, List, Mapping, Optional, Sequence, Union
 
-from .. import EVENT_LOGGER_NAME
-from .._cancellation_token import CancellationToken
-from ..tools import Tool, ToolSchema
-from ._model_client import (
+from autogen_core import EVENT_LOGGER_NAME, CancellationToken
+from autogen_core.models import (
     ChatCompletionClient,
+    CreateResult,
+    LLMMessage,
     ModelCapabilities,  # type: ignore
     ModelFamily,
     ModelInfo,
-)
-from ._types import (
-    CreateResult,
-    LLMMessage,
     RequestUsage,
 )
+from autogen_core.tools import Tool, ToolSchema
 
 logger = logging.getLogger(EVENT_LOGGER_NAME)
 
@@ -43,7 +40,8 @@ class ReplayChatCompletionClient(ChatCompletionClient):
 
         .. code-block:: python
 
-            from autogen_core.models import ReplayChatCompletionClient, UserMessage
+            from autogen_core.models import UserMessage
+            from autogen_ext.models.replay import ReplayChatCompletionClient
 
 
             async def example():
@@ -62,7 +60,8 @@ async def example():
         .. code-block:: python
 
             import asyncio
-            from autogen_core.models import ReplayChatCompletionClient, UserMessage
+            from autogen_core.models import UserMessage
+            from autogen_ext.models.replay import ReplayChatCompletionClient
 
 
             async def example():
@@ -87,7 +86,8 @@ async def example():
         .. code-block:: python
 
             import asyncio
-            from autogen_core.models import ReplayChatCompletionClient, UserMessage
+            from autogen_core.models import UserMessage
+            from autogen_ext.models.replay import ReplayChatCompletionClient
 
 
             async def example():

python/packages/autogen-ext/tests/cache_store/test_diskcache_store.py

@@ -0,0 +1,48 @@
+import tempfile
+
+import pytest
+
+diskcache = pytest.importorskip("diskcache")
+
+
+def test_diskcache_store_basic() -> None:
+    from autogen_ext.cache_store.diskcache import DiskCacheStore
+    from diskcache import Cache
+
+    with tempfile.TemporaryDirectory() as temp_dir:
+        cache = Cache(temp_dir)
+        store = DiskCacheStore[int](cache)
+        test_key = "test_key"
+        test_value = 42
+        store.set(test_key, test_value)
+        assert store.get(test_key) == test_value
+
+        new_value = 2
+        store.set(test_key, new_value)
+        assert store.get(test_key) == new_value
+
+        key = "non_existent_key"
+        default_value = 99
+        assert store.get(key, default_value) == default_value
+
+
+def test_diskcache_with_different_instances() -> None:
+    from autogen_ext.cache_store.diskcache import DiskCacheStore
+    from diskcache import Cache
+
+    with tempfile.TemporaryDirectory() as temp_dir_1, tempfile.TemporaryDirectory() as temp_dir_2:
+        cache_1 = Cache(temp_dir_1)
+        cache_2 = Cache(temp_dir_2)
+
+        store_1 = DiskCacheStore[int](cache_1)
+        store_2 = DiskCacheStore[int](cache_2)
+
+        test_key = "test_key"
+        test_value_1 = 5
+        test_value_2 = 6
+
+        store_1.set(test_key, test_value_1)
+        assert store_1.get(test_key) == test_value_1
+
+        store_2.set(test_key, test_value_2)
+        assert store_2.get(test_key) == test_value_2

python/packages/autogen-ext/tests/cache_store/test_redis_store.py

@@ -0,0 +1,53 @@
+from unittest.mock import MagicMock
+
+import pytest
+
+redis = pytest.importorskip("redis")
+
+
+def test_redis_store_basic() -> None:
+    from autogen_ext.cache_store.redis import RedisStore
+
+    redis_instance = MagicMock()
+    store = RedisStore[int](redis_instance)
+    test_key = "test_key"
+    test_value = 42
+    store.set(test_key, test_value)
+    redis_instance.set.assert_called_with(test_key, test_value)
+    redis_instance.get.return_value = test_value
+    assert store.get(test_key) == test_value
+
+    new_value = 2
+    store.set(test_key, new_value)
+    redis_instance.set.assert_called_with(test_key, new_value)
+    redis_instance.get.return_value = new_value
+    assert store.get(test_key) == new_value
+
+    key = "non_existent_key"
+    default_value = 99
+    redis_instance.get.return_value = None
+    assert store.get(key, default_value) == default_value
+
+
+def test_redis_with_different_instances() -> None:
+    from autogen_ext.cache_store.redis import RedisStore
+
+    redis_instance_1 = MagicMock()
+    redis_instance_2 = MagicMock()
+
+    store_1 = RedisStore[int](redis_instance_1)
+    store_2 = RedisStore[int](redis_instance_2)
+
+    test_key = "test_key"
+    test_value_1 = 5
+    test_value_2 = 6
+
+    store_1.set(test_key, test_value_1)
+    redis_instance_1.set.assert_called_with(test_key, test_value_1)
+    redis_instance_1.get.return_value = test_value_1
+    assert store_1.get(test_key) == test_value_1
+
+    store_2.set(test_key, test_value_2)
+    redis_instance_2.set.assert_called_with(test_key, test_value_2)
+    redis_instance_2.get.return_value = test_value_2
+    assert store_2.get(test_key) == test_value_2

python/packages/autogen-core/tests/test_chat_completion_cache.py → python/packages/autogen-ext/tests/models/test_chat_completion_cache.py

@@ -1,29 +1,17 @@
 import copy
-from typing import Any, List, Optional, Tuple, Union
+from typing import List, Tuple, Union
 
 import pytest
-from autogen_core import CacheStore
+from autogen_core import InMemoryStore
 from autogen_core.models import (
-    ChatCompletionCache,
     ChatCompletionClient,
     CreateResult,
     LLMMessage,
-    ReplayChatCompletionClient,
     SystemMessage,
     UserMessage,
 )
-
-
-class DictStore(CacheStore):
-    def __init__(self) -> None:
-        self._store: dict[str, Any] = {}
-
-    def get(self, key: str, default: Optional[Any] = None) -> Optional[Any]:
-        return self._store.get(key, default)
-
-    def set(self, key: str, value: Any) -> Optional[Any]:
-        self._store[key] = value
-        return None
+from autogen_ext.models.cache import CHAT_CACHE_VALUE_TYPE, ChatCompletionCache
+from autogen_ext.models.replay import ReplayChatCompletionClient
 
 
 def get_test_data() -> Tuple[list[str], list[str], SystemMessage, ChatCompletionClient, ChatCompletionCache]:
@@ -33,7 +21,8 @@ def get_test_data() -> Tuple[list[str], list[str], SystemMessage, ChatCompletion
     system_prompt = SystemMessage(content="This is a system prompt")
     replay_client = ReplayChatCompletionClient(responses)
     replay_client.set_cached_bool_value(False)
-    cached_client = ChatCompletionCache(replay_client, store=DictStore())
+    store = InMemoryStore[CHAT_CACHE_VALUE_TYPE]()
+    cached_client = ChatCompletionCache(replay_client, store)
 
     return responses, prompts, system_prompt, replay_client, cached_client
 

python/packages/autogen-ext/tests/models/test_reply_chat_completion_client.py

@@ -12,13 +12,8 @@
     default_subscription,
     message_handler,
 )
-from autogen_core.models import (
-    ChatCompletionClient,
-    CreateResult,
-    ReplayChatCompletionClient,
-    SystemMessage,
-    UserMessage,
-)
+from autogen_core.models import ChatCompletionClient, CreateResult, SystemMessage, UserMessage
+from autogen_ext.models.replay import ReplayChatCompletionClient
 
 
 @dataclass