http_client and http_async_client bugfix (#1454)

docs/api-reference/advanced/configuration.mdx

@@ -30,6 +30,7 @@ llm:
       response_format: 
         type: json_object
     api_version: 2024-02-01
+    http_client_proxies: http://testproxy.mem0.net:8000
     prompt: |
       Use the following pieces of context to answer the query at the end.
       If you don't know the answer, just say that you don't know, don't try to make up an answer.
@@ -89,7 +90,8 @@ cache:
       "system_prompt": "Act as William Shakespeare. Answer the following questions in the style of William Shakespeare.",
       "api_key": "sk-xxx",
       "model_kwargs": {"response_format": {"type": "json_object"}},
-      "api_version": "2024-02-01"
+      "api_version": "2024-02-01",
+      "http_client_proxies": "http://testproxy.mem0.net:8000",
     }
   },
   "vectordb": {
@@ -150,7 +152,8 @@ config = {
                 "Act as William Shakespeare. Answer the following questions in the style of William Shakespeare."
             ),
             'api_key': 'sk-xxx',
-            "model_kwargs": {"response_format": {"type": "json_object"}}
+            "model_kwargs": {"response_format": {"type": "json_object"}},
+            "http_client_proxies": "http://testproxy.mem0.net:8000",
         }
     },
     'vectordb': {
@@ -211,6 +214,8 @@ Alright, let's dive into what each key means in the yaml config above:
         - `number_documents` (Integer): Number of documents to pull from the vectordb as context, defaults to 1
         - `api_key` (String): The API key for the language model.
         - `model_kwargs` (Dict): Keyword arguments to pass to the language model. Used for `aws_bedrock` provider, since it requires different arguments for each model.
+        - `http_client_proxies` (Dict | String): The proxy server settings used to create `self.http_client` using `httpx.Client(proxies=http_client_proxies)`
+        - `http_async_client_proxies` (Dict | String): The proxy server settings for async calls used to create `self.http_async_client` using `httpx.AsyncClient(proxies=http_async_client_proxies)`
 3. `vectordb` Section:
     - `provider` (String): The provider for the vector database, set to 'chroma'. You can find the full list of vector database providers in [our docs](/components/vector-databases).
     - `config`:

embedchain/config/llm/base.py

@@ -1,7 +1,9 @@
 import logging
 import re
 from string import Template
-from typing import Any, Mapping, Optional
+from typing import Any, Mapping, Optional, Dict, Union
+
+import httpx
 
 from embedchain.config.base_config import BaseConfig
 from embedchain.helpers.json_serializable import register_deserializable
@@ -99,8 +101,8 @@ class BaseLlmConfig(BaseConfig):
         base_url: Optional[str] = None,
         endpoint: Optional[str] = None,
         model_kwargs: Optional[dict[str, Any]] = None,
-        http_client: Optional[Any] = None,
-        http_async_client: Optional[Any] = None,
+        http_client_proxies: Optional[Union[Dict, str]] = None,
+        http_async_client_proxies: Optional[Union[Dict, str]] = None,
         local: Optional[bool] = False,
         default_headers: Optional[Mapping[str, str]] = None,
         api_version: Optional[str] = None,
@@ -149,6 +151,11 @@ class BaseLlmConfig(BaseConfig):
         :type callbacks: Optional[list], optional
         :param query_type: The type of query to use, defaults to None
         :type query_type: Optional[str], optional
+        :param http_client_proxies: The proxy server settings used to create self.http_client, defaults to None
+        :type http_client_proxies: Optional[Dict | str], optional
+        :param http_async_client_proxies: The proxy server settings for async calls used to create
+        self.http_async_client, defaults to None
+        :type http_async_client_proxies: Optional[Dict | str], optional
         :param local: If True, the model will be run locally, defaults to False (for huggingface provider)
         :type local: Optional[bool], optional
         :param default_headers: Set additional HTTP headers to be sent with requests to OpenAI
@@ -181,8 +188,10 @@ class BaseLlmConfig(BaseConfig):
         self.base_url = base_url
         self.endpoint = endpoint
         self.model_kwargs = model_kwargs
-        self.http_client = http_client
-        self.http_async_client = http_async_client
+        self.http_client = httpx.Client(proxies=http_client_proxies) if http_client_proxies else None
+        self.http_async_client = (
+            httpx.AsyncClient(proxies=http_async_client_proxies) if http_async_client_proxies else None
+        )
         self.local = local
         self.default_headers = default_headers
         self.online = online

embedchain/llm/openai.py

@@ -56,7 +56,13 @@ class OpenAILlm(BaseLlm):
                 http_async_client=config.http_async_client,
             )
         else:
-            chat = ChatOpenAI(**kwargs, api_key=api_key, base_url=base_url)
+            chat = ChatOpenAI(
+                **kwargs,
+                api_key=api_key,
+                base_url=base_url,
+                http_client=config.http_client,
+                http_async_client=config.http_async_client,
+            )
         if self.tools:
             return self._query_function_call(chat, self.tools, messages)
 
@@ -69,8 +75,7 @@ class OpenAILlm(BaseLlm):
         messages: list[BaseMessage],
     ) -> str:
         from langchain.output_parsers.openai_tools import JsonOutputToolsParser
-        from langchain_core.utils.function_calling import \
-            convert_to_openai_tool
+        from langchain_core.utils.function_calling import convert_to_openai_tool
 
         openai_tools = [convert_to_openai_tool(tools)]
         chat = chat.bind(tools=openai_tools).pipe(JsonOutputToolsParser())

embedchain/utils/misc.py

@@ -442,6 +442,8 @@ def validate_config(config_data):
                     Optional("base_url"): str,
                     Optional("default_headers"): dict,
                     Optional("api_version"): Or(str, datetime.date),
+                    Optional("http_client_proxies"): Or(str, dict),
+                    Optional("http_async_client_proxies"): Or(str, dict),
                 },
             },
             Optional("vectordb"): {

tests/llm/test_openai.py

@@ -1,5 +1,6 @@
 import os
 
+import httpx
 import pytest
 from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
 
@@ -7,15 +8,27 @@ from embedchain.config import BaseLlmConfig
 from embedchain.llm.openai import OpenAILlm
 
 
-@pytest.fixture
-def config():
+@pytest.fixture()
+def env_config():
     os.environ["OPENAI_API_KEY"] = "test_api_key"
     os.environ["OPENAI_API_BASE"] = "https://api.openai.com/v1/engines/"
+    yield
+    os.environ.pop("OPENAI_API_KEY")
+
+
+@pytest.fixture
+def config(env_config):
     config = BaseLlmConfig(
-        temperature=0.7, max_tokens=50, top_p=0.8, stream=False, system_prompt="System prompt", model="gpt-3.5-turbo"
+        temperature=0.7,
+        max_tokens=50,
+        top_p=0.8,
+        stream=False,
+        system_prompt="System prompt",
+        model="gpt-3.5-turbo",
+        http_client_proxies=None,
+        http_async_client_proxies=None,
     )
     yield config
-    os.environ.pop("OPENAI_API_KEY")
 
 
 def test_get_llm_model_answer(config, mocker):
@@ -75,6 +88,8 @@ def test_get_llm_model_answer_without_system_prompt(config, mocker):
         model_kwargs={"top_p": config.top_p},
         api_key=os.environ["OPENAI_API_KEY"],
         base_url=os.environ["OPENAI_API_BASE"],
+        http_client=None,
+        http_async_client=None,
     )
 
 
@@ -93,6 +108,8 @@ def test_get_llm_model_answer_with_special_headers(config, mocker):
         api_key=os.environ["OPENAI_API_KEY"],
         base_url=os.environ["OPENAI_API_BASE"],
         default_headers={"test": "test"},
+        http_client=None,
+        http_async_client=None,
     )
 
 
@@ -110,6 +127,8 @@ def test_get_llm_model_answer_with_model_kwargs(config, mocker):
         model_kwargs={"top_p": config.top_p, "response_format": {"type": "json_object"}},
         api_key=os.environ["OPENAI_API_KEY"],
         base_url=os.environ["OPENAI_API_BASE"],
+        http_client=None,
+        http_async_client=None,
     )
 
 
@@ -136,8 +155,78 @@ def test_get_llm_model_answer_with_tools(config, mocker, mock_return, expected):
         model_kwargs={"top_p": config.top_p},
         api_key=os.environ["OPENAI_API_KEY"],
         base_url=os.environ["OPENAI_API_BASE"],
+        http_client=None,
+        http_async_client=None,
     )
     mocked_convert_to_openai_tool.assert_called_once_with({"test": "test"})
     mocked_json_output_tools_parser.assert_called_once()
 
     assert answer == expected
+
+
+def test_get_llm_model_answer_with_http_client_proxies(env_config, mocker):
+    mocked_openai_chat = mocker.patch("embedchain.llm.openai.ChatOpenAI")
+    mock_http_client = mocker.Mock(spec=httpx.Client)
+    mock_http_client_instance = mocker.Mock(spec=httpx.Client)
+    mock_http_client.return_value = mock_http_client_instance
+
+    mocker.patch("httpx.Client", new=mock_http_client)
+
+    config = BaseLlmConfig(
+        temperature=0.7,
+        max_tokens=50,
+        top_p=0.8,
+        stream=False,
+        system_prompt="System prompt",
+        model="gpt-3.5-turbo",
+        http_client_proxies="http://testproxy.mem0.net:8000",
+    )
+
+    llm = OpenAILlm(config)
+    llm.get_llm_model_answer("Test query")
+
+    mocked_openai_chat.assert_called_once_with(
+        model=config.model,
+        temperature=config.temperature,
+        max_tokens=config.max_tokens,
+        model_kwargs={"top_p": config.top_p},
+        api_key=os.environ["OPENAI_API_KEY"],
+        base_url=os.environ["OPENAI_API_BASE"],
+        http_client=mock_http_client_instance,
+        http_async_client=None,
+    )
+    mock_http_client.assert_called_once_with(proxies="http://testproxy.mem0.net:8000")
+
+
+def test_get_llm_model_answer_with_http_async_client_proxies(env_config, mocker):
+    mocked_openai_chat = mocker.patch("embedchain.llm.openai.ChatOpenAI")
+    mock_http_async_client = mocker.Mock(spec=httpx.AsyncClient)
+    mock_http_async_client_instance = mocker.Mock(spec=httpx.AsyncClient)
+    mock_http_async_client.return_value = mock_http_async_client_instance
+
+    mocker.patch("httpx.AsyncClient", new=mock_http_async_client)
+
+    config = BaseLlmConfig(
+        temperature=0.7,
+        max_tokens=50,
+        top_p=0.8,
+        stream=False,
+        system_prompt="System prompt",
+        model="gpt-3.5-turbo",
+        http_async_client_proxies={"http://": "http://testproxy.mem0.net:8000"},
+    )
+
+    llm = OpenAILlm(config)
+    llm.get_llm_model_answer("Test query")
+
+    mocked_openai_chat.assert_called_once_with(
+        model=config.model,
+        temperature=config.temperature,
+        max_tokens=config.max_tokens,
+        model_kwargs={"top_p": config.top_p},
+        api_key=os.environ["OPENAI_API_KEY"],
+        base_url=os.environ["OPENAI_API_BASE"],
+        http_client=None,
+        http_async_client=mock_http_async_client_instance,
+    )
+    mock_http_async_client.assert_called_once_with(proxies={"http://": "http://testproxy.mem0.net:8000"})