not-again-ai 0.14.0__tar.gz → 0.16.0__tar.gz

{not_again_ai-0.14.0 → not_again_ai-0.16.0}/PKG-INFO

@@ -1,12 +1,11 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.3
 Name: not-again-ai
-Version: 0.14.0
+Version: 0.16.0
 Summary: Designed to once and for all collect all the little things that come up over and over again in AI projects and put them in one place.
-Home-page: https://github.com/DaveCoDev/not-again-ai
 License: MIT
 Author: DaveCoDev
 Author-email: dave.co.dev@gmail.com
-Requires-Python: >=3.11,<4.0
+Requires-Python: >=3.11, <3.13
 Classifier: Development Status :: 3 - Alpha
 Classifier: Intended Audience :: Developers
 Classifier: Intended Audience :: Science/Research
@@ -19,26 +18,25 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Typing :: Typed
 Provides-Extra: data
 Provides-Extra: llm
-Provides-Extra: local-llm
 Provides-Extra: statistics
 Provides-Extra: viz
-Requires-Dist: azure-ai-inference (==1.0.0b5) ; extra == "llm"
-Requires-Dist: azure-identity (>=1.19,<2.0) ; extra == "llm"
-Requires-Dist: jinja2 (>=3.1,<4.0) ; extra == "local-llm"
-Requires-Dist: loguru (>=0.7,<0.8)
-Requires-Dist: numpy (>=2.1,<3.0) ; extra == "statistics" or extra == "viz"
-Requires-Dist: ollama (>=0.3,<0.4) ; extra == "local-llm"
-Requires-Dist: openai (>=1.52,<2.0) ; extra == "llm"
-Requires-Dist: pandas (>=2.2,<3.0) ; extra == "viz"
-Requires-Dist: pydantic (>=2.9,<3.0)
-Requires-Dist: pytest-playwright (>=0.5,<0.6) ; extra == "data"
-Requires-Dist: python-liquid (>=1.12,<2.0) ; extra == "llm"
-Requires-Dist: scikit-learn (>=1.5,<2.0) ; extra == "statistics"
-Requires-Dist: scipy (>=1.14,<2.0) ; extra == "statistics"
-Requires-Dist: seaborn (>=0.13,<0.14) ; extra == "viz"
-Requires-Dist: tiktoken (>=0.8,<0.9) ; extra == "llm"
-Requires-Dist: transformers (>=4.45,<5.0) ; extra == "local-llm"
-Project-URL: Documentation, https://github.com/DaveCoDev/not-again-ai
+Requires-Dist: azure-identity (>=1.19) ; extra == "llm"
+Requires-Dist: loguru (>=0.7)
+Requires-Dist: numpy (>=2.2) ; extra == "statistics"
+Requires-Dist: numpy (>=2.2) ; extra == "viz"
+Requires-Dist: ollama (>=0.4) ; extra == "llm"
+Requires-Dist: openai (>=1) ; extra == "llm"
+Requires-Dist: pandas (>=2.2) ; extra == "viz"
+Requires-Dist: playwright (>=1.49) ; extra == "data"
+Requires-Dist: pydantic (>=2.10)
+Requires-Dist: pytest-playwright (>=0.7) ; extra == "data"
+Requires-Dist: python-liquid (>=1.12) ; extra == "llm"
+Requires-Dist: scikit-learn (>=1.6) ; extra == "statistics"
+Requires-Dist: scipy (>=1.15) ; extra == "statistics"
+Requires-Dist: seaborn (>=0.13) ; extra == "viz"
+Requires-Dist: tiktoken (>=0.8) ; extra == "llm"
+Project-URL: Documentation, https://davecodev.github.io/not-again-ai/
+Project-URL: Homepage, https://github.com/DaveCoDev/not-again-ai
 Project-URL: Repository, https://github.com/DaveCoDev/not-again-ai
 Description-Content-Type: text/markdown
 
@@ -68,11 +66,9 @@ Requires: Python 3.11, or 3.12
 Install the entire package from [PyPI](https://pypi.org/project/not-again-ai/) with: 
 
 ```bash
-$ pip install not_again_ai[llm,local_llm,statistics,viz]
+$ pip install not_again_ai[data,llm,statistics,viz]
 ```
 
-Note that local LLM requires separate installations and will not work out of the box due to how hardware dependent it is. Be sure to check the [notebooks](notebooks/local_llm/) for more details.
-
 The package is split into subpackages, so you can install only the parts you need.
 
 ### Base
@@ -93,16 +89,7 @@ The package is split into subpackages, so you can install only the parts you nee
    1. Using AOAI requires using Entra ID authentication. See https://learn.microsoft.com/en-us/azure/ai-services/openai/how-to/managed-identity for how to set this up for your AOAI deployment.
       * Requires the correct role assigned to your user account and being signed into the Azure CLI.
    1. (Optional) Set the `AZURE_OPENAI_ENDPOINT` environment variable.
-1. Setup GitHub Models
-   1. Get a Personal Access Token from https://github.com/settings/tokens and set the `GITHUB_TOKEN` environment variable. The token does not need any permissions.
-   1. Check the [Github Marketplace](https://github.com/marketplace/models) to see which models are available.
-
-
-### Local LLM
- 1. `pip install not_again_ai[llm,local_llm]`
- 1. Some HuggingFace transformers tokenizers are gated behind access requests. If you wish to use these, you will need to request access from HuggingFace on the model card.
-    * Then set the `HF_TOKEN` environment variable to your HuggingFace API token which can be found here: https://huggingface.co/settings/tokens
- 1. If you wish to use Ollama:
+1. If you wish to use Ollama:
      1. Follow the instructions at https://github.com/ollama/ollama to install Ollama for your system. 
      1. (Optional) [Add Ollama as a startup service (recommended)](https://github.com/ollama/ollama/blob/main/docs/linux.md#adding-ollama-as-a-startup-service-recommended)
      1. (Optional) To make the Ollama service accessible on your local network from a Linux server, add the following to the `/etc/systemd/system/ollama.service` file which will make Ollama available at `http://<local_address>:11434`:
@@ -112,7 +99,6 @@ The package is split into subpackages, so you can install only the parts you nee
          Environment="OLLAMA_HOST=0.0.0.0"
          ```
      1. It is recommended to always have the latest version of Ollama. To update Ollama check the [docs](https://github.com/ollama/ollama/blob/main/docs/). The command for Linux is: `curl -fsSL https://ollama.com/install.sh | sh`
- 1. HuggingFace transformers and other requirements are hardware dependent so for providers other than Ollama, this only installs some generic dependencies. Check the [notebooks](notebooks/local_llm/) for more details on what is available and how to install it.
 
 
 ### Statistics
@@ -156,10 +142,8 @@ $ poetry update
 
 To install all dependencies (with all extra dependencies) into an isolated virtual environment:
 
-> Append `--sync` to uninstall dependencies that are no longer in use from the virtual environment.
-
 ```bash
-$ poetry install --all-extras
+$ poetry sync --all-extras
 ```
 
 To [activate](https://python-poetry.org/docs/basic-usage#activating-the-virtual-environment) the
@@ -215,7 +199,7 @@ Automated code quality checks are performed using
 environments and run commands based on [`noxfile.py`](./noxfile.py) for unit testing, PEP 8 style
 guide checking, type checking and documentation generation.
 
-> Note: `nox` is installed into the virtual environment automatically by the `poetry install`
+> Note: `nox` is installed into the virtual environment automatically by the `poetry sync`
 > command above. Run `poetry shell` to activate the virtual environment.
 
 To run all default sessions:

{not_again_ai-0.14.0 → not_again_ai-0.16.0}/README.md

@@ -24,11 +24,9 @@ Requires: Python 3.11, or 3.12
 Install the entire package from [PyPI](https://pypi.org/project/not-again-ai/) with: 
 
 ```bash
-$ pip install not_again_ai[llm,local_llm,statistics,viz]
+$ pip install not_again_ai[data,llm,statistics,viz]
 ```
 
-Note that local LLM requires separate installations and will not work out of the box due to how hardware dependent it is. Be sure to check the [notebooks](notebooks/local_llm/) for more details.
-
 The package is split into subpackages, so you can install only the parts you need.
 
 ### Base
@@ -49,16 +47,7 @@ The package is split into subpackages, so you can install only the parts you nee
    1. Using AOAI requires using Entra ID authentication. See https://learn.microsoft.com/en-us/azure/ai-services/openai/how-to/managed-identity for how to set this up for your AOAI deployment.
       * Requires the correct role assigned to your user account and being signed into the Azure CLI.
    1. (Optional) Set the `AZURE_OPENAI_ENDPOINT` environment variable.
-1. Setup GitHub Models
-   1. Get a Personal Access Token from https://github.com/settings/tokens and set the `GITHUB_TOKEN` environment variable. The token does not need any permissions.
-   1. Check the [Github Marketplace](https://github.com/marketplace/models) to see which models are available.
-
-
-### Local LLM
- 1. `pip install not_again_ai[llm,local_llm]`
- 1. Some HuggingFace transformers tokenizers are gated behind access requests. If you wish to use these, you will need to request access from HuggingFace on the model card.
-    * Then set the `HF_TOKEN` environment variable to your HuggingFace API token which can be found here: https://huggingface.co/settings/tokens
- 1. If you wish to use Ollama:
+1. If you wish to use Ollama:
      1. Follow the instructions at https://github.com/ollama/ollama to install Ollama for your system. 
      1. (Optional) [Add Ollama as a startup service (recommended)](https://github.com/ollama/ollama/blob/main/docs/linux.md#adding-ollama-as-a-startup-service-recommended)
      1. (Optional) To make the Ollama service accessible on your local network from a Linux server, add the following to the `/etc/systemd/system/ollama.service` file which will make Ollama available at `http://<local_address>:11434`:
@@ -68,7 +57,6 @@ The package is split into subpackages, so you can install only the parts you nee
          Environment="OLLAMA_HOST=0.0.0.0"
          ```
      1. It is recommended to always have the latest version of Ollama. To update Ollama check the [docs](https://github.com/ollama/ollama/blob/main/docs/). The command for Linux is: `curl -fsSL https://ollama.com/install.sh | sh`
- 1. HuggingFace transformers and other requirements are hardware dependent so for providers other than Ollama, this only installs some generic dependencies. Check the [notebooks](notebooks/local_llm/) for more details on what is available and how to install it.
 
 
 ### Statistics
@@ -112,10 +100,8 @@ $ poetry update
 
 To install all dependencies (with all extra dependencies) into an isolated virtual environment:
 
-> Append `--sync` to uninstall dependencies that are no longer in use from the virtual environment.
-
 ```bash
-$ poetry install --all-extras
+$ poetry sync --all-extras
 ```
 
 To [activate](https://python-poetry.org/docs/basic-usage#activating-the-virtual-environment) the
@@ -171,7 +157,7 @@ Automated code quality checks are performed using
 environments and run commands based on [`noxfile.py`](./noxfile.py) for unit testing, PEP 8 style
 guide checking, type checking and documentation generation.
 
-> Note: `nox` is installed into the virtual environment automatically by the `poetry install`
+> Note: `nox` is installed into the virtual environment automatically by the `poetry sync`
 > command above. Run `poetry shell` to activate the virtual environment.
 
 To run all default sessions:

{not_again_ai-0.14.0 → not_again_ai-0.16.0}/pyproject.toml

@@ -1,8 +1,10 @@
-[tool.poetry]
+[project]
 name = "not-again-ai"
-version = "0.14.0"
+version = "0.16.0"
 description = "Designed to once and for all collect all the little things that come up over and over again in AI projects and put them in one place."
-authors = ["DaveCoDev <dave.co.dev@gmail.com>"]
+authors = [
+    { name = "DaveCoDev", email = "dave.co.dev@gmail.com" }
+]
 license = "MIT"
 readme = "README.md"
 repository = "https://github.com/DaveCoDev/not-again-ai"
@@ -19,41 +21,47 @@ classifiers = [
     "Programming Language :: Python :: 3.12",
     "Typing :: Typed",
 ]
+requires-python = ">=3.11, <3.13"
+dependencies = [
+    "loguru>=0.7",
+    "pydantic>=2.10"
+]
+
+[project.urls]
+Homepage = "https://github.com/DaveCoDev/not-again-ai"
+Documentation = "https://davecodev.github.io/not-again-ai/"
+Repository = "https://github.com/DaveCoDev/not-again-ai"
+
+[tool.poetry]
+requires-poetry = ">=2.0.1"
+
+[tool.poetry.requires-plugins]
+poetry-plugin-export = ">=1.8"
 
-[tool.poetry.dependencies]
-# Some packages, such as scipy, constrain their upper bound of Python versions they support.
-# Without also constraining the upper bound here, Poetry will not select those versions and will
-# result in an old version being resolved/locked.
-python = "^3.11 || ^3.12"
-
-loguru = { version = "^0.7" }
-pydantic = { version = "^2.9"}
-
-
-# Optional dependencies are defined here, and groupings are defined below.
-azure-ai-inference = { version = "==1.0.0b5", optional = true }
-azure-identity = { version = "^1.19", optional = true }
-jinja2 = { version = "^3.1", optional = true }
-numpy = { version = "^2.1", optional = true }
-ollama = { version = "^0.3", optional = true }
-openai = { version = "^1.52", optional = true }
-pandas = { version = "^2.2", optional = true }
-pytest-playwright = { version = "^0.5", optional = true }
-python-liquid = { version = "^1.12", optional = true }
-scipy = { version = "^1.14", optional = true }
-scikit-learn = { version = "^1.5", optional = true }
-seaborn = { version = "^0.13", optional = true }
-tiktoken = { version = "^0.8", optional = true }
-transformers = { version = "^4.45", optional = true }
-
-[tool.poetry.extras]
-data = ["pytest-playwright"]
-llm = ["azure-ai-inference", "azure-identity", "openai", "python-liquid", "tiktoken"]
-local_llm = ["jinja2", "ollama", "transformers"]
-statistics = ["numpy", "scikit-learn", "scipy"]
-viz = ["numpy", "pandas", "seaborn"]
-
-[tool.poetry.dev-dependencies]
+[project.optional-dependencies]
+data = [
+    "playwright>=1.49",
+    "pytest-playwright>=0.7"
+]
+llm = [
+    "azure-identity>=1.19",
+    "ollama>=0.4",
+    "openai>=1",
+    "python-liquid>=1.12",
+    "tiktoken>=0.8"
+]
+statistics = [
+    "numpy>=2.2",
+    "scikit-learn>=1.6",
+    "scipy>=1.15"
+]
+viz = [
+    "numpy>=2.2",
+    "pandas>=2.2",
+    "seaborn>=0.13"
+]
+
+[tool.poetry.group.dev.dependencies]
 ipykernel = "*"
 ipywidgets = "*"
 
@@ -87,11 +95,8 @@ mkdocs-literate-nav = "*"
 [tool.poetry.group.typos.dependencies]
 typos = "*"
 
-[tool.poetry.scripts]
-not-again-ai = "not_again_ai.cli:entry_point"
-
 [build-system]
-requires = ["poetry-core"]
+requires = ["poetry-core>=2.0.0,<3.0.0"]
 build-backend = "poetry.core.masonry.api"
 
 [tool.mypy]
@@ -117,6 +122,8 @@ select = [
     "B", # flake8-bugbear
     "C4", # flake8-comprehensions
     "ISC", # flake8-implicit-str-concat
+    "PIE", # flake8-pie
+    "PT", # flake-pytest-style
     "PTH", # flake8-use-pathlib
     "SIM", # flake8-simplify
     "TID", # flake8-tidy-imports
@@ -141,7 +148,7 @@ filterwarnings = [
     # When running tests, treat warnings as errors (e.g. -Werror).
     # See: https://docs.pytest.org/en/latest/reference/reference.html#confval-filterwarnings
     "error",
-    # Add additional warning supressions as needed here. For example, if a third-party library
+    # Add additional warning suppressions as needed here. For example, if a third-party library
     # is throwing a deprecation warning that needs to be fixed upstream:
     # "ignore::DeprecationWarning:typer",
     "ignore::pytest.PytestUnraisableExceptionWarning"

not_again_ai-0.16.0/src/not_again_ai/llm/chat_completion/__init__.py

@@ -0,0 +1,4 @@
+from not_again_ai.llm.chat_completion.interface import chat_completion
+from not_again_ai.llm.chat_completion.types import ChatCompletionRequest
+
+__all__ = ["ChatCompletionRequest", "chat_completion"]

not_again_ai-0.16.0/src/not_again_ai/llm/chat_completion/interface.py

@@ -0,0 +1,32 @@
+from collections.abc import Callable
+from typing import Any
+
+from not_again_ai.llm.chat_completion.providers.ollama_api import ollama_chat_completion
+from not_again_ai.llm.chat_completion.providers.openai_api import openai_chat_completion
+from not_again_ai.llm.chat_completion.types import ChatCompletionRequest, ChatCompletionResponse
+
+
+def chat_completion(
+    request: ChatCompletionRequest,
+    provider: str,
+    client: Callable[..., Any],
+) -> ChatCompletionResponse:
+    """Get a chat completion response from the given provider. Currently supported providers:
+    - `openai` - OpenAI
+    - `azure_openai` - Azure OpenAI
+    - `ollama` - Ollama
+
+    Args:
+        request: Request parameter object
+        provider: The supported provider name
+        client: Client information, see the provider's implementation for what can be provided
+
+    Returns:
+        ChatCompletionResponse: The chat completion response.
+    """
+    if provider == "openai" or provider == "azure_openai":
+        return openai_chat_completion(request, client)
+    elif provider == "ollama":
+        return ollama_chat_completion(request, client)
+    else:
+        raise ValueError(f"Provider {provider} not supported")

not_again_ai-0.16.0/src/not_again_ai/llm/chat_completion/providers/ollama_api.py

@@ -0,0 +1,227 @@
+from collections.abc import Callable
+import json
+import os
+import re
+import time
+from typing import Any, Literal, cast
+
+from loguru import logger
+from ollama import ChatResponse, Client, ResponseError
+
+from not_again_ai.llm.chat_completion.types import (
+    AssistantMessage,
+    ChatCompletionChoice,
+    ChatCompletionRequest,
+    ChatCompletionResponse,
+    Function,
+    ToolCall,
+)
+
+OLLAMA_PARAMETER_MAP = {
+    "frequency_penalty": "repeat_penalty",
+    "max_completion_tokens": "num_predict",
+    "context_window": "num_ctx",
+    "n": None,
+    "tool_choice": None,
+    "reasoning_effort": None,
+    "parallel_tool_calls": None,
+    "logit_bias": None,
+    "top_logprobs": None,
+    "presence_penalty": None,
+}
+
+
+def validate(request: ChatCompletionRequest) -> None:
+    if request.json_mode and request.structured_outputs is not None:
+        raise ValueError("json_schema and json_mode cannot be used together.")
+
+    # Check if any of the parameters set to OLLAMA_PARAMETER_MAP are not None
+    for key, value in OLLAMA_PARAMETER_MAP.items():
+        if value is None and getattr(request, key) is not None:
+            logger.warning(f"Parameter {key} is not supported by Ollama and will be ignored.")
+
+    # If "stop" is not None, check if it is just a string
+    if isinstance(request.stop, list):
+        logger.warning("Parameter 'stop' needs to be a string and not a list. It will be ignored.")
+        request.stop = None
+
+
+def ollama_chat_completion(
+    request: ChatCompletionRequest,
+    client: Callable[..., Any],
+) -> ChatCompletionResponse:
+    validate(request)
+
+    kwargs = request.model_dump(mode="json", exclude_none=True)
+
+    # For each key in OLLAMA_PARAMETER_MAP
+    # If it is not None, set the key in kwargs to the value of the corresponding value in OLLAMA_PARAMETER_MAP
+    # If it is None, remove that key from kwargs
+    for key, value in OLLAMA_PARAMETER_MAP.items():
+        if value is not None and key in kwargs:
+            kwargs[value] = kwargs.pop(key)
+        elif value is None and key in kwargs:
+            del kwargs[key]
+
+    # If json_mode is True, set the format to json
+    json_mode = kwargs.get("json_mode", None)
+    if json_mode:
+        kwargs["format"] = "json"
+        kwargs.pop("json_mode")
+    elif json_mode is not None and not json_mode:
+        kwargs.pop("json_mode")
+
+    # If structured_outputs is not None, set the format to structured_outputs
+    if kwargs.get("structured_outputs", None):
+        # Check if the schema is in the OpenAI and pull out the schema
+        if "schema" in kwargs["structured_outputs"]:
+            kwargs["format"] = kwargs["structured_outputs"]["schema"]
+            kwargs.pop("structured_outputs")
+        else:
+            kwargs["format"] = kwargs.pop("structured_outputs")
+
+    option_fields = [
+        "mirostat",
+        "mirostat_eta",
+        "mirostat_tau",
+        "num_ctx",
+        "repeat_last_n",
+        "repeat_penalty",
+        "temperature",
+        "seed",
+        "stop",
+        "tfs_z",
+        "num_predict",
+        "top_k",
+        "top_p",
+        "min_p",
+    ]
+    # For each field in option_fields, if it is in kwargs, make it under an options dictionary
+    options = {}
+    for field in option_fields:
+        if field in kwargs:
+            options[field] = kwargs.pop(field)
+    kwargs["options"] = options
+
+    for message in kwargs["messages"]:
+        role = message.get("role", None)
+        # For each ToolMessage, remove the name field
+        if role is not None and role == "tool":
+            message.pop("name")
+
+        # For each AssistantMessage with tool calls, remove the id field
+        if role is not None and role == "assistant" and message.get("tool_calls", None):
+            for tool_call in message["tool_calls"]:
+                tool_call.pop("id")
+
+        # Content and images need to be separated
+        images = []
+        content = ""
+        if isinstance(message["content"], list):
+            for item in message["content"]:
+                if item["type"] == "image_url":
+                    image_url = item["image_url"]["url"]
+                    # Remove the data URL prefix if present
+                    if image_url.startswith("data:"):
+                        image_url = image_url.split("base64,", 1)[1]
+                    images.append(image_url)
+                else:
+                    content += item["text"]
+        else:
+            content = message["content"]
+
+        message["content"] = content
+        if len(images) > 1:
+            images = images[:1]
+            logger.warning("Ollama model only supports a single image per message. Using only the first images.")
+        message["images"] = images
+
+    try:
+        start_time = time.time()
+        response: ChatResponse = client(**kwargs)
+        end_time = time.time()
+        response_duration = round(end_time - start_time, 4)
+    except ResponseError as e:
+        # If the error says "model 'model' not found" use regex then raise a more specific error
+        expected_pattern = f"model '{request.model}' not found"
+        if re.search(expected_pattern, e.error):
+            raise ResponseError(f"Model '{request.model}' not found.") from e
+        else:
+            raise ResponseError(e.error) from e
+
+    errors = ""
+
+    # Handle tool calls
+    tool_calls: list[ToolCall] | None = None
+    if response.message.tool_calls:
+        parsed_tool_calls: list[ToolCall] = []
+        for tool_call in response.message.tool_calls:
+            tool_name = tool_call.function.name
+            if request.tools and tool_name not in [tool["function"]["name"] for tool in request.tools]:
+                errors += f"Tool call {tool_call} has an invalid tool name: {tool_name}\n"
+            tool_args = tool_call.function.arguments
+            parsed_tool_calls.append(
+                ToolCall(
+                    id="",
+                    function=Function(
+                        name=tool_name,
+                        arguments=tool_args,
+                    ),
+                )
+            )
+        tool_calls = parsed_tool_calls
+
+    json_message = None
+    if (request.json_mode or (request.structured_outputs is not None)) and response.message.content:
+        try:
+            json_message = json.loads(response.message.content)
+        except json.JSONDecodeError:
+            errors += "Message failed to parse into JSON\n"
+
+    finish_reason = cast(
+        Literal["stop", "length", "tool_calls", "content_filter"],
+        "stop" if response.done_reason is None else response.done_reason or "stop",
+    )
+
+    choice = ChatCompletionChoice(
+        message=AssistantMessage(
+            content=response.message.content or "",
+            tool_calls=tool_calls,
+        ),
+        finish_reason=finish_reason,
+        json_message=json_message,
+    )
+
+    return ChatCompletionResponse(
+        choices=[choice],
+        errors=errors.strip(),
+        completion_tokens=response.get("eval_count", -1),
+        prompt_tokens=response.get("prompt_eval_count", -1),
+        response_duration=response_duration,
+    )
+
+
+def ollama_client(host: str | None = None, timeout: float | None = None) -> Callable[..., Any]:
+    """Create an Ollama client instance based on the specified host or will read from the OLLAMA_HOST environment variable.
+
+    Args:
+        host (str, optional): The host URL of the Ollama server.
+        timeout (float, optional): The timeout for requests
+
+    Returns:
+        Client: An instance of the Ollama client.
+
+    Examples:
+        >>> client = client(host="http://localhost:11434")
+    """
+    if host is None:
+        host = os.getenv("OLLAMA_HOST")
+        if host is None:
+            logger.warning("OLLAMA_HOST environment variable not set, using default host: http://localhost:11434")
+            host = "http://localhost:11434"
+
+    def client_callable(**kwargs: Any) -> Any:
+        client = Client(host=host, timeout=timeout)
+        return client.chat(**kwargs)
+
+    return client_callable