feat(llms): option to set a default LLM factory

packages/ragbits-core/src/ragbits/core/config.py

@@ -11,5 +11,8 @@ class CoreConfig(BaseModel):
     # Pattern used to search for prompt files
     prompt_path_pattern: str = "**/prompt_*.py"
 
+    # Path to a function that returns an LLM object, e.g. "my_project.llms.get_llm"
+    default_llm_factory: str | None = None
+
 
 core_config = get_config_instance(CoreConfig, subproject="core")

packages/ragbits-core/src/ragbits/core/llms/factory.py

@@ -0,0 +1,60 @@
+import importlib
+
+from ragbits.core.config import core_config
+from ragbits.core.llms.base import LLM
+from ragbits.core.llms.litellm import LiteLLM
+
+
+def get_llm_from_factory(factory_path: str) -> LLM:
+    """
+    Get an instance of an LLM using a factory function specified by the user.
+
+    Args:
+        factory_path (str): The path to the factory function.
+
+    Returns:
+        LLM: An instance of the LLM.
+    """
+    module_name, function_name = factory_path.rsplit(".", 1)
+    module = importlib.import_module(module_name)
+    function = getattr(module, function_name)
+    return function()
+
+
+def has_default_llm() -> bool:
+    """
+    Check if the default LLM factory is set in the configuration.
+
+    Returns:
+        bool: Whether the default LLM factory is set.
+    """
+    return core_config.default_llm_factory is not None
+
+
+def get_default_llm() -> LLM:
+    """
+    Get an instance of the default LLM using the factory function
+    specified in the configuration.
+
+    Returns:
+        LLM: An instance of the default LLM.
+
+    Raises:
+        ValueError: If the default LLM factory is not set.
+    """
+    factory = core_config.default_llm_factory
+    if factory is None:
+        raise ValueError("Default LLM factory is not set")
+
+    return get_llm_from_factory(factory)
+
+
+def simple_litellm_factory() -> LLM:
+    """
+    A basic LLM factory that creates an LiteLLM instance with the default model,
+    default options, and assumes that the API key is set in the environment.
+
+    Returns:
+        LLM: An instance of the LiteLLM.
+    """
+    return LiteLLM()

packages/ragbits-core/src/ragbits/core/prompt/lab/app.py

@@ -14,8 +14,8 @@
 from rich.console import Console
 
 from ragbits.core.config import core_config
-from ragbits.core.llms import LiteLLM
-from ragbits.core.llms.clients import LiteLLMOptions
+from ragbits.core.llms import LLM
+from ragbits.core.llms.factory import get_llm_from_factory
 from ragbits.core.prompt import Prompt
 from ragbits.core.prompt.discovery import PromptDiscovery
 
@@ -30,14 +30,12 @@ class PromptState:
     Attributes:
         prompts (list): A list containing discovered prompts.
         rendered_prompt (Prompt): The most recently rendered Prompt instance.
-        llm_model_name (str): The name of the selected LLM model.
-        llm_api_key (str | None): The API key for the chosen LLM model.
+        llm (LLM): The LLM instance to be used for generating responses.
     """
 
     prompts: list = field(default_factory=list)
     rendered_prompt: Prompt | None = None
-    llm_model_name: str | None = None
-    llm_api_key: str | None = None
+    llm: LLM | None = None
 
 
 def render_prompt(index: int, system_prompt: str, user_prompt: str, state: PromptState, *args: Any) -> PromptState:
@@ -99,15 +97,17 @@ def send_prompt_to_llm(state: PromptState) -> str:
 
     Returns:
         str: The response generated by the LLM.
-    """
-    assert state.llm_model_name is not None, "LLM model name is not set."
-    llm_client = LiteLLM(model_name=state.llm_model_name, api_key=state.llm_api_key)
 
+    Raises:
+        ValueError: If the LLM model is not configured.
+    """
     assert state.rendered_prompt is not None, "Prompt has not been rendered yet."
+
+    if state.llm is None:
+        raise ValueError("LLM model is not configured.")
+
     try:
-        response = asyncio.run(
-            llm_client.client.call(conversation=state.rendered_prompt.chat, options=LiteLLMOptions())
-        )
+        response = asyncio.run(state.llm.generate_raw(prompt=state.rendered_prompt))
     except Exception as e:  # pylint: disable=broad-except
         response = str(e)
 
@@ -136,7 +136,8 @@ def get_input_type_fields(obj: BaseModel | None) -> list[dict]:
 
 
 def lab_app(  # pylint: disable=missing-param-doc
-    file_pattern: str = core_config.prompt_path_pattern, llm_model: str | None = None, llm_api_key: str | None = None
+    file_pattern: str = core_config.prompt_path_pattern,
+    llm_factory: str | None = core_config.default_llm_factory,
 ) -> None:
     """
     Launches the interactive application for listing, rendering, and testing prompts
@@ -163,9 +164,8 @@ def lab_app(  # pylint: disable=missing-param-doc
     with gr.Blocks() as gr_app:
         prompts_state = gr.State(
             PromptState(
-                llm_model_name=llm_model,
-                llm_api_key=llm_api_key,
                 prompts=list(prompts),
+                llm=get_llm_from_factory(llm_factory) if llm_factory else None,
             )
         )
 
@@ -220,14 +220,15 @@ def show_split(index: int, state: gr.State) -> None:
                                 )
                                 gr.Textbox(label="Rendered User Prompt", value=rendered_user_prompt, interactive=False)
 
-            llm_enabled = state.llm_model_name is not None
+            llm_enabled = state.llm is not None
             prompt_ready = state.rendered_prompt is not None
             llm_request_button = gr.Button(
                 value="Send to LLM",
                 interactive=llm_enabled and prompt_ready,
             )
             gr.Markdown(
-                "To enable this button, select an LLM model when starting the app in CLI.", visible=not llm_enabled
+                "To enable this button set an LLM factory function in CLI options or your pyproject.toml",
+                visible=not llm_enabled,
             )
             gr.Markdown("To enable this button, render a prompt first.", visible=llm_enabled and not prompt_ready)
             llm_prompt_response = gr.Textbox(lines=10, label="LLM response")

packages/ragbits-core/tests/unit/llms/factory/test_get_default_llm.py

@@ -0,0 +1,16 @@
+from ragbits.core.config import core_config
+from ragbits.core.llms.factory import get_default_llm
+from ragbits.core.llms.litellm import LiteLLM
+
+
+def test_get_default_llm(monkeypatch):
+    """
+    Test the get_llm_from_factory function.
+    """
+    monkeypatch.setattr(
+        core_config, "default_llm_factory", "tests.unit.llms.factory.test_get_llm_from_factory.mock_llm_factory"
+    )
+
+    llm = get_default_llm()
+    assert isinstance(llm, LiteLLM)
+    assert llm.model_name == "mock_model"

packages/ragbits-core/tests/unit/llms/factory/test_get_llm_from_factory.py

@@ -0,0 +1,22 @@
+from ragbits.core.llms.factory import get_llm_from_factory
+from ragbits.core.llms.litellm import LiteLLM
+
+
+def mock_llm_factory() -> LiteLLM:
+    """
+    A mock LLM factory that creates a LiteLLM instance with a mock model name.
+
+    Returns:
+        LiteLLM: An instance of the LiteLLM.
+    """
+    return LiteLLM(model_name="mock_model")
+
+
+def test_get_llm_from_factory():
+    """
+    Test the get_llm_from_factory function.
+    """
+    llm = get_llm_from_factory("tests.unit.llms.factory.test_get_llm_from_factory.mock_llm_factory")
+
+    assert isinstance(llm, LiteLLM)
+    assert llm.model_name == "mock_model"

packages/ragbits-core/tests/unit/llms/factory/test_has_default_llm.py

@@ -0,0 +1,20 @@
+from ragbits.core.config import core_config
+from ragbits.core.llms.factory import has_default_llm
+
+
+def test_has_default_llm(monkeypatch):
+    """
+    Test the has_default_llm function when the default LLM factory is not set.
+    """
+    monkeypatch.setattr(core_config, "default_llm_factory", None)
+
+    assert has_default_llm() is False
+
+
+def test_has_default_llm_false(monkeypatch):
+    """
+    Test the has_default_llm function when the default LLM factory is set.
+    """
+    monkeypatch.setattr(core_config, "default_llm_factory", "my_project.llms.get_llm")
+
+    assert has_default_llm() is True