∞ OpenAI Model

Documentation

Class name: LLMOpenAIModel
Category: SALT/Language Toolkit/Loaders
Output node: False

This node facilitates the integration and utilization of OpenAI's language models, enabling the loading and embedding of models for various applications such as text generation, analysis, and more. It abstracts the complexity of directly interacting with OpenAI's API, providing a streamlined interface for embedding and model management.

Input types

Required

model
- Specifies the OpenAI model to be loaded. This selection determines the capabilities and performance of the language model in use.
- Comfy dtype: COMBO[STRING]
- Python dtype: list
api_key
- The API key required for authenticating requests to OpenAI's services. It ensures secure access to the model loading functionality.
- Comfy dtype: STRING
- Python dtype: str
embedding_model
- Defines the specific embedding model to be used alongside the primary language model, enhancing its functionality with additional text embedding capabilities.
- Comfy dtype: COMBO[STRING]
- Python dtype: str

Optional

multimodal
- A boolean flag indicating whether the model to be loaded should support multimodal inputs, allowing for a broader range of input types beyond text.
- Comfy dtype: BOOLEAN
- Python dtype: bool

Output types

llm_model
- Comfy dtype: LLM_MODEL
- Returns the primary language model loaded, including its configuration and capabilities, ready for various applications.
- Python dtype: Dict[str, Any]
embed_model_only
- Comfy dtype: LLM_EMBED_MODEL
- Provides the embedding model loaded separately, detailing its specific features and how it complements the primary language model.
- Python dtype: Dict[str, Any]

Usage tips

Infra type: CPU
Common nodes: unknown

Source code

class LLMOpenAI:
    """
    @Documentation: https://docs.llamaindex.ai/en/stable/api_reference/llms/openai/
    """
    @classmethod
    def INPUT_TYPES(cls):
        return {
            "required": {
                "model": (list(ALL_AVAILABLE_MODELS),),
                "api_key": ("STRING", {
                    "multiline": False, 
                    "dynamicPrompts": False, 
                    "default": os.environ.get("OPENAI_API_KEY", "")
                }),
                "embedding_model": (
                    sorted([x.value for x in OpenAIEmbeddingModelType]),
                    {"default": "text-embedding-ada-002"},
                ),
            },
            "optional": {
                "multimodal": ("BOOLEAN", {"default": False})
            }
        }

    RETURN_TYPES = ("LLM_MODEL", "LLM_EMBED_MODEL")
    RETURN_NAMES = ("llm_model", "embed_model_only")

    FUNCTION = "load_model"
    CATEGORY = f"{MENU_NAME}/{SUB_MENU_NAME}/Loaders"

    def load_model(self, model:str, embedding_model:str, api_key:str, multimodal:bool = False) -> Dict[str, Any]:
        llm = OpenAI(model=model) if not multimodal else OpenAIMultiModal(model=model)
        llm.api_key = api_key
        embed_model = OpenAIEmbedding(model_name=embedding_model, api_key=api_key,)
        return ({"llm":llm, "llm_name": model, "embed_model":embed_model, "embed_name": embedding_model}, {"embed_model": embed_model, "embed_name": embedding_model})

3.4 KiB Raw Permalink Blame History