OpenGradient Docs

Appearance

opengradient / alphasense

Package opengradient.alphasense ​

OpenGradient AlphaSense Tools

Functions ​

create_read_workflow_tool() ​

python
def create_read_workflow_tool(tool_type: [ToolType](./types), workflow_contract_address: str, tool_name: str, tool_description: str, alpha: Optional[`Alpha`] = None, output_formatter: Callable[..., str] = <function <lambda>>) ‑> Union[`BaseTool`, Callable]

Creates a tool that reads results from a workflow contract on OpenGradient.

This function generates a tool that can be integrated into either a LangChain pipeline or a Swarm system, allowing the workflow results to be retrieved and formatted as part of a chain of operations.

Arguments

  • tool_type (ToolType): Specifies the framework to create the tool for. Use ToolType.LANGCHAIN for LangChain integration or ToolType.SWARM for Swarm integration.
  • workflow_contract_address (str): The address of the workflow contract from which to read results.
  • tool_name (str): The name to assign to the created tool. This will be used to identify and invoke the tool within the agent.
  • tool_description (str): A description of what the tool does and how it processes the workflow results.
  • alpha (Alpha, optional): The alpha namespace from an initialized OpenGradient client (client.alpha). If not provided, falls back to the global client set via opengradient.init().
  • output_formatter (Callable[..., str], optional): A function that takes the workflow output and formats it into a string. This ensures the output is compatible with the tool framework. Default returns string as is.

Returns

BaseTool: For ToolType.LANGCHAIN, returns a LangChain StructuredTool. Callable: For ToolType.SWARM, returns a decorated function with appropriate metadata.

Raises

  • ValueError: If an invalid tool_type is provided.

create_run_model_tool() ​

python
def create_run_model_tool(tool_type: [ToolType](./types), model_cid: str, tool_name: str, model_input_provider: Callable[..., Dict[str, Union[str, int, float, List, `ndarray`]]], model_output_formatter: Callable[[`InferenceResult`], str], inference: Optional[`Alpha`] = None, tool_input_schema: Optional[Type[`BaseModel`]] = None, tool_description: str = 'Executes the given ML model', inference_mode: `InferenceMode` = InferenceMode.VANILLA) ‑> Union[`BaseTool`, Callable]

Creates a tool that wraps an OpenGradient model for inference.

This function generates a tool that can be integrated into either a LangChain pipeline or a Swarm system, allowing the model to be executed as part of a chain of operations. The tool uses the provided input_getter function to obtain the necessary input data and runs inference using the specified OpenGradient model.

Arguments

  • tool_type (ToolType): Specifies the framework to create the tool for. Use ToolType.LANGCHAIN for LangChain integration or ToolType.SWARM for Swarm integration.

  • model_cid (str): The CID of the OpenGradient model to be executed.

  • tool_name (str): The name to assign to the created tool. This will be used to identify and invoke the tool within the agent.

  • model_input_provider (Callable): A function that takes in the tool_input_schema with arguments filled by the agent and returns input data required by the model.

      The function should return data in a format compatible with the model's expectations.

  • model_output_formatter (Callable[..., str]): A function that takes the output of the OpenGradient infer method (with type InferenceResult) and formats it into a string.

      This is required to ensure the output is compatible with the tool framework.

  Default returns the InferenceResult object.

  InferenceResult has attributes:
      * transaction_hash (str): Blockchain hash for the transaction
      * model_output (Dict[str, np.ndarray]): Output of the ONNX model

  • inference (Alpha, optional): The alpha namespace from an initialized OpenGradient client (client.alpha). If not provided, falls back to the global client set via opengradient.init().

  • tool_input_schema (Type[BaseModel], optional): A Pydantic BaseModel class defining the input schema.

      For LangChain tools the schema will be used directly. The defined schema will be used as
  input keyword arguments for the `model_input_provider` function. If no arguments are required
  for the `model_input_provider` function then this schema can be unspecified.

  For Swarm tools the schema will be converted to appropriate annotations.

  Default is None -- an empty schema will be provided for LangChain.

  • tool_description (str, optional): A description of what the tool does. Defaults to "Executes the given ML model".

  • inference_mode (InferenceMode, optional): The inference mode to use when running the model. Defaults to VANILLA.

Returns

BaseTool: For ToolType.LANGCHAIN, returns a LangChain StructuredTool. Callable: For ToolType.SWARM, returns a decorated function with appropriate metadata.

Raises

  • ValueError: If an invalid tool_type is provided.

Classes ​

ToolType ​

Indicates the framework the tool is compatible with.

Variables ​

  • static LANGCHAIN
  • static SWARM