Prompt Setup¶
- pydantic model ibm_watsonx_gov.entities.prompt_setup.PromptSetup¶
Bases:
BaseModel
Show JSON schema
{ "title": "PromptSetup", "type": "object", "properties": { "task_type": { "$ref": "#/$defs/TaskType", "description": "Prompt task type", "examples": [ "retrieval_augmented_generation", "classification" ] }, "question_field": { "anyOf": [ { "type": "string" }, { "type": "null" } ], "default": null, "description": "Question column name from the input", "examples": [ "question" ], "title": "Question Field" }, "context_fields": { "anyOf": [ { "items": { "type": "string" }, "type": "array" }, { "type": "null" } ], "default": null, "description": "List of context column names from the input", "examples": [ [ "context1", "context2" ] ], "title": "Context Fields" }, "prediction_field": { "anyOf": [ { "type": "string" }, { "type": "null" } ], "default": "generated_text", "description": "Prediction field name from the input", "examples": [ "generated_text" ], "title": "Prediction Field" }, "label_column": { "description": "reference output column name", "examples": [ "answer", "ground_truth" ], "title": "Label Column", "type": "string" }, "input_data_type": { "$ref": "#/$defs/InputDataType", "default": "unstructured_text", "description": "Prompt input data type", "examples": [ "unstructured_text", "structured" ] } }, "$defs": { "InputDataType": { "description": "Supported input data types", "enum": [ "structured", "unstructured_text", "unstructured_image", "multimodal" ], "title": "InputDataType", "type": "string" }, "TaskType": { "description": "Supported task types for generative AI models", "enum": [ "question_answering", "classification", "summarization", "generation", "extraction", "retrieval_augmented_generation" ], "title": "TaskType", "type": "string" } }, "required": [ "task_type", "label_column" ] }
- Config:
arbitrary_types_allowed: bool = True
- Fields:
- Validators:
require_context_and_question_columns_for_rag
»all fields
- field context_fields: Annotated[list[str] | None, FieldInfo(annotation=NoneType, required=True, description='List of context column names from the input', examples=[['context1', 'context2']])] = None¶
List of context column names from the input
- Validated by:
- field input_data_type: STRUCTURED: 'structured'>])] = InputDataType.TEXT¶
Prompt input data type
- Validated by:
- field label_column: Annotated[str, FieldInfo(annotation=NoneType, required=True, description='reference output column name', examples=['answer', 'ground_truth'])] [Required]¶
reference output column name
- Validated by:
- field prediction_field: Annotated[str | None, FieldInfo(annotation=NoneType, required=False, default='generated_text', description='Prediction field name from the input', examples=['generated_text'])] = 'generated_text'¶
Prediction field name from the input
- Validated by:
- field question_field: Annotated[str | None, FieldInfo(annotation=NoneType, required=True, description='Question column name from the input', examples=['question'])] = None¶
Question column name from the input
- Validated by:
- field task_type: CLASSIFICATION: 'classification'>])] [Required]¶
Prompt task type
- Validated by:
- validator require_context_and_question_columns_for_rag » all fields¶