Class Task

src/benchmark/types.py:9–25

Data model for a single benchmark task

from src.dynamic import dynamic_manager

class Task(BaseModel):
    """Data model for a single benchmark task"""
    model_config = ConfigDict(arbitrary_types_allowed=True, extra="allow")
    
    # Input
    task_id: str = Field(description="Unique identifier for the task")
    input: str = Field(description="The input prompt/question for the task")
    system_prompt: Optional[str] = Field(default=None, description="The system prompt for the task")
    ground_truth: Optional[Any] = Field(default=None, description="The expected correct answer")
    
    # Output
    reasoning: Optional[str] = Field(default=None, description="The reasoning process")
    result: Optional[Any] = Field(default=None, description="The final answer")
    time: Optional[float] = Field(default=0.0, description="The time taken to complete the task in seconds")
    score: Optional[float] = Field(default=0.0, description="The score of the task")
    
    extra: Optional[Dict[str, Any]] = Field(default=None, description="Additional task-specific metadata")

class Result(BaseModel):
    """Data model for the evaluation result of a single task"""