EvalResult

from modal_training_gym.common.eval import EvalResult

Saved results for one evaluation run across a dataset.

Constructor

EvalResult(**data)

Parameter	Type	Default	Description

Attributes

Attribute	Type	Default	Description
eval_id	str
eval_config_id	str
deployment_id	str
created_at	datetime.datetime
status	Literal['running', 'completed', 'failed']
rows	list[modal_training_gym.common.sample.Sample]

Methods

from_id(eval_id: 'str') -> "'EvalResult'"

list_results() -> "list['EvalResult']"

save(self) -> 'None'

to_summary(self) -> 'EvalSummary'

Source: modal_training_gym/common/eval.py