basecls.engine.evaluator# class basecls.engine.evaluator.AccEvaluator[源代码]# 基类：BaseEvaluator Classification evaluator with top-1 and top-5 accuracy. ResultType# Tuple[int, float, float] 的别名 preprocess(input_data)[源代码]# Preprocess input data per batch. 参数 input_data (Sequence[ndarray]) – input data. 返回类型 Tensor 返回 Preprocessed input data. postprocess(model_outputs, input_data)[源代码]# Postprocess model outputs with input data per batch. 参数 model_outputs (Tensor) – model outputs. input_data (Sequence[ndarray]) – input data. 返回类型 Tuple[int, float, float] 返回 A tuple that (batch size, top-1 accuracy per batch, top-5 accuracy per batch). evaluate(results)[源代码]# Evaluation function. 参数 results (Iterable[Tuple[int, float, float]]) – all results.